Effectiveness and cardiovascular safety of delamanid-containing regimens in adults with multidrug-resistant or extensively drug-resistant tuberculosis: A nationwide cohort study from Belarus, 2016-18

To address the sub-optimal treatment outcomes among patients with multidrug-resistant (MDR-TB) and extensively drug-resistant tuberculosis (XDR-TB), the National TB Programme in Belarus started using new drugs such as bedaquiline and delamanid in 2015-16. In this study, we assessed cardiovascular safety and effectiveness (culture conversion, treatment outcomes and post-treatment recurrence) of delamanid-containing regimens among adults (>18 years) with MDR-TB or XDR-TB from June 2016 to February 2018. This was a nationwide cohort study involving analysis of routinely collected programme data from the national and six regional TB hospitals. Cardiovascular adverse events (AEs) were classified as serious or not, based on international guidelines. We conducted Cox proportional hazards regression and calculated adjusted hazards ratio(aHR) and 95% confidence intervals(CI) to evaluate factors associated with AEs and unsuccessful treatment outcomes (death, failure and lost-to-follow-up). Of 125 patients enrolled (35, 28% females; mean age 43 years), 85(68%) had XDR-TB. All the patients received delamanid and 20 patients received both delamanid and bedaquiline. Cardiovascular AEs (177 episodes in total), were observed in the majority (73%) of patients but were mild and managed easily. The most common cardiovascular AEs were QTcF prolongation (64/177, 36%) and other electrocardiography (ECG) abnormalities (40/177, 23%). There were two instances of serious AEs leading to death, both of which were not related to delamanid. In multivariable analysis, male sex (aHR 0.72; 95% CI 0.51-0.99), and baseline ECG abnormalities (aHR 1.68; 95% CI 1.19-2.36) were associated with cardiovascular AEs. Median time to culture conversion was 1.1 months (interquartile range: 1.0-2.1). Culture conversion was observed in 115 (92%) patients at six months of treatment and 110 (88%) completed the treatment successfully. Loss to follow-up, failure and death were observed in 6%, 4% and 2% patients respectively. Among those assessed at 12 months post-treatment (n=33), recurrence was seen in one patient. The only factor associated with unsuccessful treatment outcomes in multivariable analysis was baseline Hepatitis C co-infection (aHR 3.61; 95% CI 1.09-11.95). In conclusion, treatment using delamanid-containing regimens was effective and had a favourable safety profile. We hope our findings inform the development of national clinical guidelines and scale-up of new drugs in other countries.