Abstract
BACKGROUND: In STAND and SimpliciTB, clinical trials for drug-susceptible TB, regimens containing pretomanid, pyrazinamide, and other agents (PaZX) had more hepatotoxicity than the standard-of-care regimen of isoniazid, rifampicin, pyrazinamide, and ethambutol (HRZE). In Nix-TB and ZeNix, clinical trials for drug-resistant TB, the regimen of bedaquiline, pretomanid, and linezolid (BPaL) demonstrated a favorable benefit-risk profile. We compare the hepatic safety of HRZE, PaZX, and BPaL in their respective populations. METHODS: In this post-hoc analysis of data from six clinical trials, rates of treatment-emergent elevations of alanine transaminase (ALT) during the first 8 weeks of treatment were estimated by Kaplan-Meier (KM) analysis and compared via log-rank testing and Cox modeling. RESULTS: The KM-estimated probabilities of treatment-emergent ALT elevations greater than 3x the upper limit of normal (>3xULN) were 5.36%, 12.7%, and 11.4% for HRZE, PaZX, and BPaL, respectively. The only significant (p < 0.05) difference was HRZE versus PaZX. The probabilities of ALT elevations >8xULN were 2.68%, 4.58%, and 1.05%, with the only significant difference being PaZX versus BPaL. CONCLUSIONS: BPaL and HRZE have similar hepatic safety profiles in their respective populations. Pretomanid and pyrazinamide should be co-administered only when the benefit outweighs the risk.