Drug-induced liver injury during the intensive phase of a fixed-dose daily regimen of anti-tuberculosis therapy: a cohort study from southern India

BACKGROUND: Anti-tuberculosis drug-induced liver injury (AT-DILI) is a major adverse event associated with significant morbidity and mortality. Limited real-world evidence exists in India on the incidence and risk factors of AT-DILI after switching over from an intermittent to a daily regimen involving the 4-drug fixed-dose combination (4FDC-ATT) in India. This study aimed to estimate the incidence and predictors of AT-DILI. METHODS: A cohort study was conducted among adults with TB (≥ 18y) initiated on 4FDC-ATT between 01 September and 15 December 2024 across 5 tertiary care centers and 13 peripheral health institutes (PHI) in 2 districts of Tamil Nadu, a southern Indian state. The AT-DILI was diagnosed according to the Indian Drug-Induced Liver Injury Network criteria, which consider higher serum concentrations of aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, total bilirubin, and the presence or absence of clinical symptoms (nausea, vomiting, abdominal pain, or loss of appetite). The Cox proportional hazards model was used to calculate the adjusted hazard ratio for predictors of AT-DILI. RESULTS: Of 435 participants recruited at the baseline, 265 were followed up till 8 weeks. Baseline characteristics of participants who completed the study or were lost to follow-up were comparable. A majority of AT-DILI cases occurred within 2 weeks, and the overall prevalence at eight weeks is 4.9% (95% CI: 3.6-6.2). A significantly higher proportion of the AT-DILI patient group reported clinical symptoms at baseline than the others. Smokeless tobacco use (aHR = 4.9, 95% CI: 1.2-20.05) and pedal edema (aHR = 13.7, 95% CI: 2.5-75.6) at baseline were observed as significant predictors of AT-DILI. CONCLUSIONS: Early onset of AT-DILI necessitates baseline and scheduled assessment of liver functions during the intensive phase, particularly among those with symptoms. Integrating LFT monitoring into the national TB program is resource-intensive, but it can be scaled up in a phased manner. Future studies should evaluate the feasibility and cost-effectiveness, as it could enhance patient safety and treatment adherence.