TB Research

Genetic and clinical risk factors for anti-tuberculosis drug-induced liver injury: insights from a prospective cohort study in central Ethiopia

Caroline Klindt, André Fuchs, Kristina Behnke, Carola Dröge, Kirsten Alexandra Eberhardt, H Orth, Frieder Pfäfflin, Andreas Schönfeld, et al. (15 authors)

Infection · 2025-09

Abstract

PURPOSE: Drug-induced liver injury (DILI) is a relevant adverse event of tuberculosis treatment (TBT) especially in sub-Saharan Africa, but data remains limited. Genetic hepatic transport proteins polymorphisms (HTPP) are potential contributors. This study aimed to assess frequency and timing of DILI, identify risk factors, and explore the association of HTPP with DILI risk in Ethiopian TBT-patients. METHODS: In this prospective study, 424 confirmed tuberculosis patients in Ethiopian were recruited before initiation of TBT. Liver function tests were conducted during the first 8 weeks of treatment. Baseline evaluations included sociodemographic-, lifestyle- and clinical data including testing for viral co-infections, and HTPP as well as liver stiffness measurement by transient elastography (TE). Multivariable logistic regression, Cox proportional hazards models, and Fine and Gray competing risks analyses were employed for statistical analysis. RESULTS: Cumulative DILI incidence was 16.0% with 4.2% classified as severe occurring most commonly within the first two weeks. Urban residence (OR 2.00, 95% CI 1.03-3.84; HR 1.80, 95% CI 1.00-3.22) was associated with increased DILI risk. In the competing risks model, urban residence (sHR 6.26, p = 0.010) and pathologic TE (sHR 5.23, p = 0.005) predicted severe DILI. The investigated HTPPs were not significantly associated with DILI. CONCLUSION: DILI is a common early complication of TBT in Ethiopian patients. Assessment of sociodemographic factors and TE before TBT may help identify high-risk individuals and offers a pragmatic approach for DILI management in resource-limited settings.

MeSH terms

  • Medicine
  • Internal medicine
  • Tuberculosis
  • Prospective cohort study
  • Logistic regression
  • Proportional hazards model
  • Transient elastography
  • Incidence (geometry)
  • Environmental health