Treatment outcomes among patients admitted to hospital with antiretroviral and/or antituberculosis drug-induced liver injury
Ruchika Mehta, Prudence Ive, Denise Evans, Colin Menezes
South African Medical Journal · 2021-04
Abstract
BACKGROUND: South Africa (SA) has among the highest rates of HIV and tuberculosis (TB) in the world. Antituberculosis and antiretroviral treatment (ART) can cause drug-induced liver injury (DILI), consequences of which are disease relapse, treatment failure and drug resistance. OBJECTIVES: To: (i) determine the demographics of patients with DILI and the proportion of patients on antituberculosis drugs v. antiretroviral therapy or both; (ii) determine the median time to DILI after starting medication, and patterns of clinical presentation; (iii) determine the numbers of patients successfully re-challenged to initial therapy as inpatients; and (iv) determine the in-hospital mortality rate and predictors of all-cause mortality. METHODS: This was a retrospective record review of adult patients with DILI admitted to a tertiary hospital in Johannesburg, SA, between October 2015 and February 2017. Data on drug history, biochemical investigations and relevant imaging were collected. RESULTS: The total sample was 129 records: 79 (61.2%) were males, 46 (35.7%) had TB DILI, 29 (22.5%) had ART DILI, and 54 (41.9%) had mixed TB/ART DILI. Only 7.4% (2/27) of those with ART DILI and 30.6% (11/36) with TB DILI were re-challenged to their original regimen by discharge. Patients were followed from admission until the earlier of death (10 with TB DILI, 2 with ART DILI and 9 with mixed DILI) or discharge (after a median (interquartile range) of 14.0 (9 - 23) days). In adjusted analysis, severe DILI at admission predicted all-cause mortality (adjusted hazard ratio 8.58; 95% confidence interval 1.13 - 65.4). CONCLUSIONS: This study is one of only a few analyses of hospitalised patients with DILI in SA. Among those with severe DILI, outcomes are poor, the majority cannot tolerate standard regimens, and mortality is high.
MeSH terms
- Medicine
- Interquartile range
- Internal medicine
- Retrospective cohort study
- Hazard ratio
- Tuberculosis
- Regimen
- Confidence interval
- Mortality rate
- Liver injury
- Pediatrics