BACKGROUND & AIMS: Drugs producing acute liver failure (ALF) are uncommon and vary geographically. Here we review the implicated drugs, clinical features, laboratory characteristics and outcome of patients with drug-induced ALF (DIALF). We analysed the predictors of mortality and their relationship with MELD, King's College criteria (KCC) and ALFSG prognostic index. METHODS: We identified DIALF patients from our drug-induced liver injury (DILI) registry (1997-2017). RUCAM was used for case adjudication. Patients who fulfilled criteria for acute liver failure and drug-induced liver injury were included. Primary outcome measure was spontaneous survival or death. RESULTS: There were 128 cases of DIALF (14%) among 905 patients with DILI. Mean age was 38 years, 68 (53%) female and 21(16.4%) children <18 years. Combination anti-TB drugs (ATD) (n = 92, 72.4%) accounted for a majority of DIALF. Others were anti-epileptic drugs (AED, n = 11, 10%), dapsone (n = 7, 5.5%), hormones (n = 2), ferrous sulphate overdose (n = 2), acetaminophen (APAP) (n = 2), antiretroviral (n = 2), CAM (N = 2), chemotherapy agents (N = 3), amoxicillin-clavulanic acid (n = 2) and others (n = 3). Forty-four patients (34%) recovered spontaneously and 84(66%) including 13 children (62%) died. Females, ascites, albumin, bilirubin, INR and MELD were significantly associated with mortality. Mortality was 79% for ATD and 100% for APAP and iron overdose. Area under ROC was 0.76 for MELD and ALFSG index and 0.51 for KCC. CONCLUSIONS: Fourteen percent of DILI resulted in DIALF. ATD, AED, dapsone and antiretroviral drugs are most common agents. Spontaneous survival was only 34% with an even higher mortality with ATD. Non-ATD and non-APAP drugs had a better survival (51%).INR and MELD predicted mortality.
BACKGROUND & AIMS: Drugs producing acute liver failure (ALF) are uncommon and vary geographically. Here we review the implicated drugs, clinical features, laboratory characteristics and outcome of patients with drug-induced ALF (DIALF). We analysed the predictors of mortality and their relationship with MELD, King's College criteria (KCC) and ALFSG prognostic index. METHODS: We identified DIALF patients from our drug-induced liver injury (DILI) registry (1997-2017). RUCAM was used for case adjudication. Patients who fulfilled criteria for acute liver failure and drug-induced liver injury were included. Primary outcome measure was spontaneous survival or death. RESULTS: There were 128 cases of DIALF (14%) among 905 patients with DILI. Mean age was 38 years, 68 (53%) female and 21(16.4%) children <18 years. Combination anti-TB drugs (ATD) (n = 92, 72.4%) accounted for a majority of DIALF. Others were anti-epileptic drugs (AED, n = 11, 10%), dapsone (n = 7, 5.5%), hormones (n = 2), ferrous sulphateoverdose (n = 2), acetaminophen (APAP) (n = 2), antiretroviral (n = 2), CAM (N = 2), chemotherapy agents (N = 3), amoxicillin-clavulanic acid (n = 2) and others (n = 3). Forty-four patients (34%) recovered spontaneously and 84(66%) including 13 children (62%) died. Females, ascites, albumin, bilirubin, INR and MELD were significantly associated with mortality. Mortality was 79% for ATD and 100% for APAP and ironoverdose. Area under ROC was 0.76 for MELD and ALFSG index and 0.51 for KCC. CONCLUSIONS: Fourteen percent of DILI resulted in DIALF. ATD, AED, dapsone and antiretroviral drugs are most common agents. Spontaneous survival was only 34% with an even higher mortality with ATD. Non-ATD and non-APAP drugs had a better survival (51%).INR and MELD predicted mortality.