Outcome and determinants of mortality in 269 patients with combination anti-tuberculosis drug-induced liver injury.

BACKGROUND AND AIM: Worldwide anti-tuberculosis (TB) drug-induced liver disease (DILI) is an important cause of hepatotoxicity, and drug-induced acute liver failure (ALF). Reported series on anti-TB DILI are limited by a mix of cases with mild transaminase elevation or adaptation. Our aim was to analyze the clinical features, laboratory characteristics, outcome, and determine predictors of 90-day mortality. METHODS: Single center analysis of consecutive cases of anti-TB DILI following combination anti-TB drugs exposure from 1997-2011. RESULTS: Of the 269 patients, 191 (71%) experienced jaundice and 69 (25.7%) accounted for ALF. The mean age and treatment duration was 41.3 years and 1.9 months, respectively; males constituted 55.7%. DILI occurred throughout the course of treatment; three-quarters occurred within the first 2 months. HIV infection was present in 21 (7.8%). The 90-day mortality was 22.7%. DILI accompanied by jaundice (n = 191), encephalopathy (n = 69) or ascites (n = 69) resulted in mortality in 30%, 69.6% and 50.7%, respectively (P < 0.001). Age, gender, transaminase levels, HIV or hepatitis B surface antigen (HBsAg) status did not influence survival. Treatment duration, encephalopathy, ascites, bilirubin, serum albumin, international normalized ratio (INR), serum creatinine and leukocyte count were associated with mortality (P < 0.001). Multivariate logistic regression model for mortality, incorporating encephalopathy, albumin, bilirubin, INR, and creatinine yielded a C-statistic of 97%. CONCLUSIONS: Anti-TB DILI occurs throughout treatment duration progressing to ALF in a quarter of patients. The overall mortality is 22.7%, which is higher when accompanied by jaundice, ascites or encephalopathy. An anti-TB DILI model, incorporating bilirubin, INR, encephalopathy, serum creatinine and albumin predicted mortality with C-statistic of 97%.

Characteristics of ascites in patients with pregnancy-specific liver diseases.

BACKGROUND & AIMS: The characteristics of ascites in patients with pregnancy-specific liver disease (PsLD), which comprise acute fatty liver of pregnancy, hemolysis, increased levels of liver enzymes, low platelet syndrome, and preeclampsia-associated liver dysfunction, are unknown. We evaluated the cellular and biochemical characteristics, and model for end-stage liver disease scores, in patients with PsLD. METHODS: We evaluated 46 consecutive patients with PsLD for the presence of ascites. We assessed cellular and biochemical characteristics of the ascites fluid from these patients. RESULTS: Ascites was observed in 35 of 46 patients with PsLD (76%). In 25 patients tested (71.4%), the ascites fluid had low levels of albumin (<0.2 g/dL) and protein (<1 g/dL) and high serum ascites albumin gradients, indicating portal hypertension. Spontaneous bacterial peritonitis was observed in 48% of patients tested and was not associated with mortality. Patients with ascites had significantly low serum levels of protein and albumin (P < .001). Model for end-stage liver disease scores did not differ between patients with or without ascites (32 vs 27; P = .1). CONCLUSIONS: Ascites occur in 76% of women with PsLD, is transient, and has characteristics of portal hypertension, based on high serum ascites albumin gradients. Almost half of patients with PsLD develop spontaneous bacterial peritonitis, which does not affect survival.