Potential benefit and lack of serious risk from corticosteroids in drug-induced liver injury: An international, multicentre, propensity score-matched analysis.

Abstract

The use of corticosteroids to treat patients with idiosyncratic drug-induced liver injury (DILI) relies on empirical clinical decisions. To investigate the relationship between corticosteroids and risk of acute liver failure (ALF) in patients with DILI and to assess if corticosteroid therapy was associated with improved outcomes in DILI patients. We analysed bona fideidiosyncratic DILI cases from the Spanish DILI Registry and Indiana University School of Medicine. Patients treated with corticosteroids were compared to those who did not receive any treatment. Nearest neighbour propensity score matching analyses were conducted. We enrolled 724 patients, 106 under corticosteroid therapy, in whom there was over-representation of more severe injury and autoimmune features, and 618 who did not receive any treatment. In an analysis of 80 pairs of propensity score-matched patients, corticosteroid administration was not associated with an increased risk of developing ALF (odds ratio=0.65; 95% confidence interval [CI]: 0.18-2.40; p=0.518). Furthermore, in an additional analysis, a Cox regression model that included 41 propensity score-matched pairs showed that patients receiving corticosteroids had a significantly higher normalisation rate of liver enzymes than untreated patients (hazard ratio [HR]=1.84; 95% CI: 1.02-3.32; p=0.043), particularly in patients with serious injury who did not resolve within 30 days (HR=2.79; 95% CI: 1.20-6.50; p=0.018). Corticosteroid therapy did not worsen outcome in DILI patients. Indeed, corticosteroid administration was associated with a greater rate of normalisation of liver enzymes in patients with serious DILI.