Appropriate timing to start and optimal response evaluation of high-dose corticosteroid therapy for patients with acute liver failure.

BACKGROUND: Corticosteroid therapy has been commonly administered to patients with acute liver injury (ALI)/acute liver failure (ALF) in Japan to prevent the development of hepatic encephalopathy, but the appropriate timing to start corticosteroid therapy has not been determined and optimal response evaluation of the therapy has not been conducted. We prospectively investigated the optimal timing to start therapy on the established severity indication: the Japan Hepatic Encephalopathy Prediction Model (JHEPM) and prothrombin time (PT).

METHODS: This prospective observational study enrolled 469 patients with ALI/ALF from 2004 to 2015. We evaluated 44 patients with ALF on high-dose corticosteroid therapy before hepatic coma development. The predictive performance for coma development was assessed using the receiver operator curve method in both PT and JHEPM probability the day before administering high-dose corticosteroid therapy.

RESULTS: Among these patients, nine developed hepatic coma after the therapy. Selection bias was adjusted using propensity score method. High-dose corticosteroid therapy tended to decrease the risk of coma development although there was no statistical significance. The cut-off value of 53%, 1.95, and 39% in JHEPM probability, PT-international normalized ratio (PT-INR), and PT activity, respectively, showed high sensitivity and specificity.

CONCLUSIONS: We propose the appropriate timing to start high-dose corticosteroid therapy in patients with ALI/ALF; 40% of JHEPM probability, 1.53 of PT-INR, and 52% of PT because these values were theoretically discriminated at 98% coverage to the patients with coma. Because the study contained selection bias, the appropriate timing for therapy should be confirmed in a future prospective study.