Monitoring for hepatotoxicity: what is the predictive value of liver "function" tests?

Abstract

Drug-induced liver injury (DILI) is a major reason drugs fail during development or are withdrawn from the market. The ability to predict, detect, and avoid DILI through appropriate patient selection and effective monitoring has proved to be an elusive goal. Many approved drugs have labeling recommendations for serum enzyme monitoring intended to detect and prevent hepatotoxicity, but such monitoring is often seen as inconvenient, uncomfortable, costly, and inefficient by both patients and doctors, and thus monitoring recommendations are poorly followed, if at all. This review considers whether monitoring works to prevent DILI, whether monitoring recommendations are derived from data or opinions, and whether any better alternatives exist.