Categorization of Drugs Implicated in Causing Liver Injury: Critical Assessment Based on Published Case Reports

An important element in assessing causality in drug-induced liver injury is whether the implicated agent is known to cause hepatotoxicity. We classified drugs into categories based on the number of published reports of convincingly documented, clinically apparent, idiosyncratic liver injury. Drugs described in the website LiverTox (http://livertox.nih.gov) were classified into five categories based on the number of published cases (category A, >= 50; category B, 12-49; category C, 4-11; category D, 1-3; category E, none). Case reports in categories C and D were individually reanalyzed using the Roussel Uclaf Causality Assessment Method. Drugs with fatal cases or with rechallenge were noted. Among 671 individual drugs or closely related agents, 353 (53%) were considered convincingly linked to liver injury in published case reports; 48 (13%) were assigned to category A, 76 (22%) were assigned to category B, 96 (27%) were assigned to category C, and 126 (36%) were assigned to category D. Another 7 (2%) were direct hepatotoxins but only in high doses and placed in a separate category (T). The remaining 318 (47%) drugs had no convincing case report of hepatoxicity in the literature (category E). All except one in category A have been available since 1999, 98% had at least one fatal case and 89% a positive rechallenge. In category B, 54% had a fatal case and 41% a rechallenge. Drugs in categories C and D less frequently had instances of fatal (23% and 7%) or rechallenge cases (26% and 11%). Conclusion: Documentation of hepatoxicity in the medical literature is variable, and many published instances do not stand up to critical review. A standardized system for categorizing drugs for hepatotoxicity potential will help develop objective and reliable, computer-based instruments for assessing causality in drug-induced liver injury.