Purpose: Analyze US rates of reported severe liver disease for the oral hypoglycemic agent troglitazone from March 1997 through February 2000 and the possible effects of publicity on reporting.
Methods: The number of troglitazone reports with liver failure and or hospitalization with jaundice or hyperbilirubinemia, made to the FDA and/or Parke-Davis are used as numerators. The denominators are numbers of patients and person-time estimates of exposure. Additionally, the amount of publicity about troglitazone during its marketing is quantified.
Results: Approximately 1.92 million patients were treated with troglitazone from March 1997 through the end of February 2000 resulting in 1.6 million person-years of exposure. Reports of 83 cases of liver failure associated with troglitazone were received (1 in 23,000 patients or 1 in 20,000 person-years). Of the 83 cases, only 49 (59%) were classified by a hepatologist to be 'possibly' or 'probably' attributed to troglitazone. For the first, second, and third years of marketing, rates of reported hepatic failure per 100,000 person years exposure to troglitazone were 8.3, 5.3, and 2.7 respectively. Rates of reported liver disease involving hospitalizations with mention of jaundice and hyperbilirubinemia per 100,000 person-years were 16.0, 6.1, and 3.6 respectively for these years. During the 3-year marketing history of troglitazone, there were 470 lay press and 158 medical literature articles with mentions of hepatotoxicity for the drug.
Conclusions: Rates of reported severe liver disease declined substantially during the second and third years of marketing of troglitazone. The decline followed increasingly stringent requirements for liver function test monitoring and may have been due to improved patient selection and management as a result of the widely publicized association between troglitazone and hepatotoxicity.