Troglitazone was removed from the U.S. market because its use was associated with an increased risk of liver failure. We evaluated the clinical features of all cases reported to the Food and Drug Administration and estimated the duration and magnitude of the risk of liver failure associated with continued use of the drug.

Data from cases of liver failure associated with troglitazone use were abstracted and analyzed. The extent of troglitazone use was determined from national marketing data, and the duration of use was estimated with data from a large, multistate, health care company. Survival analysis was performed to estimate monthly incidence rates and the cumulative risk of liver failure.

Ninety-four cases of liver failure (89 acute, 5 chronic) were reported. Of the acute cases, 58 (67%) were women and only 11 (13%) recovered without liver transplantation. Progression from normal hepatic functioning to irreversible liver injury occurred within 1 month in 19 patients who were indistinguishable clinically from the 70 patients who had an unknown time course to irreversibility, except for the post hoc observation that prior cholecystectomy was less common in those with rapid onset. The incidence of liver failure was elevated from the first through at least the 26th month of troglitazone use. Accounting for case underreporting, the number needed to harm from troglitazone use was between 600 to 1500 patients at 26 months.

The progression to irreversible liver injury probably occurred within a 1-month interval in most patients, casting doubt on the value of monthly monitoring of serum aminotransferase levels as a means of preventing troglitazone-induced acute liver failure. The cumulative risk of hepatic failure increased with continued use.