Mitochondrial dysfunction and delayed hepatotoxicity: another lesson from troglitazone

Aims/hypothesis Troglitazone was approved for treatment of type 2 diabetes mellitus, but by 2000 it had been removed from all world markets due to severe drug-induced liver injury. Even today, we still do not know how many patients sustained a long-term liver injury. No system is in place to acquire that knowledge. Regarding toxicity mechanisms, controversy persists as to which ones are class effects of thiazolidinediones (TZDs) and which are unique to troglitazone. This study aims to provide long-term outcome data and new insights on mechanisms of troglitazone-induced liver injury. Methods This case series reports the liver injuries sustained by eleven type 2 diabetic patients treated with troglitazone between 1997 and 2000. Exhaustive review of medical records was performed for all patients. Long-term outcomes were available for all the non-fatal cases. A comprehensive literature review was also performed. Results Long-term liver injury progressing to cirrhosis was identified in seven patients. All eleven cases had liver injury patterns consistent with troglitazone toxicity. Analysis of these cases and of the experimental troglitazone toxicity data points to mitochondrial toxicity as a central factor. The general clinical patterns of mitochondrial hepatotoxic events are reviewed, as are the implications for other members of the TZD family. Conclusions/interpretation This analysis enables the liver injury induced by troglitazone to be better understood. In future cases of delayed drug-induced liver injury that progresses after discontinuation, the possibility of mitochondrial toxicity should be considered. When appropriate, this can then be evaluated experimentally. Such proactive investigation may anticipate clinical risk before a large-scale therapeutic misadventure occurs. Drug-induced liver injury due to mitochondrial hepatotoxins may be less unpredictable than has previously been surmised.