Factors Associated with the Risk of Liver Enzyme Elevation in Patients with Type 2 Diabetes Treated with a Thiazolidinedione

Study Objective. To characterize frequency of liver enzyme elevation in patients with type 2 diabetes mellitus receiving troglitazone. Design. Retrospective study. Setting. Hospital-affiliated medical center. Patients. Two hundred ninety-one patients with type 2 diabetes mellitus. Intervention. Data from patients with an average troglitazone exposure of 412.7 ± 255.6 days were studied. Measurements and Main Results. Enzyme elevations more than 1.5 times the upper limit of normal (ULN) occurred in 17 patients (5.8%) and more than 3-fold elevations in 6 (2.1%). The relationship among enzyme elevation events, demographic factors, duration of troglitazone exposure, frequency of monitoring, and concurrent drugs (limited to glucose and lipid-lowering agents) was assessed by multiple logistic regression. Age was an independent predictor of risk (p=0.009), and concurrent insulin therapy approached statistical significance (p=0.051) for 1.5-fold ULN elevation in liver enzymes. Age and concurrent therapy with 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors were the only significant predictors of 3-fold ULN elevations (p=0.03 and p=0.04, respectively). Conclusion. Several factors appear to increase the risk of enzyme elevation events in patients treated with troglitazone.