Use of population modeling to define rational monitoring of amiodarone hepatic effects

Background. Amiodarone causes hepatotoxicity in experimental models, but in humans, the relationships Back between drug administration, serum concentrations, markers of liver function, and how to monitor for hepatotoxicity have not been well characterized. Methods. An open-dose, prospective study collected serum amiodarone, desethylamiodarone, ALT, AST, lactate dehydrogenase (LDH), alkaline phosphatase, total bitirubin, and albumin concentrations over a 5-year period from 125 patients. Nonlinear mixed-effects modeling (NONMEM) was used to explore the relationship between markers of hepatotoxicity and concentrations of antiodarone and desethylamiodarone. Results. No patients had clinical symptoms of hepatotoxicity during follow-up. The natural history of changes in hepatic-makers showed ALT to have the strongest independent relationship to changes in serum antiodarone (r = 0.32, P < .001). An ALT greater than 3 times the upper limit of normal developed in only 8 patients (7%). with the earliest occurrence at 55 days of therapy. A mixed-effects model relating ALT elevation to serum amiodarone was improved by the addition of an effect compartment having an equilibration half-time of 87 days (r = 0.81, P < .001). The model predicts that 6% of patients will have an ALT greater than 3 times the upper limit of normal if amiodarone concentrations are maintained at less than 2.5 mg/L, and virtually no patients will have such ALT elevations if amiodarone concentrations are maintained at less than 1.5 mg/L. Conclusions. Concentrations of amiodarone below a threshold of 1.5 mg/L are associated with a minimal risk of hepatotoxicity, whereas concentrations greater than 2.5 mg/L arc associated with a greater than 6% risk of hepatotoxicity. There is significant hysteresis between, changes in amiodarone concentration and the resulting change in ALT. The model suggests that monitoring ALT at baseline, 1, 3, and 6 months, and then semiannually would be an efficient strategy to detect amiodarone-induced hepatotoxicity.