A population model of epirubicin pharmacokinetics and application to dosage guidelines.

Abstract

To use a population approach to identify readily available clinical or biochemical characteristics that influence the pharmacokinetics of epirubicin and to develop new dosage guidelines based on these results. Data were available from 109 patients with advanced breast cancer, 72 of whom were known to have liver metastases. They were treated with single-agent epirubicin 12.5 to 120 mg/m(2). Analysis was performed using the software package NONMEM and a three-compartment model was fitted to the data. Individual clearance (CL) estimates ranged from 4 to 86 l/h and the final model included CL as a function of aspartate aminotransferase (AST): CL (l/h)=72.9-(72.9x0.135xlnAST). Inclusion of this factor reduced the interindividual variability in CL from 49% to 39%. Using a target AUC of 4000 ng.h/ml, the following doses were predicted to achieve this exposure with the greatest precision: AST <150 IU/l 125 mg; AST 150-250 IU/l 90 mg; AST 250-500 IU/l 60 mg; AST >500 IU/l 30 mg. These new guidelines were compared with three other guidelines based on serum bilirubin or AST concentrations and body surface area (BSA). The new guidelines achieved the target with greater precision (root mean squared error, rmse, 39.0%) than the current UK guidelines, current USA guidelines or an earlier equation based on AST (rmse 63%, 62% and 59%, respectively). The proposed dosing guidelines should reduce variability in systemic exposure to epirubicin more effectively than traditional approaches. In addition, as they do not require adjustment according to BSA, they could reduce dosage preparation time and the potential for prescribing and dispensing errors.