Background: Epirubicin is a widely used anthracycline drug. It undergoes rapid reduction to epirubicinol, extensive hepatic conjugation of epirubicin and epirubicinol with glucuronid acid by UDP-glucuronosyltransferase 2B7 (UGT2B7), and subsequent elimination mainly through biliary excretion, mediated by the membrane-located drug efflux transporters ABCB1 (MDR1, P-glycoprotein), ABCC2 (MRP2) and possibly ABCG2 (BCRP). Dose individualization based on PK parameters, clinical and genetic covariates appears to be an obvious means of optimizing treatment, but data on such correlations are very limited. The aim of this study was to quantitatively assess the influence of anthropometric, biochemical parameters, and epirubicin metabolic pathway-associated gene polymorphisms on drug PKs in solid tumor patients.

Materials and Methods: 33 patients with advanced gastroesophageal cancer were included in the study. They received 3-weekly epirubicin 50 mg/m² i.v., cisplatin 60 mg/m² i.v., and capecitabine 1250 mg/m² bid over 2 weeks (ECC-schedule). Extensive plasma sampling over 24 hours after the begin of drug treatment was performed for epirubicin (13 concentration-time data points per patient). Concentrations of epirubicin and epirubicinol were quantitatively assessed by a validated HPLC method, and subjected to population analysis using nonlinear mixed-effect modeling (NONMEM). DNA was isolated from peripheral blood and submitted to sequencing of UGT2B7 (5UTR, exons 1,2,4), MDR1 (exons 12,21,26), ABCC2 (5UTR, exons 10, 29, 31) and ABCG2 (exons 2, 5, 15). An extensive covariate model was build with anthropometric, biochemical and genetic data. Statistical discrimination between hierarchical models was based on the log-likelihood ratio test, using the minimal function of the objective value (MFOV) and an inclusion criterium of p ≤ 0.01, which corresponds to a decrease in the MFOV of 6.6 (degrees of freedom (DF) = 1) or 9.2 (DF = 2) .

Results: A four-compartment model with linear elimination and distribution was optimal to simultaneously describe concentration-time data of epirubicin and epirubicinol. Body-surface area (BSA) was significantly correlated with epirubicin clearance (CL_EPI) and intercompartmental clearance between epirubicin central and peripheral compartments (Q_1/2). A novel UGT2B7 polymorphism within the promoter region (-66T>C) was found at an allele frequency of 9.1% (4 patients heterozygous, one patient homozygous). This mutation was significantly and negatively correlated with epirubicinol clearance (CL_EOL), and population modelling suggested a 38% and 58% lower CL_EOL in patients being heterozygous or homozygous for -66T>C, respectively (p=0.009). The homozygous nonsynonymous polymorphism G2677T in the MDR1 gene (MDR1*7) was significantly associated with a decreased  CL_EOL (-42%, p=0.01), but this correlation did not resist multivariate testing. No other relationships between anthropometric, biochemical parameters, gene polymorphisms and PK parameters reached statistical significance. No association was found between genetic markers and drug toxicity. Two patients developed reversible cardiac failure CTC3 after 2 cycles of ECC and discontinued epirubicin. They both had no UGT2B7 -66T>C mutations.

Conclusions: The described model was successful in assessing gene-dose effects on epirubicin PKPDs. In vitro functionality of UGT2B7 -66T>C should be tested. This is the first study suggesting dependency of epirubicin PKs on UGT2B7 genetic variability, and potential clinical significance should be studied in a more extended patient cohorts.