The objectives of this study were to investigate the disposition of gemcitabine, epirubicin, paclitaxel, 2',2'-difluorodeoxyuridine and epirubicinol, and characterize the pharmacokinetic and pharmacodynamic profile of treatment in patients with breast cancer. PATIENTS AND METHODS: The drug dispostion in 15 patients who received gemcitabine 1000 mg/m2, epirubicin 90 mg/m2 and paclitaxel 175 mg/m2 (GEP) on day 1 of a 21-day cycle, was compared with that of patients treated with epirubicin 90 mg/m2 and paclitaxel 175 mg/m2 (EP, n = 6) and epirubicin 90 mg/m2 alone (n = 6). Drug and metabolite levels in plasma and urine were assessed by high-performance liquid chromatography and parameters of drug exposure were related to hematological toxicity by a sigmoid-maximum effect (Emax) model. RESULTS: Paclitaxel administration significantly increased the epirubicinol area under the concentration-time curve, from 357+/-146 (epirubicin) to 603+/-107 (EP) and 640+/-81 h x ng/ml (GEP), and reduced the renal clearance of epirubicin and epirubicinol by 38 and 52.2% and 34.5 and 53% in GEP- and EP-treated patients, respectively, compared with epirubicin alone. Gemcitabine had no apparent effect on paclitaxel and epirubicin pharmacokinetics, and renal clearance of epirubicin and epirubicinol. The only pharmacokinetic/pharmacodynamic relationship observed was between neutropenia and the time spent above the threshold plasma level of 0.1 micromol/l (tC0.1) of paclitaxel, with the time required to obtain a 50% decrease in neutrophil count (Et50) of GEP being 7.8 h, similar to that of EP. CONCLUSIONS: Paclitaxel and/or its vehicle, Cremophor EL, interferes with the disposition and renal excretion of epirubicin and epirubicinol; gemcitabine has no affect on epirubicin and paclitaxel plasma pharmacokinetics and renal excretion of epirubicin, while neutropenia is not enhanced by gemcitabine.