It has been hypothesized that the paclitaxel vehicle Cremophor EL (CrEL) is responsible for nonlinear drug disposition by micellar entrapment. To gain further insight into the role of CrEL in taxane pharmacology, we studied the pharmacokinetics of paclitaxel in the presence and absence of CrEL after i.p. and i.v. dosing. Patients received an i.p. tracer dose of [G-(3)H]paclitaxel in ethanol without CrEL (100 microCi diluted further in isotonic saline) on day 1, i.p. paclitaxel formulated in CrEL (Taxol; 125 mg/m(2)) on day 4, an i.v. tracer of [G-(3)H]paclitaxel on day 22, and i.v. Taxol (175 mg/m(2)) on day 24. Four patients (age range, 54-74 years) were studied, and serial plasma samples up to 72 h were obtained and analyzed for total radioactivity, paclitaxel, and CrEL. In the presence of CrEL, i.v. paclitaxel clearance was 10.2 +/- 3.76 liters/h/m(2) (mean +/- SD), consistent with previous findings. The terminal disposition half-life was substantially prolonged after i.p. dosing (17.0 +/- 11.3 versus 28.7 +/- 8.72 h), as was the mean residence time (7.28 +/- 2.76 versus 40.7 +/- 13.8 h). The bioavailability of paclitaxel was 31.4 +/- 5.18%, indicating insignificant systemic concentrations after i.p. treatment. CrEL levels were undetectable after i.p. dosing (<0.05 microl/ml), whereas after i.v. dosing, the mean clearance was 159 +/- 58.4 ml/h/m(2), in line with earlier observations. In the absence of CrEL, the bioavailability and systemic concentrations of i.p. paclitaxel were significantly increased. This finding is consistent with the postulated concept that CrEL is largely responsible for the pharmacokinetic advantage for peritoneal cavity exposure to total paclitaxel compared with systemic delivery.