Oral paclitaxel is not inherently bioavailable because of the overexpression of P-glycoprotein by intestinal cells and the significant first-pass extraction by cytochrome P450-dependent processes. This study sought to simulate the toxicological and pharmacological profile of a clinically relevant schedule of paclitaxel administered on clinically relevant i.v. dosing schedules in patients with advanced solid malignancies using oral paclitaxel administered with cyclosporin A, an inhibitor of both P-glycoprotein and P450 CYP3A. Nine patients were treated with a single course of oral paclitaxel in its parenteral formulation at a paclitaxel dose level of 180, 360, or 540 mg. Cyclosporin A was administered at a dose of 5 mg/kg p.o. 1 h before and concurrently with oral paclitaxel. Blood sampling was performed to evaluate the pharmacokinetics of paclitaxel, 6-alpha-hydroxypaclitaxel, 3-p-hydroxypaclitaxel, and cyclosporin A. The pharmacokinetic behavior of paclitaxel was characterized using both compartmental and noncompartmental methods. Model-estimated parameters were used to simulate paclitaxel concentrations after once daily and twice daily oral administration of paclitaxel and cyclosporin A. Aside from an unpleasant taste, the oral regimen was well tolerated, and there were no grade 3 or 4 drug-related toxicities. The systemic exposure to paclitaxel, as assessed by maximum plasma concentration (Cmax) and area under the plasma concentration versus time curve (AUC) values, did not increase as the dose of paclitaxel was increased from 180 to 540 mg, and there was substantial interindividual variability (4-6-fold) at each dose level. Mean paclitaxel Cmax values approached plasma concentrations achieved with clinically relevant parenteral dose schedules, averaging 268+/-164 ng/ml. AUC values averaged 3306+/-1977 ng x h/ ml, which was significantly lower than AUC values achieved with clinically relevant i.v. paclitaxel dose schedules. However, computer simulations using pharmacokinetic parameters derived from the present study demonstrated that pharmacodynamically relevant steady-state plasma paclitaxel concentrations of at least 0.06 microM would be achieved after protracted once daily and twice daily dosing with oral paclitaxel and cyclosporin A. Paclitaxel metabolites were detectable in three patients, and the 6-alpha-hydroxypaclitaxel: paclitaxel and 3-p-hydroxypaclitaxel:paclitaxel AUC ratios averaged 0.63 and 0.86, respectively; these values were substantially higher than values reported in patients treated with i.v. paclitaxel. Oral paclitaxel was bioavailable in humans when administered in combination with oral cyclosporin A 5 mg/kg 1 h before and concurrently with paclitaxel treatment, and plasma paclitaxel concentrations achieved with this schedule were biologically relevant and approached concentrations attained with clinically relevant parenteral dose schedules. However, treatment of patients with oral paclitaxel using a single oral dose administration schedule failed to achieve sufficiently high systemic drug exposure and pharmacodynamic effects. In contrast, computer simulations demonstrated that clinically relevant pharmacodynamic effects are likely to be achieved with multiple once daily and twice daily oral paclitaxel-cyclosporin A dosing schedules.