Background:The aim of this study was to quantitatively assess the effect of anthropometric and biochemical variables and third-space effusions on paclitaxel pharmacokinetics in solid tumor patients. Materials and Methods: Plasma concentration-time data of paclitaxel were collected in patients with non^small cell lung cancer (n = 84), ovarian cancer (n = 40), and various solid tumors (n = 44), totaling 168 patients. Paclitaxel was given as a 3-hour infusion (n = 163) at doses ranging from 100 to 250 mg/m 2 , or as a 24-hour infusion (n = 5) at a dose of 135 or 175 mg/m 2 . Data were analyzed using nonlinear mixed-effect modeling. Results: A three-compartment model with saturable elimination and distribution was used to describe concentration-time data. Male gender and body surface area were positively correlated with maximal elimination capacity of paclitaxel (VM EL ); patient age and total bilirubin were negatively correlated withVM EL (P < 0.005 for all correlations).Typically, male patients had a 20% higher VM EL ; a 0.2 m 2 increase of body surface area led to a 9% increase of VM EL ; a 10-year increase of patient age led to a 5% decrease of VM EL ; and a 10-Amol increase of total bilirubin led to a 14% decrease of VM EL . Third-space effusions were not correlated with paclitaxel pharmacokinetics. Conclusions: This extended retrospective population analysis showed patient gender to significantly and independently affect paclitaxel distribution and elimination. Body surface area, total bilirubin, and patient age were confirmed to affect paclitaxel elimination. This pharmacokinetic model allowed quantification of the covariate effects on the elimination of paclitaxel and may be used for covariate-adapted paclitaxel dosing.Paclitaxel formulated in Cremophor EL is regularly administered to patients with non -small cell lung cancer (NSCLC), ovarian cancer, and breast cancer. The pharmacokinetics of paclitaxel were initially believed to be linear, but Sonnichsen and Gianni reported that paclitaxel pharmacokinetics were best described with a two-and three-compartment model, respectively, with saturable elimination and saturable distribution to the tissues (1, 2). Later, the formulation vehicle of paclitaxel, Cremophor EL, was reported to be the cause of the (apparent) nonlinear plasma behavior of paclitaxel, probably by entrapment of paclitaxel into micelles (3 -7). Generally, substantial interpatient variability of paclitaxel pharmacokinetic variables has been noted (8).Hepatic metabolism and biliary excretion are the most important elimination routes of paclitaxel and its metabolites (9). Paclitaxel has been shown to bind extensively to plasma proteins (from 95% to < 97%), with high central (mean = 13.8 L/m 2 ) and steady-state (mean = 182 L/m 2 ) volumes of distribution (10-13). However, data on specific tissue and compartment distribution of paclitaxel in humans following i.v. administration are limited. Paclitaxel concentrations in ascites have been analyzed in single patients, where the concentration ...