OBJECTIVE To characterize the population pharmacokinetics of pegylated-liposomal doxorubicin in patients with acquired immunodeficiency disease (AIDS)-related Kaposi's sarcoma and to explore the relationship between response of the cutaneous Kaposi's sarcoma lesions to treatment and measures of drug exposure. METHODS Forty-three male patients (median age, 40 years; age range, 28 to 50 years), body surface area, 1.89 m2; range, 1.5 to 2.3 m2) with AIDS and at least five biopsy-proven cutaneous Kaposi's sarcoma lesions were randomized to receive either a 10 or 20 mg/m2 dose of study drug for their first cycle and the alternate dose 3 weeks later. Patients continued to receive the study drug at a dose of 20 mg/m2 every 3 weeks. Serial blood samples were obtained after the first two doses and analyzed by HPLC for determination of total plasma doxorubicin concentration. Kaposi's sarcoma lesion response was categorized as either progressive disease, stable disease, partial response, or complete response. Classification and regression tree (CART) analysis was used to determine the relationship between drug exposure and categorical lesion response. Iterative two-stage analysis was used to characterize both the pharmacokinetics of pegylated-liposomal doxorubicin and to model the probabilities of achieving a specific lesion response. RESULTS The pharmacokinetics of pegylated-liposomal doxorubicin were best described by a two-compartment linear structural model. Lesion response was significantly related to both the average daily maximum doxorubicin concentration (Cmax,avg) and dose intensity. CONCLUSIONS The pharmacokinetics of pegylated-liposomal doxorubicin are strikingly different from conventional doxorubicin. Identification of both Cmax,avg and dose intensity as predictors of lesion response will provide guidelines for future dosing regimen designs.