Phase I and Pharmacokinetic Study of the Oral Fluoropyrimidine Capecitabine in Combination With Paclitaxel in Patients With Advanced Solid Malignancies

Abstract

PURPOSE: To evaluate the feasibility of administering the oral fluoropyrimidine capecitabine in combination with paclitaxel, to characterize the principal toxicities of the combination, to recommend doses for subsequent disease-directed studies, and to determine whether significant pharmacokinetic interactions occur between these agents when combined. PATIENTS AND METHODS: Sixty-six courses of capecitabine and paclitaxel were administered to 17 patients in a two-stage dose-escalation study. Paclitaxel was administered as a 3-hour intravenous (IV) infusion every 3 weeks, and capecitabine was administered continuously as two divided daily doses. During stage I, capecitabine was escalated to a target dose of 1,657 mg/m2/d, whereas the paclitaxel dose was fixed at 135 mg/m2. In stage II, paclitaxel was increased to a target dose of 175 mg/m2, and the capecitabine dose was the maximum established in stage I. Pharmacokinetics were characterized for each drug when given alone and concurrently. RESULTS: Myelosuppression, predominately neutropenia, was the principal dose-limiting toxicity (DLT). Othertoxicities included hand-foot syndrome, diarrhea, hyperbilirubinemia, skin rash, myalgia, and arthralgia. Two patients treated with capecitabine 1,657 mg/m2/d and paclitaxel 175 mg/m2 developed DLTs, whereas none of six patients treated with capecitabine 1,331 mg/m2/d and paclitaxel 175 mg/m2 developed DLTs during course 1. Pharmacokinetic studies indicated that capecitabine and paclitaxel did not affect the pharmacokinetic behavior of each other. No major antitumor responses were noted. CONCLUSION: Recommended combination doses of continuous capecitabine and paclitaxel are capecitabine 1,331 mg/m2/d and paclitaxel 175 mg/m2/d IV every 3 weeks. Favorable preclinical mechanistic interactions between capecitabine and paclitaxel, as well as an acceptable toxicity profile without clinically relevant pharmacokinetic interactions, support the performance of disease-directed evaluations of this combination.