Phase I Trial, Including Pharmacokinetic and Pharmacodynamic Correlations, of Combination Paclitaxel and Carboplatin in Patients With Metastatic Non–Small-Cell Lung Cancer

PURPOSE: To determine the maximum-tolerated dose of paclitaxel with carboplatin with and without filgrastim support in patients with metastatic non–small-cell lung cancer (NSCLC) and to investigate the pharmacokinetics of paclitaxel and carboplatin and correlate these with the pharmacodynamic effects. PATIENTS AND METHODS: Thirty-six chemotherapy-naive patients with metastatic NSCLC were entered into this phase I dose-escalation and pharmacokinetic study. Paclitaxel was initially administered as a 24-hour infusion at a fixed dose of 135 mg/m 2 , and the carboplatin dose was escalated in cohorts of three patients, using Calvert's formula [dose(mg) = area under the concentration time curve (glomerular filtration rate + 25)], to target areas under the concentration time curve (AUCs) of 5, 7, 9, and 11 mg/mL × minute. A measured 24-hour urinary creatinine clearance was substituted for the glomerular filtration rate. Once the maximum-tolerated AUC (MTAUC) of carboplatin was reached, the paclitaxel dose was escalated to 175, 200, and 225 mg/m 2 . When the paclitaxel dose escalation began, the AUC of carboplatin was reduced to one level below the MTAUC. RESULTS: Myelosuppression was the major dose-limiting toxicity. Thrombocytopenia was observed at a carboplatin AUC of 11 mg/mL × minute after course 2 and thereafter. End-of-infusion plasma paclitaxel concentrations and median duration of time above 0.05 μM were similar in course 1 versus course 2 at the 135 and 175 mg/m 2 dose levels. The neutropenia experienced by patients was consistent with that observed in patients who had received paclitaxel alone. Measured carboplatin AUCs were approximately 12% (20% v 3% with course 1 v course 2, respectively) below the desired target, with a standard deviation of 34% at all dose levels. A sigmoid-maximum effect model describing the relationship between relative thrombocytopenia and measured free platinum exposure indicated that patients who received the combination of carboplatin with paclitaxel experienced less severe thrombocytopenia than would be expected from carboplatin alone. Of the 36 patients entered onto the study, one experienced a complete response and 17 had partial responses, for an overall response rate of 50%. The recommended doses of paclitaxel (24-hour infusion) and carboplatin for future phase II studies of this combination are (1) paclitaxel 135 mg/m 2 with a carboplatin dose targeted to achieve an AUC of 7 mg/mL × minute without filgrastim support; (2) paclitaxel 135 mg/m 2 with a carboplatin dose targeted to achieve an AUC of 9 mg/mL × minute with filgrastim support; and (3) paclitaxel 225 mg/m 2 with a carboplatin dose targeted to achieve an AUC of 7 mg/mL × minute with filgrastim support. CONCLUSION: The regimen of paclitaxel and carboplatin is well-tolerated and has promising activity in the treatment of NSCLC. There is no pharmacokinetic interaction between paclitaxel and carboplatin, but there is a pharmacodynamic, platelet-sparing effect on this dose-limiting toxicity of carboplatin.