Optimal modeling for phase I design of a two drug combination-results of a phase I study of cisplatin with 9-nitrocamptothecin.

Combination of platinum with topoisomerase-I inhibitors are synergistic. The objectives of this study were to determine MTD range and toxicity profile of combinations of oral 9-nitrocamptothecin (9NC) and intravenous cisplatin in patients with refractory solid tumors. Each course was 28 days starting on day 1 with cisplatin, and then 9NC daily for 5 days/week for three weeks. A new two arm crossing design was created: patients in arm 1 were treated with at the single agent recommended dose of cisplatin (50 mg/m(2)), and increasing doses of 9NC and in arm 2 with the single agent recommended dose of 9NC (1.5 mg/day) and increasing dose of cisplatin. Once a dose limiting toxicity was observed, the dose of the escalated drug was decreased by one level, and the fixed-dose drug was then escalated. A 3 + 3 design was used. Eligibility criteria were standard for a phase I trial. Pharmacokinetics was performed. Eighteen patients were treated on Arm 1, 3 at the crossing level, and 33 on Arm 2. Dose limiting toxicities were gastrointestinal at the crossing dose level. After crossing, prolonged grade 3 thrombocytopenia was the DLT in arm 1, and grade 4 neutropenia in Arm 2. Only one patient with ovarian cancer had a partial remission, and 12 patients had disease stabilization (24% of clinical benefit). A Bayesian optimal dose finding was tested post-facto. The recommended doses for phase II studies by the 3 + 3 design are cisplatin 60 mg/m(2) and 9NC 1.25 mg/day and cisplatin 40 mg/m(2) and 9NC 2.0 mg/day. The Bayesian optimal dose finding suggested a different solution, closest to that of the latter dosing which may be less toxic.