Randomized trial of filgrastim vs. sequential filgrastim and molgramostim after dose-intensified carboplatin, cyclophosphamide, and etoposide: a phase I pilot study.

This phase I randomized study was designed in order to verify if the sequential administration of filgrastim, a granulocyte colony-stimulating factor (G-CSF), and molgramostim, a granulocyte-macrophage colony-stimulating factor (GM-CSF), was superior to filgrastim alone in improving tolerance of dose-intensified carboplatin (CBDCA), cyclophosphamide (CTX), and etoposide (VP-16). A group of 10 heavily pretreated patients with stage IV disease and no therapeutic option were enrolled into the study. They received two courses of the same chemotherapy with CTX and VP-16 at doses of 1,500 mg/m2 and 400 mg/m2, respectively. CBDCA doses were escalated from 450 to 600 mg/m2. After chemotherapy each patient was allocated randomly to receive either 14 days of G-CSF (arm A) or 7 days of G-CSF followed by 7 days of GM-CSF (arm B). Crossover in the second chemotherapy course was accomplished. Both G-CSF and GM-CSF were given 5 microg/kg/day, subcutaneously. Twenty chemotherapy courses are evaluable, 10 in each arm. Absolute neutrophil count < 1 x 10(3)/microl was observed for 54 days in arm A vs. 68 days in arm B (P < 0.02); platelet (PLT) count < 20 x 10(3)/microl, 57 days vs. 30 days (P < 0.01); days of hospitalization 35 vs. 16 (P < 0.38); PLT transfusion, 107 vs. 58 (P < 0.01); packed red blood cell unit transfusions, 15 vs. 5 (P < 0.13). Seven patients had responses. These data indicate that dose-intensified chemotherapy may be delivered without bone marrow or peripheral stem cell support, with acceptable toxicity, and that, while G-CSF alone shortens days of neutropenia, the combination of the two cytokines shortens the time of thrombocytopenia and decreases the number of PLT transfusions.