PURPOSE: The primary objective of this phase I trial was to determine the feasibility of administering a combination of paclitaxel, cisplatin, and doxorubicin with or without granulocyte colony-stimulating factor (G-CSF) in patients with advanced endometrial and other gynecologic cancers.

PATIENTS AND METHODS: Patients were chemotherapy-naive. Doxorubicin was administered as a brief infusion, paclitaxel for 3 hours, and cisplatin for 60 minutes. Treatments were repeated every 3 weeks. For most dose levels, the cisplatin and doxorubicin were fixed at 60 mg/m(2) and 45 mg/m(2), whereas the paclitaxel was escalated in successive cohorts from 90 to 250 mg/m(2). Patients who had received previous radiotherapy to the whole pelvis were escalated separately from those who had not.

RESULTS: Eighty patients received 320 cycles of therapy. When G-CSF was not used, myelosuppression prevented escalation beyond the starting dose for patients with or without previous pelvic radiotherapy. When G-CSF was added, neurotoxicity became dose-limiting for both groups. Ten patients were removed from the study for asymptomatic declines in ejection fraction, but no symptomatic congestive heart failure was observed. Major antitumor responses occurred in 46% of patients (six of 13) with measurable endometrial carcinoma and 50% of patients (eight of 16) with measurable cervical carcinoma.

CONCLUSION: The combination of paclitaxel, doxorubicin, and cisplatin at relevant single-agent doses is active and feasible with the addition of G-CSF. A regimen of cisplatin 60 mg/m(2), doxorubicin 45 mg/m(2), and paclitaxel 160 mg/m(2) with G-CSF support is recommended for further testing.