Background: Because carcinomas of unknown primary origin are highly malignant tumors with a bad prognosis (median survival, 6-12 months) and no current optimal therapy, the authors designed a prospective dose-dense chemotherapy regimen with the objective of improving the results observed in patients who receive conventional treatment.
Methods: Eighty-two patients received alternative bimonthly cycles of doxorubicin 50 mg/m(2) with cyclophosphamide 1000 mg/m(2) (AC) and etoposide 300 mg/m(2) with cisplatin 100 mg/m(2) (EP). Cycles were given at 2-week intervals with granulocyte-macrophage-stimulating factor support (5 microg/kg per day) from Day 4 to Day 10. Patients without measurable lesions were included, because the major end point was survival.
Results: The median number of alternative cycles of AC and EP was 4 cycles (range, 1-12 cycles). An objective response was observed in 24 of 62 patients (39%; 95% confidence interval, 30-48%) with measurable lesions, including 6 patients who achieved a complete response. Among 20 patients with nonmeasurable disease, 9 patients (45%) had no evidence of progressive disease at the end of chemotherapy. The overall median survival of 82 patients was 10 months, with 5 patients surviving clinically disease free at 17 months, 29 months, 45 months, 64 months, and 70 months after the end of treatment. Myelosuppression was the most common toxicity. Two toxic deaths occurred.
Conclusions: Using these doses and schedules, a dose-dense chemotherapy regimen did not appear to improve the outcome of patients with carcinoma of unknown primary site. Alternative studies dealing with news drugs will be required.
Copyright 2002 American Cancer Society. DOI 10.1002/cncr.10264