Epirubicin, cisplatin, and prolonged or brief infusional 5-fluorouracil in the treatment of carcinoma of unknown primary site

The cytotoxic regimen of epirubicin, cisplatin, and continuous infusional 5-fluorouracil (ECF) has demonstrated activity in a range of malignancies, including gastroesophageal, breast, and pancreatic cancers. Prolonged infusional central venous catheter (CVC) mediated therapy is not always feasible and modifications of the 5-fluorouracil (5FU) schedule have been reported. We reviewed our experience of both the standard and a modified ECF regimen in patients diagnosed with carcinoma of unknown primary site (CUPS). A retrospective analysis of all patients diagnosed with CUPS (31 adenocarcinoma and 5 poorly differentiated carcinoma) and treated with ECF between June 1994 and June 1998 was undertaken. Thirty-six patients, median age 56 (range: 24-74), were treated thrice-weekly with 50 mg/m2 epirubicin, 60 mg/m2 cisplatin, and 5-FU administered either by continuous infusion 200 mg/m2/d via a CVC (standard ECF) or as a 6-h infusion 600 mg/m2 through a peripheral venous catheter (modified ECF). Thirteen patients were treated with standard ECF and 23 received modified ECF. The median number of cycles administered was 4 (range: 1-10). Thirty-two patients had evaluable disease, seven (22%; 95% confidence interval: 8-36%) demonstrated a partial response, including three patients that received standard ECF and four treated with modified ECF. There were no complete responses. The median survival for all 36 patients was 9.0 mo. Median survival for patients treated with standard ECF was 11.7 mo as compared to 5.1 mo for the modified ECF schedule (p=0.052). Liver involvement and elevation of serum CA19.9 were identified as possible adverse prognostic factors. Both regimens were well tolerated, with the only grade 3/4 toxicity recorded being leukopenia (four patients), nausea/vomiting (seven patients), and diarrhea (one patient). CVC complications in the standard ECF group were thrombosis (one patient) and infection (three patients). There were no treatment-related deaths. We conclude that ECF, whether modified or standard, has modest activity in the setting of CUPS. Patient survival is comparable to survival documented in previous reports of CUPS treatment. The apparent survival difference between the two ECF schedules may be the result of patient selection factors. The optimal treatment of CUPS remains unknown and can only be determined through randomized controlled trials.