Hepatic transcatheter arterial chemoembolization alternating with systemic protracted continuous infusion 5-fluorouracil for gastrointestinal malignancies metastatic to liver: a phase II trial of the Puget Sound Oncology Consortium (PSOC 1104).
We assessed a regimen of alternating regional and systemic therapy in patients with gastrointestinal malignancies with liver-dominant metastases for feasibility, toxicity, response rate, response duration, patterns of progression, and progression-free and overall survival. Regional therapy comprised selective hepatic transcatheter arterial chemoembolization (TACE) using a suspension of cisplatin and particulate polyvinyl alcohol. This procedure was delivered between cycles of protracted continuous infusion 5-fluorouracil (PCI-5FU) as systemic chemotherapy. Patient eligibility criteria included: (a) having histologically documented adenocarcinoma arising from a gastrointestinal primary site with unresectable liver metastases bidimensionally measurable on computerized tomography scan; (b) age greater than 18 years; and (c) performance status 0-2 (Zubrod). PCI-5FU (250 mg/m2/day) was administered i.v. for 28 days, followed by the first TACE (TACE 1) delivered to the hepatic artery supplying the lobe with the greatest tumor burden. Restaging was performed before TACE 2 and TACE 3, which followed at monthly intervals. PCI-5FU for 21 days was sandwiched between each of the TACE treatments. After the final TACE, maintenance PCI-5FU was given for 28 days of each 35-day cycle until toxicity or progression. Between December 23, 1991, and January 19, 1995, 32 patients were registered in this trial, of whom 27 were eligible; 20 completed one or more treatment cycles and were evaluable for radiographic response. Patients with colorectal liver metastases predominated (74%). Twelve (44%) of 27 patients had failed one or more prior treatment regimens. There were no treatment-related deaths, and hematological and hepatic toxicities were generally manageable and reversible. Two patients, however, developed hepatic abscesses requiring drainage, and one patient developed an infarcted gallbladder, which necessitated cholecystectomy. There were no patients with complete responses; there were 8 (40%) with partial responses, 4 (20%) with minor responses, 2 (10%) with stable disease, and 6 (30%) who progressed on the treatment. The median duration of response for partial responders was 4.2 months (127 days; range, 56-245 days). The median reduction in carcinoembryonic antigen for responders was 87.5%. Two patients underwent subsequent resection of residual metastases; one of them is still alive at 58.4 months follow-up. The predominant site of disease progression was the liver; 25% of the patients progressed in extrahepatic sites. The median overall survival for the whole group is 14.3 months (95% confidence interval, 7.2-16.2). Actuarial overall survival for the whole group at 1 year and 2 years is 57 and 19%, respectively. Alternating systemic PCI-5FU and regional TACE (cisplatin/polyvinyl alcohol) is an active and feasible regimen with manageable toxicities in patients with metastatic gastrointestinal malignancies with liver-dominant disease and merits further investigation. The complications seen were in line with those reported at other specialized centers.