Comparison of portal vein chemotherapy with hepatic artery chemotherapy in the treatment of liver micrometastases.

This study was conducted in a rat model of hepatic micrometastases generated by the intraportal injection of a colonic carcinoma cell line bound to polystyrene microspheres. Thirty-two animals received continuous infusions of 5-fluorouracil (5-FU) into either the portal vein or hepatic artery for a period of 7 days. Drug infusions were begun at 0, 2, 4, and 6 days after the time of tumor inoculation in four groups of animals, respectively. Subsequent tumor growth in these animals at 1 month was compared with tumor growth in 11 control animals that did not receive chemotherapy. When 5-FU was administered via the portal vein on the same day as tumor inoculation, liver metastases were reduced by approximately 91% (p = 0.004). Portal vein chemotherapy administered 6 days after tumor inoculation, when macroscopic nodules had a mean diameter of 0.33 +/- 0.03 mm (SE), produced no tumor response (p = 0.36). Histologic examination of these lesions revealed early invasion outside distended portal venules. In contrast, hepatic artery 5-FU infusions administered at 0, 2, 4, and 6 days after tumor implantation all reduced the subsequent development of hepatic metastases by approximately two thirds of that observed in the untreated group (p = 0.004). We conclude that hepatic artery chemotherapy may have an important complementary role to play as an adjuvant treatment for gastrointestinal cancer.