Treatment of cancer of the liver. Twenty years' experience with infusion and resection in 414 patients.

The factors influencing survival for patients with cancer of the liver were studied by reviewing the records of 414 patients operated on in a private oncology practice. Approximately half (47%) had colorectal metastasis; 17% had metastatic breast carcinoma, 14% had malignant hepatoma, 5% had metastatic melanoma, and the remainder had a variety of primary cancers. Eighty-two per cent of all patients had advanced liver disease when first diagnosed. One quarter of the patients had some type of resection; the remainder had abdominal exploration plus insertion of an infusion catheter into the hepatic artery. The postoperative mortality rate after liver resection for 108 patients was 6.5%. After resection, the most important prognostic factor influencing survival was the presence or absence of extrahepatic metastases. When possible, resection was by far the best treatment available, and the best results were seen in patients who had resection of a solitary lesion. For advanced disease, when resection was not possible, intra-arterial chemotherapy, primarily with 5-fluorouracil (5-FU), was associated with response rates of 36% for colorectal cancer, 45% for breast cancer, 13% for hepatocellular cancer, 12% for melanoma, and 14% for metastases from other primary sites. The patients who responded to infusion lived longer than those who did not respond. For example, at 18 months, 26% of the responders with colorectal cancer were alive, as were 50% of the responders with breast cancer and 40% of the responders with hepatocellular cancer. In contrast, at 18 months, there were no survivors among the nonresponders with colorectal, breast, or hepatocellular cancer. For those patients treated solely by infusion chemotherapy, the extent of disease in the liver was the most reliable factor in predicting the length of survival. However, very few patients treated in this manner lived longer than 3 years.