Multivariate Analysis of a Personal Series of 247 Patients with Liver Metastases from Colorectal Cancer

Abstract

Patients (117) with colorectal hepatic metastases has insertion of catheters for infusional chemotherapy. The 2 yr survival estimate of patients with < 50% hepatic replacement and no other adverse factors was 37%. Of 39 patients in this group 9 are alive at 24 mo. Of the catheters placed into the hepatic artery (HA), 23 were inserted; into the portal venous system (PV), 18; into both HA and PV, 64; or into an accessory HA following ligation, 12. Patients (59) had ligation of the common HA also. The 30-day postoperative mortality rate was 1.7% (2/117) and morbidity was 37.6%. The majority of complications were related to fever (61%, 27/44). Over the past 2 yr, 87% of patients were discharged within 10 days following surgery. Preoperative CEA [carcinomembryonic antigen] ranged from 0.5-12,150 ng/ml (median 165 ng/ml); 93% (78/84) had plasma CEA levels > 5 ng/ml. All patients had careful intraoperative staging: per cent hepatic replacement (PHR) ranged from 5-95% (median 60%) portal, celiac or periaortic lymph node metastases were observed in 31% (36/117). Initial intrahepatic chemotherapy programs consisted of either CAMP [cytoxan, actinomycin methotrexate and fluorouracil (5FU)] (9 patients), MAFL [actinomycin D, 5FU, methotrexate and levamisole] (60 patients), BFS [1,3-bis-(2-chloroethyl)-1-nitrosourea, 5FU and streptozotocin] (22 patients), continuous infusion FUDR [5-fluoro-2-deoxyuridine] (14 patients) or miscellaneous drugs (4 patients). Median survival time of 109 evaluable patients was 11.5 mo. The effect of 20 variables on the observed survival time was analyzed using a multivariate proportional hazard model. Three varibles influenced survival: PHR emerged as the most significant, P = 0.000001. Increased PHR was associated with decreased survival time. Lymph node metastases and prior chemotherapy were prognostic factors also, P = 0.0006 and P = 0.03, respectively. No patient with PHR > 80% lived > 8 mo. Utilization of these variables would appear to be necessary for accurate stratification and evaluation of future chemotherapy trials in patients with colorectal hepatic metastases.