Background and objectives: There is no appropriate grading system for prediction of survival of patients with metastatic liver cancer (MLC) from colorectal carcinoma. We propose the modified grading system for MLC from the present Japanese system.
Methods: We compared predictive accuracies of survival of 121 Japanese MLC patients of five systems, including clinical risk score (CRS) proposed by Memorial-Sloan-Kettering-Cancer-Center, original H-number (OHN) by Japanese Society for Cancer of the Colon and Rectum, revised H-number (RHN) and Grade by the same society (GJSCCR), and our modified Grade (MGJSCCR) based on OHN and presence of primary lymph node metastasis.
Results: Univariate analysis showed that discrimination of both disease-free and overall survival rates was significant for CRS, OHN and MGJSCCR (P < 0.05) but not for RHN and GJSCCR. Multivariate analysis showed CRS and MGJSCCR as the best systems for predicting disease-free and overall survival according to disease stage, for which Akaike information criteria (AIC) value was the lowest (423.7 and 313.9, hazard ratio 1.73 and 1.47, respectively for CRS, 423.9 and 313.5, hazard ratio 1.75 and 1.69, respectively for MGJSCCR; P < 0.05).
Conclusions: The simpler system of MGJSCCR is a better predictive grading system of prognosis of MLC patients who had undergone hepatic resection.
(c) 2008 Wiley-Liss, Inc.