Background: The aim of this study was to compare the risk scores of Fong et al., Nordlinger et al., and the TNM classification of colorectal liver metastases proposed by the UICC.
Methods: Data from 282 consecutive patients undergoing 303 liver resections for metastatic colorectal cancer between 1995 and 2006 at the Department of Surgery, University of Erlangen were analyzed. The median follow-up time was 34 months. A curative (R0) resection was performed in 92% of the patients.
Results: Applying the clinical risk score of Fong with preoperative data identified three risk groups. The survival rates between "low risk" (n = 22) and "intermediate risk" (n = 222) diverged (P = 0.073). The survival rates between "intermediate risk" and "high risk" (n = 59) differed significantly (P = 0.030). Using the risk scoring system of Nordlinger, patients were divided into two risk groups (i.e., "low risk" (n = 218) and "intermediate risk" (n = 68)). Significant differences in survival between the groups were noted (P = 0.012). Applying the clinical TNM classification of colorectal liver metastases revealed no significant differences in survival between the risk groups.
Conclusions: Our study found the clinical risk score developed by Fong et al. to be a reliable preoperative prognostic tool for selecting patients for surgical resection of colorectal liver metastases.
(c) 2009 Wiley-Liss, Inc.