ABSTRACT
BACKGROUND/AIMS: Little is known about the ability of the inflammation-based Glasgow prognostic score (GPS). METHODOLOGY: 106 patients who underwent curative resection for colorectal liver metastasis (CRLM) were analyzed. Patients with an elevated Creactive protein concentration (>10 mg/L) and hypoalbuminemia (<35 g/L) at admission were assigned a GPS 2, those with only 1 of these biochemical abnormalities were assigned a GPS 1, and those without either abnormality were assigned a GPS 0. RESULTS: Multivariate analysis showed that 2 variables, carcinoembryonic antigen (CEA) concentration > 30 ng/mL and a GPS 1 or 2, were independently prognostic of survival. Patients were classified into 3 groups on the basis of these 2 variables. Patients with GPS 1 or 2 and CEA concentration > 30 ng/mL were assigned a new score of 2, those with either 1 factor were assigned a new score of 1, and those with neither factors were assigned a new score of 0. The 5-year overall survival rates of new scores of 0, 1, 2 were 71.5%, 31.6%, and 0%, respectively (P < 0.0001). CONCLUSIONS: This simple staging system may be able to identify a subgroup of patients who are eligible for curative resection but show poor prognosis.