Evaluation of an inflammation-based prognostic score in patients with inoperable pancreatic cancer

Background/Aims: Patients with pancreatic cancer have one of the poorest survival rates and selection of patients for active treatment remains problematical. The present study assesses the value of an inflammation-based score ( Glasgow Prognostic Score, GPS) in patients with inoperable pancreatic cancer. Methods: The GPS was constructed as follows: patients with both an elevated C-reactive protein (> 10 mg/l) and hypoalbuminaemia (< 35 g/l) were allocated a score of 2. Patients in whom only 1 or none of these biochemical abnormalities was present were allocated a score of 1 or 0, respectively. Results: One hundred and eighty-seven patients were studied and 49 (26%) underwent an operative palliative bypass procedure. At the end of follow-up, 181 (97%) patients died, 17% of patients were alive at 12 months. On univariate analysis, age ( p < 0.01), TNM stage ( p < 0.001) and the GPS ( p < 0.001) were significant predictors of survival. On multivariate survival analysis, stratified for bypass procedure, age ( hazard ratio 1.53, 95% CI 1.12 - 2.10, p = 0.008), TNM stage ( hazard ratio 1.70, 95% CI 1.33 - 2.18, p < 0.001) and the GPS ( hazard ratio 1.72, 95% CI 1.40 - 2.11, p < 0.001) remained independent significant predictors of survival. Conclusion: At diagnosis, the presence of a systemic inflammatory response ( as measured by the GPS) appears to be a useful indicator of poor outcome, independent of TNM stage, in patients with inoperable pancreatic cancer. Copyright (C) 2006 S. Karger AG, Basel and IAP.