Prognostic value of preoperative inflammatory markers in resectable biliary tract cancer - Validation and comparison of the Glasgow Prognostic Score and Modified Glasgow Prognostic Score in a Western cohort.

INTRODUCTION: Established preoperative prognostic factors for risk stratification of patients with biliary tract cancer (BTC) are lacking. A prognostic value of the inflammation-based Glasgow Prognostic Score (GPS) and Modified Glasgow Prognostic Score (mGPS) in BTC has been indicated in several Eastern cohorts. We sought to validate and compare the prognostic value of the GPS and the mGPS for overall survival (OS), in a large Western cohort of patients with BTC. MATERIAL AND METHODS: We performed a retrospective single-center study for the period 2009 until 2017. 216 consecutive patients that underwent surgical exploration with a diagnosis of perihilar cholangiocarcinoma (PHCC), intrahepatic cholangiocarcinoma (IHCC), or gallbladder cancer (GBC) were assessed. GPS and mGPS were calculated where both CRP and albumin were measured pre-operatively (n = 168/216). Survival was analyzed by Kaplan-Meier estimate and uni-/multivariate Cox regression. RESULTS: GPS and mGPS were negatively associated with survival (p < 0.001/p < 0.001), and the association was significant in all three subgroups. GPS, but not the mGPS, identified an intermediate risk group: with GPS = 1 having better OS than GPS = 2 (p = 0.003), but worse OS than GPS = 0 (p = 0.008). In multivariate analyses of resected patients, GPS (p = 0.001) and mGPS (p = 0.03) remained significant predictors of survival, independent of postoperatively available risk factors. CONCLUSIONS: Preoperative GPS and mGPS are independent prognostic factors in BTC. The association to OS was shown in all patients undergoing exploration, in resected patients only, and in both cholangiocarcinoma and gallbladder cancer. Furthermore, GPS - which weights hypoalbuminemia higher - could identify an intermediate risk group.