Prognostic efficacy of preoperative mGPS, SIS and LCS in patients with gastric cancer.

Background: Systemic inflammation and interactions with host-tumor are currently identified as a hallmark of cancer. The purpose of this study was to assess the prognostic value of preoperative modified Glasgow Prognostic Score (mGPS), systemic inflammation score (SIS) and "lymphocyte C-reactive protein score" (LCS) in gastric cancer (GC) patients.

Methods: 358 GC patients were enrolled in this retrospective study. Kaplan-Meier method, multivariate Cox regression analysis, time-dependent receiver operating characteristics analysis (ROC), concordance index (C-index), and Akaike information criterion (AIC) were applied for assessments of the prognostic values.

Results: Preoperative increased mGPS, SIS and LCS were all significantly linked with unfavorable overall survival using the Kaplan-Meier method (P < 0.001). Multivariate analysis proved that SIS was the only independent indicator among these three scoring systems. At the 4th-month point postoperatively, the time-dependent ROC curves of SIS and LCS crossed the curve of mGPS and were consistently superior to that of mGPS thereafter. The model incorporating SIS had higher C-index and smaller AIC than did the model without SIS or the models with mGPS or LCS.

Conclusion: Preoperative SIS exceeded both the mGPS and LCS and was the most clinically promising and feasible prognostic scoring system for GC patients.