BackgroundWe sought to investigate the prognostic value of complete blood count (CBC)-based biomarkers for patients with resectable gastric cancer (GC).MethodsPatients with GC who underwent primary surgical resection between December 2008 and December 2013 were included. The estimated area under the curve (AUC) and multivariate Cox regression models were used to identify the best CBC-based biomarker. Time-dependent receiver operating characteristic (t-ROC) curve analysis was used to predict overall survival and compare the prognostic impact.ResultsIn the 1810 patients analyzed, the median follow-up period was 51.0 months (range 1–101 months). Based on multivariate analysis, the lymphocyte-monocyte ratio (LMR) and hemoglobin (Hb) level were independent prognostic factors (both P < 0.05). Based on the LMR and Hb level, we established the CBC-based inflammatory score (CBCS). A higher CBCS was associated with older age, female sex, higher American Society of Anesthesiologists (ASA) score, proximal tumor location, larger tumor size, later stage and vascular involvement (all P < 0.05). Univariate analyses showed that a higher CBCS was also associated with worse overall survival (OS), which was consistent in each stage (all P < 0.05). Multivariate analysis revealed that the CBCS was a significant independent biomarker (P < 0.05). The AUC for the CBCS (0.627) was significantly higher than the AUCs for the LMR (0.573) and Hb level (0.605) (both P < 0.05). Furthermore, the t-ROC curve of the CBCS was superior to that of the prognostic nutritional index (PNI), systemic immune-inflammation index (SII), modified Glasgow prognostic score (mGPS) and C-reactive protein/albumin ratio (CRP/Alb) throughout the observation period.ConclusionThe preoperative LMR and Hb level were optimal CBC-based biomarkers for predicting OS in GC patients after curative resection. Based on the LMR and Hb, we developed a novel and easily obtainable prognostic score called the CBCS, which may improve the prediction of clinical outcomes.