Novel prognostic indicator combining inflammatory indicators and tumor markers for gastric cancer.

ABSTRACT

BACKGROUND: Gastric cancer (GC) is one of the most common malignant tumors worldwide, and we hope to identify an economical but practical prognostic indicator. It has been reported that inflammatory indicators and tumor markers are associated with GC progression and are widely used to predict prognosis. However, existing prognostic models do not comprehensively analyze these predictors. METHODS: This study retrospectively reviewed 893 consecutive patients who underwent curative gastrectomy from January 1, 2012, to December 31, 2015, in the Second Hospital of Anhui Medical University. Prognostic factors predicting overall survival (OS) were analyzed using univariate and multivariate Cox regression analyses. Nomograms including independent prognostic factors were plotted for predicting survival. RESULTS: Ultimately, 425 patients were enrolled in this study. Multivariate analyses demonstrated that the neutrophil-to-lymphocyte ratio (NLR, total neutrophil count/lymphocyte count × 100%) and CA19-9 were independent prognostic factors for OS (p=0.001, p=0.016). The NLR-CA19-9 score (NCS) is constructed as the combination of the NLR and CA19-9. We defined NLR<2.46 and CA19-9≤37 U/ml as an NCS of 0, NLR≥2.46 or CA19-9>37 U/ml as an NCS 1, and NLR≥2.46 and CA19-9>37 U/ml as an NCS of 2. The results showed that higher NCS was significantly associated with worse clinicopathological characteristics and OS (p<0.05). Multivariate analyses revealed that the NCS was an independent prognostic factor for OS (NCS1 p<0.001, HR=3.172, 95% CI=2.120-4.745; NCS2 p<0.001, HR=3.052, 95% CI=1.928-4.832). Compared with traditional predictive indices, the NCS had the highest AUC for a 12-month survival, a 36-month survival, a 60-month survival, and OS (AUC= 0.654, 0.730, 0.811, 0.803, respectively). The nomogram had a higher Harrell's C-index than the TNM stage alone (0.788 vs. 0.743). CONCLUSIONS: The NCS provides more accurate predictions of the prognosis of GC patients, and its predictive value is significantly better than that of traditional inflammatory indicators or tumor markers. It is an effective complement to existing GC assessment systems.