Abstract
BACKGROUND: Several studies investigated the utility of inflammation and nutritional markers in predicting the prognosis in patients with gastric cancer; however, the markers with the best predictive ability remain unclear. This retrospective study aimed to determine inflammation and nutritional markers that predicted prognosis in elderly patients over 75 years of age undergoing curative gastrectomy for gastric cancer. METHODS: Between January 2005 and December 2015, 497 consecutive elderly gastric cancer patients aged over 75 years underwent curative gastrectomy in 12 institutions. The geriatric nutritional risk index (GNRI), prognostic nutritional index, neutrophil/lymphocyte ratio, platelet/lymphocyte ratio, and C-reactive protein/albumin ratio were examined as prognostic markers for overall survival (OS) and disease-specific survival (DSS) using area under the curve (AUC) using receiver operating characteristic (ROC) curve analysis. RESULTS: The GNRI had the highest AUC and predictive value for both OS (0.637, p < 0.001) and DSS (AUC 0.645, p < 0.001). The study cohort was categorized into the high and low GNRI groups based on the optimal GNRI cut-off values for OS (97.0) and DSS (95.8) determined with the ROC analysis. For both OS and DSS, there was a significant correlation between the GNRI and several clinicopathological factors including age, body mass index, albumin, American Society of Anesthesiologists physical status score, depth of tumor invasion, lymph node metastasis, lymphatic invasion, pathological stage, operation duration, bleeding, procedure, approach, death due to primary disease, and death due to other disease. The GNRI remained a crucial independent prognostic factor for both OS (Hazard ratio [HR] = 1.905, p < 0.001) and DSS in multivariate analysis (HR = 1.780, p = 0.043). CONCLUSIONS: Among a panel of inflammation and nutritional markers, the GNRI exhibited the best performance as a prognostic factor after curative gastrectomy in elderly patients with gastric cancer, indicating its utility as a simple and promising index for predicting OS and DSS in these patients.