The pretreatment Controlling Nutritional Status (CONUT) score is an independent prognostic factor in patients with resectable thoracic esophageal squamous cell carcinoma: results from a retrospective study

BACKGROUND: The purpose of this study was to investigate the impact of the Controlling Nutritional Status (CONUT) score on survival compared with the platelet to lymphocyte ratio (PLR), the neutrophil to lymphocyte ratio (NLR), and the Glasgow Prognostic Score (GPS) in patients with resectable thoracic esophageal squamous cell carcinoma (ESCC). METHODS: One hundred eighty-five consecutive patients who underwent subtotal esophagectomy with curative intent for resectable thoracic ESCC were retrospectively reviewed. Time-dependent receiver operating characteristic curve analyses for 3-year overall survival (OS) as the endpoint were performed, and the maximal Youden indices were calculated to assess discrimination ability and to determine the appropriate cut-off values of CONUT, PLR, and NLR. The patients were then classified into high and low groups based on these cut-off values. Correlations between CONUT and other clinicopathological characteristics were analyzed. Prognostic factors predicting overall survival (OS) and relapse-free survival (RFS) were analyzed using Cox proportional hazards models. RESULTS: The areas under the curve predicting 3-year OS were 0.603 for CONUT, 0.561 for PLR, 0.564 for NLR, and 0.563 for GPS. The optimal cut-off values were two for the CONUT score, 193 for PLR, and 3.612 for NLR. The high-CONUT group was significantly associated with lower BMI, high-PLR, high-NLR, and GPS1/2 groups. On univariate analysis, high-CONUT, high-PLR, high-NLR, and GPS 1/2 groups were significantly associated with poorer OS and RFS. Of these factors, multivariate analysis revealed that only the CONUT score was an independent prognostic factor for OS (HR 2.303, 95 % CI 1.191–4.455; p = 0.013) and RFS (HR 2.163, 95 % CI 1.139–4.109; p = 0.018). CONCLUSIONS: The CONUT score was an independent predictor of OS and RFS before treatment and was superior to PLR, NLR, and GPS in terms of predictive ability for prognosis in patients with resectable thoracic ESCC.