Predictive potential of preoperative Naples prognostic score-based nomogram model for the prognosis in surgical resected thoracic esophageal squamous cell carcinoma patients: A retrospective cohort study

Abstract

The present study aimed to establish an effective prognostic nomogram model based on the Naples prognostic score (NPS) for resectable thoracic esophageal squamous cell carcinoma (ESCC). A total of 277 patients with ESCC, who underwent standard curative esophagectomy and designated as study cohort, were retrospectively analyzed. The patients were divided into different groups, including NPS 0, NPS 1, NPS 2, and NPS 3 or 4 groups, for further analysis, and the results were validated in an external cohort of 122 ESCC patients, who underwent surgery at another cancer center. In our multivariate analysis of the study cohort showed that the tumor-node-metastasis (TNM) stage, systemic inflammation score, and NPS were the independent prognostic factors for the overall survival (OS) and progression-free survival (PFS) durations. In addition, the differential grade was also an independent prognostic factor for the OS in the patients with ESCC after surgery (all P < .05). The area under the curve of receiver operator characteristics for the PFS and OS prediction with systemic inflammation score and NPS were 0.735 (95% confidence interval [CI] 0.676–0.795, P < .001) and 0.835 (95% CI 0.786–0.884, P < .001), and 0.734 (95% CI 0.675–0.793, P < .001) and 0.851 (95% CI 0.805–0.896, P < .001), respectively. The above independent predictors for OS or PFS were all selected in the nomogram model. The concordance indices (C-indices) of the nomogram models for predicting OS and PFS were 0.718 (95% CI 0.681–0.755) and 0.669 (95% CI 0.633–0.705), respectively, which were higher than that of the 7th edition of American Joint Committee on Cancer TNM staging system [C-index 0.598 (95% CI 0.558–0.638) for OS and 0.586 (95% CI 0.546–0.626) for PFS]. The calibration curves for predicting the 5-year OS or PFS showed a good agreement between the prediction by nomogram and actual observation. In the external validation cohort, the nomogram discrimination for OS was better than that of the 7th edition of TNM staging systems [C-index: 0.697 (95% CI 0.639–0.755) vs 0.644 (95% CI 0.589–0.699)]. The calibration curves showed good consistency in predicting the 5-year survival between the actual observation and nomogram predictions. The decision curve also showed a higher potential of the clinical application of predicting the 5-years OS of the proposed nomogram model as compared to that of the 7th edition of TNM staging systems. The preoperative NPS-based nomogram model had a certain potential role for predicting the prognosis of ESCC patients.