AUTHOR=Lin Yanhong , Chen Dinghang , Lu Jieming , Huang Yicheng , Han Ziyang , Kang Mingqiang 

TITLE=Prognostic value of the log odds of negative lymph nodes/T stage ratio (LONT) in postoperative esophageal cancer: a SEER-based study

JOURNAL=Frontiers in Oncology

VOLUME=Volume 15 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2025.1619106

DOI=10.3389/fonc.2025.1619106

ISSN=2234-943X

ABSTRACT=IntroductionSurgery remains the primary treatment for patients with esophageal cancer (EC), yet postoperative prognosis is often unsatisfactory. Accurate prediction of cancer-specific survival (CSS) can assist clinicians in personalized treatment planning. This study aimed to develop an interactive web-based tool to estimate CSS in patients with T1~3N0~2M0 EC after surgery, based on the log odds of negative lymph nodes/T stage ratio (LONT).MethodsA total of 2,221 patients with T1~3N0~2M0 EC were identified from the Surveillance, Epidemiology, and End Results (SEER) database. Patients were randomly divided into training and testing sets. Univariate Cox regression analysis was conducted to identify factors associated with CSS. Cox regression and random survival forest (RSF) models were used to compare the predictive performance of LONT and N stage. Model performance was evaluated using receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration curves. An interactive web-based tool was then constructed for individualized survival prediction.ResultsUnivariate analysis revealed that age, sex, T stage, N stage, chemotherapy, and LONT were significantly associated with CSS. ROC curve comparisons showed that LONT outperformed N stage in predictive accuracy, particularly for 1-year CSS. DCA and calibration curves indicated that the model had high predictive accuracy in both training and testing sets.DiscussionThe developed interactive web-based tool provides effective estimation of 1-, 3-, and 5-year CSS, as well as survival trends, in postoperative patients with T1~3N0~2M0 EC. This tool may aid clinical decision-making by enabling more accurate individualized prognosis prediction.