A combined tumor burden–based nomogram for synchronous esophageal and gastric cancers: development and external validation
Résumé
Background: Synchronous dual-primary esophageal and gastric cancers (SEGC) are rare upper gastrointestinal malignancies presenting substantial challenges for prognostic assessment and individualized clinical evaluation. Conventional single-organ staging systems inadequately capture the cumulative tumor burden of synchronous cancers, limiting accurate prognostic assessment when tumors with different extents coexist. This study aimed to develop and externally validate a prognostic nomogram incorporating a Combined EG_Stage framework for exploratory risk stratification and post-treatment prognostic counseling in actively treated SEGC patients. Methods: A retrospective development cohort was assembled from a high-volume tertiary cancer center over 10 years (n = 46), with overall survival (OS) as the primary endpoint. A simplified Combined EG_Stage framework was developed to characterize cumulative anatomical tumor extent. Candidate prognostic predictors included age, treatment modality, and Combined EG_Stage. A multivariable Cox proportional hazards model was used to construct the nomogram. External validation used an independent cohort identified from 6,896 patients with multiple primary malignancies in the Surveillance, Epidemiology, and End Results (SEER) database, yielding 88 eligible patients. Model performance was assessed by discrimination, calibration, risk stratification, and decision curve analysis. Results: = 0.022). Decision curve analysis suggested limited and exploratory net benefit across selected threshold probabilities, which should be interpreted cautiously given the small sample size. Conclusions: This study developed and externally evaluated a clinically interpretable but modestly discriminative prognostic nomogram for actively treated SEGC patients. By integrating age, treatment modality as a real-world prognostic indicator, and combined anatomical tumor burden, the model may support exploratory risk stratification in this rare malignancy. Given the limited sample size and only modest discriminative performance, the model should be considered hypothesis-generating rather than definitive, and further multicenter validation is required to assess its robustness and generalizability.