Novel Methylation‐Based Model for Accurate Diagnosis of Upper Mediastinal Lymph Node Metastasis in Esophagogastric Junction Adenocarcinoma

Even after surgical resection, the prognosis is poor for patients with esophagogastric junction adenocarcinoma who develop upper mediastinal lymph node metastasis. Moreover, the current preoperative diagnostic accuracy for upper mediastinal lymph node metastasis remains inadequate, which complicates treatment planning. To enhance diagnostic precision and facilitate personalized therapeutic strategies, we aimed to develop a novel diagnostic model that integrates a DNA methylation panel with conventional clinical parameters. First, genome-wide methylation profiling of 69 tumors identified seven CpG sites significantly associated with upper mediastinal lymph node involvement. We then evaluated a methylation panel based on these markers in two independent cohorts comprising a total of 133 treatment-naïve patients. The methylation panel alone achieved moderate diagnostic performance (AUC 0.71, 95% CI 0.52-0.91), whereas its integration with clinical data substantially improved accuracy (training cohort: AUC 0.87, 95% CI 0.75-0.99). Applying the same model formula to an independent pretreatment validation cohort confirmed its robustness (AUC 0.85, 95% CI 0.70-1.00). Furthermore, the methylation panel was identified as an independent prognostic factor for overall survival (p < 0.01). Finally, to facilitate clinical implementation, the integrated model was formalized as a nomogram. In summary, this novel diagnostic model enables highly accurate detection of upper mediastinal lymph node metastasis in esophagogastric junction adenocarcinoma, with the potential to optimize treatment decisions and reduce the need for invasive procedures, thereby contributing to precision oncology.