Different lymph node staging systems for patients with adenocarcinoma of esophagogastric junction

In addition to the traditional TNM N staging system, lymph node ratio (LNR) and log odds of metastatic lymph nodes (LODDS) staging methods were developed in cancers. This study aimed to examine their relative prognostic performance in patients with adenocarcinoma of the esophagogastric junction (AEG).

Patients who underwent surgical resection for AEG were identified from the Surveillance, Epidemiology, and End Results (SEER) Program and the First Affiliated Hospital of Xi'an Jiaotong University as the training and validation sets, respectively. The Akaike's Information Criterion (AIC), Harrell's C statistic, and ROC curves were utilized for comparison.

A total of 735 patients were involved in the training set. LODDS and LNR staging systems had better prognostic performance than the TNM N staging systems (when considered as a categorical variable: C index = 0.728 and 0.712 vs 0.671; AIC: 6247.537 and 6265.996 vs 6320.045; AUC: 0.762 and 0.719 vs 0.692. For the continuous model: C index = 0.675 and 0.686 vs 0.658; AIC = 6243.740 and 6261.027 vs 6355.077; AUC = 0.778 and 0.733 vs 0.693). In the validation set of 183 patients, the TNM N staging scheme outperformed the LODDS and LNR staging systems (C index = 0.788 vs 0.779 and 0.767; AIC = 1014.702 vs 1026.899 and 1025.288; AUC = 0.806 vs 0.787 and 0.791) when considered a categorical variable. However, when considered a continuous variable, the LODDS and LNR staging systems were better than the TNM N staging system (C index = 0.724 and 0.733 vs 0.747; AIC = 1018.075 and 1025.803 vs 1026.085; AUC = 0.811 and 0.810 vs 0.806).

The LNR and LODDS staging schemes could be considered new options for prognostic prediction of AEG with respect to lymph node status, especially when considered as continuous variables.