Optimizing gastric cancer treatment: the role of LODDs in lymph node staging

Gastric cancer is one of the most common malignancies worldwide and is associated with poor prognosis, placing a considerable burden on public health. Overall treatment outcomes remain unsatisfactory, and accurate lymph node staging is essential for optimizing therapeutic strategies and improving survival. Although the prognostic value of log odds of positive lymph nodes (LODDs) has been investigated in gastric cancer, its comparative performance against AJCC N stage and lymph node ratio (LNR) in gastric adenocarcinoma, particularly with external institutional validation and sensitivity analyzes based on adequate lymph node evaluation, remains insufficiently characterized. This study aimed to compare the prognostic value of AJCC-N, LNR and LODDs in gastric adenocarcinoma and to develop an externally validated LODDs-based prognostic nomogram.

We included 4,054 patients with gastric adenocarcinoma from the SEER database (2015-2019) who underwent gastrectomy and had available regional lymph node examination data, as well as 383 patients from the First Affiliated Hospital of Hainan Medical University who underwent gastrectomy with operatively and pathologically confirmed D2 lymphadenectomy. Because anatomical D2 lymphadenectomy status cannot be directly identified in SEER, examined lymph node count was used as an indicator of the adequacy of lymph node evaluation, and sensitivity analyzes were performed in SEER patients with ≥16 examined lymph nodes. Clinicopathological variables included sex, age, race, tumor size, T stage, AJCC N stage (AJCC-N), lymph node ratio (LNR), and log odds of positive lymph nodes (LODDs). The primary endpoint was overall survival (OS), defined as the time from the date of surgery to death from any cause or last follow-up. Between-group comparisons were performed using the chi-square test. Optimal cut-off values were determined with X-tile software. Survival differences were evaluated by Kaplan-Meier curves. Time-dependent receiver operating characteristic (ROC) curves and corresponding time-dependent areas under the curve (AUCs) were used to compare the predictive performance of AJCC-N, LNR and LODDs for 1-, 3- and 5-year OS. Cox regression models were applied to identify independent prognostic factors, which were then incorporated into a nomogram. Nomogram performance was assessed using calibration curves, time-dependent C-index analysis and decision curve analysis (DCA).

AJCC-N, LNR and LODDs were strongly and positively correlated in all three datasets (P < 0.001). Time-dependent ROC analysis showed that LODDs had slightly larger areas under the curve than LNR and AJCC-N for predicting 1-, 3- and 5-year survival. Multivariable Cox regression confirmed that LODDs, together with sex, age, race, T stage and tumor size, were independent risk factors for overall survival (P < 0.05). The nomogram constructed from these factors showed good agreement between predicted and observed outcomes on calibration curves, and DCA indicated meaningful clinical net benefit across a broad range of threshold probabilities.

LODDs sensitively reflects metastatic tumor burden and showed favorable prognostic performance compared with AJCC-N and LNR in gastric adenocarcinoma. The LODDs-based nomogram may serve as a useful tool for individualized prognostic assessment, although further validation in well-defined lymphadenectomy cohorts is warranted.