A Multidimensional CT-Based Tumor Location Metric for Predicting Lymph Node Metastasis and Occult Lymph Node...

IntroductionLymph node (LN) metastasis of lung adenocarcinoma, particularly occult lymph node metastasis (OLNM), remains difficult to predict using conventional imaging and basic positional parameters, adversely affecting prognosis. This study aimed to: (1) evaluate a new CT index (a/A + b/B) for predicting OLNM in cN0 patients, and (2) create and validate a nomogram for predicting LN metastasis probability in CT-suspected cases.MethodsThis single-center retrospective study involved 387 lung adenocarcinoma patients (225 non-LN, 84 OLNM, and 78 with CT-suspected/pathologically confirmed LN metastasis) who had surgery from 2016 to 2024. A spatial index (a/A + b/B) was developed to combine two perpendicular anatomical axes, vertical proximity to the interlobar fissure and horizontal proximity to the mediastinum/hilum, into one standardized metric.ResultsThe combined ratio a/A + b/B was significantly lower in the OLNM and metastatic groups. A lower a/A + b/B was significantly associated with OLNM (OR = 0.175, 95% CI: 0.048-0.638, P = 0.007), suggesting a 0.1-unit decrease in the a/A + b/B ratio increased OLNM risk by 19%. A nomogram utilizing a/A + b/B differentiates CT-suspected LN metastasis from node-negative cases, showing AUCs of 0.83 (95% CI: 0.76-0.90) in training, 0.85(95% CI: 0.76-0.94) in internal validation, and 0.70(95% CI: 0.50-0.91) in external validation.ConclusionA ratio of a/A + b/B < 0.6 indicates a high OLNM risk in cN0 patients. This index can predict OLNM when LNs are small on CT, and a nomogram can differentiate CT-suspected LN metastasis from node-negative cases, meriting further study.