Clinical and CT characteristics for predicting lymph node metastasis in patients with synchronous multiple primary lung adenocarcinoma.

Purpose: This study aims to investigate the risk factors for lymph node metastasis (LNM) in synchronous multiple primary lung cancer (sMPLC) using clinical and CT features, and to offer guidance for preoperative LNM prediction and lymph node (LN) resection strategy.

Materials and methods: A retrospective analysis was conducted on the clinical data and CT features of patients diagnosed with sMPLC at the Third Affiliated Hospital of Kunming Medical University from January 1, 2018 to December 31, 2022. Patients were classified into two groups: the LNM group and the non-LNM (n-LNM) group. The study utilized univariate analysis to examine the disparities in clinical data and CT features between the two groups. Additionally, multivariate analysis was employed to discover the independent risk variables for LNM. The diagnostic efficacy of various parameters was evaluated using the receiver operating characteristic (ROC) curve.

Results: Among the 688 patients included in this study, 59 exhibited LNM. Univariate analysis revealed significant differences between the LNM and n-LNM groups in terms of gender, smoking history, CYFRA21-1 level, CEA level, NSE level, lesion type, total lesion diameter, main lesion diameter, spiculation sign, lobulation sign, cavity sign, and pleural traction sign. Logistic regression identified CEA level (OR = 1.042, 95%CI: 1.009-1.075), lesion type (OR = 9.683, 95%CI: 3.485-26.902), and main lesion diameter (OR = 1.677, 95%CI: 1.347-2.089) as independent predictors of LNM. The regression equation for the joint prediction was as follows: logit(p)= -7.569+0.041*CEA level +2.270* lesion type +0.517* main lesion diameter.ROC curve analysis showed that the AUC for CEA level was 0.765 (95% CI, 0.694-0.836), for lesion type was 0.794 (95% CI, 0.751-0.838), for main lesion diameter was 0.830 (95% CI, 0.784-0.875), and for the combine predict model was 0.895 (95% CI, 0.863-0.928).

Conclusion: The combination of clinical and imaging features can better predict the status of LNM of sMPLC, and the prediction efficiency is significantly higher than that of each factor alone, and can provide a basis for lymph node management decision.