Risk Stratification Of Mediastinal Lymph Node Metastasis In Lung Adenocarcinoma

Purpose:To investigate the risk factors of mediastinal lymph node(LN)metastasis in patients with primary lung adenocarcinoma.Then,we performed a prediction model to stratify the probability of mediastinal LN metastasis on the basis of our surgical experience,which can provide theoretical basis for thoracic surgeons in mediastinal lymph node dissection.Methods:A total of 245 patients with primary lung adenocarcinoma were divided into the metastatic group(66 patients)and the non-metastatic group(179 patients)according to the occurrence of mediastinal LN metastasis.The clinical data of the two groups were compared,and the risk factors for mediastinal LN metastasis in patients with lung adenocarcinoma were analyzed by the multivariate Logistic regression.The nomogram prediction model was constructed based on the independent risk factors of mediastinal LN metastasis in patients with lung adenocarcinoma of which calibration curve and decision curve were drawed.Next,the prediction models was constructed based on the concordance index(c-index).According to the total score of the prediction of lung adenocarcinoma mediastinal LN metastasis based on the nomogram model,the decision-making tree model was further established by using the recursive partition analysis method.Results:1.Between the two groups showed that age,the diameter of tumor,the histological classification of tumor,and the degree of tumor differentiation were the risk factors for metastasis of mediastinal LN,and the difference were statistically significant(all P<0.05).The age,tumor diameter and tumor differentiation are independent risk factors for mediastinal LN metastasis in patients with lung adenocarcinoma(OR=0.366,95%CI:0.189～0.706,P=0.003;OR=1.416,95%CI:1.059～1.893,P=0.019;OR=5.151,95%CI:2.720～9.755,P=0.001).2.Based on the age,tumor diameter and degree of tumor differentiation,a predictive model for mediastinal LN metastasis of lung adenocarcinoma was established,and the concordance index(c-index)of the nomogram was 0.819.The correction curve showed that the predicted probability of mediastinal LN metastasis of lung adenocarcinoma was in good agreement with the actual probability of LN metastasis.When the decision curve showed that the probability threshold of mediastinal LN metastasis of lung adenocarcinoma was 0.001～0.713,it was better to use the nomogram to predict the model.3.we stratified the cohort into four risk subgroups using recursive partitioning analysis(RPA).The low-risk group was patients with well or mid differentiated lung adenocarcinoma;the intermediate low-risk group was patients with poorly differentiated lung adenocarcinoma+the diameter of tumor ≤4.0cm+ age≥60 years old;the intermediate high-risk group was patients with poorly differentiated lung adenocarcinoma+the diameter of tumor ≤4.0cm+age≤59 years old;while the high-risk group was patients with poorly differentiated lung adenocarcinoma+ the diameter of tumor>4.0cm.The risk of mediastinal LN metastasis was 15%,39%,58%,and 77%for the low-risk,the intermediate low-risk,the intermediate high-risk,and the high-risk groups,respectively.Conclusion:Age,the diameter of tumor,and the degree of tumor differentiation were independent risk factors for mediastinal LN metastasis.We developed the first nomogram and RPA stratification system,which can provide theoretical basis for clinicians to conduct mediastinal LNs dissection for primary lung adenocarcinoma.