Prediction of Mediastinal Station IV Lymph Node Metastasis in Non-small Cell Lung Cancer

Background: Lung cancer is one of the malignant tumors with the highest morbidity and mortality in the world. non-small cell lung carcinoma (NSCLC) accounts for about 85% of lung cancer, and its 5-year survival rate is about 19%. Mediastinal lymph node metastasis is a common metastasis pathway in non-small cell lung carcinoma (NSCLC), and its occurrence is closely related to lymphatic drainage pattern. The lymphatic drainage pattern of different lung lobe tumors is also different. Many studies have shown that the fourth and seventh stations of mediastinal lymph nodes are the areas with high incidence of lymph node metastasis. In particular, lymph node metastasis at station 4 was associated with poorer patient outcomes. Although systemic lymph node dissection usually includes at least three sets of mediastinal lymph nodes, including station 7, there is no uniform protocol for station 4 dissection. This situation has a negative impact on the stage and prognosis assessment of lung cancer patients. CT examination is an important tool to evaluate mediastinal lymph node status, but the accuracy is not high. The emerging CT radiomics has shown great application prospect in the accuracy of diagnosis of lymph node metastasis. The use of radiomics to evaluate the station lymph nodes is helpful to improve the accuracy of the diagnosis of lymph node metastasis, and it is also expected to provide a more scientific basis for determining the scope of lymph node dissection. Objective: To predict the lymph node metastasis of the fourth mediastinal group by CT imaging, and to help determine the scope and stage of lymph node dissection by comparing with the pathological gold standard. Study design: The clinical and pathological data of newly diagnosed non-small cell lung cancer patients admitted to the Department of Cardiothoracic Surgery of Qilu Hospital from 2017 to March 2024 were retrospectively collected. Clinicopathologic data and imaging data of 150 patients are expected to be collected. Inclusion criteria included patients who had undergone pathological examination of the fourth group of lymph nodes at initial visit and enhanced CT scan within two weeks prior to surgery. Lymph node Region of Interest (ROI) was sketched for all enrolled patients, and all lymph nodes were divided into metastatic and non-metastatic groups. The purpose of the study was to analyze the imaging data of these patients and integrate the corresponding clinical and pathological information, such as clinical factors: age, lung lobe, gender, image signs, etc, and pathological factors: pathological type, histological type, etc. In addition, the short diameter of each lymph node was measured to determine the metastasis rate under different short diameter criteria. Using machine learning technology to construct prediction model. The purpose of the model was to identify and extract the features of the fourth group of lymph nodes with and without metastasis. By careful comparison with the final pathological results, the investigators will evaluate the accuracy and effectiveness of the model in predicting the status of lymph node metastasis. The model can quantify the risk of lymph node metastasis and help doctors develop more personalized treatment plans.