肺癌患者 z-bo 2020-10-22 405

　　In the past two decades, significant progress has been made in the development and selection of personalized treatments for lung cancer patients. Non-small cell lung cancer is still the main type of lung cancer and the leading cause of cancer-related deaths worldwide. Currently, there are two treatment strategies, tyrosine kinase inhibitors and immune checkpoint inhibitors, but it may not be easy to choose the best treatment for patients with non-small cell lung cancer. Maybe. During the treatment, the biomarkers in the patient's body changed and the treatment failed. Therefore, scientists at Mofit Cancer Research Center have developed a non-invasive and accurate method to analyze tumors in patients, mutations and biomarkers help determine the best treatment for patients.

　　Related research results have been published in the international journal Nature Communications. In this article, the researchers used a deep learning model to show how to use positron emission imaging/computed tomography (PET/CT) for recognition. Which patients with non-small cell lung cancer are sensitive to tyrosine kinase therapy, and which patients can benefit from immune checkpoint inhibitor therapy; the model is a radioactive tracer element 18F-fluorodeoxyglucose (18F-fluorodeoxy) glucose, a kind of Carbohydrate molecules used for PET/CT imaging using 18F-FDGPET/CT imaging describe abnormal glucose metabolism sites, which helps to accurately analyze the patient’s tumor characteristics. Researcher Dr. Matthew Schabath said that this imaging technique, called 18F-FDGPET/CT, is widely used to stage patients with non-small cell lung cancer. The use of glucose radiotracer elements is usually affected by EGFR activation and inflammation in patients. EGFR (epidermal growth factor receptor) is the most common mutation in patients with non-small cell lung cancer. The mutation status of EGFR is considered to be a predictor of patient treatment. Patients with mutations in EGFR activity usually require tyrosine kinase therapy.

　　In this study, researchers used retrospective data from patients with non-small cell lung cancer from two Chinese institutions to develop a deep learning model based on 18F-FDGPET/CT. These two institutions are Shanghai Dragon. This hospital model and the Affiliated Hospital of Hebei Medical University 4 can classify EGFR mutation status in patients by generating EGFR deep learning scores for each patient. After creation, researchers can use patients from the other two institutions. Data validation model. The other two institutions are the Fourth Affiliated Hospital of Harbin Medical College and Mofit Cancer Research Center.

　　Researcher WeiMu said in a previous study that he used radioactive mixing as a non-invasive method to predict EGFR mutation status in patients, but compared with other studies, our researcher’s analysis can more accurately predict EGFR mutations in patients . At the same time, it can produce many benefits, such as training, validation and deep learning score detection of multi-cohort data from the above four institutions, which can increase the diversity of the model. According to the researchers, deep learning scores are generally positively correlated with longer progression-free survival in patients treated with tyrosine kinases, while continued clinical practice in patients receiving checkpoint inhibitor immunotherapy is positively correlated. We found a negative correlation between survival rate and longer progression-free survival rate. Later, researchers evaluated this new model with different treatment methods in practice. We want to conduct a more detailed study to confirm that it can be used as a clinical decision support tool.