Exploration Of Indicators For Predicting The Efficacy Of Immune Checkpoint Inhibitors In Non-small Cell Lung Cancer Based On Differential Genes Screened By Radiology

Purpose Lung cancer is a disease with high morbidity and mortality among all malignant tumors in the world,among which non-small cell lung cancer((NSCLC))is the most common histological subtype.Checkpoint inhibitors targeting PD-(L)-1 improve the long-term survival of patients with advanced NSCLC.However,PD-1/PD-L1 antibody monotherapy is beneficial to only a small number of patients,and the only clinically approved level of PD-L1 expression is not ideal for predicting the efficacy of PD-1/PD-L1 checkpoint inhibitors.Biomarkers with high specificity and detection rate are needed to predict the efficacy of PD-1/PD-L1 immune checkpoint inhibitors.Radiology is to extract high-throughput imaging features from CT and other images,use computer analysis technology to fully mine the hidden information in medical images,and analyze different clinical phenotypes of diseases,such as genotyping,disease diagnosis,therapeutic efficacy and clinical outcomes.Compared with the shortcomings of pathological biopsy in many aspects,such as invasiveness,limitation,sampling error and so on,radiology can comprehensively analyze the focus information through noninvasive methods,and help the clinic to make a better diagnosis and treatment plan,so as to achieve accurate medical treatment.In recent years,with the rapid development of radiology,there has been some progress in predicting the efficacy of immunosuppressive agents.In this study,three-dimensional tumor stereoscopic cutting and radiological feature extraction were performed on the preoperative CT images of patients with non-small cell lung cancer by the combination of radiomics and genomics.Clustering according to the radiological features,the differential genes among different imaging subtypes were analyzed,and the correlation between candidate genes and CD8+T lymphocyte infiltration was analyzed,so as to provide a new choice for predicting the efficacy of immunotherapy for non-small cell lung cancer.Further guide the immunotherapy of non-small cell lung cancer and further analyze the mechanism of differential genes predicting the efficacy of PD-1/PD-L1 immune checkpoint inhibitors.Materials and methods The CT scan images and clinical information of 89 patients with non-small cell lung cancer((NSCLC))who received surgery and treatment were downloaded from the cancer imaging database(TCIA,The Cancer Imaging Archive),and the corresponding gene expression data(micro array)was downloaded from GEO database(Gene Expression Omnibus).The NVIDIA AI Assisted Annotation expansion package of 3D-Slicer(4.11.0)was used to cut the primary lesions of preoperative CT images of 89 patients with non-small cell lung cancer((NSCLC)).The region of interest((Region Of Interest,ROI),)was obtained and the PyRadiomics expansion package in 3D-Slicer was used to extract the radiological features of the region of interest.Cluster score(R language NbClust package)was used to analyze all radiological characteristics of 89 ROI,and radiological subtypes were obtained.the differential genes among different imaging subtypes were analyzed by R language Limma package,and these genes were used as candidate genes.The correlation between candidate genes and CD8+T lymphocyte infiltration was analyzed in tumor immune assessment database(Tumor Immune Estimation Resource,TIMER).Positive correlation genes are expected to become indicators for predicting the efficacy of PD-1/PD-L1 immune checkpoint inhibitors in non-small cell lung cancer,and radiation subgroups with high expression of differential genes may benefit from immunotherapy.Finally,protein interaction and functional enrichment analysis were carried out by using STRING online analysis tool for the mechanism of differential genes predicting the efficacy of PD-1/PD-L1 immune checkpoint inhibitors.Results In this study,pre-processing,tumor cutting and radiological feature extraction were performed on the preoperative CT images of 89 patients with non-small cell lung cancer.2 radiological subtypes were obtained by cluster analysis according to radiological features,and 19 statistically significant differential genes were obtained by radiological subtype differential gene analysis.Among them,CCL19 was closely related to CD8+T lymphocyte infiltration in non-small cell lung cancer,and CCL19 was found in lung adenocarcinoma(cor=0.3346,).There was a moderate positive correlation between CCL19 and CD8+T lymphocyte infiltration in lung squamous cell carcinoma(P <0.0001)and lung squamous cell carcinoma(cor=0.4224,P <0.0001).CCL19 is expected to be used as an index to predict the efficacy of PD-1/PD-L1 immunocheckpoint inhibitors in non-small cell lung cancer.Group-2 radiological subtypes with high expression of CCL19 may benefit from immunotherapy.In addition,through protein interaction network and functional enrichment analysis,it was found that the biological processes involved in differential genes mainly included immune reaction,B cell proliferation,lymphocyte chemotaxis,lymphocyte activation and so on.Cell composition analysis showed that the related molecules were concentrated in the extracellular region,and the molecular functions involved were cytokine activity,chemokine activity,CCR chemokine receptor binding and so on.The results show that these differential genes are related to the function of immune cells and tumor immune microenvironment..Conclusion The differentially expressed genes in non-small cell lung cancer can be effectively identified based on the radiological characteristics,and the differentially expressed genes are correlated with the function of immune cells and tumor immune microenvironment,and the genes positively correlated with CD8+T lymphocytes are expected to become indicators for predicting the efficacy of PD-1/PD-L1 immunosuppressive checkpoint inhibitors in non-small cell lung cancer.In this study,the radiogenomics method of screening differentially expressed genes based on radiology provides a new idea for the study of the molecular mechanism of non-small cell lung cancer.In addition,it provides a promising,non-invasive and reproducible method for predicting the efficacy of immunotherapy..