肿瘤微环境特异性CT影像组学标签预测非小细胞肺癌免疫治疗疗效

打开文本图片集

Abstract：ObjectiveToconstructanomogramforpredictingtheeficacyofimmunecheckpointinhibitors（ICIs）inadvanced non-small cellungcancer （aNSCLC）byintegratingchest CTradiomics signature thatreflects the tumormicroenvironment （TME）andclinical parametersofthepatients.MethodsTranscriptomicandCTimagingdatafromTCGA，GEOandTCIA databases were integrated for weighted geneco-expresionnetwork analysis （WGCNA）ofthe GEOcohort toidentify the immunotherapy-related genes （IRGs）asociated with ICIs response. A prognostic model was built using these IRGs in the TCGA cohort to assessimmune microenvironment featuresacrossdiferent risk groups.Radiomics features were extracted from TCIA lung_3 cohort using PyRadiomics，and 94 features showing strong association with IRGs （ |r|>0.4） were selected. A retrospectivecohortconsistingof210aNSCLCpatientsreceivingfirst-lineICIsatGuangdongProvincialPeople'sHospitalwas analyzedand divided into training （n=147）and validation （n=63） groups. Leastabsolute shrinkage and selection operator was usedforradiomicfeaturesselectionandlogisticegressionwasappliedtoonstructacombinediical-radiomicmodeland nomogram for predicting ICIs therapyresponse.Theperformanceof the model wasevaluatedusingROCcurve，calibration curve，and decision curve analysis.Results WGCNA identified84 IRGsenriched in immune activation pathways.The combined model outperformed individual models in both the training （AUC=0.725， 95% CI： 0.644-0.807） and validation cohorts （AUC=0.706， 95% CI：0.577-0.836）.Calibration curveand decision curve analyses confirmed the clinical efcacy of the nomogram forpredicting ICIs therapy response inaNSCLC patients. Conclusion Thegenomic-radiomic-clinical multidimensionalpredictiveframework establishedinthisstudyprovidesaninterpretablebiomarkercombinationandclinical decision-making tool for evaluating ICIs eficacyin aNsCLC， potentiall facilitating personalized immunotherapydecisionmaking.

Keywords：non-smallcellungcancer;immunechckpoint inhibitors;tumor microenvironment; machinelearning;radiomics

肺癌在全球范围内发病率和病死率居高不下，其中非小细胞肺癌（NSCLC）占肺癌病例的 85% ，晚期患者的总体生存预后较差1。（剩余18841字）