Decoding Tumor Phenotypes for ALK, ROS1, and RET Fusions in Lung Adenocarcinoma Using a Radiomics Approach

摘要: Quantitative imaging using radiomics can capture distinct phenotypic differences between tumors and may have predictive power for certain phenotypes according to specific genetic mutations. We aimed identify the clinicoradiologic predictors of with ALK (anaplastic lymphoma kinase), ROS1 (c-ros oncogene 1), or RET (rearranged during transfection) fusions in patients lung adenocarcinoma.A total 539 pathologically confirmed adenocarcinomas were included this retrospective study. The baseline clinicopathologic characteristics retrieved from patients' medical records ALK/ROS1/RET fusion status was reviewed. computed tomography (CT) positron emission evaluated a approach. Significant features fusion-positive tumor prediction model extracted all features, used calculate diagnostic performance predicting 3 fusions' positivity. compared versus ROS1/RET similarity 2 groups.The combination younger age, advanced stage, solid on CT, higher values SUV(max) mass, lower kurtosis inverse variance 3-voxel distance than those fusion-negative (sensitivity specificity, 0.73 0.70, respectively). significantly different central location, SUV(max), homogeneity 1-, 2-, distances, sum mean 2-voxel tumors.ALK/ROS1/RET possess clinical that enable good discrimination adenocarcinomas.