P60.05 Radiomic Signature to Predict Outcomes in EGFR-Mutant Non-Small Cell Lung Cancer

MINATTA, J.N.; DEZA, D.; AINESEDER, M.; MESTAS NUÑEZ, M.; MOSQUERA, C.; LUPINACCI, L.; BENITEZ, S.; SEEHAUS, A.; LUNA, D.; BERESÑAK, A.; DIAZ, F.

JOURNAL OF THORACIC ONCOLOGY

LIPPINCOTT WILLIAMS & WILKINS

Año: 2021 vol. 16

1556-0864

Radiomics has become an area of interest for tumor characterization in18F-Fluorodeoxyglucose positron emission tomography/computed tomography(18F-FDG PET/CT) imaging. The aim of the present study was to demonstrate howimaging phenotypes was connected to somatic mutations through an integrated analysisof 115 non-small cell lung cancer (NSCLC) patients with somatic mutation testingsand engineered computed PET/CT image analytics. A total of 38 radiomic featuresquantifying tumor morphological, grayscale statistic, and texture features were extractedfrom the segmented entire-tumor region of interest (ROI) of the primary PET/CT images.The ensembles for boosting machine learning scheme were employed for classification,and the least absolute shrink age and selection operator (LASSO) method was usedto select the most predictive radiomic features for the classifiers. A radiomic signaturebased on both PET and CT radiomic features outperformed individual radiomic features,the PET or CT radiomic signature, and the conventional PET parameters including themaximum standardized uptake value (SUVmax), SUVmean, SUVpeak, metabolic tumorvolume (MTV), and total lesion glycolysis (TLG), in discriminating between mutant-typeof epidermal growth factor receptor (EGFR) and wild-type of EGFR- cases with anAUC of 0.805, an accuracy of 80.798%, a sensitivity of 0.826 and a specificity of0.783. Consistently, a combined radiomic signature with clinical factors exhibited afurther improved performance in EGFR mutation differentiation in NSCLC. In conclusion,tumor imaging phenotypes that are driven by somatic mutations may be predicted byradiomics based on PET/CT images.