Yun Seong Kim
Pusan National University School of Medicine, South Korea
Title: Texture Analysis Predicting EGFR Mutation and Recurrence in Lung Adenocarcinoma
Biography
Biography: Yun Seong Kim
Abstract
Objectives: We aimed to investigate the discriminative value of texture feature in EGFR mutation of lung adenocarcinoma and prognostic value of texture features using 18Fluorine-Fluorodeoxyglucose (18F-FDG) Positron Emission Tomography/Computed Tomography (PET/CT).
Methods: 63 lung adenocarcinoma patients with preoperative 18F-FDG PET/CT between January 2010 and December 2014 were included. Texture features are extracted automatically by using LIFEx software (University of Paris-Saclay, France), which provided texture features of gray level co-occurrence matrix, neighborhood gray-level different matrix, gray-level run length matrix and gray-level zone length matrix.
Results: Contrast (p=0.0179), dissimilarity (p=0.024), entropy (p=0.0097), HGRE (p=0.0093), HGZE (p=0.0044), LRHGE (p=0.0076), RLNU (p=0.0249), SRHGE (p=0.0105), SZHGE (p=0.014), ZLNU (p=0.011), SUVmax (p=0.0087), SUVmean (p=0.0084), SUVpeak (p=0.0105), TLG (p=0.0138) were lower in adenocarcinoma with mutant EGFR, while energy (p=0.102), homogeneity (p=0.0318), LGRE (p=0.0079), LGZE (p=0.0055), LRLGE (p=0.0084), LZLGE (p=0.037), SRLGE (p=0.0059), SZLGE (p=0.0417) were higher in adenocarcinoma of mutant EGFR. Entropy (odd ratio 0.2548, 95% CI 0.09-0.7209, p=0.01) was the independent predictor of EGFR mutation. In addition, LRHGE (hazard ratio 1.0017, 95% CI 1.0008-1.0026, p=0.0002) predicted the recurrence in lung adenocarcinoma.
Conclusion: Texture features predicted EGFR mutation in lung adenocarcinoma. In addition, LRHGE was an independent predictor of recurrence in patients with lung adenocarcinoma.