18 F-fluorodeoxyglucose positron emission tomography/computed tomography-based radiomic features for prediction of epidermal growth factor receptor mutation status and prognosis in patients with lung adenocarcinoma

• Published in: Translational lung cancer research Vol. 9; no. 3; p. 563
• Main Authors: Yang, Bin, Ji, Heng-Shan, Zhou, Chang-Sheng, Dong, Hao, Ma, Lu, Ge, Ying-Qian, Zhu, Chao-Hui, Tian, Jia-He, Zhang, Long-Jiang, Zhu, Hong, Lu, Guang-Ming
• Format: Journal Article
• Language: English
• Published: China 01.06.2020
• Subjects: radiomics; epidermal growth factor receptor (EGFR); Lung adenocarcinoma; positron emission tomography; prognosis
• ISSN: 2218-6751

To investigate whether radiomic features from ( F)-fluorodeoxyglucose positron emission tomography/computed tomography [( F)-FDG PET/CT] can predict epidermal growth factor receptor ( ) mutation status and prognosis in patients with lung adenocarcinoma. One hundred and seventy-four consecutive patients with lung adenocarcinoma underwent ( F)-FDG PET/CT and gene testing were retrospectively analyzed. Radiomic features combined with clinicopathological factors to construct a random forest (RF) model to identify mutation status. The mutant/wild-type model was trained on a training group (n=139) and validated in an independent validation group (n=35). The second RF classifier predicting the 19/21 mutation site was also built and evaluated in an mutation subset (training group, n=80; validation group, n=25). Radiomic score and 5 clinicopathological factors were integrated into a multivariate Cox proportional hazard (CPH) model for predicting overall survival (OS). AUC (the area under the receiver characteristic curve) and C-index were calculated to evaluate the model's performance. Of 174 patients, 109 (62.6%) harbored mutations, 21L858R was the most common mutation type [55.9% (61/109)]. The mutant/wild-type model was identified in the training (AUC, 0.77) and validation (AUC, 0.71) groups. The 19/21 mutation site model had an AUC of 0.82 and 0.73 in the training and validation groups, respectively. The C-index of the CPH model was 0.757. The survival time between targeted therapy and chemotherapy for patients with mutations was significantly different (P=0.03). Radiomic features based on ( F)-FDG PET/CT combined with clinicopathological factors could reflect genetic differences and predict mutation type and prognosis.