Second Clinical School, Lanzhou University, Lanzhou, China; Key Laboratory of Medical Imaging, Gansu Province, China; Department of Radiology, Lanzhou University Second Hospital, Lanzhou, China.
Second Clinical School, Lanzhou University, Lanzhou, China; Department of Neurosurgery, Lanzhou University Second Hospital, Lanzhou, China.
Clin Radiol. 2021 Jun;76(6):473.e17-473.e24. doi: 10.1016/j.crad.2021.02.012. Epub 2021 Mar 14.
The purpose of this study was to investigate the relationship between epidermal growth factor receptor (EGFR) mutation status and computed tomography (CT) features in patients with lung adenocarcinoma.
A total of 483 patients with lung adenocarcinoma diagnosed between January 2015 and April 2020 were included in this study. All patients underwent a preoperative chest CT, and a total of 31 detailed CT features were quantified. The mutation status of EGFR exon 18-21 was detected by a polymerase chain reaction (PCR)-based amplified refractory mutation system. Student's t and Fisher's exact or chi-square tests were used to compare continuous and categorical variables, respectively. Least absolute shrinkage and selection operator (LASSO) regularisation was used to determine the optimal combination of CT features and clinical characteristics to predict the EGFR mutation status. The model was tested using a validation set consisting of 120 patients.
EGFR mutations were found in 249 (51.6%) of 483 patients with lung adenocarcinoma. Univariate analysis showed that 14 CT features and two clinical characteristics correlated significantly with the EGFR mutation status. Smoking history, long-axis diameter, bubble-like lucency, pleural retraction, thickened bronchovascular bundles, and peripheral emphysema were independent predictors of the EGFR mutation status, according to LASSO regularisation. In the training and verification cohorts, the areas under the curve of the prediction model were 0.766 and 0.745, respectively.
CT features of patients with lung adenocarcinoma can help predict the EGFR mutation status.
本研究旨在探讨表皮生长因子受体（EGFR）突变状态与肺腺癌患者 CT 特征之间的关系。
本研究纳入了 2015 年 1 月至 2020 年 4 月期间诊断为肺腺癌的 483 例患者。所有患者均接受了术前胸部 CT 检查，并对 31 个详细的 CT 特征进行了量化。EGFR 外显子 18-21 的突变状态通过聚合酶链反应（PCR）扩增耐药突变系统检测。采用 Student's t 检验和 Fisher's 确切检验或卡方检验分别比较连续变量和分类变量。最小绝对收缩和选择算子（LASSO）正则化用于确定预测 EGFR 突变状态的 CT 特征和临床特征的最佳组合。该模型使用包含 120 例患者的验证集进行测试。
在 483 例肺腺癌患者中，发现 EGFR 突变 249 例（51.6%）。单因素分析显示，14 个 CT 特征和两个临床特征与 EGFR 突变状态显著相关。根据 LASSO 正则化，吸烟史、长轴直径、泡状透亮影、胸膜回缩、增厚的支气管血管束和周围肺气肿是 EGFR 突变状态的独立预测因素。在训练集和验证集中，预测模型的曲线下面积分别为 0.766 和 0.745。
肺腺癌患者的 CT 特征有助于预测 EGFR 突变状态。
