LAUSR

AIM To investigate the differential diagnosis value of preoperative computed tomography (CT) features between pre/minimally invasive and invasive adenocarcinoma in pulmonary mixed ground glass nodules (mGGNs). MATERIALS AND METHODS The histopathological data and CT images of 146 mGGNs in 141 patients were reviewed retrospectively. Logistic regression analysis and receiver operating characteristic (ROC) curve analysis were performed to identify the CT features differentiating between pre/minimally invasive and invasive adenocarcinoma and to evaluate their accuracy. RESULTS In univariate analysis, there were significant differences (p<0.05) in the nodule diameter, volume, density, mass, solid portion volume, shape, margin, air bronchogram, and pleural retraction between pre/minimally invasive and invasive adenocarcinoma. Multivariate logistic regression analyses revealed that nodule mass and volume were statistically significant independent differentiators. ROC curve analysis was performed to evaluate the differentiators. According to the corresponding ROC curve, the optimal cut-off mass to differentiate pre/minimally invasive adenocarcinoma from invasive adenocarcinoma was 254.87 mg, with a sensitivity of 84.52%, a specificity of 88.71%, and an accuracy of 86.30%. Compared with the area under the ROC curve (AUC) for mass, volume, and diameter, the differential diagnosis value of mass was higher than those of volume and diameter. CONCLUSION Nodule mass and volume were significant differentiators of pre/minimally invasive adenocarcinoma from invasive adenocarcinoma in mGGN, and mass had a higher differential diagnosis value.