Three-dimensional mean CT attenuation value of pure and part-solid ground-glass lung nodules may predict invasiveness in early adenocarcinoma

•Pre/minimally invasive adenocarcinoma shows GGO-dominant nodule on chest CT. •Pre/minimally invasive adenocarcinomas could be candidate for sublobar resection. •Accurate radiological prediction is crucial to determine the surgical procedure. •Three-dimensional densitometric evaluation is helpful to predict tumor invasiveness. •Mean CT value calculated by 3D-CT was well correlated with pathological features. AIM This study evaluated the relationship between three-dimensional (3D) mean computed tomography (CT) attenuation values of ground-glass nodules (GGN) and pathological invasiveness in early lung adenocarcinoma. The diagnostic accuracy of 3D CT attenuation values was compared with that of two-dimensional (2D) CT attenuation values and standardised uptake value on positron-emission tomography (PET). MATERIALS AND METHODS Surgical and radiological data from 96 pure or part-solid GGNs of <20 mm were analysed retrospectively. Mean 2D and 3D CT attenuation values of the tumours were obtained with semi-automated volumetric software. Pathological invasiveness was diagnosed according to the International Association for the Study of Lung Cancer (IASLC))/American Thoracic Society (ATS)/European Respiratory Society (ERS) classification. Pre-invasive lesions and minimally invasive adenocarcinomas were classified as non-invasive adenocarcinoma. Univariate and multivariate analyses determined relationships between pathological invasiveness and clinical/radiological findings. Receiver operating characteristic (ROC) analysis was performed to determine the optimal cut-off value for detecting invasive adenocarcinoma. RESULTS A total of 66 non-invasive and 30 invasive adenocarcinoma cases between 2010 and 2016 were analysed. Univariate analysis revealed four tumour invasiveness-associated predictors: maximum diameter, SUVmax, mean 2D CT attenuation value, and mean 3D CT attenuation value (p<0.05). Multivariate analysis revealed that the maximum diameter, SUVmax, and mean 3D CT attenuation value were significant predictors of pathological invasiveness (p=0.023, 0.022, 0.004). The area under the ROC curve to predict invasive adenocarcinoma for mean 3D CT attenuation value was 0.838 and the cut-off value was –489 HU. CONCLUSION The mean 3D CT attenuation value could distinguish pre-invasive lesions and minimally invasive adenocarcinoma from invasive adenocarcinoma. This study evaluated the relationship between three-dimensional (3D) mean computed tomography (CT) attenuation values of ground-glass nodules (GGN) and pathological invasiveness in early lung adenocarcinoma. The diagnostic accuracy of 3D CT attenuation values was compared with that of two-dimensional (2D) CT attenuation values and standardised uptake value on positron-emission tomography (PET). Surgical and radiological data from 96 pure or part-solid GGNs of <20 mm were analysed retrospectively. Mean 2D and 3D CT attenuation values of the tumours were obtained with semi-automated volumetric software. Pathological invasiveness was diagnosed according to the International Association for the Study of Lung Cancer (IASLC))/American Thoracic Society (ATS)/European Respiratory Society (ERS) classification. Pre-invasive lesions and minimally invasive adenocarcinomas were classified as non-invasive adenocarcinoma. Univariate and multivariate analyses determined relationships between pathological invasiveness and clinical/radiological findings. Receiver operating characteristic (ROC) analysis was performed to determine the optimal cut-off value for detecting invasive adenocarcinoma. A total of 66 non-invasive and 30 invasive adenocarcinoma cases between 2010 and 2016 were analysed. Univariate analysis revealed four tumour invasiveness-associated predictors: maximum diameter, SUVmax, mean 2D CT attenuation value, and mean 3D CT attenuation value (p<0.05). Multivariate analysis revealed that the maximum diameter, SUVmax, and mean 3D CT attenuation value were significant predictors of pathological invasiveness (p=0.023, 0.022, 0.004). The area under the ROC curve to predict invasive adenocarcinoma for mean 3D CT attenuation value was 0.838 and the cut-off value was –489 HU. The mean 3D CT attenuation value could distinguish pre-invasive lesions and minimally invasive adenocarcinoma from invasive adenocarcinoma.