Lung Adenocarcinoma Manifesting as Pure Ground-Glass Nodules: Correlating CT Size, Volume, Density, and Roundness with Histopathologic Invasion and Size

Abstract

Objectives: The aims of this study were to quantify the relationship between computed tomography (CT) size, volume, density, and roundness of lung adenocarcinomas (ACs) manifesting as pure ground-glass nodules (pGGNs) on CT images and to correlate these parameters with histologic features of invasiveness.

Methods: From 2005 to 2015, 63 ACs manifesting as pGGNs on CT images were surgically resected at our institution. CT size was measured, and roundness, volumes and densities were computed. CT parameters were correlated to age and sex, as well as to size and number of invasive foci and histologic AC subcategories. Correlations were quantified with Spearman rank correlation coefficients.

Results: Of 63 ACs, 28 (44%) were AC in situ, 25 (40%) were minimally invasive AC, and 10 (16%) were invasive AC. Six of 35 nodules with invasive foci (17%) were smaller than 10 mm. Correlations between age and CT size, volume, density, and roundness were not significant (range r = -0.061 to 0.144, p = 0.285 to 0.902). Correlations between size and number of invasive foci with CT size (r = 0.417, p < 0.001 and r = 0.389, p = 0.003, respectively) were similar to the correlations with volume (r = 0.401, p = 0.001 and r = 0.350, p = 0.005, respectively) and stronger than the correlation with density (r = 0.237, p = 0.062 and r = 0.222, p = 0.081, respectively) and roundness (r = 0.059, p = 0.648 and r = -0.030, p = 0.831, respectively).

Conclusions: In ACs manifesting as pGGNs on CT images, nodule size is positively related to size and number of histologically invasive foci. However, invasive foci can be found in pGGNs smaller than 10 mm. Measuring volume and density of pGGNs provides no advantage over two-dimensional size measurements, which appear sufficient for risk estimation in clinical practice.

Keywords: Computed tomography; Lung adenocarcinoma; Pure ground-glass nodule; Radiologic-pathologic correlation.