Visual Emphysema as a Category Modifier in Lung-RADS Using a Secondary Analysis of National Lung Screening Trial.

Abstract

The Lung CT Reporting and Data System (Lung-RADS) does not consider emphysema, a lung cancer risk factor detectable on CT, when assessing nodule risk. This study aimed to evaluate the impact of incorporating emphysema into Lung-RADS on lung cancer diagnosis. In this secondary analysis of the National Lung Screening Trial data, CT arm participants with noncalcified nodules were assigned to Lung-RADS categories, and their emphysema severity was visually dichotomized. Lung cancer rates within each Lung-RADS category were compared based on emphysema severity. A modified Lung-RADS, reclassifying nodules with significant emphysema into a higher category, was evaluated against standard Lung-RADS. A study of 9,444 participants (782 [8.3%] with lung cancer) revealed difference in lung cancer rates across Lung-RADS categories based on visual emphysema severity: category 2 (2.6% versus 4.9%; P = .007), 3 (4.9% versus 9.0%; P < .001), 4A (9.2% versus 15.5%; P = .01), 4B (16.1% versus 24.1%; P = .12), and 4X (25.3% versus 33.2%; P = .008) without or with significant emphysema. Compared with standard Lung-RADS, modified Lung-RADS demonstrated a comparable area under the curve (0.73 versus 0.74, P = .009), increased sensitivity (61.3% versus 67.6%, P < .001), decreased specificity (77.2% versus 71.4%, P < .001), and improved goodness of fit (P = .008) for predicting lung cancer. Lung cancer rates differ by emphysema severity within Lung-RADS categories. Using the visual emphysema severity as a category modifier in Lung-RADS increased sensitivity while achieving comparable area under the curve for lung cancer.