Real-World Use and Outcomes of Lung-RADS v1.1 Category 4B Versus Lung-RADS v2022 Category 0 for Suspected Infectious...

BACKGROUND. Lung-RADS version 2022 (v2022) updated the reporting of suspected infectious/inflammatory findings warranting short-term follow-up, replacing the option in Lung-RADS version 1.1 (v1.1) to classify such findings as Lung-RADS category 4B with an option to classify such findings as Lung-RADS category 0. OBJECTIVE. The purpose of this study was to compare real-world use and outcomes of Lung-RADS v1.1 category 4B and Lung-RADS v2022 category 0 for suspected infectious/inflammatory findings on lung cancer screening (LCS) examinations. METHODS. This retrospective study included 14,478 patients (mean age, 63.2 ± 7.0 [SD] years; 7801 men, 6677 women) who underwent LCS CT from January 2019 to December 2023, interpreted using either v1.1 or v2022. Radiology reports were reviewed to identify examinations reported as Lung-RADS 4B using v1.1 or Lung-RADS 0 using v2022 due to suspected infectious/inflammatory findings and to characterize primary findings. Cases were classified as benign or malignant on the basis of EMR data and follow-up imaging. Examinations classified as Lung-RADS 4B and Lung-RADS 0 were compared. RESULTS. The frequency of Lung-RADS 4B classification due to suspected infectious/inflammatory findings using v1.1 was 0.4% (55/14,478) and Lung-RADS 0 classification due to suspected infectious/inflammatory findings using v2022 was 0.8% (88/11,348) (p < .001). The primary findings for Lung-RADS 4B examinations were most commonly solid nodule (65%) or part-solid nodule (24%) and for Lung-RADS 0 examinations were most commonly ground-glass nodule/opacity (43%), solid nodule (25%), or tree-in-bud nodules (13%) (p < .001). Of Lung-RADS 4B examinations, 76% were benign, 22% were malignant, and 2% had insufficient follow-up; of Lung-RADS 0 examinations, 98% were benign, none were malignant, and 2% had insufficient follow-up (p < .001). No examination with a primary finding of ground-glass nodule/opacity or tree-in-bud nodules was malignant. CONCLUSION. For suspected infectious/inflammatory findings, Lung-RADS 0 for v2022 was used more frequently than Lung-RADS 4B for v1.1. However, no Lung-RADS 0 examination was malignant. Lung-RADS 0 designations were commonly for findings (i.e., ground-glass or tree-in-bud nodules) for which v2022 indicates short-term follow-up may not be warranted. CLINICAL IMPACT. Adherence to Lung-RADS guidance regarding types of suspected infectious/inflammatory findings warranting Lung-RADS 0 classification is important for avoiding potentially unnecessary follow-up recommendations.