Assessment of Small Airways with Computed Tomography: Mosaic Attenuation or Lung Density?

BACKGROUND Computed tomography (CT) assessment of air trapping has been considered useful as a measure of small airway disease. Mean lung density (MLD) and the percentage of the lung field occupied by low attenuation area (LAA%) can be evaluated automatically, and their expiratory/inspiratory (E/I) ratios correlate with asthma severity and spirometry parameters. However, mosaic attenuation, another indicator of air trapping, has been assessed visually, and its functional relevance remains controversial. OBJECTIVES This retrospective study was conducted to correlate mosaic attenuation, which was assessed visually and automatically, and the E/I ratios of MLD and LAA% (defined as areas <-960 Hounsfield units) with clinical and physiological variables, including impulse oscillometry (IOS) indices. MATERIAL AND METHODS In 36 nonsmoking patients with stable asthma, the lungs were scanned at full inspiration and full expiration. Mosaic attenuation was measured visually and automatically, by counting areas with CT values higher than the surrounding areas. MLD and LAA% were measured using our validated method. Spirometry, IOS, exhaled NO and the sputum eosinophil count were evaluated. RESULTS The automatic results and visual scores of mosaic attenuation correlated well on expiratory scans (r = 0.894) and to a lesser degree on inspiratory scans (r = 0.629; p < 0.0001 for both). However, only the E/I ratios of MLD and LAA% correlated with forced expiratory volume in 1 s/forced vital capacity of spirometry and the IOS indices of resistance from 5 to 20 Hz and the integrated area of low-frequency reactance. CONCLUSIONS Our automatic method for analysis of mosaic attenuation is likely useful, but the results themselves may not be reflecting small airway involvement of asthma, unlike the E/I ratios of MLD and LAA%.