Visual and Quantitative Interstitial Lung Abnormality Progression in COPDGene

Interstitial lung abnormalities (ILA) are visually identified changes on chest computed tomography (CT) scans that may represent early/mild pulmonary fibrosis. Quantitative interstitial abnormalities (QIA) measure potential parenchymal lung injury on chest CT scans using an automated algorithm. It is not known if combining these visual and quantitative assessments improves prediction of imaging progression. To assess the utility of quantitative imaging to predict imaging progression of ILA and adverse clinical outcomes in a cohort of smokers. ILA presence, subtypes, and progression, as well as QIA were assessed on chest CT scans from participants ~5 years apart in COPDGene. Multivariable logistic regression assessed associations with ILA progression, Cox proportional hazards assessed the relationship between ILA progression and mortality. 4373 participants had serial CT scans, 544 (12%) had ILA on at least one; of those 391 (72%) had imaging visual progression and 153 (28%) did not. Specific imaging features were associated progression, (e.g. traction bronchiectasis, OR=3.1, 95% CI 1.3-7.3, P=0.003). Among those with ILA, baseline quantitative measures (QIA and forced vital capacity [FVC]) were not associated with progression, however, visual imaging progression was associated with increased longitudinal change of QIA (mean difference 6.5%, 95% CI 4.9%-8.1%, P<0.0001). In ILA, QIA increase was associated with an increased rate of mortality independent of FVC decline, (HR=1.05, 95% CI 1.01-1.09, P=0.009). Baseline quantitative measures (QIA and FVC) were not associated with visual ILA progression, however longitudinal change in QIA was correlated with imaging progression and adverse clinical outcomes.