Beyond the binary: rethinking the ILA to ILD continuum

In medicine, we define disease with the goal of differentiating illness from health. However, health and disease are often on a spectrum, and important clinical entities exist between these states. Interstitial lung abnormalities (ILAs), specific imaging findings on chest computed tomography (CT) scans in older adults, are on the spectrum between lung health and interstitial lung disease (ILD). ILAs have been associated with important clinical outcomes including increased respiratory symptoms such as cough and shortness of breath, decreased lung function, decreased functional capacity, and increased mortality.1-3 ILAs also share genetic4 and environmental5 risk factors with advanced forms of ILD, including pulmonary fibrosis. It is important to note that while ILA and ILD share risk factors and clinical associations, the clinical findings are generally less severe among those with ILA compared to those with ILD. We also know that not everyone with ILA will progress to ILD.6 To identify those at highest risk for progression to ILD and adverse outcomes, the definition of ILA has changed over time. The initial definition was intentionally broad, including a range of interstitial features from inflammatory (eg, centrilobular nodules) to advanced fibrotic changes such as traction bronchiectasis (TBe) and honeycombing.7 As data evolved demonstrating which imaging patterns and features were most likely to progress,8 the Fleischner Society refined this definition in 2020 to exclude more inflammatory features, with the goal of identifying the subset of those with ILA at highest risk of progression.9 With this change, the distinction between ILA and ILD at times felt ambiguous, particularly when advanced fibrotic changes were present. In such cases, classification as an abnormality or disease depended on the context in which the changes were discovered, whether incidental or part of a clinical evaluation. To address this ambiguity, the recent American Thoracic Society (ATS) clinical statement further refined the ILA definition and, importantly, provided a framework for differentiating ILA from ILD using a combination of imaging findings, patient symptoms, physiology, and pathology (if available).10 From an imaging perspective, among those with ILA, ILD was defined as a visual estimate of greater than 5% of the total lung volume being occupied by advanced fibrosis (honeycombing and/or traction bronchiectasis with reticulation), progressive fibrosis on serial imaging, or the presence of any major fibrotic ILD pattern (eg, usual interstitial pneumonia) on chest CT.10 Though traction bronchiolectasis has been studied as part of a continuum of fibrotic disease, it was not included in the advanced fibrotic features definition due to lack of convincing evidence for its prognostic significance.11 In this issue of the American Journal of Respiratory and Critical Care Medicine, Dr. Cheng and colleagues evaluate the prognostic significance of the extent of traction bronchiolectasis (TBe) and its ability to identify early fibrotic ILD.12 Using the presence and extent of TBe to define evidence of fibrosis, the group generated a classification criterion with 4 categories: no ILA, nonfibrotic ILA (NF-ILA), fibrotic ILA (F-ILA), and undiagnosed fibrotic ILD (U-ILD). These criteria and those from the recent ATS clinical statement were then applied in two lung cancer screening cohorts, and concordance between the two methods and clinical outcomes were assessed across categories. As expected, including TBe in the definition of fibrosis led to a higher number of participants with U-ILD—344 participants (4%) compared to 185 (2.6%) using the ATS classification. Using the TBe classification, the authors found that compared to participants without ILA, those with U-ILD had more respiratory hospitalizations and an increased risk of death. Including TBe upweighted the severity of the ILA classification compared to the ATS definition and still identified a subset at higher risk for adverse outcomes. It is important to note that the prevalence of ILA identified here, including U-ILD, was 17%, which is higher than what has been reported in most studies,2 underscoring the need for replication to ensure findings are not specific to these populations. As imaging definitions of ILD continue to evolve, it is important to consider that a single line between abnormality and disease may not exist. However, with clinical outcomes consistently worsening with increasing fibrotic changes, quantitative imaging techniques could be an attractive possibility to help differentiate ILA from ILD. Several quantitative imaging techniques have been used in ILA and ILD populations. These have ranged from density-based measures, assessing the volume of high-attenuation areas,13 to methods based on regions of interest combined with machine learning,14,15 and artificial intelligence–based methods that quantify fibrosis,16 among others. Despite differences in technique, all have shown that increasing amounts or volume of measured interstitial changes are associated with adverse outcomes. As studies have shown that imaging progression may be better than forced vital capacity at predicting certain clinical outcomes, including transplant-free survival, in pulmonary fibrosis,16 quantitative measures have been used as endpoints in clinical trials. For example, quantitative lung fibrosis was used in the INBUILD trial evaluating nintedanib in progressive pulmonary fibrosis and e-Lung is now being used as a co-primary endpoint in the DROP-FPF trial, which is a phase 3 trial of nerandomilast in patients with ILA and family history of pulmonary fibrosis. Quantitative imaging measures offer the possibility of continuous measures or thresholds of severity that could more easily adapt as research progresses and would allow for thresholds to vary by disease type or imaging pattern. Dr. Cheng and colleagues present a novel method, utilizing the presence and extent of traction bronchiolectasis to help classify ILA severity, identify ILD among those with ILA, and associate findings with important clinical outcomes. The ability of both TBe and the classification system from the ATS clinical statement on ILA to identify participants at risk for adverse outcomes demonstrates that there is more than one way to define ILD on imaging. This study and others highlight the need for ongoing research to help refine the definition of ILA and determine how to best integrate imaging findings in the care of patients with ILA and ILD. Supplementary material is available at American Journal of Respiratory and Critical Care Medicine online. Please see the ICMJE disclosure forms, which have been provided as supplementary material.