CT imaging has a key role to play in the management of patients with stable chronic obstructive pulmonary disease (COPD), said R. Graham Barr, MD, DrPH, in a presentation at the GOLD COPD International Conference.

Use of CT is proven to save lives in patients eligible for lung cancer screening, which includes most patients with COPD, said Barr, professor of medicine and epidemiology at Columbia University Irving Medical Center, New York City. CT imaging also helps define COPD etiotypes and identifies airway vs emphysema as the predominant phenotype, he said.

The GOLD 2025 guidelines outline the use of CT for patients with COPD; these include use in differential diagnosis, lung volume reduction, and lung cancer screening. Lung cancer screening with CT is especially valuable in COPD patients, Barr said.

Imaging can be used to classify COPD into one of the COPD etiotypes, which were introduced in the 2023 GOLD report. These etiotypes are based on the underlying cause of COPD, rather than symptoms alone, with the goal of guiding more targeted treatment approaches, Barr said. The seven types according to cause are COPD-G (genetic), COPD-C (cigarette or cannabis smoke exposure), COPD-P (indoor or outdoor pollution), COPD-I (infection), COPD-D (developmental), COPD-A (asthma-related), and COPD-U (cause unknown). The etiotypes are not mutually exclusive, Barr noted in his presentation.

CT helps identify airway vs emphysema as the main phenotype in COPD, Barr said. Quantitative CT assessment of emphysema, defined as the percentage of lung voxels with attenuation < -950 Hounsfield units (HU), has been shown to correlate with pathologic emphysema and can serve as a simple guide, he noted. In one study of moderate-to-severe COPD, airway predominant disease was associated with percent emphysema_{-950 HU} less than 5% on CT, whereas, emphysema-predominant COPD was associated with percent emphysema_{-950 HU} of at least 10% on CT, he added.

Airway wall thickness (AWT) measurement on CT is more complicated, said Barr. Although CT scans can show thickened airways, they do not reliably measure the smallest airways where disease is most severe, which makes AWT less reliable in clinical practice. 

In a 2023 study published in Thorax focusing on emphysema, Barr and colleagues examined the ability of unsupervised machine learning to measure COPD phenotypes based on CT scans from approximately 3000 individuals. The machine learning algorithm identified six reproducible CT emphysema subtypes, notably combined bronchitic-apical emphysema and diffuse emphysema, which supports the potential value of CT for more personalized COPD care, Barr said.

Why Etiotypes Matter, and Imaging’s Impact

COPD has long been considered primarily as a consequence of tobacco smoking, but the GOLD framework broadens this perspective by recognizing multiple etiotypes, said Stephen R. Dachert, MD, assistant professor of clinical thoracic medicine at the Lewis Katz School of Medicine at Temple University, Philadelphia. “Identifying these etiotypes underscores that COPD results from complex gene/environment interactions rather than a single exposure,” Dachert told Medscape Medical News. “Lung injury can arise through diverse biological pathways, and defining the causal mechanism in a given patient creates opportunities for more targeted prevention strategies and etiotype-specific therapeutic development,” he said.

Patterns of distribution of emphysema on CT are also clinically meaningful, said Dachert. “For example, upper lobe-predominant emphysema is linked to more rapid lung function decline and structural progression,” he said. “Imaging therefore provides objective characterization of emphysema severity and distribution, which is critical when evaluating candidacy for interventions such as bronchoscopic or surgical lung volume reduction,” he noted.

Several challenges limit routine use of imaging-based phenotyping, Dachert said. “Quantitative CT analysis is not universally available in clinical practice, and there is variability across scanners, acquisition protocols, and analysis software that can affect reproducibility,” he said. “Cost and access remain additional constraints; however, continued advances in artificial intelligence, image processing, and standardization efforts are likely to improve scalability and consistency in the near future,” he added.

“Each of the proposed COPD etiotypes (COPD-G, COPD-D, COPD-C, COPD-P, COPD-I, COPD-A, COPD-U) requires focused investigation to clarify underlying mechanisms and to determine whether tailored prevention or treatment approaches improve outcomes,” Dachert told Medscape Medical News. More research is needed to define specific risk factors, establish reliable diagnostic criteria for etiotype classification, and evaluate differential treatment responses, he said. Identifying and validating etiotype-specific and prognostically meaningful biomarkers would help guide individualized care, Dachert emphasized.

Barr disclosed receiving funding from the National Institutes of Health, the COPD Foundation, and the American Lung Association. 

Dachert disclosed having no financial conflicts of interest.