Asthma clusters from quantitative computed tomographic measurements in the lungs in CBIOPRED cohort

<bold>Background</bold> Asthma is a complex heterogeneous disease characterized by airway inflammation, airway remodeling and airflow obstruction. Computed tomography (CT) has been used to assess airway dimensions and air trapping in asthma. <bold>Objectives</bold> To obtain clusters of asthma from quantitative CT (qCT) measurements from high-resolution chest CT scans. <bold>Methods</bold> 239 patients with non-smoking severe asthma (n=132), smoking/ex-smoking severe asthma (n=49) and mild-moderate asthma (n=58) and non-smoking healthy controls (n=68) underwent chest CT scan after clinical and physiologic characterization. qCT measurements of proximal airway remodeling and lung densitometry and lung volumes were used in a consensus cluster analysis. <bold>Results</bold> Four qCT-based phenotypes were obtained:. Group 3 had middle proximal airway thickening nor any severe air trapping while Group 1 with no proximal airway remodeling showed severe air trapping and reduced lung density. Group 2 showed airway wall thickening and some dilatation but no air trapping while Group 4 with middle airway wall changes had air trapping with reduction in lung density. Group 4 had the greatest degree of airflow obstruction compared to other Groups. Groups 2 and 3 that had near normal lung function and Group 1 and 4 had most smokers/ex-smokers. Sputum eosinophils was highest in Groups 1 and 4. <bold>Conclusion</bold> qCT clustering reveals the different degrees of airway wall remodeling with air trapping and hypodense lungs, with the most severely airflow obstructed with eosinophilia to have all the features of airway remodeling with air trapping. Patients with severe eosinophilic asthma possess the most severe air trapping.