Lung-volume controlled computerised tomography in real-life acute severe asthma

Abstract

Objective: It is not known how airway structure is altered during real-life acute asthma exacerbations. The aim of this study was to examine changes in airway structure during acute asthma exacerbations and at convalescence by using lung-volume controlled high resolution computerised tomography (HRCT).

Methods: Eight subjects with acute asthma exacerbation admitted to hospital were recruited. HRCT was performed within 72 h of admission (n = 8) and repeated after 8 weeks of convalescence (n = 7). Individual airways were carefully matched on acute and convalescent CT data sets for comparisons of airway parameters. A novel methodology was employed for standardisation of lung volumes to permit valid comparisons of lung imaging. Measurements of bronchial cross sectional airway area (Aa) and bronchial luminal area (Ai) for each matched airway were obtained using a validated program.

Results: The airway wall thickness was analysed as wall area (WA) calculated as a percentage: WA% = WA/Aa × 100. Wilcoxon signed-rank testing was used to compare acute and convalescent asthma and Spearman's correlation to examine associations. Airway lumen (Ai) areas were similar in both acute and stable asthma phases (6.6 ± 3.1 mm(2) versus 7.2 ± 3.8 mm(2) p = 0.8). However, the airway wall was significantly thickened during acute asthma exacerbations compared to convalescence (62 ± 4% versus 55 ± 7%; p = 0.01). There was no correlation between airway structure dimensions and lung function measurements.

Conclusions: This is the first study to demonstrate an increase in airway wall thickness during real-life acute asthma exacerbation. However, narrowing of the airway lumen area was variable and will require larger studies able to detect small differences. These results suggest that airway wall thickening linked to mucosal inflammation is likely to characterise acute asthma in vivo but that changes in the airway lumen accompanying bronchoconstriction may be more heterogeneous.