Qualitative analysis of high-resolution CT scans in severe asthma

High-resolution CT (HRCT) scanning is part of the management of severe asthma, but its application varies between centers. We sought to describe the HRCT scan abnormalities of a large severe asthma cohort and to determine the utility of clinical features to direct the use of HRCT scanning in this group of patients.

Subjects attending our Difficult Asthma Clinic (DAC) between February 2000 and November 2006 (n = 463) were extensively re-characterized and 185 underwent HRCT scan. The HRCT scans were analyzed qualitatively and the interobserver variability was assessed. Using logistic regression we defined clinical parameters that were associated with bronchiectasis (BE) and bronchial wall thickening (BWT) alone or in combination.

HRCT scan abnormalities were present in 80% of subjects and often coexisted with BWT (62%), BE (40%), and emphysema (8%). The interobserver agreement for BE (kappa = 0.76) and BWT (kappa = 0.63) was substantial. DAC patients who underwent HRCT scanning compared with those who did not were older, had longer disease duration, had poorer lung function, were receiving higher doses of corticosteroids, and had increased neutrophilic airway inflammation. The sensitivity and specificity of detecting BE clinically were 74% and 45%, respectively. FEV(1)/FVC ratio emerged as an important predictor for both BE and BWT but had poor discriminatory utility for subjects who did not have airway structural changes (FEV(1)/FVC ratio, >or= 75%; sensitivity, 67%; specificity, 65%).

HRCT scan abnormalities are common in patients with severe asthma. Nonradiologic assessments fail to reliably predict important bronchial wall changes; therefore, CT scan acquisition may be required in all patients with severe asthma.