In this issue of the European Respiratory Journal, VIGNOLA et al.[1] report an association between high-resolution computed tomography (HRCT) scan abnormalities of the conducting airways and an analysis of markers of the tissue remodelling process. They demonstrate an association between the level of HRCT abnormality in the airways and increased levels of active elastase with overproduction of tissue-inhibitor metalloproteinase-1 over metalloproteinase-9 in the sputum. The study design is cross-sectional in nature and based on small numbers of patients collected into three groups, according to whether they had newly diagnosed and untreated asthma, treated asthma or chronic obstructive pulmonary disease (COPD). The design does not include a control group with normal lung function and lacks longitudinal data collection to ensure that the sputum markers and HRCT abnormalities change in the expected directions with disease progression. After the first reading, the notion that editorial writers have been described as" the sort of people who come down from the hills after the battle is over to shoot the wounded" bubbled up in my thoughts. However, the temptation to take this approach passed with more careful evaluation of their manuscript.

Airflow limitation during forced expiration is a defining feature of both asthma, where it is reversible, and COPD, where an irreversible component remains after bronchodilator therapy. The emphysema phenotype of COPD destroys lung tissue beyond the terminal bronchioles and reduces the recoil force required to drive air out of the lung [2, 3], whereas the airway-obstructive phenotype is produced by remodelling of the tissue in the walls of the smaller bronchi and bronchioles v2 mm in diameter in the adult lung [4]. The question is: how can sputum collected at the mouth and HRCT images that encompass lesions produced by both phenotypes of the remodelling process contribute to our understanding of such a complex process? The Edinburgh group was the first to use computed tomography (CT) to diagnose emphysema in living patients with COPD [5], and improvements in this area of technology currently allow a fairly accurate assessment of the extent and severity of emphysema present in individuals with COPD [6–9]. Therefore, patients with COPD without radiological evidence of emphysema might exclusively have the small airway-obstructive phenotype of airflow limitation. Unfortunately, HRCT is not suited to the direct assessment of the site of obstruction in the small conducting airways because the walls of these airways provide too diminutive a target [10]. However, a Japanese study has shown that patients with COPD can be separated into emphysematous, airwayobstructive and mixed phenotypes, using estimates of CT lung density to determine the severity of emphysema and the