Cigarette smokers with asthma have a distinct clinical phenotype from non-smokers with asthma. This may reflect altered airway inflammation although how cigarette smoking directs this is unclear. We employed exploratory factor analysis to examine the impact of smoking on airway inflammation.Methods
22 smokers (sm), 10 ex-smokers (ex-sm) and 21 never smokers (ns) with asthma performed spirometry, induced sputum and completed asthma control questionnaires pre and post an oral corticosteroid trial. Sputum fluid cytokines were quantified using a 25-plex bead system (Invitrogen, Paisley). Factor analysis was performed (SPSS V.17) using principal component analysis and Varimax rotation. Factors were identified according to; visual inspection of the scree plot, eigenvalues >1.1, minimum of three cytokines loading >0.4. Sequential removal of cytokines was performed in stages according to; moderate to strong (>0.4) loadings, followed by requirement for ‘strong’ loading (>0.6) to only one factor then removal of cytokines that reduced the reliability of the data set. In a final step cytokines with the lowest loadings were removed if a factor had >3.Results
The subjects were well matched except for higher asthma control questionnaires scores and inhaled corticosteroid dose in sm. Sm failed to demonstrate a lung function response to oral corticosteroids in contrast to nsm. No sputum cell differential differences were evident between smokers and non-smokers with asthma. A number of pre-steroid sputum cytokines were elevated in sm compared to nsm. The greatest difference present for interleukin 6 (sm 34.4 pg/ml (IQR 14.1, 72.4), nsm 8.1 pg/ml (4.4, 11.1), p<0.001). Factor analysis of the pre-steroid cytokines demonstrated that three factors explained 90% of the variance in the data. Sequential processing revealed three cytokines per factor (Abstract S122 table 1).Discussion
Sputum cytokine profiling of subjects with asthma with differing smoking histories reveals distinct groupings when examined by exploratory factor analysis providing insight into airway inflammation in asthma and the impact of smoking. Larger cohorts of patients with asthma should be examined to confirm these preliminary findings.