Prediction of asthma exacerbations in children: results of a one-year prospective study

Background Underdiagnosis and low levels of asthma control are frequent occurring problems in patients with asthma. Objective The study aim was to evaluate the ability of non-invasive inflammatory markers in exhaled breath to predict exacerbations of childhood asthma, and to assess the time course of changes in these exhaled markers before, during and after exacerbations. Methods The design was a prospective one-year longitudinal study. Regular two-month visits at the outpatient clinic were performed. Forty children with asthma (aged 6-16 years) participated. The primary outcome measure was the occurrence of an exacerbation. Assessment was made of the presence and severity of pulmonary symptoms, use of medication, and measurements of forced expiratory volume in 1 s using home monitor. The following independent parameters were assessed during outpatient visits: (1) exhaled nitric oxide, (2) inflammatory markers in exhaled breath condensate: acidity, nitrite, hydrogen peroxide, interleukin-1 alpha, -5, -13, interferon-gamma, (3) lung function, (4) asthma control score. Results Thirty-eight of 40 children completed the study. Sixteen children developed exacerbations, of which ten were moderate and six severe. Univariate Cox regression analysis revealed that condensate acidity, interleukin-5 and asthma control score were significant predictors of an asthma exacerbation (P < 0.05). In the multivariate Cox regression analysis, exacerbations were best predicted by the asthma control score and by the level of interleukin-5 in exhaled breath condensate (Wald scores of 7.19 and 4.44, P = 0.007 and P = 0.035 respectively). The predicted survival curve of this multivariate model showed a two times reduced risk on exacerbations in the category of children with the 10% most optimal values of IL-5 and asthma control score. Conclusions and Clinical Relevance Both exhaled breath condensate interleukin-5 level and asthma control score were significant predictors of asthma exacerbations. These findings open up the possibility of assessing the potential of such parameters to titrate asthma treatment in future studies.