To the Editors:
We read with great interest the recent article by Michils et al. 1 demonstrating that sequential changes in exhaled NO fraction (FeNO) are associated with asthma control, even in smokers who present lower levels of FeNO compared to nonsmokers. Besides tobacco smoking, we have recently shown that a significant confounding factor in the diagnosis of asthma using FeNO is the presence of allergic rhinitis, since patients with allergic rhinitis have been found to have elevated FeNO levels [2]. This effect may well be attributed to the underlying atopic status of those patients, since elevated FeNO has been shown to be associated with a phenotype characterised by atopy and increased airway responsiveness [3]. In the present study we attempted to evaluate whether the coexistence of allergic rhinitis may affect the validity of sequential changes of FeNO in predicting control in asthmatic patients.
We performed a post-hoc analysis of a prospectively collected database of all consecutive patients evaluated in the Asthma Clinic of the Respiratory Medicine Dept, University of Thessaly Medical School (Larissa, Greece), between June 2007 and February 2009. Of the 259 patients in the database we excluded 82 smokers, 26 patients with difficult-to-treat asthma in whom total control was never achieved during that time period, and 41 patients who were not seen at least twice. Patients were treated according to Global Initiative for Asthma (GINA) guidelines and were all receiving inhaled corticosteroids. All patients were submitted to FeNO measurements using a portable analyser (NiOX MiNO; Aerocrine, Solna, Sweden) as previously described [2] and to spirometry with a dry spirometer (KoKo Legend; Ferraris Respiratory Ltd, Hertford, UK). The assessment of asthma control was based on the evaluation of an experienced physician (K. Kostikas) according to the GINA guidelines and the use of the original Asthma Control Questionnaire (ACQ) [4]. The diagnosis of allergic rhinitis was based on history and appropriate assessment of atopic status, as previously described [5].
Comparisons among groups at baseline were performed with one-way ANOVA with Bonferroni post-hoc tests. Comparisons of FeNO levels between the two visits were performed using Wilcoxon signed rank tests. For the assessment of the performance of FeNO percentage changes compared to baseline (ΔFeNO%) in the identification of loss of control, receiver operating characteristics (ROC) curves were created by plotting sensitivity against 1-specificity. The area under the ROC curve (AUC) with 95% confidence intervals (CI) and its difference from 0.5 were calculated. Additionally, sensitivities, specificities, positive and negative predictive values (PPV and NPV, respectively) were calculated for specific cut-off points. Statistical analysis was performed with GraphPad Prism 5 (GraphPad Software Inc, La Jolla, CA, USA) and MedCalc 9 (MedCalc Software, Mariakerke, Belgium).
110 patients with two consecutive visits, of whom in the first they had well-controlled asthma, were included in the analysis. Of those patients, 61 lost control in the subsequent visit (32 with rhinitis and 29 without rhinitis) and 49 remained well-controlled (26 with rhinitis and 23 without rhinitis). Demographic characteristics are presented in table 1⇓. Study groups did not differ in demographic and spirometric characteristics, FeNO levels or ACQ scores at baseline. In both groups where control was lost in visit 2 (with and without rhinitis), FeNO increased significantly (p<0.001 for both comparisons). In contrast, patients who maintained control at visit 2 did not present significant differences in FeNO levels between visits (p = 0.100 and p = 0.146 for patients with and without allergic rhinitis, respectively).
In ROC analysis, ΔFeNO% was associated with loss of control, since both ROC curves differ significantly from 0.5 (fig. 1⇓, table 2⇓). The diagnostic performance of ΔFeNO% did not differ significantly between patients without and with allergic rhinitis, despite a trend in favour of the former (AUC 0.893 versus 0.786; p = 0.152). In patients without rhinitis, an increase in FeNO >30% from baseline was highly indicative of loss in asthma control (PPV 0.89), whereas an increase <20% was unlikely to be related to loss of control (NPV 0.81). Additionally, an increase in FeNO >40% in patients with rhinitis was also highly suggestive of loss in asthma control (PPV 0.92). In contrast, an increase in FeNO <20% was not as strong a predictor of maintenance of control (NPV 0.72) in patients with asthma and rhinitis as in those without rhinitis.
Our data support our previous observations suggesting that the presence of atopy, and especially allergic rhinitis, may impair the diagnostic value of FeNO in asthmatics [2]. The poorer diagnostic performance of FeNO in patients with concomitant rhinitis may plausibly represent cases of deterioration in rhinitis symptoms (related to increased nasal inflammation) without concurrent impairment of asthma control. Interestingly, our results independently confirm the data of Michils et al. [1] and further support the observation that an increase in FeNO >40% from baseline is highly indicative of loss of asthma control, even in patients with underlying atopy, as expressed by the presence of allergic rhinitis.
Statement of interest
None declared.
- © ERS Journals Ltd