- Top of page
- Material and methods
Background: Allergic asthma is consistently associated with increased FENO levels whereas divergence exists regarding the use of exhaled nitric oxide (NO) as marker of inflammation in nonallergic asthma and in asthmatic smokers. The aim of this study is to analyze the effect of having allergic or nonallergic asthma on exhaled nitric oxide levels, with special regard to smoking history.
Methods: Exhaled NO measurements were performed in 695 subjects from Turin (Italy), Gothenburg and Uppsala (both Sweden). Current asthma was defined as self-reported physician-diagnosed asthma with at least one asthma symptom or attack recorded during the last year. Allergic status was defined by using measurements of specific immunoglobulin E (IgE). Smoking history was questionnaire-assessed.
Results: Allergic asthma was associated with 91 (60, 128) % [mean (95% CI)] increase of FENO while no significant association was found for nonallergic asthma [6 (–17, 35) %] in univariate analysis, when compared to nonatopic healthy subjects. In a multivariate analysis for never-smokers, subjects with allergic asthma had 77 (27, 145) % higher FENO levels than atopic healthy subjects while subjects with nonallergic asthma had 97 (46, 166) % higher FENO levels than nonatopic healthy subjects. No significant asthma-related FENO increases were noted for ex- and current smokers in multivariate analysis.
Conclusions: Both allergic and nonallergic asthma are related to increased FENO levels, but only in never-smoking subjects. The limited value of FENO to detect subjects with asthma among ex- and current smokers suggests the predominance of a noneosinophilic inflammatory phenotype of asthma among ever-smokers.
There has been a continually increased interest for the use of exhaled breath as noninvasive assessment of airways inflammation in asthma. Exhaled nitric oxide (NO) is the most promising exhaled breath marker and has gone a long way from the observational study reporting increased exhaled NO levels in asthmatics (1) to current investigations regarding a potential role in titrating steroids in asthma management, predicting asthma exacerbations and outcome after steroid withdrawal in stable asthma (2).
Allergic asthma is associated with increased exhaled NO levels (3–5) whereas the effect of nonallergic asthma on the levels of fraction of NO in the exhaled air (FENO) is unclear with small-sized studies (4, 6) reporting similar FENO levels in nonallergic asthma and healthy nonatopic controls while larger studies are contradictory with two studies reporting higher levels (7, 8) and one study reporting similar FENO levels in nonallergic asthmatics and healthy controls (9).
Although a certain caution is advocated regarding the use of FENO measurements in smokers (2) due to FENO levels decreased with up to 60% in current smokers, the current ATS/ERS recommendations state that the value of FENO as asthma marker is not affected by smoking (10). However this appears to be contradicted by newer studies, where similar FENO levels have been noted in smokers with and without asthma (11, 12).
The aim of the present study was to analyse the effect of having allergic or nonallergic asthma on exhaled nitric oxide levels, with special regard to smoking history.
- Top of page
- Material and methods
The main finding in this study is that both allergic and nonallergic asthma are related to increased FENO levels, but only among never-smoking subjects. The use of FENO measurements in order to identify asthmatic subjects in a population appears to be of limited value in ex-smokers and of no value in the case of current smokers.
The strength of our study resides in the fact that this is the first population-based multicenter study investigating the value of FENO in detecting subjects with asthma when special regard is taken both to the type (allergic vs nonallergic) of asthma and the subjects’ smoking history. The use of different techniques and/or equipment for measuring NO in different study centers represents the main weakness of our study as it lead to different absolute FENO values and therefore the necessity of reporting the results as percentual increases of FENO. However, the extrapolation of our results might prove to be easier when using the percentual change of FENO. Similar effects in all study centers of known determinants of FENO were observed by meta-analysis, suggesting valid data. The heterogeneity regarding the effects of age and previous smoking is in line with results from the two big available population-based studies looking at determinants of FENO levels (12, 23).
A higher exhalation flow rate was used in Turin than in the other centers. A higher exhalation flow rate leads to a more peripheral NO sampling, but the contribution from central airways to measured exhaled NO is still predominant (24). The asthma-related increase of FENO is in a majority of cases related to the increase production of NO in the central airways (24), that will be reflected in exhaled NO measured at flow-rates between 50 and 200 ml/s with online methods and 50–350 ml/s with offline methods (25). Furthermore our main findings remained significant even when excluding Turin from the analyses (data not shown).
Allergic and nonallergic asthma
In our material, both allergic and nonallergic asthma were associated with increased FENO levels, but only in never-smokers. This is in disagreement with some of the previous studies that report similar levels of FENO in nonallergic asthmatic and nonatopic controls (4–6), but in agreement with two other studies (7, 8). Multiple explanations may lie behind the divergent findings of these studies. The FENO increase reported by one of the studies was small (8) so in order to detect a significant difference between nonallergic asthmatics and nonatopic nonasthmatics it is essential to have adequate power. One important aspect regarding nonallergic asthma is the heterogeneity of the inflammation patterns with divergent reports regarding the extent of the eosinophilic inflammation with no differences (26) or lower eosinophil involvement (27) when compared to allergic asthma. Therefore differences between studies regarding the proportion of subjects with eosinophilic pattern of inflammation among the subjects with nonallergic asthma could to some extent explain the different results.
The association of allergic asthma with increased levels of FENO is a finding in line with the literature (3–5). The mechanism behind this increase appears to be the activation of iNOS in the bronchial epithelial cells (28) as iNOS in lung epithelial cells has been reported be the main determinant of the FENO levels in recent studies (29, 30). The link between the eosinophil activation and epithelial iNOS activation was demonstrated in a murine model of allergic asthma (31) and human studies support this by demonstrating good correlations between exhaled nitric oxide and blood, lung tissue and sputum eosinophils (9, 32–34).
Both current (35, 36) and previous smoking (37) are associated with decreased FENO levels. Although the ERS/ATS guidelines (10) state, based on the study by Horvath et al. (38), that exhaled NO is still reflecting the presence of asthma in current smokers, this appears to be contradicted by two more recent studies where no difference was found between smokers with and without asthma (11, 12). The difference in the results between studies may be partially explained by the lack of adjustment of results for atopy in the study by Horvath et al. where all the subjects with asthma included in the study were polysensitized while the controls were probably mostly nonsensitized as no allergic status is reported for the control group.
The results concerning the limited value of FENO as asthma marker in ex-smokers are in accordance with a previous population-based study (12) where no relation of FENO to physician-diagnosed asthma and asthma symptoms was found among ex-smokers, although ex-smokers had similar FENO levels as never-smokers in the respective study.
The main explanation for the lack of association between exhaled NO and asthma in current smokers is probably the dominance of the neutrophilic phenotype of inflammation in asthmatic smokers as increased neutrophilic inflammation and lower eosinophil number in induced sputum have been documented in smokers with asthma (39, 40). This would limit the use of exhaled NO as a marker of asthmatic inflammation in smokers, since exhaled NO is a marker of eosinophilic inflammation (41–43). It is unclear if there is any shift of the inflammatory phenotype of asthma in ex-smokers as studies with long follow-up time after smoking cessation are lacking, in contrast with chronic obstructive pulmonary disease where such studies are available and the results point consistently towards the persistence of the neutrophilic pattern of inflammation after smoking cessation (44). No relation between time since quitting smoking and exhaled NO could be found in our material, suggesting no shift in the inflammation pattern. The only available smoking cessation studies in asthmatics have shown better response to oral steroids at 1 year (45), but not at 8 weeks (46), after smoking cessation when assessing PEF, but no effect on FEV1 or asthma control. Similarly a lower number of neutrophils in induced sputum has been reported both at 8 weeks and 1 year (45, 46) after smoking cessation, but no decrease in neutrophil mediators (46).
In the present study, atopy was associated with similar increases of FENO levels both in never- and ever-smokers. Atopy is characterized by subclinical eosinophilic inflammation (47) and increased exhaled NO levels (48). There is also well studied association between exhaled NO levels and eosinophilic inflammation (2). Taken together this indicates that smoking does not affect the association between atopic eosinophilic inflammation and FENO. The lack of association between asthma and FENO among ever-smokers is therefore more likely explained by a predominant noneosinophilic asthma phenotype in ever-smokers.
In conclusion, both allergic and nonallergic asthma were associated with increased FENO levels in never-smoking subjects, confirming an independent effect of having asthma on increased FENO levels. The limited use of FENO measurements in ex- and current smokers for detecting subjects with asthma may be explained by the dominance of noneosinophilic inflammation in ever-smoking asthmatics.