Abstract
Introduction Partitioning parameters measured from exhaled nitric oxide, such as the alveolar concentration of nitric oxide (CalvNO), may provide better predictors of future asthma exacerbation than exhaled nitric oxide fraction at an expiratory flow rate of 50 mL·s−1 (FENO50). We aimed to determine whether any partitioned nitric oxide parameters were more closely associated than FENO50 with subsequent asthma exacerbations.
Methods 68 asthmatic children (mean±sd age 9.0±2.4 years) were followed prospectively (134 visits) and exacerbations were recorded. Childhood Asthma Control Test (cACT), spirometry, FENO50, CalvNO, bronchial flux of nitric oxide (JawNO), transfer factor of nitric oxide (DawNO) and airway wall concentration of nitric oxide (CawNO) were measured.
Results No exacerbation was recorded in 99 visits (Group 1) and an exacerbation was recorded in 35 visits (Group 2). The median (range) FENO50, JawNO, CalvNO, DawNO and CawNO of Group 1 versus Group 2: 12.7 (4–209) versus 13.5 (3.8–149.9) ppb, 715 (10–12 799) versus 438 (40–7457) pL·s−1, 3.4 (0.2–10.8) versus 5.2 (1.7–23.6) ppb, 38.3 (0.2–113.3) versus 38 (1.3–144.5) pL·s−1·ppb−1 and 26.8 (4.1–2163) versus 29.9 (5.5–3054) ppb, respectively. Other than for CalvNO (p<0.001), there was no difference between the two groups. CalvNO >7 ppb predicted asthma exacerbation with specificity 90.9% and positive likelihood ratio (LR) 3.1. Conversely, CalvNO <4 ppb excluded an exacerbation with sensitivity 71.4% and negative LR 0.48. An increase of CalvNO by 0.5 ppb between visits could also predict an exacerbation with sensitivity 92%, specificity 92%, positive LR 11.8 and negative LR 0.08.
Conclusions Assessment of CalvNO improved prediction of subsequent exacerbation, highlighting the importance of distal inflammation in asthma outcomes in children.
Abstract
Partitioning exhaled nitric oxide allows improved prediction of risk of an asthma attack in the subsequent 4 months. CalvNO>7 ppb was highly specific for a subsequent exacerbation, while CalvNO <4 ppb excluded risk of an attack with high specificity. https://bit.ly/3zWZWYp
Introduction
The function of the distal airway generations has in the past been difficult to assess and, in particular, spirometry is insensitive to small airway disease. Novel physiological tools such as the forced oscillation technique and multiple breath washout have demonstrated that distal disease makes an important contribution to asthma severity. Distal obstruction may result from many different factors, including remodelling, inflammation, airway instability due to loss of alveolar tethering attachments and mucus plugging. Distal airway disease is known to independently contribute to the severity of airway hyperresponsiveness in asthma [1], but unsurprisingly there is a poor correlation between distal inflammation measured directly and physiological parameters [2–4]. Distal inflammation can be studied directly using transbronchial biopsy and has been implicated in the severity of asthma [5], but this invasive technique is not suitable for routine monitoring and especially not in children.
Measurement of exhaled nitric oxide fraction at an expiratory flow rate of 50 mL·s−1 (FENO50) has long been used as a marker of (mainly eosinophilic) inflammation, nitrosative stress and altered nitrogen redox physiology of the airways in asthma [6]. Measurement of FENO at multiple flow rates allows the contributions of distal airways (alveolar concentration of nitric oxide (CalvNO)) and proximal airways (bronchial flux of nitric oxide (JawNO)) to be determined [7]. Cohen et al. [8] demonstrated that a fine-particle inhaled corticosteroid (ICS), ciclesonide, which would be expected to be deposited distally, improved CalvNO and reduced gas trapping on computed tomography scanning. A study of apparently steroid-refractory asthma, also in adults, demonstrated the presence of untreated distal inflammation and that asthma control improved after treatment with ciclesonide [9]. Thus, there is evidence that distal airways inflammation may be an independent contributor to poor adult asthma outcomes.
Measurement of future risk is an increasingly important part of the assessment of asthma, with the realisation that good control does not exclude the possibility of a high risk of subsequent asthma exacerbations. For example, sputum eosinophilia and persistent elevation of FENO50 are markers of risk of asthma exacerbations in apparently well-controlled asthmatic subjects [10, 11]. We hypothesised that measurement of distal airway inflammation (CalvNO) using variable flow measurements of FENO would be a better marker of future risk than FENO50 or JawNO, which measure more proximal inflammation. We recruited asthmatic children to a prospective follow-up study to determine which nitric oxide measurements were most closely associated with subsequent asthma exacerbations.
Methods
Subjects
We recruited 68 asthmatic children (mean±sd age 9.0±2.4 years; 45 males); all also had allergic rhinitis and were sensitised to aeroallergens (table 1). None was a self-reported current or ex-smoker. Inclusion criteria were: 1) clinical diagnosis of asthma (a history of at least two of cough, shortness of breath, recurrent wheeze and chest tightness) and 15% increase in forced expiratory volume in 1 s (FEV1) after administration of 400 µg short-acting β2-agonist; 2) exclusion of other diseases mimicking asthma; and 3) age >6 years to be able to perform FENO measurements at incremental flows. All subjects were recruited from the asthma outpatient clinic of the Paediatric Respiratory Unit of the University Hospital of Alexandroupolis (Alexandroupolis, Greece). The study was approved by the Ethics Committee of the University Hospital of Alexandroupolis. Consent was obtained from parents and age-appropriate assent was obtained from child participants.
Study design
The study design is summarised in figure 1. A detailed medical history including baseline medication and asthma control, physical examination, and specific IgE tests (radioallergosorbent tests) for 10 common aeroallergens were recorded. All subjects were stratified according Global Initiative for Asthma guidelines to controlled, partially controlled and uncontrolled at enrolment (visit 1) and at visits 2 (4 months after visit 1) and 3 (4 months after visit 2). We recorded Childhood Asthma Control Test (cACT), spirometry pre- and post-bronchodilator administration, and FENO50, and calculation of JawNO and CalvNO was performed; the transfer factor of nitric oxide (DawNO) and the airway wall concentration of nitric oxide (CawNO) were also computed. Any moderate or severe exacerbation in the previous 4 months was recorded in all participants at visits 2 and 3. Moderate exacerbation was defined as including one or more of deterioration in symptoms, deterioration in lung function and increased rescue bronchodilator use, lasting at least 2 days, but not severe enough to warrant systemic corticosteroid prescription and/or hospitalisation. Severe exacerbation was defined as requiring any of high-dose oral corticosteroids for at least 3 days, increase in maintenance oral corticosteroid dose, emergency department visit or hospitalisation [12].
Pulmonary function testing
A dry rolling seal spirometer was used for pulmonary function testing (Spirodoc; MIR, Rome, Italy) based on European Respiratory Society/American Thoracic Society (ERS/ATS) criteria [13].
Nitric oxide testing
Exhaled nitric oxide was measured prior to spirometry using an analyser (CLD 88 sp; Eco Medics, Dürnten, Switzerland) [14] and according to ERS/ATS guidelines [15]. FENO was initially measured at a flow rate of 50 mL·s−1 (FENO50), followed by three measurements at 30, 100 and 300 mL·s−1; the latter were used to calculate JawNO, CalvNO, DawNO and CawNO (Högman–Meriläinen algorithm) [14]. To ensure the highest possible success rate, a three-step approach was adopted: 1) a specialised and experienced nurse demonstrated the procedure; 2) two test measurements (at flow rates of 50 and 300 mL·s−1) were performed to familiarise with the technique (lower and higher flows) and the device; and 3) the child performed the measurement at incremental flows. A single trial was performed per each flow rate; the trial was repeated only in the case of technical or quality issues (according to device quality control algorithms and to investigators’ experience). The duration of the whole procedure (FENO50 and FENO at incremental flow rates) was 10–15 min.
Statistics
We have previously reported a 40% higher CalvNO in children with poorly controlled asthma compared with those with controlled disease [16]. Assuming 25% of asthmatic children would experience an exacerbation, 65 children examined twice (i.e. 130 visits) would be required to obtain a similar CalvNO difference at a 5% α-level with 85% power. Sample size estimation was performed using G*Power software [17].
Continuous variables were compared with the t-test and categorical variables were compared with the Chi-squared test. Linear mixed modelling (LMM) with adjustment for repeated observations (i.e. study visits) was used to compare FENO50, JawNO, CalvNO, DawNO and CawNO (log-transformed values) between visits followed by an exacerbation and those not followed by an exacerbation. LMM was also applied to explore differences of log-transformed bronchial inflammation parameters in relation to the severity of exacerbations. Repeated measures (mixed effects) logistic regression was used to explore predictors of an asthma exacerbation. Spearman's correlation and Cox survival analysis was used to explore the relationship between FENO50, JawNO, CalvNO, DawNO and CawNO z-score values and time to asthma exacerbation. Receiver operating characteristic curve analysis was applied to calculate the overall predictive ability of bronchial inflammation markers; sensitivity, specificity, and positive and negative likelihood ratio (LR) were calculated for different CalvNO levels. The lower CalvNO value with positive LR >3 was considered as the high-risk cut-off. All analyses were performed using SPSS version 25.0 (IBM, Armonk, NY, USA). p<0.05 was considered significant.
Results
The characteristics of the study population are presented in table 1. 69 children were enrolled in the study (visit 1), of whom 21 reported at least one exacerbation prior to the start of the study; at visit 2, 19 out of 68 children (one child was lost to follow-up) reported that they had suffered an exacerbation, while 16 out of 66 children (two further children were lost to follow-up) reported an exacerbation at visit 3 (figure 1). In total, no exacerbation was reported in 99 visits (Group 1) and an exacerbation was reported in 35 visits (Group 2); 10 (28.6%) of the exacerbations were severe. There were no differences at baseline in age, height, weight, treatment for allergic rhinitis, asthma control and cACT between the two groups (table 2). Participants in the exacerbation group were more frequently treated with ICSs or leukotriene receptor antagonists (LTRAs). There were also no differences in spirometric parameters between the two groups, including post-bronchodilation reversibility of FEV1 (table 2).
FENO measurement at incremental flows was successful at the first attempt in 117 of the 134 visits (87.3%). A repeated measurement was required in 17 visits, mainly (15 out of 17 (88.2%)) due to expiratory flow instability at 300 mL·s−1 that resulted in poor equation fitting [7] and even negative CalvNO values. In the participants who did not exacerbate (Group 1), the median (range) values were: FENO50 12.7 (4–209) ppb, JawNO 715 (10–12 799) pL·s−1, DawNO 38.3 (0.2–113.3) pL·s−1·ppb−1, CawNO 26.8 (4.1–2163) ppb and CalvNO 3.4 (0.2–10.8) ppb (figure 2). Compared with Group 1, those who experienced an exacerbation (Group 2) had significantly higher median (range) CalvNO (5.2 (1.7–23.6) ppb; p<0.001), but similar FENO50 (13.5 (3.8–149.9) ppb; p=0.744), JawNO (438 (40–7457) pL·s−1; p=0.708), DawNO (38 (1.3–144.5) pL·s−1·ppb−1; p=0.431) and CawNO (29.9 (5.5–3054) ppb; p=0.399) (figure 2) on the visit preceding the exacerbation. Participants who experienced severe exacerbations had significantly higher CalvNO, FENO50 and CawNO on the preceding visit compared with controls (figure 3).
Predictors and time to exacerbations
Repeated measures logistic regression revealed that CalvNO was the only bronchial inflammation parameter that was associated with increased risk of asthma exacerbation within the following 4 months (table 3). The effect of CalvNO was independent of the spirometric parameters (including reversibility to bronchodilation), FENO50 levels, gender and use of controller therapy (ICSs and LTRAs) (table 3). With the exception of DawNO, all other bronchial inflammation indices were negatively correlated with time to asthma exacerbation. The strongest correlation was observed for CalvNO and the weakest for JawNO. Cox survival analysis corroborated these results (figure 4).
Predictive ability of bronchial inflammation parameters
The area under the curve (AUC) values of FENO50 0.507 (95% CI 0.390–0.623), JawNO 0.516 (95% CI 0.407–0.641), DawNO 0.521 (95% CI 0.420–0.652) and CawNO 0.565 (0.448–0.681) reflect the low ability of the these parameters to identify children at risk. The AUC of CalvNO was 0.690 (95% CI 0.585–0.794), indicating moderate overall ability to differentiate children at risk for asthma exacerbation. The predictive characteristics of different CalvNO levels are shown in figure 5. CalvNO >7 ppb predicted asthma exacerbation with high accuracy (specificity 90.9%, positive LR 3.1), while CalvNO>10 ppb had specificity 99% and positive LR 19.8, but sensitivity 20% and negative LR 0.8. Conversely, CalvNO <4 ppb, for example, had lower ability to exclude an exacerbation within the next 4 months (sensitivity 71.4% and negative LR 0.48).
Children with CalvNO >7 ppb had a high probability of experiencing an exacerbation within the next 4 months (table 4); they also had more exacerbations (50.0% versus 21.9%; p<0.001), more severe exacerbations (90.0% versus 0.9%; p<0.001) and shorter time to exacerbation (4.2±2.9 versus 12.5±2.1 weeks; p<0.001). The corresponding differences in cACT and FEV1 were not significant, but participants with CalvNO >7 ppb presented lower forced mid-expiratory flows (table 4).
The predictive characteristics of a change in CalvNO between visits were assessed in children with an exacerbation after visit 2 but without an exacerbation between visits 1 and 2, since treatment with systemic corticosteroids could have affected nitric oxide measurements. The median (range) CalvNO change was −0.1 (−3.6–5.8) ppb or −5.4% (−56.3–150%) in those who did not exacerbate versus 1.2 (0.2–4) ppb or 60% (3.6–480%) in those who experienced an asthma exacerbation (p<0.001) (figure 6). CalvNO increase was a better risk predictor of future exacerbation (AUC 0.939 (95% CI 0.832–0.987)) than a single CalvNO measurement (AUC 0.650 (95% CI 0.507–0.776); p<0.001). An increase of CalvNO of 0.5 ppb from visit 1 to visit 2 had sensitivity 92%, specificity 92%, positive LR 11.8 and negative LR 0.08 for the identification of an exacerbation within the next 4 months. Similarly, the AUC for the relative CalvNO change between the two visits was also informative (0.931 (95% CI 0.817–0.984)); a 10% increase of CalvNO from the previous visit had sensitivity 92.3%, specificity 88.2%, positive LR 7.9 and negative LR 0.09 in predicting a future exacerbation. Given the small numbers, we did not subdivide further to explore whether the predictive power was greater within 1 month of the measurement.
Discussion
In this prospective study which recruited children with a wide range of asthma severities, we have shown that partitioning exhaled nitric oxide allows improved prediction of risk of an asthma exacerbation in the subsequent 4 months. Specifically, a marker of distal inflammation, CalvNO, was the best predictor of risk of all the parameters measured. We found decreased forced mid-flows in children with high CalvNO levels (i.e. >7 ppb), which seems to support the value of CalvNO as a marker of small airway disfunction in asthma [7]. More important, CalvNO >7 ppb was highly specific but not very sensitive for a subsequent exacerbation and CalvNO <4 ppb excluded risk of an attack also with high specificity but low sensitivity. Even more sensitive, in a small subgroup, was an increase in CalvNO of >0.5 ppb between visits 4 months apart. As expected, neither symptoms (cACT) nor FEV1 or airway reversibility tests were useful in the assessment of risk.
This is the first study prospectively evaluating CalvNO as a marker of risk in children. The strengths of the study include its prospective design and the fact that children were assessed with varying levels of asthma severity, although all participants had well-controlled asthma at recruitment. The use of the improved Eco Medics device for measuring exhaled nitric oxide enabled us to measure CalvNO and JawNO in all our subjects, unlike our previous experience [16]. However, inevitably there are some weaknesses. There were some potential markers of risk that we did not include; in particular, well-known risk factors for an asthma exacerbation include a previous severe exacerbation, over-use of short-acting β2-agonists and under-use of ICSs [18]. We did not have any objective assessments of inhaler use. We also did not measure other markers of risk such as induced sputum and peripheral blood eosinophil count, and we did not attempt to see if fine-particle ICSs reduced the risk of exacerbations in those with high CalvNO. We had no data on lung diffusion capacity, and we could not assess the effect of other factors such as cardiac output, haemoglobin concentration and airway lining fluid pH on CalvNO levels [6]. Finally, our findings ideally need to be validated in a second cohort.
This study has mechanistic and clinical implications for children with asthma. The risk of an asthma exacerbation associated with small airways inflammation, as measured by CalvNO, underscores the importance of distal as well as proximal airways disease in the pathophysiology of asthma. As a clinical test, partitioning nitric oxide is likely only to be useful in specialist settings. The equipment is expensive and the test is time consuming. There is also considerable overlap between children who did and did not relapse. However, those children who have a really high CalvNO in the present data form a subgroup who are at high risk and need a focused reassessment of risk factors. Those children with an increase in CalvNO between visits also merit this approach, although the usefulness of a change in CalvNO is limited by the need to make measurements at two time-points. Conversely, those with a very low CalvNO and no change over time would not be expected to have exacerbations, and if exacerbations are reported, the paediatrician might consider whether symptoms are being over-reported [19].
Ultimately, whether partitioning nitric oxide has clinical value depends on whether taking action on the results improves outcomes; there is little point in making measurements clinically if no useful action results. The obvious next study is to determine whether fine-particle ICSs improve outcomes when added to the treatment regime of those suffering acute exacerbations and have an elevated or increasing CalvNO.
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-01684-2021.Shareable
Footnotes
This article has an editorial commentary: https://doi.org/10.1183/13993003.00802-2022
Conflict of interest: None declared.
- Received June 14, 2021.
- Accepted January 5, 2022.
- Copyright ©The authors 2022. For reproduction rights and permissions contact permissions{at}ersnet.org