Exhaled Nitric Oxide and Clinical Phenotypes of Childhood Asthma

Bruno Mahut; Séverine Peyrard; Christophe Delclaux

Respiratory Research. 2011;12(65)

Discussion

The main results of our study is the demonstration that single-flow rate exhaled NO (FENO_0.05) is independently associated with two main asthma pathophysiological characteristics, namely airway inflammation and airway tone, but that FENO_0.05 does not help to distinguish a relevant clinical phenotype of childhood asthma in a cross-sectional assessment.

Design Issues

We specifically assessed different components of asthma control definition, such as symptoms and exacerbations because whether severe exacerbation constitutes the ultimate expression of loss of control and/or a more unpredictable even remains controversial.^[12,14] Table 2 further showed that control and exacerbations were two different dimensions of asthma, which could be related to our pediatric population.^[14] Only two levels of control were assessed (controlled versus partially/uncontrolled patients) because the achievement of control is the main clinical issue. Our percentage of controlled children (33%) is in accordance with a French cross-sectional study in childhood asthma.^[26] Since a dose effect of ICS on exhaled NO can be demonstrated,^[2] we used it as an indirect index of airway inflammation, which is an obvious short-cut inasmuch as the explained variance of FE_NO by ICS is only ~3% in our study. Sputum eosinophilia can be regarded as the gold standard measure of inflammation but is not routinely assessed in most centres. The recent study of Schleich and colleagues demonstrates that FE_NO is able to identify a sputum eosinophil count ≥ 3% with reasonable accuracy and thresholds which vary according to dose of ICS.^[27] Finally, we deliberately chose to add two variables in the cluster analyses, namely post-bronchodilator FEF_50%/TLC and bronchodilator response. The former is an index of airway size/lung size.^[28] We hypothesized that such an index, which has been linked to the risk of airway responsiveness,^[29] may influence symptoms. The latter may also be an important pathophysiological characteristic of asthma associated with poor clinical outcomes in childhood asthma.^[30]

Four clusters were identified in our pediatric population. The first two clusters can be considered similar when excluding gender and corresponds to the most prevalent group of asthmatic children in an out-hospital specialized clinic (73%, 123/169) with 75/123 children having partially/uncontrolled asthma. The remaining two clusters (27%, 46/169) were constituted of boys and girls with a more severe (or undertreated) disease (38/46 with partially/uncontrolled asthma, 20/46 with a recent exacerbation). Interestingly, these two clusters only differ by their underlying severity factors, namely increased airway tonus (cluster 3, 6% of the population) and parental tobacco exposure while having small airway/lung size ratio (cluster 4, 21%). The prevalence of cluster 3 is similar to that of difficult-to-treat patients (~5%), and may also constitute an asthma phenotype characterized by increased airway tone and lability.^[14,30] The fourth cluster segregates children exposed to passive smoking that have a more severe disease, which is in line with the results of a French cross-sectional study in 3431 children demonstrating that unacceptable asthma control was associated with passive exposure to parental tobacco smoke.^[26] Overall, the phenotypes that have been identified by the cluster analysis can be a posteriori explained, which further validate the statistical approach to assess exhaled NO usefulness. It has to be emphasized that the degree of asthma control did not clearly differentiate the clusters in our study. Several explanations can be discussed. Firstly, our study deals with childhood asthma, this specific population is often partially controlled because some degree of exercise limitation (or symptoms) is often present and exacerbations are often unpredictable events, mostly related to viral infections.^[31,32] Secondly, normal lung function (under treatment) is the rule in asthmatic children^[15,17] and a « phenotype » of children exhibiting a decline in lung function is almost impossible to isolate.^[17] Thirdly, therapeutic compliance (and inability to use the inhaler properly) in children may be difficult to obtain^[33] that may further explain the presence of mild symptoms.

Usefulness of FENO_0.05 Measurement

We confirm by our first statistical approach that FE_NO is linked to its classical modifiers such as height and atopy. The link with ICS dose is more controversial, but we previously evidenced such a link with a plateau effect of ICS.^[6] We show that exhaled NO and bronchodilator response are linked, a result that was previously obtained by different research groups.^[2,3,9] More interestingly, we demonstrate for the first time that the relationships between exhaled NO and both ICS dose (an indirect marker of airway inflammation) and bronchodilator response (a marker of airway tonus) are independent. Exhaled NO measurement is claimed to be an allergic inflammometer, but it is also a marker of airway smooth muscle tone.^[34]

The recent study of Dweik and colleagues has shown that a high FE_NO "phenotype" (FE_{NO 0.05} > 35 ppb) was characterized by greatest airway reactivity, airflow limitation, hyperinflation, sputum eosinophilia and levels of symptoms.^[9] Consequently, our results (first statistical approach) are in agreement with their data, but we further suggest that FE_NO is not specifically associated with a clinically relevant phenotype in asthmatic children. Haldar and colleagues elegantly demonstrated that eosinophilic inflammation helps to characterize adult asthma phenotypes.^[22] Consequently our results may seem at variance, but exhaled NO and airway eosinophilia could be discordant in childhood.^[14] In summary, to our best knowledge, this is the first study in childhood asthma, showing that exhaled NO does help to describe a clinically useful phenotype despite its ability to describe underlying pathophysiology (inflammation and modified airway smooth muscle function).

Limitations of the Study

Principal among these is the cluster analysis methodology. The use of an algorithm that separates the population into discrete clusters may not be realistic. The limited size of our out-hospital population, the restricted analysis of childhood asthma (that could be a more homogeneous disease per se), our choice of clustering parameters and the loss of clinical material through attrition of the data set may have introduced some bias, but we hypothesized that the clinical relevance of exhaled NO should be "easily" demonstrated. It has to be stated that, in a cross-sectional design mainly assessing asthma control, FENO0.05 was not associated with a specific cluster, which is in accordance with recent trials failing to demonstrate the clinical usefulness of this measure for asthma control.^[35]

Perspectives

It has to be emphasized that, in a cross-sectional design mainly assessing asthma control, FENO_0.05 was not associated with a specific cluster, which is in accordance with recent trials failing to demonstrate the clinical usefulness of this measure for asthma control.^[35] Whether peripheral airway/alveolar NO concentration after correction for axial NO back-diffusion, which is elevated in a subset of asthmatic patients (~25%), could help to identify a specific "phenotype" of asthma warrants further studies.^[4,36,37]

In conclusion, FENO_0.05 is independently linked to two pathophysiological characteristics of childhood asthma (ICS-dependant inflammation and bronchomotor tone) but does not help to identify a clinically relevant phenotype of asthmatic children in a cross-sectional analysis of routinely recorded parameters.

1 2 3 4

Next Section

Competing interests
The authors declare that they have no competing interests.

Authors' contributions
BM carried out the measurements, participated in the design of the study and drafted the manuscript. SP performed the statistical analysis and helped to draft the manuscript. CD2 conceived of the study, participated in its design and helped to draft the manuscript. All authors read and approved the final manuscript.

Cite this: Exhaled Nitric Oxide and Clinical Phenotypes of Childhood Asthma - Medscape - May 20, 2011.

Table 1. Clinical and physiological characteristics of the asthmatic children
Characteristics	N = 169
Sex (male, %)	104 (61%)
age, years	10.5 ± 2.6
height, cm	142 ± 15
weight, kg	38 ± 15
BMI, kg.m⁻²	18.0 ± 3.5
Atopic status (skin prick test)
negative	27 (16%)
1 positive	41 (24%)
> 1 positive	101 (60%)
Tobacco exposure, n (%)
maternal	25 (15%)
paternal	24 (14%)
paternal and maternal	36 (21%)
Clinical events within past 3 months
controlled, n (%)	56 (33%)
partially or uncontrolled	113 (67%)
number of days with symptoms, median [IQ]	4 [0–12]
severe exacerbation	42 (31%)
number of days with systemic steroid, median [IQ]	0 [0–2]
Treatment
beta-agonist on demand, n (%)	82 (48%)
low ICS dose, n (%); mean ± SD dose, μg	45 (27%); 154 ± 51
medium ICS dose, n (%); mean ± SD dose, μg	28 (17%); 357 ± 50
high ICS dose, n (%); mean ± SD dose, μg	14 (8%); 707 ± 154
LABA, n (%)	73 (43%)
Pulmonary function tests	Pre BD	Post BD
sRaw, % predicted	204 ± 45	126 ± 30
FEV₁, % predicted	97 ± 13	107 ± 12
FEV₁/FVC, %	78 ± 8	84 ± 6
FVC, % predicted	105 ± 13	108 ± 12
FEF_75–25%, % predicted	71 ± 19	91 ± 20
FEF_50%, % predicted	71 ± 18	91 ± 19
TLC, % predicted	104 ± 11	103 ± 10
FRC, % predicted	103 ± 17	100 ± 15
RV, % predicted	107 ± 28	93 ± 22
RV/TLC	0.25 ± 0.06	0.22 ± 0.05
FEF_50%/TLC	0.71 ± 0.14	0.80 ± 0.18
FENO_0.05 ppb, median [IQ]	29 [15–48]

BD: denotes bronchodilator
Results are provided as mean ± SD or median [25^th – 75^th percentiles: IQ] or absolute number with percent age (%).

Table 2. Orthogonal varimax rotation results
28 variables	factors
	1	2	3	4	5	6	7	8	9
Clinical
gender								- 0.765
age		−0.879
height		- 0.889
weight		−0.905
BMI		−0.751
early wheezing (< 2 years)					−0.339		0.565		−0.368
atopy (positive SPT)					0.800
parental smoking									0.764
Treatment
LABA			0.840
ICS treatment			0.872
ICS dose			0.898
ICS dose > 200 μg/d			0.814
Clinical events
partially or uncontrolled				0.439			0.626
days with symptoms				0.089			0.777
exacerbation				0.942			0.054
days with oral steroid				0.940			0.085
Baseline PFT values
FEV₁						0.843
FEV1/FVC	0.843
sRaw	−0.609
RV/TLC		0.645				0.454
FEF_50%/TLC	0.834							−0.050
Post bronchodilator PFT values
FEV₁						0.885
FEV1/FVC	0.757
sRaw	−0.391
RV/TLC		0.664
FEF_50%/TLC	0.797							0.286
FEV₁ response to BD %	−0.569				0.397			0.392
Exhaled NO	- 0.061	- 0.132	- 0.119	- 0.059	0.789	- 0.149	0.064	0.030	0.076

PFT: denotes pulmonary function tests
All results for exhaled NO are shown in italic for information.
Factor analysis (PCA in our study) is based on the procedure for obtaining a new set of uncorrelated (orthogonal) variables, usually fewer in number than the original set (9 instead of 28 in our study), that reproduces the co-variability observed among a set or original variables. Then we requested a rotation of the resulting factors which follows completion of the analysis of the data. The most common rotation performed is the Varimax rotation [23]. This tends to produce "simple structure", that is, factors which have very high (that are provided in the table) or very low (provided for exhaled NO) loadings for the original variables and thus simplifies the interpretation of the resulting factors.

Table 3. Clinical and physiological characteristics according to the four clusters
Characteristics	Cluster 1 n = 79	Cluster 2 n = 44	Cluster 3 n = 11	Cluster 4 n = 35	P value*
Gender, girls/boys ^#	0/79	44/0	5/6	16/19	< 0.001
Age, years	10.6 ± 2.9	11.0 ± 2.6	10.0 ± 1.3	10.0 ± 2.3	0.34
Height, cm ^#	144 ± 17	143 ± 14	139 ± 9	139 ± 14	0.33
Weight, kg	39 ± 16	38 ± 16	35 ± 7	35 ± 13	0.46
BMI, kg.m⁻²	18.2 ± 3.2	18.0 ± 4.2	17.8 ± 2.3	17.7 ± 3.4	0.92
Early wheezing, n	33	19	4	12	0.86
Atopy, n (%)	67 (85)	39 (89)	8 (73)	28 (80)	0.55
Tobacco exposure, both ^#	0	0	1	35	< 0.001
maternal	0	0	0	25	< 0.001
paternal	0	0	1	23	< 0.001
ICS, n (%)	37 (47)	25 (57)	5 (45)	20 (57)	0.62
ICS dose, BED μg/d ^#	135 ± 208	180 ± 236	149 ± 199	188 ± 217	0.57
LABA, n	33 (42)	22 (50)	4 (36)	15 (43)	0.58
Clinical events
Partially or uncontrolled, n	46 (58)	29 (66)	10 (91)	28 (80)	0.028
Days with symptoms ^#	3 [0–8]	4 [0–12]	3 [3–7]	7 [3–15]	0.045
With exacerbation, n	16 (20)	6 (14)	6 (55)	14 (41)	0.006
Days with oral steroid ^#	0 [0–0]	0 [0–0]	3 [0–3]	0 [0–3]	0.021
Pulmonary function tests
Before bronchodilation
FE_NO, ppb ^#	25 [14–45]	34 [19–51]	21 [9–49]	30 [14–52]	0.58
sRaw, % pred	201 ± 40	200 ± 54	211 ± 37	212 ± 47	0.68
FEF_50%/TLC	0.70 ± 0.13	0.74 ± 0.13	0.62 ± 0.15	0.69 ± 0.14	0.5
FEV₁, % pred	98 ± 13	100 ± 13	91 ± 14	97 ± 14	0.28
FVC, % pred	104 ± 14	106 ± 13	106 ± 15	106 ± 12	0.88
FEV₁/FVC ^#	78 ± 8	80 ± 6	74 ± 8	77 ± 9	0.41
FEF_25–75%, % pred	74 ± 18	72 ± 18	59 ± 19	70 ± 20	0.1
TLC, % pred	105 ± 12	104 ± 9	103 ± 12	106 ± 10	0.32
FRC, % pred	103 ± 17	102 ± 17	101 ± 22	105 ± 18	0.22
RV, % pred	107 ± 28	104 ± 27	101 ± 20	115 ± 32	0.23
RV/TLC	0.24 ± 0.06	0.25 ± 0.06	0.24 ± 0.04	0.27 ± 0.06	0.14
After bronchodilation
sRaw, % pred	125 ± 28	127 ± 31	117 ± 22	133 ± 34	0.22
FEF_50%/TLC ^#	0.78 ± 0.17	0.86 ± 0.15	0.80 ± 0.22	0.75 ± 0.22	0.025
FEV₁, % pred ^#	107 ± 12	110 ± 12	105 ± 14	107 ± 12	0.47
FVC, % pred	107 ± 13	108 ± 12	107 ± 14	110 ± 12	0.69
FEV₁/FVC, %	84 ± 6	86 ± 4	83 ± 8	82 ± 7	0.05
FEF_25–75%, % pred	92 ± 20	95 ± 16	86 ± 25	84 ± 21	0.17
TLC, % pred	103 ± 11	102 ± 9	101 ± 13	104 ± 10	0.73
FRC, % pred	101 ± 16	100 ± 14	95 ± 19	101 ± 13	0.3
RV, % pred	95 ± 22	90 ± 22	91 ± 25	96 ± 23	0.78
RV/TLC	0.22 ± 0.05	0.22 ± 0.05	0.23 ± 0.04	0.23 ± 0.05	0.6
Bronchodilator response, % ^#	8 [5–12]	10 [5–17]	15 [9–20]	8 [3–16]	0.036

# : variables included in the cluster analysis
* : Comparison between clusters using analysis of variance for continuous variables or Kruskal Wallis Rank test (non normal values) and χ² test for proportions. Significance values for variables included in the cluster analysis are a product of the cluster algorithm and are provided for illustrative purposes only.
Results are provided as mean ± SD or median [25^th - 75^th percentiles] or absolute number with percent age (%).

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.