Multicentre trial evaluating alveolar NO fraction as a marker of asthma control and severity.

Abstract

Exhaled NO can be partitioned in its bronchial and alveolar sources, and the latter may increase in the presence of recent asthmatic symptoms and in refractory asthma. The aim of this multicentre prospective study was to assess whether alveolar NO fraction and FE(NO) could be associated with the level of asthma control and severity both at the time of measurement and in the subsequent 3 months.
Asthma patients older than 10 years, nonsmokers, without recent exacerbation and under regular treatment, underwent exhaled NO measurement at multiple constant flows allowing its partition in alveolar (with correction for back-diffusion) and bronchial origins based on a two-compartment model of NO exchange; exhaled NO fraction at 50 ml/s (FE(NO,0.05)) was also recorded. On inclusion, severity was assessed using the four Global initiative for asthma (GINA) classes and control using Asthma Control Questionnaire (ACQ). Participants were followed-up for 12 weeks, control being assessed by short-ACQ on 1st, 4th, 8th and 12th week.
Two-hundred patients [107 children and 93 adults, median age (25th; 75th percentile) 16 years (12; 38)], 165 receiving inhaled corticosteroid, were included in five centres. The two-compartment model was valid in 175/200 patients (87.5%). Alveolar NO and FE(NO,0.05) did not correlate to control on inclusion or follow-up (either with ACQ /short-ACQ values or their changes), nor was influenced by severity classes. Alveolar NO negatively correlated to MEF(25-75%) (rho = -0.22, P < 0.01).
Alveolar and exhaled NO fractions are not indexes of control or severity in asthmatic children and adults under treatment.

Full text

Page 1

ORIGINAL ARTICLE
AIRWAY DISEASES
Multicentre trial evaluating alveolar NO fraction as a
marker of asthma control and severity
B. Mahut1,2, L. Trinquart3,4,5, M. Le Bourgeois6, M.-H. Becquemin7, N. Beydon8, F. Aubourg9,
M. Jala4, B. Bidaud-Chevalier2,4, A.-T. Dinh-Xuan3,9, O. Randrianarivelo7, A. Denjean8, J. de Blic3,6
& C. Delclaux2,3,10
1Cabinet La Berma, 4 avenue de la Providence, 92 160 Antony, France;2Assistance Publique-Ho ˆpitaux de Paris, Ho ˆpital Europe ´en Georges
Pompidou, Service de Physiologie, Clinique de la Dyspne ´e, Paris, France;3Paris Descartes University, Paris, France;4Unite ´ de Recherche Cli-
nique et d’Epide ´miologie, Assistance Publique-Ho ˆpitaux de Paris, Ho ˆpital Europe ´en Georges Pompidou, Paris, France;5INSERM CIE 4, Paris,
France;6Assistance Publique-Ho ˆpitaux de Paris, Ho ˆpital Necker Enfants Malades, Service de Pneumo-Pe ´diatrie, Paris, France;7Assistance
Publique-Ho ˆpitaux de Paris, Ho ˆpital de la Pitie ´, Salpetrie `re, Service de Physiologie, Explorations Fonctionnelles, Paris, France;8Assistance
Publique-Ho ˆpitaux de Paris, Ho ˆpital Robert Debre ´, Service de Physiologie, Explorations Fonctionnelles, Paris, France;9Assistance Publique-
Ho ˆpitaux de Paris, Ho ˆpital Cochin, Service de Physiologie, Explorations Fonctionnelles, Paris, France;10CIC Plurithe ´matique, 9201 de l’Ho ˆpital
Europe ´en Georges Pompidou, Paris, France
To cite this article: Mahut B, Trinquart L, Le Bourgeois M, Becquemin M-H, Beydon N, Aubourg F, Jala M, Bidaud-Chevalier B, Dinh-Xuan A-T, Randrianarivelo
O, Denjean A, de Blic J, Delclaux C. Multicentre trial evaluating alveolar NO fraction as a marker of asthma control and severity. Allergy 2010; 65: 636–644.
Several studies have evaluated the relationships between
exhaled NO and asthma control. An increase in exhaled NO
fraction (FENO) has been demonstrated in acute asthma and
in presence of recent asthmatic symptoms (1–4), which clearly
emphasizes the ability of exhaled NO to indicate major loss
of asthma control. It has also been demonstrated that the
sequential measurementof FENO
degrees of loss of control (5). The usefulness of a single
could predict milder
measurement to describe subtle degrees of asthma control at
the time of measurement or its ability to predict further loss
of control would constitute two more clinically relevant
end-points, but remains controversial (6, 7).
Partitioning exhaled NO in its bronchial and alveolar
sources deserves scrutiny. We and others previously demon-
strated that FENO measured at 50 ml/s mainly assesses
the bronchial origin of exhaled NO, whereas alveolar NO
Keywords
alveolar NO; asthma; control; exhaled NO;
severity.
Correspondence
Prof C. Delclaux, Service de Physiologie –
Clinique de la Dyspne ´e, Ho ˆpital Europe ´en
Georges Pompidou, 20, rue Leblanc, 75015
Paris, France.
Accepted for publication 8 September 2009
DOI:10.1111/j.1398-9995.2009.02221.x
Edited by: Marc Humbert
Abstract
Background: Exhaled NO can be partitioned in its bronchial and alveolar sources,
and the latter may increase in the presence of recent asthmatic symptoms and in
refractory asthma. The aim of this multicentre prospective study was to assess
whether alveolar NO fraction and FENO could be associated with the level of
asthma control and severity both at the time of measurement and in the subsequent
3 months.
Methods: Asthma patients older than 10 years, nonsmokers, without recent exacerba-
tion and under regular treatment, underwent exhaled NO measurement at multiple
constant flows allowing its partition in alveolar (with correction for back-diffusion)
and bronchial origins based on a two-compartment model of NO exchange; exhaled
NO fraction at 50 ml/s (FENO,0.05) was also recorded. On inclusion, severity was
assessed using the four Global initiative for asthma (GINA) classes and control using
Asthma Control Questionnaire (ACQ). Participants were followed-up for 12 weeks,
control being assessed by short-ACQ on 1st, 4th, 8th and 12th week.
Results: Two-hundred patients [107 children and 93 adults, median age (25th; 75th
percentile) 16 years (12; 38)], 165 receiving inhaled corticosteroid, were included in
five centres. The two-compartment model was valid in 175/200 patients (87.5%).
Alveolar NO and FENO,0.05did not correlate to control on inclusion or follow-up
(either with ACQ /short-ACQ values or their changes), nor was influenced by severity
classes. Alveolar NO negatively correlated to MEF25–75%(rho = )0.22, P < 0.01).
Conclusion: Alveolar and exhaled NO fractions are not indexes of control or sever-
ity in asthmatic children and adults under treatment.
Allergy
636
Allergy 65 (2010) 636–644 ª 2009 John Wiley & Sons A/S

Page 2

concentration, measured in the compressible compartment,
implicates deep lung inflammation, a process that is of
importance in asthma (1, 4, 8). An increased alveolar NO
concentration has been reported in steroid-resistant asthma
(9) and in presence of nocturnal symptoms (10), which sup-
ports the hypothesis that its assessment could be linked to
severity and control.
The aim of this multicentre prospective study was to assess
whether alveolar NO fraction could be associated with the
level of asthma control and severity both at the time of
measurement and in the subsequent 3 months of follow-up in
a hospital cohort of asthmatic children and adults. We evalu-
ated whether an elevated alveolar NO concentration is associ-
ated with milder degrees of loss of asthma control and/or
severe asthma, while a normal NO concentration indicates
better control and/or a nonsevere asthma. Only patients
under treatment and without overt symptoms of loss of con-
trol were included as we wanted to demonstrate the clinical
relevance of alveolar NO in daily practice. A secondary
objective was to assess whether FENOis also associated with
milder degrees of loss of asthma control and/or severe
asthma.
Methods
Participants
Patients older than 10 years were eligible if they were
addressed for pulmonary function testing for the follow-up of
an established asthma (diagnosis of asthma recorded in their
general practitioner’s notes and at least one prescription for
any anti-asthma medication in the last 12 months) in one of
the five participating centres. The study was restricted to nev-
ersmokers. Recent (within 2 weeks) exacerbation or anti-
asthma treatment modification and other chronic respiratory
disease were the noninclusion criteria.
Approval for the study was given by our local ethics
committee; all participants (and/or parents) gave written,
informed consent.
Study design
This design is summarized in Fig. 1. On inclusion, patients
underwent pulmonary function tests (see below) and an
assessment of both asthma severity and control according to
recent AmericanThoracic Society/European Respiratory
Society (ATS/ERS) Statement (11). Asthma control was
assessed at inclusion using Asthma Control Questionnaire
[ACQ, including forced expiratory volume in 1 s (FEV1)] (12)
and on subsequent weeks using both phone interviews
(ACQs-short without FEV1) (13) and the Asthma Control
Diary (ACD) (12). A peak flow meter (MiniWright?, Clem-
ent Clark International, Harlow, UK) was provided to each
included patient for the whole follow-up. Severity is defined
as the difficulty in controlling asthma with treatment and was
assessed using the four-class of GINA guidelines (14). Two
independent GINA classifications were combined (14), the
first one based on symptoms and FEV1and the second one
based on current medication, as previously done (15). Atopy
(self-reported) was defined based on the knowledge of posi-
tive skin prick test, or a history of seasonal rhinitis /conjunc-
tivitis.Participantswerefollowed
12 weeks or until the occurrence of (i) a frank modification
of treatment [twofold increase or decrease of inhaled cortico-
steroid (ICS) for more than 3 days] or (ii) severe exacerbation
(episode of increasing asthma symptoms requiring a course
of oral steroids for more than 3 days) (11).
up prospectivelyfor
Pulmonary function tests
At inclusion, exhaled NO fraction was measured before
spirometry at multiple constant flows (between 50 and 250 ml/s,
four to six expiratory flows of which the 50 ml/s flow:
FENO,0.05) using the same chemiluminescence analyser (END-
ONO 8000; SERES, Aix en Provence, France) in all centres.
The alveolar NO fraction (Calv,NO, ppb), the maximum NO
flux from conducting airways (J’awNO, nl/min) based on a
two-compartment model of NO exchange dynamics (4, 16) and
the Pearson correlation coefficient between NO output and
expiratory flow were calculated automatically. The relationship
between NO output and expiratory flow was deemed linear
when r ‡ 0.80, as previously described (17); when linearity was
not observed, only FENO,0.05was recorded. Validity criteria of
each FENO and analyser calibration were as recommended
by international guidelines (18). Thus, we obtained the follo-
wing exhaled NO variables: Calv,NO, J’aw,NO, FENO,0.05 and
Calv-cor,NO, the latter being the value of alveolar NO concen-
tration corrected for back-diffusion of NO from conducting
airways (19). Spirometry was performed according to interna-
tional guidelines (20), under anti-asthma treatment.
Statistical analysis
We hypothesized that half of the patients would be con-
trolled. We calculated that a sample size of 100 would have
more than 90% power to detect a difference between the
expected Calv-cor,NOmean of 5 ppb among stable patients and
7 ppb among the nonwell-controlled patients, assuming a
standard deviation of 3 ppb and using a two-sided two-group
t-test with a significance level of 0.05. We chose a priori a
target sample size of 150 patients, to take into account an
expected 50% of patients who may be lost to follow-up at
3 months.
Because exhaled NO variables and asthma control scores
were not distributed normally, we used nonparametric meth-
ods. Quantitative variables were compared between groups
using Kruskal–Wallis tests. First, we compared exhaled NO
variables measured at inclusion between the different levels
of asthma control and severity at that time. Second, we com-
pared the exhaled NO variables measured at inclusion
between (i) well-controlled and nonwell-controlled patients at
each follow-up visit and (ii) the groups of patients whose
control improved, worsened or did not change at each
follow-up visit. Two cut-off values for control were assessed
using an ACQ or ACQs score < 0.75 (to be confident that a
patient has well-controlled asthma) or 1.5 (to be confident
Mahut et al.
Alveolar NO and asthma control severity
Allergy 65 (2010) 636–644 ª 2009 John Wiley & Sons A/S
637

Page 3

that the patient has inadequately controlled asthma) as rec-
ommended (21). Improvement of asthma control was defined
by an absolute decrease in the ACQs score of 0.5 or more,
while worsening was defined by an increase of 0.5 or more,
as a 0.5 change in ACQ is considered to be the minimum
change that is clinically relevant (5).
Agreement between the daily diary (ACD) and the weekly
telephonic asthma control (ACQs) was assessed using Bland
and Altman methodology, including the difference plot and
calculation of 95% limits of agreement (22). Two-tailed
P-values were deemed significant when inferior to 0.05. Data
analysis was performed using sas v9.1 software (SAS France,
Figure 1 Study design and flow-chart of the participants. The
upper panel describes the study design. *Exhaled NO was
measured before spirometry at multiple constant flow rates allow-
ing, when the linearity of the relationship between NO output and
expiratory flow rate was confirmed (Pearson correlation coefficient
r ‡ 0.80), its partition in alveolar NO concentration (Calv,NO), which
was further corrected for back-diffusion (Calv-cor,NO) and maximum
flux of NO from conducting airways (J’aw,NO). When r value was
<0.80, the sole FENO,0.05, obtained in all patients, was recorded.
ACQ denotes Asthma Control Questionnaire, ACQs denotes its
shortened version (without FEV1measurement) and ACD denotes
Asthma Control Diary. GINA severity denotes the four classes
of asthma severity (intermittent, persistent mild, moderate, severe)
as previously used by Liard et al. (14) The middle panel shows the
flow of the 175/200 participants who had alveolar NO measure-
ment (primary objective). Reasons for study exit are detailed.
Only four asthma exacerbations (0.09 per patient per year (95%
CI 0.02–0.23) were observed. The lower panel shows the flow of
the 200 participants who had FENO,0.05 measurement (secondary
objective).
Alveolar NO and asthma control severity
Mahut et al.
638
Allergy 65 (2010) 636–644 ª 2009 John Wiley & Sons A/S

Page 4

Gre ´ gy sur Yerres, France) by independent investigators
(L. T. and M. J.).
Results
Description of the population
From June 2006 to December 2007, 200 patients were
included. Figure 2 shows the flow-chart of patients through-
out the study. The clinical and functional characteristics of
the patients at the time of inclusion are described in
Table 1.
Validity of exhaled NO values
The Pearson correlation coefficient between NO output and
expiratory flow was superior to 0.80 in 175 out of 200
patients (87.5%). Patients (n = 25) in whom the linearity of
the relationship between NO output and expiratory flow was
not observed were predominantly adults (adults: 17/25 vs 76/
175, P = 0.02) with a FENO,0.05of 40.0 ppb [21.4; 80.0] vs
29.1 ppb [13.9; 46.4] for the other 175 patients (P = 0.01),
which does not seem related to atopy as a trend toward the
opposite result was evidenced (with atopy: 17/25 vs 145/175,
P = 0.09).
Consequently, flow-independent exhaled NO parameters
were calculated for 175 patients (Table 1). A significant
relationshipwas observed between
(rho = 0.35, 95% CI 0.21–0.47). When alveolar NO was
corrected for back-diffusion the statistical significance of this
relationship was lost (J’awNO and Calv-cor,NO: rho = )0.02,
95% CI )0.17 to 0.12).
Atopic asthmatic patients (n = 145) vs nonatopic patients
(n = 27) had lower (trend) Calv-cor,NO(3.5 ppb [1.7; 6.2] vs
5.8 ppb [2.8; 11.3], P = 0.06), higher J’awNO (60.1 nl/min
[29.8; 103.0] vs 36.6 nl/min [15.4; 64.5], P = 0.05) and similar
FENO,0.05 (30.9 ppb [16.2; 48.2] vs 24.2 ppb [12.8; 45.4],
P = 0.29).
Patients receiving ICS (n = 140) vs those without steroid
treatment (n = 32) had similar Calv-cor,NO(3.5 ppb [2.0; 6.3]
J’awNO
and Calv,NO
Figure 2 Bland–Altman analyses of ACD and ACQs over time.
Asthma control was assessed using several criteria for two reasons.
As we expected that some patients would be lost of follow-up, we
planned to have both shortened version of ACQ (ACQs) and ACD to
verify internal consistency of control assessment and to ensure that
at least one mean of control assessment will be obtained. A good
agreement between the two questionnaires is demonstrated. A
constant difference over time is evidenced, ACQs scores being
higher than ACD scores. The number of patients evaluated at each
time-point is given in Fig. 1.
Mahut et al.
Alveolar NO and asthma control severity
Allergy 65 (2010) 636–644 ª 2009 John Wiley & Sons A/S
639

Page 5

vs 4.6 [0.2; 7.2], P = 0.86) but lower values of both J’awNO
(48.6 nl/min [24.0; 89.6] vs 95.7 [49.9; 154.8], P = 0.002)
and FENO,0.05 (28.0 ppb [13.5; 42.0] vs 47.5 [24.3; 62.2],
P = 0.002).
Exhaled NO measurements and asthma control
There was a good agreement between the daily diary (ACD)
and the weekly telephonic asthma control (ACQs) question-
naires when considering weekly assessments separately as well
as the multiple assessments per patient (mean bias: )0.19;
95% limits of agreement: )1.43 to 1.05). Figure 2 shows the
Bland-Altman analyses for the two scores obtained at the
different time-points. It has to be noted that ACQs was
systematically higher than ACD.
On inclusion, well-controlled patients represented 106/200
(53%) when choosing ACQ < 0.75 and 147/200 (73%)
when choosing an ACQ < 1.5. Figure 3 shows that all
exhaled NO parameters were non significantly different at
inclusion and during the whole follow-up for well- (ACQ/
ACQs < 1.5) and non well-controlled (ACQ/ACQs ‡ 1.5)
asthmatic patients. There was no significant relationship
between modifications of ACQ/ACQs scores over time
(improvement or deterioration) and initial exhaled NO
parameters (Fig. 4).
When choosing a cut-off value of 0.75 for ACQ/ACQs to
define well-controlled vs nonwell-controlled patients, similar
results were observed, for instance on inclusion: Calv-cor,NO
(well-controlled: 3.6 ppb [1.9; 6.4], n = 106 vs 3.8 [1.7; 6.8],
n = 69, P = NS).
Exhaled NO measurements and asthma severity
Exhaled NO values at inclusion were similar for the four
classes of asthma severity (GINA) (Fig. 5).
Nevertheless, a significant relationship was observed between
Calv-cor,NOand MEF25–75%(rho = )0.22, 95% CI )0.36 to
)0.06, P < 0.01).
Discussion
We and others suggested that alveolar NO fraction could be
a noninvasive marker of both asthma control and severity.
Nevertheless, these results had mostly been obtained in a lim-
ited number of selected patients, and strenuous criteria of
asthma control (loss of control) and severity (refractory
asthma) were often used. Consequently, we designed the first
multicentre longitudinal study that specifically addressed the
clinical potential of partitioning exhaled NO in a large hospi-
tal cohort of asthmatic children and adults under treatment,
mainly with persistent asthma. Our study demonstrates that
alveolar NO concentration and FENO,0.05are neither associ-
ated with asthma control evaluated by Asthma Control Diary
and Questionnaire or severity evaluated by GINA classes in
this specific sample of asthma patients [mainly (?80%)
receiving ICS plus long acting b2-agonist].
Validity of the exhaled NO measures
The five centres used the same analyser, which has specifically
been designed for the multiple constant flow rate measure-
Table 1 Clinical and functional characteristics of the patients
CharacteristicChildren (n = 107)Adults (n = 93)Whole population (n = 200)
Gender ratio, F/M
Age, years
Duration of asthma, years
Self-reported atopy, n (%)*
Asthma medications, n ?
short acting b2-agonist ?
ICS
ICS + long acting b2-agonist
leukotriene modifier
BEDD < 500 lg
BEDD ‡ 500 lg
Pulmonary function tests
PEF, % predicted
FEV1, % predicted
MEF25–75%, % predicted
MEF50%, % predicted
Exhaled NO variables
Calv,NO, ppb (175 patients)
Calv-cor,NO, ppb (175)
J’aw,NO, nl/min (175)
FENO,0.05, ppb (200)
43/64
12 [11; 14]
9.0 [6.0; 11.0]
89 (85)
52/41
42 [29; 55]
18.0 [9.0; 27.0]
73 (79)
95/105
16 [12; 38]
10.0 [7.0; 17.0]
162 (82)
55
4
89
31
50
43
43
5
67
14
20
52
98 (49)
9 (4.5)
156 (78)
45 (22.5)
70 (35)
95 (47.5)
95 [85; 106]
100 [92; 109]
85 [70; 103]
84 [71; 101]
95 [81; 110]
93 [82; 101]
68 [49; 87]
70 [50; 86]
95 [84; 106]
97 [87; 106]
80 [61; 92]
78 [60; 93]
5.2 [3.4; 8.3]
3.1 [1.6; 5.5]
49.8 [25.6; 111.4]
28.0 [12.8; 49.9]
7.3 [4.1; 12.1]
4.8 [2.1; 10.3]
60.5 [31.8; 93.5]
31.8 [19.6; 46.8]
5.8 [3.7; 9.4]
3.6 [1.7; 6.5]
58.4 [28.2; 103.1]
30.0 [15.9; 48.1]
Data are summarized using median (25% percentile; 75% percentile) BEDD, beclomethasone equipotent daily dose; ICS, inhaled corticosteroid.
*Three missing data (one adult, two children). ?Three missing data (two adults, one child). ?Either on demand or systematic administration.
Alveolar NO and asthma control severity
Mahut et al.
640
Allergy 65 (2010) 636–644 ª 2009 John Wiley & Sons A/S