INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The Global Initiative on Asthma (GINA) recommends the future development of noninvasive tests to reflect asthmatic airway inflammation for use in diagnosis, monitoring, and evaluation of treatment. The need for methods to detect early asthma, especially in infants and young children, is emphasized (1). Wheezing in young children is reported in one of five of all children younger than age 6 (2). Diagnosis, treatment, and monitoring are hampered by lack of objective measures. It is generally recognized that children younger than 6 yr can rarely cooperate in obtaining reproducible spirometric measurements (3).

Nitric oxide (NO) measurements may provide a noninvasive method to detect asthma in young children. The fraction of NO in exhaled air (FeNO) of children and adults with asthma is increased compared with nonatopic healthy control subjects (4), and is increased during pollen season in children with pollen-allergic asthma (9). FeNO increases during acute asthma (10) and is reduced by steroid (4, 13) and leukotriene antagonists (16).

However, there is a need to develop applicable FeNO measurement techniques that are independent of verbal instruction and do not require the active cooperation of the child. ATS recommendations on FeNO measurements advise on-line measurements of NO during exhalation for least 6 s with a flow of 45-55 ml/s against a resistance causing airway pressure > 5 cm H₂O (17). The ERS task force recommended an exhalation flow of 100 ml/s (18). Such single-breath on-line measurements (SBOL) of FeNO are probably even more demanding than spirometry and are rarely successful in young children. Use of various biofeedback systems and animated applications will certainly increase the acceptability in children older than 6 yr and even in some younger than 6 yr but will still require more active cooperation than can be expected from the average young child.

It has been suggested that NO concentration be measured in mixed exhaled air collected from repeated tidal exhalations into an inert bag (FeNO[mixed]) (19). This method was applied in young children younger than 6 yr of age with recurrent wheeze and showed increased levels of NO during acute wheezy attacks and decreased levels after treatment with oral steroids (20). However, the method may be biased from the marked flow dependency of FeNO, which is inversely correlated with the exhalation flow (21). Flow varies in proportion to the breathing frequency and NO concentration is biased from such flow variations during tidal breathing, that is, FeNO collected during tidal breathing is biased from the breathing frequency. This may hamper repeatability and sensitivity to reflect FeNO from lower airways. Any contamination from nasal air will not be discovered as the NO profiles are not monitored on-line. Finally, the influence of NO in the dead space cannot be accounted for (22).

We have therefore developed a technique to measure FeNO at fixed flow rates during controlled tidal breathing (FeNO [controlled]). This method is acceptable in awake young children 2 yr or older. In the present study we have validated this method against the reference method FeNO(SBOL) in school children with asthma. In addition, we compared it with the method previously used in this age group, the FeNO(mixed) method.

Subsequently, we validated the methods indirectly through comparison of exhaled NO in healthy and wheezy young children of various severities as well as the response to dose titration of inhaled corticosteroids (ICS) in young children with asthma aged 2-5 yr.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Equipment

FeNO was measured with the Aerocrine NO system with a CLD 77 AM chemiluminescence analyzer from EcoPhysics AG (Duernten, Switzerland). The sensitivity of this analyzer is 0.1 parts per billion (ppb), the rise time (0 to 90%) is 0.1 s, and the sample flow is 110 ml/min. On-line measurements of NO, flow, and pressure were collected at a rate of 10 times/s. The analyzer was calibrated at least once daily using certified NO gas (100 ppb) from AGA (Gothenburg, Sweden). Flow and pressure were measured on-line by a heated pneumotachometer (Hans Rudolf Inc., Kansas City, MO).

FeNO at Single Breath On-line Measurement: FeNO(SBOL)

Measurements of FeNO were performed in accordance with the recommendations of the European Respiratory Society Task Force Report (18).

Practical procedure. The child inhaled to total lung capacity from NO-free air and exhaled subsequently for 10 s against a linear resistor of 20 kPa/L/s. Noseclips were not used. The exhalation flow rate was displayed on-line on a computer screen together with the predefined limits of acceptance of 0.09-0.11 L/s, allowing the child to navigate within these flow limits.

Data selection. The measurement was rejected if a stable flow was not sustained for the last 5 s of exhalation, and if a steady level of NO was not maintained during this period.

Data end point. NO concentration was calculated as the mean value from 50 to 90% of the whole breath. FeNO(SBOL) was selected as the lowest of three technically acceptable measurements.

FeNO at Fixed Flow during Controlled Tidal Breathing: FeNO (Controlled)

Practical procedure (Figure 1). FeNO was measured on-line during spontaneous breathing. The flow measured by the pneumotach was displayed on-line on the computer screen, which allowed the operator to control this flow by continuously adapting the outlet resistance to target the flow at the desired level. The child was breathing slowly and regularly through a mouthpiece. No active cooperation was required of the child. Noseclips were not used.

View larger version (24K): [in this window] [in a new window]   Figure 1.   Schematic illustration of NO measurement during controlled tidal breathing.

A mouthpiece was connected to a two-way valve (Hans Rudolf, Inc.). NO-free air (AGA, Gothenburg, Sweden) was continuously flushed (1 L/min) through the inlet of the valve to allow the child to inhale NO-free air and to prevent collection of NO in the dead space of the valve system and sample tube from the previous exhalation. During the exhalation phase the valve from the NO-free air supply closes. Air was sampled at the outlet for on-line NO analyses (Figure 1).

Resistance at the exhalation valve was continuously adjusted by the operator with an adjustable valve aiming to target the exhalation flow within the preset flow limits of 40-60 ml/s (Figures 1 and 2). The resistance was continuously, quickly, and smoothly adjusted during the exhalation phase. This provided a minimum mouth pressure of 5 cm H₂O and restricted the flow rate within the predefined limits.

View larger version (16K): [in this window] [in a new window]   Figure 2.   Measurement in a 3-yr-old boy during controlled tidal breathing. Five consecutive NO samplings at the last part of the exhalation flow curve within flow range and corresponding NO values are embossed.

Data analysis. The software used to display, store, and calculate the measurements was a prototype version developed in collaboration with Aerocrine AB (Stockholm, Sweden). The software analyzes NO only when the exhalation flow curve is within the target interval.

The NO and flow signals are staggered as the NO signal has a lag time due to the time required for the analyzer to collect and analyze the air sample. The volume of the sample tube used is approximately 1.2 ml. As the sample rate is 110 ml/min, the lag time is approximately 0.7 s. NO and flow signals are adjusted for such lag time to obtain corresponding values. The software displays the corresponding flow and NO measurement points superimposed on the flow and NO concentration curves, allowing on-line quality assurance of the data analysis (Figure 2).

Data selection and data end point. Exhalations with sudden NO peaks were excluded, as contamination from the nasal cavity could be assumed. Exhalations with a sudden drop in pressure and/or flow were excluded as this reflected leakage. After at least 1 s of exhalation within the target range it was assumed that the rinse volume had been flushed and the steady state of NO was obtained. The subsequent last 0.5 s of exhalation within the target flow was used to estimate the mean NO FeNO(controlled). Concomitant mouth pressure above 5 cm H₂O was required.

Three repeat runs were performed, and FeNO(controlled) was chosen as the lower value of three acceptable curves. The lowest of three technically acceptable methods was chosen, because false high values are more likely to occur than false low values, that is, the bias is not random but biased toward high values. This choice is analogous with the recommendations on spirometry where the highest of three values is chosen because the bias is toward falsely low values.

FeNO in Mixed Exhaled Air: FeNO(mixed)

Practical procedure. The child breaths through a mouthpiece to fill an NO impermeable gas sample bag of 750 ml (Quintron, Milwaukee, WI). Noseclips were not used. NO concentration in the bag was measured less than 15 min after filling. The procedure was performed in duplicate to test for reproducibility.

A mouthpiece was connected to a two-way valve (Hans Rudolf, Inc., Kansas City, MO) inhaling NO-free air and exhaling into the bag.

Resistance at the valve outlet was fixed (2-mm tube). In pilot studies this pressure was always above 5 cm H₂O, sufficient to close the soft palate and prevent nasal contamination of exhaled air.

Patients

1. School children, 7-14 yr, with asthma were included if they presented FeNO(SBOL) > 15 ppb on two separate days.

2. Healthy children 2-5 yr of age were recruited if they had no history of wheezing or atopic disorders, had no history of airway infection for the previous 2 wk, lived in nonsmoking home, and had a normal pulmonary auscultation at the day of NO measurement.

3. Children with asthma aged 2-5 yr were recruited if they had a diagnosis based on typical asthma symptoms, symptom relief from ICS or inhaled ₂-agonist, and relapse during interruption of such treatment. The children were studied in a stable clinical condition and presented with normal pulmonary auscultation and no signs of airway infection, and had been without symptoms or change in medication in the preceding 2 wk. Treatment with ₂-agonist was allowed up to 4 h before measuring FeNO. Two subgroups of children with asthma were compared:

a. Children with mild episodic symptoms of wheeze requiring only inhaled ₂-agonist as needed and no controller treatment.

b. Children with asthma with recurrent symptoms treated with regular ICS.

The local ethics committee approved the study (KF 01-050/99) and written informed consent was obtained from the parents, and verbal consent was obtained from the child.

Study Design

1. School children performed measurements using the following three techniques.

a. Forced single-breath exhalation maneuvre at a flow of 0.09-0.11 L/s.

b. Controlled tidal breathing with a target flow of 0.09-0.11 L/s.

c. Collection of mixed exhaled air by tidal breathing.

Correlation and agreement between methods were calculated.

2. School children repeated measurements of FeNO(controlled) at a further three different flow levels:

a. 0.04-0.06 L/s

b. 0.14-0.16 L/s

c. 0.19-0.21 L/s.

3. Healthy children, children with wheeze, and children with asthma aged 2-5 yr performed duplicate measurements of FeNO(controlled) and FeNO (mixed).

Correlation and agreement within methods were calculated.

4. FeNO(controlled) was measured in children with asthma aged 2-5 yr during dose down-titration of their inhaled corticosteroid maintenance treatment. FeNO(controlled) was measured 2-3 wk after dose reduction.

Statistics

FeNO values are logarithmically transformed because the data deviate significantly from normality. FeNO values are expressed by their geometric mean and 95% confidence interval. Unpaired and paired t tests are used to compare data between and within different groups of children. Two-ways ANOVA analyses are used to analyze the effect of flow and budesonide dose on FeNO concentration. The methods are compared using the mean ratio difference, correlation coefficient (r), and limits of agreement.

The limits of agreement between and within methods are estimated from the mean difference ± 1.96 times the standard deviation of the difference (22). The limits of agreement are expressed as a factor difference between two methods because of the log-transformation of FeNO data.

A computer package (MedCalc; MedCalc Software, Mariakerke, Belgium) was used for statistical analysis. A two-tailed p value < 0.05 is considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sixty-seven children completed the study: 16 school children and 51 younger than 5 yr old. The mean age of the school children (range) was 9 yr (7). All children were on daily ICS with a dose of 400 µg budesonide and ₂-agonist as needed via a Turbohaler. The group of young children consisted of 14 healthy young children, mean age (range) 3 11/12 yr (2), 15 children with wheeze 4 7/12 yr (2) treated with inhaled terbutaline via a metal spacer (Nebuchamber; AstraZeneca, Lund) as needed, and 22 children with asthma 4 4/12 yr (2) treated with inhaled budesonide in a mean (range) daily dose of 313 µg (100-800) as well as terbutaline as needed, both inhaled via a metal spacer.

FeNO(controlled) was attempted in 60 children and succeeded in 51. Of the 9 children who failed to complete the measurements, 7 had asthma and 2 had episodic wheeze (4 of 2 yr, 3 of 3 yr, and 2 of 5 yr).

FeNO(mixed) was completed in 35 of the 51 children completing the controlled tidal breathing method. No one failed the FeNO(mixed) method.

Correlation and agreement between the methods were estimated with FeNO(SBOL) as the reference (Table 1). Limits of agreement between FeNO(controlled) and FeNO(SBOL) range from 30% to +40%, whereas limits of agreement between FeNO(mixed) and FeNO(SBOL) range from 50% to 530%.

FeNO(controlled) is closely related to flow within the range of normal tidal flow as measured in 9 school children (Figure 3). Children with asthma and healthy children separated significantly as measured by FeNO(controlled) as well as FeNO (mixed) (p < 0.05). Children with wheeze and healthy control children separated significantly as measured by FeNO(mixed) (p < 0.05), but not by FeNO(controlled) (p = 0.057) (Figure 4 and Table 2).

View larger version (11K): [in this window] [in a new window]   Figure 3.   Mean (SEM) FeNO(controlled) at different flow levels measured in school children with asthma.

View larger version (15K): [in this window] [in a new window]   Figure 4.   Geometric mean (± SEM) FeNO(controlled) in healthy children, children with wheeze and children with asthma aged 2-5 yr.

Two children with wheeze and 13 children with asthma had a positive skin allergy or RAST test, whereas there were no positive findings in the other 12 young children with wheeze and 8 young children with asthma. In one girl with asthma no allergy test was performed. FeNO(controlled) = 5.8 ppb (3.4- 9.8) in the allergic group versus 4.7 ppb (3.6-6.1) in the nonallergic group (p = 0.42). Furthermore, no difference in FeNO (controlled) was found within the children with asthma when they were divided into an allergic and a nonallergic group: 5.8 ppb (3.2-10.7) versus 5.7 ppb (4.3-7.7) (p = 0.68), respectively.

ICS dose was tapered in 9 young children with asthma. The daily treatment dose was reduced from a mean (range) of 388 µg budesonide/d (100-800 µg) to 110 µg/d (0-400 µg). Mean FeNO (controlled) (95% CI) increased significantly from 4 ppb (2-7 ppb) to 13 ppb (10-18 ppb) (p < 0.0001) (Figure 5), whereas FeNO(mixed) increased from 6 ppb (4-10 ppb) to 9 ppb (6-13 ppb) (p = 0.036).

View larger version (15K): [in this window] [in a new window]   Figure 5.   ICS tapering in 9 children with asthma aged 2-5 yr.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have outlined and validated a method for on-line measurements of FeNO at a fixed exhalation flow rate in children older than the age of 2 yr. The method allows on-line measurements during spontaneous breathing while the exhalation flow is adjusted by changing the exhalation resistance. This method provided clinically relevant measurements, allowing separation of wheezy symptoms in young children into different severities and demonstrated a dose relation of steroid treatment in young children with asthma.

On-line measurements of FeNO(controlled) allow breath-to-breath profiles to be scrutinized to ensure a stable, calm, and reproducible breathing pattern. The exhalation flow is targeted at 50 ml/s (range 40-60) by continuously adapting the exhalation resistance. Adjustments of exhalation resistance to target a fixed exhalation flow can be done manually or by automatic flow controllers and servomechanisms. The imposed exhalation resistance ensures increased mouth pressures helping to close off the soft velum and avoiding contamination from nasal air.

The time that the child is required to exhale within the target flow is obviously essential to the success of this method. The method assumes rapid-response analyzers. The time required for the applied analyzer to process the analysis and provide a full response is around 0.1 s from a sample flow of 2 ml/s. This delay is therefore of negligible importance; analyzers with a longer response time and higher sampling flow rates are unable to track the NO changes during spontaneous breathing. The minimal time required to achieve steady state is primarily determined by the dead space of the upper airways and equipment (rinse volume), which approximates 2-3 ml/kg. Approximately 1 s of exhalation at the target flow of 50 ml/s is therefore required to flush the rinse volume. We used the mean NO concentration within the target flow range in the subsequent part of the exhalation to reflect the NO concentration of the conducting airways.

The method still requires passive cooperation inasmuch as the child needs to breath slowly and regularly through a tightly fitted facemask, or close the mouth around a mouthpiece, which is often the limiting factor, as well as to tolerate breathing against resistance. No other active cooperation is needed.

Measurements during spontaneous breathing may be slightly biased, as there is no control over the lung volume at which the flow is measured. The starting lung volume affects the exhaled NO concentration, and the functional residual capacity will probably increase due to the imposed expiratory resistance. In addition, the time within the target flow may be too short to ensure a true equilibration with airway FeNO. For these reasons NO levels measured during spontaneous breathing may not equate SBOL measurements. Still, this does not invalidate the method but requires separate characterization including definition of normal values in healthy children specific to this method.

Wider use of this method requires the availability of dedicated software, which needs to adjust for the lag time (staggering of flow and NO signal) and the individual rinse volume.

The method was acceptable in most young children. In the present study 51 of 60 young children completed repeated measurements, of whom 21 were 2-3 yr old. NO output follows the concentration gradient and is therefore influenced by flow rate (21, 24). ERS (18) and ATS (17) recently recommended on-line measurement of NO during a fixed expiratory flow of 100 ml/s (90-110 ml/s) and 50 ml/s (45-55 ml/s), respectively. Discrimination between children with asthma and healthy children is better at low flow rates (25). We therefore adopted the lower flow target from the ATS recommendations for our measurements of FeNO in young children.

NO concentration is several decades higher in the nasal cavity, which may confound measurements of NO from lower airways. The soft palate seals the nasal cavity when mouth pressure increases above 3-4 cm H₂O (27). In accordance with ERS and ATS recommendations we therefore adopted an expiratory resistance ensuring mouth pressure above 5 cm H₂O. In addition we excluded the occasional measurements with lower mouth pressures and the sporadic measurements with obvious NO peaks, suspecting nasal contamination.

Atmospheric NO levels of exhaled air in the apparatus dead space may confound FeNO during tidal breathing (22, 28). The dead space was therefore continuously flushed with NO-free air and the child inhaled NO-free air.

We validated the FeNO(controlled) directly through comparison with measurements of FeNO(SBOL) in school children able to perform both methods. There was no significant difference between the two methods (p > 0.05) and the limits of agreement ranged within a factor of 0.7-1.4, that is, FeNO (controlled) can be expected to be between 30% and +40% of the concomitant FeNO(SBOL) measurement.

Indirect validation was provided in this study through comparison of FeNO(controlled) in young children with asthma versus healthy control children and through measurements of FeNO(controlled) during steroid tapering in young children with asthma from the age of 2. FeNO(controlled) was significantly increased in young children with a clinical diagnosis of asthma despite current treatment with inhaled corticosteroids. Corticosteroids are believed to control the inflammatory component of asthma, though certain aspects of the inflammation such as the leukotriene pathway may not be checked (29, 30). The significantly increased FeNO could be due to such an unchecked inflammatory component or to underdosing of the steroid.

FeNO(controlled) seemed to be increased in young children with mild episodic wheeze treated with inhaled ₂-agonist as needed, though the difference was borderline significant (p = 0.057). The tendency to increased FeNO(controlled) probably reflects a mild inflammatory component in these young children with mild episodic wheeze. This observation is supported by a previous report in infants with wheeze, in which sedated infants with episodic wheeze had raised exhaled NO as compared with healthy control infants (31).

Upper airway infections are very common in association with wheezy episodes and may bias the FeNO measurement. In the present study, the children had no signs of infection for at least 2 wk prior to the measurements.

FeNO(controlled) in healthy children was 3 (2) ppb, which is in agreement with previous reports on FeNO(SBOL) in healthy children (32, 33). NO concentration measured during endotracheal intubation was 5-6 ppb in two healthy infants (20).

FeNO(controlled) increased during dose tapering of inhaled corticosteroids in young children with asthma from the age of 2, which is in accordance with a previous study in adults (34). Several studies have shown that steroid decreases FeNO in children (4, 11, 12, 35) and adults (15, 36). In some cases a plateau response of exhaled NO was found at a dose of budesonide of 400 µg daily (14). The dose-related change in FENO further validates such measures of lower airway inflammation in young children.

We did not find a statistical difference in FeNO between children who were allergic and children who were nonallergic. A positive correlation between FeNO and a positive skin test has been reported (32). It is also observed that children with asthma sensitive to grass pollen also presented increased FeNO values out of the grass pollen season during symptom-free intervals (9). A family history of atopy was recently reported to be associated with increased FeNO levels, independent on symptoms of wheezing (31).

FeNO was previously reported in young children by measurement of NO concentration in exhaled air collected in inert bags after collection of a defined volume (20). Recommendations for standardized off-line collection have been made and are identical for both adults and school children (17). While performing tidal breathing the child inhales NO-free air and exhales against resistance through a two-way valve into an inert bag. The mouth pressure may be maintained through the use of a restrictor at the exhalation outlet. The method has some inherent inaccuracies. Any contamination from nasal air will not be apparent as the NO profiles are not monitored, but will add to the NO level measured in the mixed air. Second, the variable flow pattern and breathing frequency will affect the NO concentration measured. NO concentration is clearly flow dependent within the flow interval of tidal breathing of young children (Figure 3). Increased breathing frequency will increase the flow rate and reduce the NO concentration. Therefore, the NO concentration in a defined volume collected from rapid breathings is reduced as compared with the NO concentration collected from slow breathings. Children often hyperventilate when presented with a facemask (36). Finally, the NO concentration measured in the mixed air will be confounded NO from the previous exhalation remaining in the anatomical and apparatus dead space (22, 28). Measurements of NO concentrations in exhaled air collected from uncontrolled tidal breathing into a mixed air reservoir are therefore prone to bias. In accordance, the limits of agreement between FeNO(SBOL) and FeNO(mixed) ranged from 0.5 to 5.3, that is, FeNO (mixed) can be expected to be between 50% and +530% of the FeNO(SBOL) measured concomitantly. Such poor repeatability reduces the sensitivity needed to separate the healthy from the diseased by FeNO. On-line measurements of NO are therefore preferable.

In conclusion, FeNO(controlled) is a feasible method for measurement of exhaled NO in young children from the age of 2. The measurements seem valid and of acceptable repeatability. On-line measurements of NO during controlled exhalation presently seem to be the method of choice for measuring FeNO in young children.