INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cysteinyl leukotrienes (Cys-LT) are mediators pivotal to asthma pathophysiology. In addition to their potent bronchoconstrictory properties, they have several proinflammatory properties (1, 2). Leukotriene receptor antagonists (LTRAs) may accordingly control part of the airway inflammation in asthma. Such possible anti-inflammatory properties of LTRAs are essential to their positioning in asthma management.

NO in exhaled air (FENO) is increased in patients with asthma and correlates with asthma severity, sputum eosinophils, methacholine reactivity, as well as peak flow variability, probably related to airway inflammation (3). Although the precise mechanism that links NO with airway inflammation remains to be elucidated, it has been suggested as a useful marker of asthma control, and as a noninvasive measure of the effect of pharmacologic intervention on airway inflammation in asthma. We have therefore studied the effect of the LTRA montelukast on FENO in a randomized, placebo-controlled, crossover trial in asthmatic children.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthmatic children 6 to 15 yr of age were eligible for the study if considered stable and well controlled. Inhaled glucocorticosteroid treatment was allowed if maintained at a regular dose for 4 wk before the run-in period and throughout the study. Inhaled ₂-agonist, either short-acting (terbutaline or salbutamol) or long-acting (formoterol), was used only as rescue treatment as required. No other antiasthma medication was allowed.

Asthmatic children were allocated into a 2-wk run-in period, where FENO was measured at the start (Baseline 1) and at the end (Baseline 2). Children were randomized into controlled treatments provided FENO was above 20 ppb at both baseline measurements, and they reported no or only mild symptoms during the run-in, i.e., no daily asthma symptoms.

Subsequent to the run-in the children were randomized using a computer generated schedule into a crossover, double-blind study of montelukast chewable tablet 5 mg in the evening versus corresponding placebo. Treatment periods were 14 d. The child visited the clinic 2 d into each treatment period, at the time of crossover, and at the end of the last treatment period. A 2-wk open run-out period was scheduled where all children were given additional budesonide 200 µg from Turbuhaler in the evening irrespective of baseline treatment. The children visited the clinic after completion of the 2-wk run-out period.

Unused medication was returned to the clinic at the end of each of the blinded treatment periods, and the compliance was estimated by pill counts.

FENO and spirometry were measured at each clinic visit followed by inhalation of terbutaline Turbuhaler 0.5 mg × 2 and measurements were repeated 20 min later.

The local Ethics Committee (KF02-125/98) and the Health Authorities (2612-586) approved the study. Informed consent was obtained from the children and their parents.

Measurements of FENO were performed in accordance with the recommendations of the European Respiratory Society Task Force Report (11). The child inhaled to total lung capacity from NO-free air, and exhaled subsequently (without noseclip) for 10 s against a linear resistor of 10 kPa/L/s (Hans Rudolf Inc., Kansas City, MO). The exhalation flow rate was displayed on-line on a computer screen together with the limits of acceptance of 100 to 130 ml/s, allowing the child to navigate within these flow limits. 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. NO concentration was calculated as the mean value from 50 to 90% of the whole breath. FENO was selected as the lowest of three technically acceptable measurements.

FENO was measured with the Aerocrine NO system, including CLD 77 AM chemiluminescence analyzer from Eco Physics 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, the sample flow 110 ml/min and the lag-time from mouthpiece were less than 1 s. The software used to display, store, and calculate the measurement is the EBA-Software version 0.10 from Aerocrine AB (Stockholm, Sweden), running on an IBM pentium-2, 150 MHz PC. Flow measurements was performed by a heated pneumotachometer (Hans Rudolf Inc.).

Calibration of the NO analyzer was done daily using certified NO gas 20 ppm ± 2% with guaranteed stability from AGA, Lidingö, Sweden. Zero-point determination used instrument-filtered zero-air. In addition a "biologic standard" was established from four members of the staff who had no respiratory or atopic disorders. FENO was determined on every study day in this group to further assure stable measurements.

Children without atopic symptoms and without atopy in first- degree relatives were included to define the normal value of FENO.

Spirometry was performed with a Master Screen unit (E. Jäeger GmbH, Würzburg, Germany) in accordance with ATS guidelines, and the best of three technically acceptable measurements was chosen. Maximal midexpiratory flow (MMEF) was measured between 75 and 85% of the forced vital capacity.

Formoterol and terbutaline were withheld for 12 and 4 h, respectively, before the measurements of FENO and spirometry.

Data were collected according the principles of Good Clinical Practice including external data monitoring and drug accountability control. The data manager broke the blinded code after clean file had been declared and data files had been locked in the computer.

All children completing measurements after first and second treatment period were included in the analyses of treatment effects based on the intention-to-treat principle.

The reproducibility was estimated from the coefficient of variation calculated as SD/mean. The within-subject standard deviation (SDw) was calculated as the SD of the differences between paired measurements (Baselines 1 and 2) from all subjects divided by 2. The within-subject coefficient of variation (CVw) was calculated from the SDw divided by the mean. The difference between paired measurements was plotted against their mean to examine whether variability was independent of the magnitude of the measurement (12).

Analyses of treatment effects were done by two-tailed paired t tests with 0.05 as level of significance. Central tendency was expressed by the mean and 95% confidence internal (95% CI).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Twenty-nine asthmatic children (mean age, 12 yr; range, 6 to 15), were included into the study. One child was withdrawn early in the second treatment period because of deterioration of asthma. No other asthma exacerbations were reported during the study. Two children were withdrawn after the first treatment period because of noncompliance. Twenty-six children completed the study and were included in the analysis (Table 1).

One child had taken only three tablets during the first treatment period but six tablets too many in the subsequent period. Noncompliance was less extreme in the other children (Table 1). In three cases more tablets were used than required for the treatment period. These were all accounted for by the patients as being lost. All children were included into the outcome analysis according to the intention-to-treat principle.

FEV₁ was 101% predicted (95% CI, 97 to 106) at first study visit defining the baseline. All children except one had a positive skin allergy relevant to the asthma history. Eleven of 26 children were receiving regular treatment with inhaled budesonide (0.1 to 0.8 mg daily dose) or fluticasone propionate (0.1 mg daily dose) from dry powder inhalers; overall mean daily dose was 273 µg. Eighteen children used formoterol, seven used terbutaline, and one used salbutamol from dry powder inhalers as as-required rescue treatment. FENO at baseline was 35.3 ppb (95% CI, 30.6 to 39.9). In the subgroup of children without concurrent inhaled steroid treatment mean FENO at baseline was 36.8 (95% CI, 30.5 to 43.0), whereas in children receiving regular treatment with inhaled steroids mean FENO at baseline was 33.2 (95% CI, 25.2 to 41.2), with no significant difference between the two groups.

The mean FENO over all study days in the biologic pool was 4.9 (95% CI, 4.5 to 5.2). One subject from this biologic pool performed 37 measurements on different days with a coefficient of variation of 16%. Within-subject coefficient of variation (CVw) of FENO from first to second baseline measurement in the 26 children was 38%. This variability was independent of the FENO level.

Twelve normal nonatopic children 7 to 15 yr of age recorded a mean FENO of 8.6 (95% CI, 2.8 to 14.14).

FENO was significantly reduced by 20% after the 2-wk treatment with montelukast as compared with placebo (Figure 1). This effect occurred rapidly, with a 15% fall in FENO within 2 d (Table 2). FENO at baseline and at the end of placebo were similar (Table 2). Accordingly, similar conclusions were achieved when comparing FENO after montelukast treatment with FENO at baseline. A carryover effect of montelukast into the early part of the placebo period was apparent as FENO at Day 2 of the placebo treatment was 31 (95% CI, 25.7 to 36.2), but it increased significantly back to the baseline level at Day 14. 

View larger version (34K): [in this window] [in a new window]   Figure 1.   FENO was significantly reduced after treatments with montelukast and budesonide as compared with placebo. Such treatment effects were independent of concurrent treatment with inhaled steroid (circles, solid lines) or no concurrent treatment with inhaled steroid (crosses, dashed lines).

Children receiving regular topical steroid treatment (n = 11) showed a 22% fall in FENO, from 30.1 (95%CI, 16 to 45.8) with placebo to 23.6 (95% CI, 17.4 to 29.8) with montelukast, and children without steroid treatment (n = 15) exhibited an 18% fall in FENO, from 40.3 (95% CI, 29.9 to 50.7) with placebo to 33.1 (95% CI, 25.4 to 40.8) with montelukast. The fall in FENO in children with or without concomitant steroid treatment did not differ significantly.

FENO decreased significantly from steroid treatment during the open run-out period as compared with placebo within the double-blind study (Table 2) as well as compared with baseline.

FENO increased highly significantly, by 6% after inhalation of terbutaline from an overall mean of 31.8 to 33.7 (n = 161; p < 0.0001). This increase was independent of study treatment. FENO measured after inhalation of terbutaline showed similar treatment effects from montelukast and budesonide as FENO measured before terbutaline.

Lung function exhibited a tendency to improve after montelukast and after budesonide as compared with the placebo treatment period, but this was not statistically significant (Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

FENO in asthmatic children was significantly decreased from treatment with the LTRA montelukast. This effect is supported in a study of adults with severe asthma in whom the LTRA ONO-1078 (Pranlukast) reduced the increase in FENO subsequent to steroid reduction (13). Such findings corroborates anti-inflammatory properties of LTRA.

Cys-LTs, in addition to their potent bronchoconstrictory properties, have several proinflammatory properties. Cys-LTs are potent mediators of eosinophil influx into the bronchial airway mucosa (14, 15). Correspondingly, Cys-LT receptor antagonists (LTRA) reduce peripheral eosinophil number in asthmatic children (16) and in adults (17), and reduce eosinophils in induced sputum (18) and bronchoalveolar lavage of adult asthmatics (19). Eosinophilia in the airway mucosa is a hallmark of the airway inflammation in bronchial asthma, and the antieosinophilic effect of LTRA may contribute to inflammatory control in bronchial asthma. In addition Cys-LTs increase microvascular blood flow (20) and venular permeability (21), allowing the escape of macromolecules into the tissue, which provides the source for plasma protein-derived inflammatory mediators, including the kinins and complement and clotting systems, which may fuel the inflammatory process (22, 23). LTRA would therefore be expected to control important aspects of the inflammatory process in bronchial asthma.

There is increasing evidence that endogenous NO is a principal signaling molecule. NO derived from constitutive NO synthase is involved in physiologic regulation of airway function, whereas NO derived from inducible NO synthase (iNOS) is involved in inflammatory diseases of the airways and in host defense against infection. The NO may amplify and perpetuate asthmatic inflammation (24).

iNOS is synthesized during inflammation de novo by increased gene transcription (25) and very high expression of messenger ribonucleic acid for iNOS has been found in the lower airways (26). Thus, it seems likely that the respiratory epithelium is an important source of exhaled NO. Furthermore, it seems to originate primarily in the conducting airways since it occurs with an early peak appearing before carbon dioxide has reached its maximum (27).

FENO is increased in asthmatic patients. It increases during the late but not during the early part of an allergic asthmatic reaction (28), and during the pollen season in allergic asthmatic children, with a subsequent fall out of season (29). Airway epithelial cells from asthmatic patients, but not from normal subjects, immunostain strongly for NO synthesis (30). The increased FENO in asthmatics is thought to reflect an upregulated iNOS induced by a variety of proinflammatory cytokines, which have been shown in vitro to increase the expression of iNOS in cultured human airway epithelial cells (25). The reduced FENO from the LTRA montelukast may reflect a reduced load of such inflammatory cytokines. Clinical studies have reported increased FENO in patients with asthma correlating with asthma severity, sputum eosinophils, and methacholine reactivity, as well as peak flow variability. FENO is therefore suggested to reflect aspects of airway inflammation, although the precise mechanism remains to be elucidated (3).

FENO decreased significantly from a low dose of inhaled budesonide. Although this was not part of the controlled study, the finding is in agreement with a dose-related reduction in FENO from steroid in previous studies, with a maximal effect apparently occurring within 1 to 2 wk (8, 31, 32). FENO was lower after a budesonide 200 µg daily dose as compared with montelukast treatment. However, the relative potency of the effect of these treatments on FENO should be determined from controlled dose-response comparisons. The effect of steroids is probably through a direct effect on iNOS opposed to the indirect effect of LTRA.

The effect of montelukast on FENO was comparable in the 11 children receiving regular inhaled steroids (mean daily dose, 273 µg) and the remaining 15 children receiving only as-needed rescue bronchodilator treatment (Figure 1). The addition of 0.2 mg budesonide reduced FENO in children receiving a maintenance treatment with inhaled steroids as well as those with no steroid treatment, yet without normalizing FENO in either group. It may be that further increased doses of steroid would eventually normalize FENO, or this may reflect that the Cys-LT component of the airway inflammation is unabated by steroid treatment. Previous studies have corroborated that the Cys-LTs levels are unaffected by steroid treatment (33, 34). This may argue the case for the use of LTRAs as add-on treatment in children who are not well controlled on a moderate dose of inhaled steroids, as LTRA may target inflammatory pathways insensitive to steroids.

Montelukast showed a tendency of an improved lung function when compared with placebo. However, this was not significant, which is in keeping with the normal lung function of the children included into the study. With little room to move little improvement would be possible, and effect on lung function was not an aim of this study. Interestingly, this exemplifies the role of FENO as a biochemical indicator of the respiratory system, which can indicate changes in ongoing airway inflammation without concomitant detectable changes in classic lung function tests, notwithstanding that the within-subject coefficient of variation of FENO over 2 wk was 38%.

The effect of montelukast was probably not related to a change in airway diameter since FENO increased slightly, but definitely from terbutaline inhalation. Previous reports were unable to find an effect from ₂-stimulation (8, 35) probably because of the very small effect, which requires a large number of tests to reveal its significance. The reason for the increased FENO may be the ventilatory-perfusion mismatch caused by ₂-stimulation (36). Alternatively it may relate to the airway diameter. Previous studies have reported a reduced FENO from bronchoconstrictor stimuli (37). This suggests that airway diameter and the effect of ₂-agonists on FENO should be observed in clinical studies as well as in clinical practice utilizing FENO measurements.

The effect of montelukast was rapid, with most of the effect apparent within 2 d. This is probably different from the kinetic of the effect of glucocorticosteroids on FENO, which improves within 1 to 2 wk (31). This difference in kinetics suggests that the mechanism of action of LTRA is not similar to that of glucocorticosteroids, but they act on airway inflammation through different mechanisms.

LTRA treatment did not normalize FENO, nor was eosinophilia normalized in previous studies on their anti-inflammatory effect. LTRA treatment therefore seems unable to obtain full control of the asthma airway inflammation. However, the documented anti-inflammatory effects of LTRA provide an added value as compared with other bronchodilator treatments. It has previously been suggested that treatment able to reduce NO from the lung might reduce the requirement for inhaled corticosteroids (24). The LTRA montelukast seems to provide such a treatment contributing to the control of airway inflammation and therefore reducing the requirement for inhaled steroids.