INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inhalation of sulfur dioxide (SO₂) causes bronchoconstriction in most people with asthma (1). In animal studies, SO₂-induced bronchoconstriction appears to be reflexly mediated through parasympathetic pathways, for it is blocked by interruption of afferent or efferent neural pathways, and the introduction of SO₂ into the anatomically separated upper airways of cats produces constriction in the lower airways (2). Parasympathetic pathways appear to account for only part of SO₂-induced bronchoconstriction in people with asthma (1, 3, 4). Neither the nature of the nonparasympathetic mechanisms nor the mechanisms that activate parasympathetic pathways are known. Some researchers have suggested that SO₂ activates mast cells in the airways, causing release of mediators that induce both direct and parasympathetically mediated reflex bronchoconstriction (5, 6). This hypothesis is partially supported by the inhibition of SO₂-induced bronchoconstriction by cromolyn sodium (7) and nedocromil sodium (8), agents known to inhibit mediator release from mast cells. However, the effect of these agents on the response to SO₂ could also be explained by a direct inhibitory effect on neural reflex pathways.

Mast cells release various chemical mediators, both preformed and newly generated. Important among the latter are the cysteinyl leukotrienesLTC₄, LTD₄, and LTE₄. These mediators are potent constrictors of airway smooth muscle and are at least 1,000 times more potent than histamine on a molar basis (9, 10). Elevated levels of leukotrienes have been detected in the bronchoalveolar lavage fluid, urine, and blood of subjects with asthma after experimentally induced or spontaneously occurring bronchospasm (11). Clinical trials with new pharmacologic agents that antagonize specific leukotriene receptors or inhibit leukotriene synthesis have shown that LTC₄, LTD₄, and LTE₄, acting through a common LTD₄ receptor, are important in the pathogenesis of asthma. Moreover, studies with the oral leukotriene receptor antagonist zafirlukast (Accolate; Zeneca Pharmaceuticals, Wilmington, DE) have shown that the drug inhibits both the early and late airway response to inhaled allergen (15, 16) and the bronchoconstrictor response to exercise (17). We hypothesize that cysteinyl leukotrienes contribute to the bronchoconstriction seen in SO₂-sensitive patients during exposure to SO₂; therefore, we conducted this trial to determine whether zafirlukast inhibits SO₂-induced bronchoconstriction in subjects with asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Population and Study Design

Seven men and five women, between the ages of 25 and 43 yr, with mild-to-moderate asthma were enrolled in a randomized, double-blind, placebo-controlled, two-period crossover trial. All subjects were nonsmokers and had reversible airway disease demonstrated by a  15% increase in FEV₁ after inhaled ₂-agonist or a positive response to methacholine (a 20% decrease in FEV₁ caused by a methacholine concentration of < 8 mg/ml). Subjects were free of clinical signs or symptoms of airway obstruction and had an FEV₁ of  70% of predicted at screening and a positive response to inhaled SO₂ (an 8-unit increase in specific airway resistance with an SO₂ concentration of  4 ppm). Women of child-bearing potential used a double-barrier, nonhormonal birth control method during the trial. The protocol for this study was approved by the Committee on Human Research of the University of California, San Francisco, and informed written consent was obtained from all subjects.

Key exclusion criteria included the following: acute illness or disease; any history of seasonal asthma; respiratory tract infection within 6 wk of screening; history of drug or alcohol abuse; use of intravenous, oral, nasal, or inhaled corticosteroids, nedocromil sodium, or cromolyn sodium within 4 wk of screening; or a positive test to hepatitis B surface antigen.

Each subject was studied on 3 separate d within a 3-wk period. On the screening day, subjects provided a complete medical history and underwent a comprehensive physical examination, 12-lead electrocardiographic examination, vital signs assessment, routine clinical laboratory tests, pulmonary function tests, and a baseline SO₂ challenge. On two treatment days 5 to 14 d apart, subjects received either a single oral 20-mg dose of zafirlukast or placebo. Treatment was assigned in a random sequence, and administration was double-blind. Subjects were challenged with increasing concentrations of SO₂ at 2 and 10 h after treatment.

SO₂ Challenge

To assess response to SO₂, we measured specific airway resistance (SRaw). Subjects sat in a constant-volume body plethysmograph (18) while we measured airway resistance (Raw) and thoracic gas volume (Vtg) at 30-s intervals to obtain baseline values. Raw was multiplied by Vtg to calculate SRaw. Subjects then inhaled SO₂ (75% relative humidity, 22° C) at doubling concentrations of 0.25, 0.5, 1.0, 2.0, 4.0, and 8.0 ppm for 4 min at each concentration while performing eucapnic hyperventilation at 20 L/min. The concentration of SO₂ causing an increase in SRaw of at least 8 units (PC₈SRaw) was determined after each challenge.

For the SO₂ challenges, a known flow from a calibrated tank of medical-grade SO₂ (500 ppm; Liquid Carbonic Specialty Gas Corporation, San Carlos, CA) was mixed with air delivered at 2.8 L/s from a compressed air source to a 3-L glass mixing chamber. Before entering the mixing chamber, air was passed through a high-efficiency particulate adsorption filter (HEPA) and then humidified by a bubble humidifier. All tubing in contact with the gas mixture was made of Teflon (E. I. DuPont de Nemours & Company, Wilmington, DE). To document that temperature and relative humidity of inspired air were relatively constant, we measured temperature and dew point continuously with a digital humidity analyzer equipped with a platinum temperature probe and a mirrored dew-point hygrometer (Model No. 911; E. G. and G. Dew All, Waltham, MA) and calculated relative humidity from standard tables. We measured SO₂ concentrations continuously from a port just proximal to the subject's mouthpiece with a pulsed fluorescent SO₂ analyzer (Model No. 43; Thermo-Electron Corporation, Bohemia, NY). End-tidal CO₂ percentage was measured by a Beckman LB-1 medical gas analyzer (Beckman Instructions, Palo Alto, CA) that sampled air from the exhalation circuit. Exhaled CO₂ was maintained constant by adding a metered flow of 100% CO₂ to the inspired gas mixture. Tidal volume and minute ventilation were determined by electrically integrating the output from a pneumotachograph (Fleisch No. 3; A. Fleisch, Lausanne, Switzerland) attached to a differential pressure transducer (Validyne MP-45; Validyne Engineering Corp., Northridge, CA). A signal proportional to volume was displayed on an electrical bar graph. Subjects achieved their target minute ventilation (20 L/min) by breathing in time with a metronome at a tidal volume set on the bar graph.

SRaw was measured immediately before the SO₂ challenge (before initiating isocapnic hyperventilation) and after each SO₂ challenge concentration. FEV₁ and FVC were measured before administration of trial medication and before and after 2- and 10-h challenges.

Blood samples were collected before subjects received trial medications and before the 2- and 10-h challenges to determine any association between zafirlukast plasma concentrations and effect. Plasma concentrations of zafirlukast were measured using a high-performance liquid chromatography method with fluorescence detection and a lower quantification limit of 0.75 ng/ml (19).

Safety Assessments

Adverse effects were monitored throughout the trial, and subjects were interviewed for subjective symptoms. Safety was also assessed from the results of clinical laboratory tests, vital signs, and electrocardiographic and physical examinations.

Statistical Methods

On the basis of results from a trial by Bigby and Boushey (8), a sample size of 12 subjects was deemed sufficient to determine statistical significance (p  0.05) between treatments. Treatment with zafirlukast and placebo was compared with respect to the primary efficacy end point, PC₈SRaw, using an analysis of variance (ANOVA) in the framework of a crossover design (20). The PC₈SRaw end point was calculated from challenge data by linear interpolation between two concentrations bracketing the 8-unit increase in SRaw. The ANOVA model included the following factors: treatment sequence, subject number within a treatment sequence, challenge day, and treatment. If the SO₂ challenge failed to produce an 8-unit increase in SRaw, the PC₈SRaw was assumed to be 8 ppm (a conservative estimate).

A regression analysis was used to determine the relationship between zafirlukast plasma concentration and PC₈SRaw, using changes in individual subject's PC₈SRaw and zafirlukast plasma concentration.

Adverse events were summarized by the most recent treatment received using COSTART (Coding Symbols for Thesaurus of Adverse Reaction Terms) terminology. ANOVA was used to determine changes in clinical laboratory tests, vital signs, and electrocardiographic and physical examinations.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All 12 subjects (five women, seven men) completed the trial. The mean age of subjects was 31.3 yr (range: 24 to 43 yr), and their mean prebronchodilator FEV₁ at screening was 86.5% of predicted.

Two hours after treatment with zafirlukast, the SO₂ concentration-response curves were shifted to the right in 10 of 12 subjects (Figure 1). The observed pattern of this shift, showing an increase in the threshold dose but no change in maximum response, is typical for a competitive receptor antagonist (21). Ten hours after treatment, curves were shifted to the right in 10 of 12 subjects (Figure 2). Mean PC₈SRaw for all subjects was significantly higher 2 h after treatment with zafirlukast compared with placebo (3.1 versus 1.5 ppm; p = 0.02). The mean PC₈SRaw remained higher 10 h after treatment, although the difference between treatments was no longer significant (2.7 versus 1.9 ppm; p = 0.09) (Figure 3).

View larger version (27K): [in this window] [in a new window]   Figure 1.   Individual subject responses (change in SRaw) to SO2 challenges conducted 2 h after treatment with zafirlukast (20 mg) and placebo.

View larger version (27K): [in this window] [in a new window]   Figure 2.   Individual subject responses (change in SRaw) to SO2 challenges conducted 10 h after treatment with zafirlukast (20 mg) and placebo.

View larger version (19K): [in this window] [in a new window]   Figure 3.   Comparison of mean SO2 concentration causing an 8-unit increase in specific airway resistance (PC8SRaw) in 12 subjects with asthma 2 and 10 h after treatment with zafirlukast (20 mg) or placebo.

All subjects who received zafirlukast had quantifiable plasma concentrations. Two hours after dosing, plasma levels were > 100 ng/ml in 11 of 12 subjects (in the remaining subject, the level was 11.8 ng/ml). By 10 h after dosing, plasma levels fell, as expected. In a previous pharmacokinetic trial using a single 20-mg dose of zafirlukast, the median time to maximum plasma concentration (T_max) was 3 h (range: 1.5 to 6 h) after dosing (unpublished observations). In the present study, there was a poor correlation (r ² = 0.44; p = 0.07) between plasma concentration of zafirlukast and PC₈SRaw at 2 h, which may reflect the nearly uniform high plasma concentrations of zafirlukast or the fact that this time point was during the absorption phase of the concentration-time profile and may not reflect the amount of drug available in the lung. At 10 h, when plasma concentrations approximate trough levels, there was an association between plasma concentrations and PC₈SRaw (r² = 0.76; p = 0.001). Smith and associates described a similar relationship between zafirlukast plasma concentration 12 to 14 h after dosing and inhibition of bronchoconstriction produced by inhaled LTD₄ (22).

Safety

Zafirlukast was well tolerated by all subjects. Four subjects (most recent treatment: two zafirlukast, two placebo) had a total of five adverse events (back pain, pelvic pain, and three cases of pharyngitis) during the trial; none of these events were serious. No clinically significant changes were observed in results of clinical laboratory tests, vital signs, or electrocardiographic and physical examinations.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A single, oral, 20-mg dose of zafirlukast increased the concentration of SO₂ needed to produce an 8-unit increase in SRaw at both 2 and 10 h after administration, with a statistically significant difference between zafirlukast and placebo 2 h after treatment. The lack of statistical significance at 10 h after administration of zafirlukast, despite an increased mean PC₈SRaw, probably reflects not only the lower plasma concentrations of zafirlukast at 10 h but also the normal diurnal variation in bronchial responsiveness. Because both challenges were performed on the same study day, the 2-h challenge was completed early in the morning, when bronchial reactivity is greatest. Conversely, the 10-h challenge was completed in late afternoon, when bronchial reactivity is lower. In fact, the mean PC₈SRaw after placebo was greater at 10 h (1.9 ppm) than at 2 h (1.5 ppm), demonstrating this diurnal pattern.

The concentration of SO₂ in urban areas has been linked to respiratory symptoms and function in several epidemiologic studies (23). Dramatic increases in mortality occurred during severe episodes of air pollution in the Meuse Valley, Belgium; Donora, Pennsylvania; and London, England. Estimates suggest that the ambient concentrations of SO₂ during these episodes were very high (23, 27, 28).

The ability of inhaled SO₂, a common air and industrial pollutant, to produce bronchoconstriction has been recognized for decades. Brief exposure to SO₂ concentrations of 5 ppm or greater increases airway resistance in healthy human volunteers (1, 2, 29). The high concentrations of SO₂ required to induce bronchoconstriction in normal volunteers are greater than concentrations found in the most polluted urban environments.

Sheppard and colleagues reported in 1980 that subjects with mild asthma develop bronchoconstriction at a lower threshold concentration of SO₂ and with greater magnitude than do nonasthmatic subjects (1). Additionally, Sheppard and colleagues suggested that because of their increased responsiveness to inhaled SO₂, subjects with asthma may develop clinically significant bronchoconstriction at ambient SO₂ concentrations considerably lower than those allowed by regulations governing the workplace and environment. Subsequent studies have demonstrated that moderate exercise increases the bronchomotor effect of SO₂ in subjects with asthma, so that concentrations as low as 0.1 ppm can cause significant bronchoconstriction (30, 31). Our study was designed to model a realistic "real-world" exposure to SO₂. Using eucapnic hyperventilation to mimic increased minute ventilation, as occurs with exercise, subjects breathed concentrations of SO₂ commonly equaled or exceeded in polluted urban or occupational environments.

The mechanism by which SO₂ produces bronchoconstriction has not been confirmed. Early studies demonstrated the importance of parasympathetic reflex pathways in the bronchoconstrictor response to SO₂ (1, 2), and more recent studies with agents that inhibit the release of mediators from mast cells (7, 8) suggest that SO₂ acts by stimulating mast cell secretion. In challenge experiments similar to those in the present study, Bigby and Boushey reported a 2-fold increase in PC₈SRaw after pretreatment with nedocromil sodium (8). In contrast, Wiebicke and coworkers reported that inhaled beclomethasone, at a dose that reduces histamine and methacholine responsiveness, has no effect on SO₂-induced bronchoconstriction (32). The effects of cromolyn sodium and nedocromil sodium can be explained by their effect on mast cells or by a direct inhibitory effect on neural reflex pathways. In fact, nedocromil sodium has been shown to have an effect on chloride transport in sensory nerves (33). Our results, demonstrating the inhibition of SO₂-induced bronchoconstriction by a leukotriene receptor antagonist, suggest that release of leukotrienes contributes to the response to SO₂. However, our results do not allow us to assess the relative importance of the direct effect of leukotrienes on airway smooth muscle versus an indirect effect on muscarinic responsiveness and neuropeptide release, as has been reported in other species (34, 35). Now that experimental data support the involvement of leukotrienes, studies can be designed with both an anticholinergic agent and a leukotriene receptor antagonist, to evaluate the contribution of each.

In summary, we have shown that a single oral dose of the leukotriene receptor antagonist zafirlukast inhibits SO₂-induced bronchoconstriction in subjects with mild-to-moderate asthma. Therefore, we conclude that SO₂-induced bronchoconstriction involves leukotriene release.