INTRODUCTION

TOP ABSTRACT INTRODUCTION CONCLUSION DISCUSSION REFERENCES

Zafirlukast is a potent and selective cysteinyl leukotriene (cysLT) antagonist that has been approved for the prophylactic treatment of asthma. It is the first member of a new drug family available for the management of asthma in more than 20 years.

It is well established that treatment of the underlying inflammation of asthma is critical to successful management (1). By blocking the pro-inflammatory actions of the cysLTs, zafirlukast targets disease components that are untreated by drugs that only relieve asthma symptoms. Further, zafirlukast does not have the limitations seen with other classes of anti-asthma drugs. For example, poor patient compliance is often a problem with anti-asthma drugs that are administered by inhalation (2). Because zafirlukast is an orally active drug that is administered twice daily, problems with patient compliance are expected to be minimized.

Clinical trials with zafirlukast can be divided into three types: (1) studies to explore its pharmacokinetic-pharmacodynamic relationships; (2) studies to define its activity in clinical models of asthma in which cysLTs are endogenously produced; and (3) multicenter trials to establish the agent's safety and efficacy in chronic, stable asthma. The objective of this paper is to review clinical trials performed with zafirlukast to date.

	PHARMACOKINETICS AND PHARMACODYNAMICS OF ZAFIRLUKAST

Single-dose Pharmacokinetics and Pharmacodynamics

The single-dose pharmacokinetics of zafirlukast were determined in several trials involving patients who received 5-, 10-, 20-, or 40-mg tablets. Zafirlukast was rapidly and completely absorbed, reaching maximal plasma concentrations (t_max) by 3 h.

Relative to an oral solution, the bioavailability of the tablets was 100%, based on the area under the curve (AUC) of the plasma zafirlukast concentration-versus-time curve. Maximum plasma concentrations (C_max) were proportional to the dose administered, with the plasma zafirlukast concentration-versus-time curve best fitted by a two-compartment model (Figure 1). Zafirlukast was widely distributed, with a volume of distribution (V_d) of 293 L. More than 99% of zafirlukast was protein-bound when plasma concentrations were 0.25 to 10 µg/ml.

View larger version (26K): [in this window] [in a new window]   Figure 1.   Single-dose pharmacokinetics of zafirlukast. Plotted is the zafirlukast plasma concentration versus time curve for normal subjects who received 5-, 10-, 20-, 40-, or 50-mg tablets of zafirlukast.

In these single-dose pharmacokinetic studies, the elimination half-life (t_1/2) was 8.7 h, which might suggest the need for three times per day dosing. However, the pharmacodynamic effects of zafirlukast were longer lasting. In a double-blind, placebo-controlled, crossover study, normal subjects were challenged with aerosol leukotriene D₄ (LTD₄) at 2, 12, or 24 h after administration of a single 40-mg dose (3). The dose of LTD₄ that reduced specific airway conductance (SGaw) by 35% was measured. Zafirlukast increased the LTD₄ dose required to reduce SGaw by 35% 117-fold, from 34 µg/ml to 3,965 µg/ml. With 12-h and 24-h pretreatments, zafirlukast increased in the LTD₄ dose that produced a 35% decrease in SGaw by ninefold and fivefold, respectively. Plasma zafirlukast levels correlated with efficacy across the pretreatment groups, but imperfectly within each group.

A double-blind, placebo-controlled, two-period, crossover trial was performed in asthma patients who received zafirlukast doses of 5, 10, 20, 40, or 100 mg 12 h prior to LTD₄ challenge (4). Mean baseline FEV₁ for each group of subjects ranged from 86 to 95% of predicted. The provocative concentration of LTD₄ that reduced FEV₁ by 20% (LTD₄ PC₂₀FEV₁) was measured. For patients receiving 10, 40, or 100 mg zafirlukast, the LTD₄ PC₂₀FEV₁ increased 10-fold or more. An association was observed between the plasma zafirlukast concentration and the drug's protective effect. Thus, although the pharmacokinetic t_1/2 was approximately 8 h, the pharmacodynamics demonstrated drug activity 12 h after dosing, consistent with the twice-daily dosing regimen.

Multiple-dose Pharmacokinetics

During twice-daily dosing with zafirlukast, steady-state plasma concentrations were proportional to the dose and were predictable from the single-dose pharmacokinetic data. In these studies, the elimination t_1/2 was 10 h (5).

The pharmacokinetics of zafirlukast in children (ages 7 to 11 yr), adolescents (ages 12 to 17 yr), and adults with asthma were similar to those of normal adult subjects. In addition, the pharmacokinetics of zafirlukast in men and women were similar after adjusting for body weight differences and were not affected by race. In patients older than 65 yr of age and in patients with biopsy-proven cirrhosis, there was an approximately twofold increase in the C_max and AUC when compared with normal subjects. The pharmacokinetics were unaffected in patients with renal impairment relative to normal control subjects (5).

When zafirlukast was administered with food, its bioavailability was more variable. In 75% of subjects, there was a net reduction in bioavailability of approximately 40%. Thus, zafirlukast should not be taken with meals (5).

Disposition and Metabolism

Zafirlukast is extensively metabolized in humans. It undergoes hydroxylation, hydrolysis, and N-acetylation to yield plasma metabolites that are at least 90-fold less potent than the parent drug as cysLT₁ receptor antagonists. The CYP2C9 isozyme of the cytochrome P450 system is responsible for the methyl hydroxylation of zafirlukast.

Following administration of radiolabeled zafirlukast, approximately 10% of the dose was excreted in the urine, with the remainder excreted in the feces. Zafirlukast itself was not detected in the urine. Zafirlukast is excreted in breast milk; following 40 mg twice per day dosing (two times the labeled dosage), the average steady-state zafirlukast concentration in breast milk was 50 nanograms (ng)/ml, compared with 255 ng/ ml in plasma. Zafirlukast is therefore contraindicated in nursing mothers (5).

Drug-Drug Interactions

The CYP2C9 isozyme of P450 metabolizes zafirlukast as well as nonsteroidal anti-inflammatory drugs, warfarin (primarily S-warfarin), and tolbutamide. The CYP1A2 isozyme metabolizes theophylline and caffeine; the CYP3A4 isozyme metabolizes terfenadine. The potential for drug-drug interactions between these drugs and zafirlukast was assessed in a series of studies.

• When aspirin was coadministered with zafirlukast, plasma zafirlukast concentrations increased by approximately 45%, whereas plasma aspirin levels were not significantly affected. In contrast, co-administration of zafirlukast with erythromycin produced an approximately 40% decrease in plasma zafirlukast concentration (5).
• Several studies were performed to study potential interactions between zafirlukast and terfenadine. Zafirlukast did not affect the pharmacokinetics of terfenadine nor did it affect electrocardiograms or increase the QT interval (5).
• When zafirlukast and racemic warfarin were administered in combination, the AUC of the prothrombin time curve was increased, and mean prothrombin times increased 35%, compared with warfarin plus placebo. Product labeling therefore warns that, if the two drugs are coadministered, prothrombin times should be monitored and warfarin dosing modified accordingly (5, 8, 9).
• Zafirlukast may be safely administered with a number of other therapies that are routinely used in the treatment of asthma and allergy. These drugs include -agonist and anticholinergic bronchodilators, antihistamines, and inhaled corticosteroids. When administered with theophylline, plasma zafirlukast levels declined approximately 30%. More importantly, however, zafirlukast did not affect plasma theophylline levels (5).
• Zafirlukast did not affect serum levels of estradiol, progesterone, follicle-stimulating hormone, or luteinizing hormone in women using oral contraceptives over a 30-d period (5).

	EFFECT OF ZAFIRLUKAST IN CLINICAL MODELS OF ASTHMA

Early clinical studies established that zafirlukast can potently antagonize aerosolized LTD₄-induced bronchconstrictiona model in which exogenous LTD₄ is administered. In the next series of clinical studies, a primary objective was to ascertain whether zafirlukast would antagonize the effects of endogenously released cysLTs triggered by a variety of physiologic and pharmacologic stimuli.

Allergen-induced Bronchoconstriction in Patients with Atopy and Asthma

A double-blind, placebo-controlled, crossover trial was performed in 10 atopic patients with asthma who exhibited at least a 15% fall in FEV₁ and a doubling-dose fall in the PC₂₀FEV₁ following allergen challenge (10). These patients were allergic to either grass pollen or house-dust mite. The patients received a 40-mg dose of zafirlukast 2 h before allergen challenge. Zafirlukast significantly reduced allergen-induced early- and late-phase bronchoconstriction and suppressed the allergen-induced increase in histamine reactivity.

A second trial was performed in 12 atopic individuals with sensitivity to cat dander (11). In this study, the dose of inhaled allergen that produced a 20% decrease in FEV₁ (allergen PD₂₀FEV₁) was measured. In placebo-treated patients, the mean allergen PD₂₀FEV₁ was 460 AU/ml. However, when subjects were pretreated with 40 mg zafirlukast, the allergen PD₂₀FEV₁ was increased to 6,996 AU/ml. Of the 12 patients, eight were able to tolerate 3- to 30-fold more allergen. Since higher allergen challenge doses would be expected to release greater amounts of cysLTs, this trial design is a more rigorous test than studies performed at a single allergen dose.

Similarly, a baseline allergen PD₂₀FEV₁ was determined in 10 males with mild allergic asthma; patients were then randomized into a double-blind, placebo-controlled, crossover trial (12). The allergen PD₂₀FEV₁ after placebo treatment was similar to that determined in the initial screening. However, after 20 mg zafirlukast, the median allergen PD₂₀FEV₁ was 5.5-fold higher (Figure 2). In addition, zafirlukast shortened the recovery time to reach within 90% of baseline FEV₁ following allergen challenge from 60 min to 40 min.

View larger version (19K): [in this window] [in a new window]   Figure 2.   Patients with atopy and asthma treated with zafirlukast tolerate more allergen. The provocative dose of allergen that reduced FEV1 by 20% (PD20FEV1) was determined in 10 patients receiving placebo (white bar) or 20 mg zafirlukast (black bar). The PD20FEV1 at a control challenge is also depicted. The group geometric means ± SD are plotted, showing statistically significant improvement (p = 0.01) provided by zafirlukast. Reprinted from Reference 12 by permission.

Exercise-induced Asthma

The bronchoconstrictor response following exercise challenge was evaluated in eight asthma patients treated with either placebo or 20 mg zafirlukast in a double-blind, crossover trial (13). The challenge consisted of 6 min of exercise on a treadmill at constant speed and gradient while the patient inspired dry air at room temperature. Zafirlukast did not affect baseline FEV₁, which was in the normal range for most of the subjects before the exercise challenge. However, zafirlukast blunted the maximum fall in FEV₁ following exercise; the mean decrease in FEV₁ was 36% after placebo and 21.6% after zafirlukast (Figure 3).

View larger version (18K): [in this window] [in a new window]   Figure 3.   Zafirlukast reduces exercise-induced bronchoconstriction. Eight patients with asthma received either placebo (open squares) or 20 mg zafirlukast (closed squares) before a 6-min exercise challenge. Exercise-induced bronchoconstriction is plotted as a percentage of the pre-exercise baseline FEV1 for a 30-min period following the start of exercise. Reprinted from Reference 13 by permission.

Cold Air-induced Bronchoconstriction

The use of zafirlukast in preventing bronchoconstriction in response to cold-air exposure was investigated in a randomized, double-blind, placebo-controlled, three-period crossover trial in a population of 24 patients with asthma (14). Baseline criteria included FEV₁ greater than or equal to 60% predicted and evidence of a 20% reduction in FEV₁ on cold air bronchoprovocation. During each treatment period, subjects received a single dose of zafirlukast (20 or 40 mg) or placebo; dosing was followed 2 and 8 h later by bronchoprovocation with cold dry air. At the 8-h time point, statistically significant differences were observed between drug and placebo among treatments in the quantities of cold air required to decrease FEV₁ by 10%. Improvements relative to placebo were 29% and 32% for the 20 and 40 mg doses, respectively (p  0.05). No statistically significant differences were observed at the 2-h time point.

In a second trial, a single 80-mg dose of zafirlukast was administered to 10 patients with mild-to-moderate asthma (15), showing that treatment significantly shortened recovery times after cold-air bronchoprovocation measured at 30 min and 4 h (p < 0.05 and p < 0.01, respectively) and attenuated cold air- induced bronchoconstriction 24 h after treatment (p < 0.05).

Other Clinical Models

Platelet-activating factor (PAF) elicits bronchoconstriction in humans via the release of endogenous cysLTs (16). Following the administration of either 40 mg zafirlukast or placebo, eight male subjects were challenged with 45 µg of nebulized PAF; reductions in specific airway conductance (SGaw) were then measured. The mean maximal PAF-induced fall in SGaw was 43.1% after placebo, but only 17.5% after zafirlukast. This represented a 59% inhibition of PAF-induced bronchoconstriction. PAF-induced neutropenia was not affected by zafirlukast and is not likely to be mediated by cysLTs.

In several additional studies, zafirlukast was administered by inhalation before challenge with exercise (17), antigen (18, 19), or cold-air bronchoprovocation (20). In each trial, inhaled zafirlukast reduced the bronchoconstrictor response to challenge. However, data obtained with oral zafirlukast suggest that this mode of administration is capable of attenuating both the early- and late-phase bronchoconstrictive response to allergen. For example, the bronchoconstrictive response to allergen was significantly reduced following single-dose oral zafirlukast administration (40 mg) during both the early- (0- 120 min) and late-phase (120-360 min) response periods (10). This result was supported by a recent study in which allergic patients with mild asthma were challenged with allergen via direct segmental instillation. When patients were treated with oral zafirlukast for 7 d (160 mg twice per day) prior to challenge, the influx of basophils, mast cells, and eosinophils was significantly reduced 48 h later, compared with placebo treatment (21). The failure of inhaled zafirlukast to reduce the late-phase response is unexplained and may be the result of experimental variables that were not considered or a difference in the potency of inhaled and oral formulations.

	MULTICENTER TRIALS OF ZAFIRLUKAST IN CHRONIC, STABLE ASTHMA

The ability of zafirlukast to inhibit allergen- and exercise- induced bronchoconstriction proved that the drug could inhibit the effects of cysLTs released endogenously by physiologic stimuli in a controlled, clinical laboratory setting. The next clinical studies involved patients with mild-to-moderate asthma who were treated for 6 wk or longer with zafirlukast or placebo. The goal of these trials was to assess the efficacy and safety of zafirlukast in the management of asthma.

13-Week Trial in Mild Asthma: Effect on Daytime Asthma Symptoms

A 13-wk, randomized, double-blind, placebo-controlled, multicenter trial was designed to ascertain if 20 mg zafirlukast administered twice per day could reduce daytime asthma symptom scores (22). Patients with mild asthma who received -agonists on an as-needed basis and had FEV₁ values of at least 55% of predicted 6 h after -agonist use comprised the study population. Study criteria dictated no upper limit for percent predicted FEV₁. Inclusion criteria further dictated that patients had weekly symptom scores of 8 or greater. The daytime asthma symptom score was based on a daily score ranging from 0 (no symptoms) to 3 (severe symptoms); thus, the maximal weekly score was 21. Following randomization, the zafirlukast and placebo groups had similar baseline characteristics. The average weekly asthma symptom score was 11.2 or approximately 50% of the maximum response, and baseline FEV₁ values for the zafirlukast and placebo groups were 76% and 78% of predicted, respectively (22).

Both placebo and zafirlukast reduced daytime asthma symptom scores; however, the effects of zafirlukast were significantly greater than the observed placebo effect at each week of the trial (Figure 4A). Maximal effects were achieved after 3 to 5 wk of therapy and were maintained for the duration of the trial. At the trial endpoint, the symptom score decreased 26.5% relative to baseline in the zafirlukast group and 13.3% in the placebo group (22). In addition, zafirlukast significantly reduced the frequency of nighttime awakenings (p < 0.05) and morning asthma symptoms (p < 0.01).

View larger version (21K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   Figure 4.   Zafirlukast reduces daytime asthma symptoms and decreases -agonist use. Patients with mild asthma were treated with either placebo or 20 mg zafirlukast twice per day for 13 wk. The effect of treatment on daytime symptom score (A) and 2-agonist use (B) are plotted on a weekly basis. Reprinted from Reference 22 by permission.

For each week of treatment and at the endpoint of the 13-week trial, zafirlukast reduced -agonist use relative to placebo (3.91 puffs per day versus 3.14 puffs per day). Among patients in the zafirlukast group, -agonist use decreased 22.3% (p < 0.01) over the course of the trial; in contrast, -agonist use increased 7% among patients receiving placebo (Figure 4B). Pulmonary function was quantitated by measuring FEV₁ and morning peak expiratory flow rate (PEFR). In these patients, who had only mild decrements in baseline pulmonary function, zafirlukast produced small, but statistically significant, improvements in both indices (p  0.05 and p < 0.01, respectively) (22).

Adverse events were comparable in the zafirlukast-treated and placebo groups (68% versus 65%). The most commonly reported events were pharyngitis and headache. The only adverse event that reached statistical significance compared with placebo (p < 0.05) was asthma exacerbation, which occurred in 6.5% of patients in the placebo group but only 2.7% of patients who received zafirlukast (22). There were not statistically significant differences between the two treatment groups in any of the clinical parameters monitored. Inasmuch as elevated hepatic enzymes have been noted in a number of trials with antileukotriene agents, alanine transaminase levels were measured as part of the safety monitoring for this trial. Clinically silent liver function test abnormalities were observed in both treatment groups (4% of patients receiving placebo versus 3.3% of patients receiving zafirlukast); the frequency of occurrence of alanine transaminase elevations to levels higher than two times the upper limit of the normal range was similar in the two groups (2.8% of patients who received placebo versus 3.1% of patients receiving zafirlukast).

In a subprotocol of this trial, additional data on medical examinations, patient questionnaires, and daily diaries were collected on 146 patients (103 receiving zafirlukast, 43 receiving placebo) in an effort to determine clinical effectiveness (23). Clinical outcomes measures included days per month without asthma symptoms, activity limitations. -agonist usage, sleep disturbance, asthma episodes, and adverse events; economic outcome measures included frequency and type of unscheduled health care contacts, use of -agonists, use of nonasthma medications, and days absent from work or school. Relative to patients who received placebo, patients receiving zafirlukast (20 mg twice daily) had nearly twice as many days without symptoms (89%; p = 0.03), -agonist usage (89%; p = 0.001), or an episode of asthma (98%; p = 0.003). Furthermore, they missed only about half as many work or school days (55%; p = 0.04), had 55% fewer health care contacts (p = 0.007), and used 19% less nonasthma medications (p > 0.2) than did patients who received placebo. Of interest was the fact that these improved outcomes occurred in the face of decreased "rescue" -agonist usage; 17% fewer canisters of inhaled -agonists were used by patients taking zafirlukast (p = 0.17).

6-Week Dose-ranging Trial in Mild-to-Moderate Asthma

A 6-wk, placebo-controlled, dose-ranging study was designed to assess the effects of zafirlukast on a number of clinical endpointsdaytime and nighttime asthma symptoms, -agonist use, and pulmonary function in patients with mild-to-moderate asthma (24). The study population received -agonists on an as-needed basis and had FEV₁ values between 40 and 75% of predicted. Their weekly daytime asthma symptom scores at baseline were 10 or higher. In this study, zafirlukast was administered twice per day at doses of 5, 10, or 20 mg; each treatment group had 66 or 67 patients for analysis of drug efficacy. The average baseline FEV₁ values in the placebo and 40 mg/ day groups were 69% and 66%, respectively. These patients had had asthma for a mean duration of about 20 yr.

Of the three zafirlukast doses, 40 mg/d had the most consistent effect on the overall clinical endpoints (24). By the 6-wk endpoint, zafirlukast treatment had significantly reduced the asthma symptom score by 27%, morning asthma symptoms by 28%, and nighttime awakenings by 46%. In contrast, the asthma symptom score and morning asthma symptoms were reduced by 13% and 10%, respectively, in placebo-treated patients. Nighttime awakenings increased 4% in the placebo group.

-Agonist use declined 31% in the zafirlukast group and 15% in the placebo group by the end of the trial. While a placebo effect was apparent, zafirlukast was still significantly better than placebo at week 1 and for each week thereafter (Figure 5A). Zafirlukast also improved pulmonary function in these patients, an effect that was evident after 2 wk and was maintained until the end of the trial (Figure 5B). At the 6-wk endpoint, zafirlukast had significantly increased FEV₁ by 11%, whereas placebo was without effect (p  0.01). The effects of zafirlukast on morning and evening PEFR were less pronounced.

View larger version (17K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 5.   Zafirlukast reduces -agonist and improves pulmonary function. Patients with mild-to-moderate asthma were treated with either placebo (n = 15) or 20 mg zafirlukast (n = 15) twice per day for 6 wk. Plotted are the changes in -agonist use (A) and FEV1 (B) during the double-blind treatment period.

No serious adverse events occurred during the 6-wk trial; adverse events such as headache, gastritis, pharyngitis, and rhinitis were reported in 63% of subjects who received placebo, but only 53% of those receiving active treatment with zafirlukast (24). Among the 19 patients in the trial (10 in the treatment group and nine in the placebo group) who had events that were related or possibly related to treatment, only one was judged to be treatment-relateda patient in the placebo group who had an episode of asthma exacerbation. With respect to clinical laboratory parameters, there were no differences among the treatment groups, except for elevations in SGPT levels, which occurred in 4% of subjects in the 40-mg and placebo groups (24).

In comparing this study with the 13-wk trial, zafirlukast produced similar improvements in asthma symptoms and -agonist use, both in patients with mild asthma who participated in the 13-wk study and in patients with more severe symptoms. However, zafirlukast had a greater quantitative effect on FEV₁ among patients who initially demonstrated more severe airway obstruction (22, 24).

13-Week Dose-ranging Trial in Mild Asthma

A 13-week, double-blind, placebo-controlled, multicenter, dose-ranging study was designed to evaluate several zafirlukast dosing regimens (25). The population was similar to that of the other 13-wk trialpatients who had mild asthma, who received -agonists on an as-needed basis, and who had FEV₁ values of at least 55% of predicted and weekly asthma symptom scores of 8 or greater. There were six treatment groups; patients were randomized to receive twice daily administration of zafirlukast at doses of 4, 10, 20, 40, or 80 mg or placebo.

In patients receiving twice-daily zafirlukast, there was a dose-dependent reduction in as-needed -agonist use, with patients randomized to the higher doses showing the greatest reductions (25) (Figure 6). An improvement in morning symptoms was even observed at the lowest doses of 4 and 10 mg twice per day; increasing the dose produced additional decreases in symptoms without increasing adverse events or abnormalities in clinical laboratory tests (25).

View larger version (20K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 6.   Dose-response relationship of zafirlukast on -agonist use and morning asthma symptoms. Patients with mild asthma were treated with either placebo or the indicated dose of zafirlukast for 13 wk. Changes from baseline in 2-agonist use (A) and morning symptoms (B) are plotted.

Retrospective analysis of an early clinical trial suggested that patients with more severe asthma experience greater benefit from leukotriene receptor antagonism than those with less severe disease (Figure 7). A recently published report appears to corroborate these results (26). Subgroup analysis was conducted on integrated data from three 13-wk multicenter trials with comparable designs, entry criteria, and clinical parameters, as well as demographics and baseline asthma characteristics. The trials included more than 1,000 patients with mild-to-moderate asthma. Daytime and nocturnal symptoms, -agonist usage, and pulmonary function (PEFR, FEV₁, and percent predicted FEV₁) were assessed in the various subgroups of patients receiving 20 mg zafirlukast twice daily or placebo. Particularly with respect to nocturnal symptoms and the various measures of pulmonary function, larger effect sizes were seen for patients receiving zafirlukast who had greater room for improvement, i.e., patients whose nocturnal symptoms were more severe and those with more compromised pulmonary function.

View larger version (21K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Figure 7.   Effect of zafirlukast after grouping patients by disease severity. Patients were stratified into three groups according to their baseline FEV1: <65%, 65-80%, and >80% of predicted (A). The mean change in FEV1 after zafirlukast or placebo treatment is plotted. Patients were also stratified according to their baseline -agonist use (B). The mean change in -agonist use by patients treated with zafirlukast or placebo is shown.

Zafirlukast, Cromolyn, and Placebo

Two published abstracts have reported the results of 13-wk trials comparing the use of zafirlukast, cromolyn, or placebo in patients with mild-to-moderate asthma (27, 28). The two studies shared similarities in their designs, but specific protocols, study parameters, entry criteria, and cromolyn dosages differed in the two trials. Nevertheless, each trial showed that active treatment was superior to placebo with respect to several endpoints and interim measurements (e.g., symptom scores, ₂-agonist usage). Neither of the studies was sufficiently powered, however, to detect differences between active treatment groups. It is therefore not surprising that statistically significant differences between zafirlukast and cromolyn were not observed. Additionally, methodologic differences precluded more definitive direct comparisons between the two agents.

Possible Anti-inflammatory Effects of Zafirlukast

A 7-d, double-blind, placebo-controlled, crossover trial was designed to probe possible anti-inflammatory effects of 20 mg zafirlukast administered twice per day (29). The 11 patients included in the trial had mild allergic asthma with at least two positive skin tests and baseline FEV₁ values greater than 70% of predicted. After 5 d of therapy, the patients underwent segmental antigen challenge (SAC) followed immediately, and again 48 h later, by bronchoalveolar lavage (BAL). Fluid taken from the first BAL immediately after SAC revealed no effect of zafirlukast on total or differential cell counts or the levels of histamine, cysLTs, tumor necrosis factor alpha (TNF-), or interferon gamma (IFN-) in the BAL fluid. The BAL albumin levels after zafirlukast (8.7 µg/ml) treatment were lower than the levels observed after placebo (12.1 µg/ml), but there was considerable intersubject variability.

On day 7, 48 h after SAC, interesting effects were noted. Zafirlukast significantly reduced the number of basophils and lymphocytes, while exhibiting a trend for reducing the number of alveolar macrophages, eosinophils, and total leukocytes present in the BAL fluid. While the eosinophil count exhibited a downward trend, there was no effect on levels of major basic protein and eosinophil cationic protein, two products that cause airway epithelial desquamation (29). Zafirlukast significantly blunted the increase in TNF- in BAL fluid observed at 48 h after SAC, possibly due to decreased alveolar macrophage activation. Finally, phorbol ester-induced superoxide release from ex vivo alveolar macrophages isolated from BAL fluid was significantly blunted by zafirlukast treatment, again suggesting an effect on decreasing macrophage activation. Thus, in this study, zafirlukast reduced several cellular and mediator indices of inflammation following allergen challenge.

A recent trial using a similar design found significant decreases in the influx of inflammatory cells at 48 h postchallenge among patients treated with zafirlukast (21). Basophils again declined with zafirlukast treatment (160 mg twice daily) relative to the values obtained from patients receiving only placebo (from 0.30 thousand/ml to 0.13 thousand/ml [p < 0.001]). Eosinophil numbers also decreased, from 70.5 thousand/ml to 39.0 thousand/ml (p < 0.05) when patients received zafirlukast. Furthermore, superoxide release from alveolar macrophages also decreased significantly (p < 0.05). This report thereby confirms and extends earlier data; zafirlukast exhibits anti-inflammatory activity in that it reduces inflammatory cellular influx into the airways and decreases cellular (macrophage) activation.

Lastly, a very recent preliminary report described the use of zafirlukast in a population of 386 patients with asthma who remained symptomatic despite treatment with high doses of inhaled corticosteroids (mean dosage of 1,600 µg/d) (30). Patients were randomized to receive either zafirlukast (80 mg twice daily) or placebo in addition to their usual therapy. After 6 wk of treatment, patients who received zafirlukast exhibited significantly higher morning PEFR than did patients who received placebo (18.7 L/min versus 1.5 L/min, respectively; p < 0.001), and there was evidence of incremental improvements each week of the trial. There was no evidence of any increases in adverse events in the zafirlukast group, and only about half as many patients in the zafirlukast-treated group as in the placebo group (8 versus 15% of patients) experienced worsening asthma over the course of the trial that necessitated a change in therapy.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION CONCLUSION DISCUSSION REFERENCES

Clinical studies demonstrate that zafirlukast is safe and effective for the prophylactic treatment of asthma. Zafirlukast has been well tolerated, demonstrating rates of adverse events comparable to placebo. The most commonly reported adverse events have been headache, infection, nausea, and diarrhea, each occurring with only slightly greater frequency among patients who received zafirlukast than among those receiving placebo (5). Clinical laboratory abnormalities have been rare. At the approved dosage, elevations in hepatic enzymes have also been comparable to the values obtained from patients receiving placebo, although elevated hepatic enzyme levels have been observed among patients receiving very high dosages (80 mg twice daily), and rare cases of hepatitis and hyperbilirubinemia in patients receiving zafirlukast have been reported (5).

In the months since zafirlukast became widely available, a concern gradually arose in response to several reported cases of eosinophilic conditions, including Churg-Strauss syndrome, that came to light during routine postmarketing surveillance (5). Although no cases of Churg-Strauss syndrome or eosinophilic conditions had been found among any of the patients in the zafirlukast clinical trials, it should be noted that the patient populations differedthe trials were conducted among patients with mild-to-moderate asthma, not among those with more severe disease who required systemic corticosteroids. Only when "real world" usage of the drug commenced (i.e., use in patients with severe, oral steroid-dependent asthma), did these highly unusual patient presentations emerge.

To date, 12 cases of probable Churg-Strauss syndrome have been reported; of these, four have histologic confirmation. Because the incidence of Churg-Strauss syndrome in the general population is thought to be extremely low, any reports of the condition in association with a new pharmaceutical need to be taken seriously. In a recently published report, Wechsler and colleagues (31) described eight cases of Churg-Strauss syndrome among patients who received zafirlukast; these patients were previously being treated with high-dose systemic corticosteroids, and zafirlukast had been introduced as a means by which their corticosteroid dosage might be tapered. The current product label for zafirlukast stipulates that "caution is required when oral steroid reduction is being considered" (5). Importantly, postmarketing surveillance has identified no cases of Churg-Strauss syndrome among patients with mild-to-moderate asthma, the approved indication for zafirlukast.

There are no data to suggest that the incidence of Churg-Strauss syndrome among patients receiving zafirlukast is any different than that among patients receiving other asthma therapies. Continuing postmarketing surveillance and focused analysis of existing data may shed additional light on this critical question.

With respect to the effects and efficacy of zafirlukast, pharmacodynamic studies demonstrate that it antagonized LTD₄-induced bronchoconstriction, as evidenced by more than a 100-fold shift of the LTD₄ dose-response curve. Significant antagonism was still evident 12 h after administration of zafirlukast, consistent with the drug's twice-daily dosing regimen. Zafirlukast was effective in a number of clinical models of asthma, including allergen-induced bronchoconstriction in atopic asthma and exercise-induced asthma. Thus, zafirlukast effectively blocks the effects of cysLTs, whether administered exogenously or released endogenously in response to physiologic or pharmacologic stimuli.

The pharmacokinetics of zafirlukast were independent of both time and dose, with little drug accumulation in plasma after twice-daily dosing. Peak plasma drug concentrations occurred by 3 h after administration, and the bioavailability of zafirlukast tablets was 100%, relative to an oral solution.

The results of 6- to 13-wk trials demonstrate that zafirlukast effectively reduces asthma symptoms, decreases as-needed -agonist use, and improves pulmonary function. The latter effect is especially evident in patients with significant airway obstruction at baseline. From the dose-ranging studies, it was determined that the 20-mg twice-daily dose is efficacious, as judged by each of these clinical endpoints.

In order to ascertain whether zafirlukast has anti-inflammatory activity, SAC followed by BAL was performed in patients with both allergies and asthma. Zafirlukast produced significant effects on BAL cell counts and appeared to decrease the activation of alveolar macrophages purified from BAL fluid. These preliminary results suggest that zafirlukast does have anti-inflammatory properties.

The recently revised "Guidelines for Asthma Diagnosis and Management" published by the National Institutes of Health recommend that antileukotriene agents can be used as an alternative primary controller agent in patients 12 yr of age or older with mild persistent asthma (32). These agents may be particularly useful in patients who do not respond well to or who cannot use standard asthma therapies or in whom inhaled steroids are associated with local adverse effects. Further clinical experience will be necessary to fully delineate the place of antileukotrienes in clinical practice.

	DISCUSSION

TOP ABSTRACT INTRODUCTION CONCLUSION DISCUSSION REFERENCES

Busse: What do other compounds, such as inhaled corticosteroids and -agonists, do after segmental bronchoprovocation? I understand that it has been studied with inhaled corticosteroids. I think that you need some landmarks for comparison of the zafirlukast data. In general, this form of challenge is a major stimulus in the airway, and it takes an awful lot to block it.

Calhoun: There are data available with 50 mg oral prednisone. It blocks the influx of both basophils and eosinophils to about the same degree.

Cohn: In addition, Dr. Steven Peters used rhinitis and asthma patients in a segmental antigen challenge study with zileuton. He showed a significant drop in basophil influx. I'd like to also mention some anecdotal information from a zileuton study. They compared zileuton plus low-dose inhaled steroid with a double dose of inhaled steroid. There were equivalent outcomes in lung function data, symptomatology, and -agonist use. This suggests that adding a leukotriene-modulating drug to low-dose inhaled corticosteroid can give you effects equivalent to a high dose of inhaled steroid. This would be beneficial to patients.