James P. Kemp, MD; Margaret C. Minkwitz, PhD; Catherine M. Bonuccelli, MD; and Marshelle S. Warren, MD

Therapeutic Effect of Zafirlukast as Monotherapy in Steroid-Naive Patients With Severe Persistent Asthma* James P. Kemp, MD; Margaret C. Minkwitz, PhD; Catherine M. Bonuccelli, MD; and Marshelle S. Warren, MD Study objectives: We evaluated the efficacy of the leukotriene receptor antagonist zafirlukast (Accolate ), 20 mg twice daily, as monotherapy in patients with severe persistent asthma (defined by an FEV 1 < 60% of predicted before treatment and frequent night-time symptoms). Design: Data were analyzed from a subgroup of 261 steroid-naive patients (zafirlukast, n 149; placebo, n 112) from four randomized, double-blind, placebo-controlled, 13-week trials with similar experimental designs, entry criteria, and clinical assessments. Patients: These patients were mostly men (57%) older than 30 years (56%) with pulmonary obstruction, ie, FEV 1 /FVC ratio < 0.7 (79%), and reversible airway disease demonstrated by a 15% increase in FEV 1 after inhaled bronchodilator use. Results: At end point, patients who received zafirlukast monotherapy had significant (p < 0.05) improvements from baseline, and compared with placebo, in FEV 1, morning and evening peak expiratory flow (PEF), daytime asthma symptoms, nighttime awakenings, and 2 -agonist use. A stratified analysis based on the FEV 1 /FVC ratio showed an interaction between treatment and the amount of airflow obstruction for nighttime awakenings and mornings with asthma. Moreover, 37% of patients in both treatment groups had PEF variability > 20% (an indirect measure of airway inflammation). Zafirlukast patients with PEF variability > 20% had increases from baseline in the morning and evening PEF of approximately 40 and 11 L/min, respectively. For patients who take zafirlukast and who have a PEF variability of < 20%, the morning and evening PEF increased by 25 and 30 L/min, respectively. Regardless of the degree of PEF variability, zafirlukast significantly (p < 0.05) increased morning and evening PEF compared with placebo. Conclusion: Patients with severe persistent asthma who received zafirlukast as monotherapy had clinically significant improvements across all efficacy measures compared with placebo and significant reductions in PEF variability. (CHEST 1999; 115:336 342) Key words: asthma symptoms; pulmonary function; severe persistent asthma; zafirlukast Abbreviations: NAEPP National Asthma Education and Prevention Program; PEF peak expiratory flow Asthma is defined as a chronic inflammatory disorder of the airways, which involves various cells and mediators in its immunohistopathology. In susceptible individuals, chronic inflammation causes For editorial comment see page 313 recurrent episodes of wheezing, breathlessness, chest tightness, and coughing, especially at night or * From the University of California School of Medicine (Dr. Kemp), San Diego, CA; and Zeneca Pharmaceuticals (Drs. Minkwitz, Bonuccelli, and Warren), Wilmington, DE. Dr. Kemp is a consultant to Zeneca Pharmaceuticals and a participating investigator in zafirlukast clinical trials. Financial support for this study was provided by Zeneca Pharmaceuticals, Wilmington, DE. Manuscript received March 10, 1998; revision accepted August 26, 1998. Correspondence to: Dr. James P. Kemp, Allergy and Asthma Medical Group and Research Center, 9610 Granite Ridge Dr, Suite B, San Diego, CA 92123 in the early morning. These episodes are usually associated with widespread but variable airflow obstruction that is often reversible either spontaneously or through treatment. 1 The new National Asthma Education and Prevention Program (NAEPP) guidelines for the diagnosis and management of asthma classify asthma into four categories according to severity, ie, mild intermittent, mild persistent, moderate persistent, and severe persistent. 2 Patients diagnosed with severe persistent asthma have one or more of the following characteristics: continual symptoms, symptoms that limit physical activity, or frequent exacerbation of asthma. Other distinguishing features include frequent nighttime symptoms, an FEV 1 or peak expiratory flow (PEF) of 60% of predicted or lower, or PEF variability greater than 30%. For patients with severe persistent asthma, the recommended therapeutic approach for long-term 336 Clinical Investigations

control involves treatment with an anti-inflammatory agent (ie, a high-dose inhaled corticosteroid), a long-acting bronchodilator (either a long-acting 2 - agonist, sustained-release theophylline, or long-acting inhaled 2 -agonist tablets), and long-term treatment with corticosteroid tablets or syrup (2 mg/kg/d, not to exceed 60 mg/d). 2 The goals of treatment are to prevent symptoms, restore and maintain normal pulmonary function, help patients regain normal activity levels, prevent recurrent exacerbations, provide optimal pharmacotherapy with minimal side effects, and meet both patient and family expectations of satisfactory care. 2 Because clinical manifestations of asthma vary among patients and between seasons, treatment must be selected to suit individual patients. At routine office visits, the severity of asthma can be underestimated. For example, in mild-to-moderate asthma, it is not unusual for patients with symptoms and exacerbations to have normal spirometry results at office visits. In such cases, regular monitoring with peak flowmeters helps evaluate the effectiveness of therapy, avoiding inadequate treatment. Monitoring the frequency of nighttime symptoms and the extent of activity limitation also provides important information for determining the severity of asthma and an appropriate treatment regimen. The class of compounds known as leukotriene receptor antagonists is currently recommended as an alternative to inhaled corticosteroids or anti-inflammatory medications in patients with mild-to-moderate persistent asthma. 2 In reviewing baseline characteristics of patients involved in the zafirlukast (Accolate; Zeneca Pharmaceuticals; Wilmington, DE) clinical trial program, we discovered that a number of patients who were diagnosed as having mild-to-moderate asthma on the basis of their treatment history therapy with 2 -agonist only had an FEV 1 60% of predicted at trial entry. As such, these patients would be classified by the most recent NAEPP guidelines as having severe persistent asthma. This finding prompted our investigation of the efficacy of the recommended dosage of zafirlukast as monotherapy with that of placebo in the subgroup of steroid-naive patients (ie, patients who had not received corticosteroid therapy) classified as having severe persistent asthma. Although data from patients in this subgroup were available from four placebo-controlled trials, the number of such patients in each trial was too small to permit separate analyses. Because the four trials had similar experimental designs, entry criteria, and clinical assessments, we pooled data from patients who had an FEV 1 of 60% of predicted at baseline and frequent nighttime symptoms (ie, two or more awakenings per week). Trial Design Materials and Methods This analysis evaluated a subset of 261 men and women who received treatment with a 20-mg twice daily (bid) dose of zafirlukast (n 149) or placebo (n 112) in four randomized, double-blind, placebo-controlled, 13-week trials with similar study designs, entry criteria, and clinical assessments. All of the trials had three periods, including a 1-week screening, a 2- to 3-week, single-blind, placebo run-in, and 13 weeks of doubleblind treatment. All four trials were conducted under the auspices of one sponsor (Zeneca Pharmaceuticals), which had responsibility for site selection, trial monitoring, and data management, review, and analysis. All of these factors contributed to the ability to combine data across these trials for the present analysis. Appropriate institutional review boards approved each trial, and all patients gave their written, informed consent. Patient Population Entrance criteria varied only slightly among the 13-week trials and are summarized in Table 1. For this analysis, we included only patients from the trials who had an FEV 1 60% of predicted (a minimum of 6 h after 2 -agonist use). Patients in this subgroup were nonsmokers 12 years or older who had a documented history of asthma, reversible airway disease demonstrated by a 15% increase in FEV 1 after inhaled bronchodilator use, and previous treatment with 2 -agonist alone at trial entry. Patients were excluded from the trials if they had an acute illness or disease, a history of alcohol or drug abuse, or an upper or lower respiratory tract infection or hepatitis B vaccination within 6 weeks of screening. Patients were also excluded if they used cromolyn sodium or corticosteroids (oral or inhaled) within 4 weeks of the screening or astemizole within 3 months of the screening. Screening and Placebo Run-in Periods During the screening period, patients recorded morning and evening PEF, daily asthma symptoms, and 2 -agonist use on diary cards. Other evaluations included a medical history, interviews about subjective symptoms, a physical examination, vital signs measurements, a 12-lead ECG, clinical laboratory tests, and a urine drug screen. All procedures, except the medical history and urine drug screen, were repeated at the end of the placebo run-in period. FEV 1 was measured at least 6 h after 2 -agonist use during screening and run-in periods. Albuterol was provided for use as rescue medication throughout the trial. Double-blind Treatment In each trial, only symptomatic patients, as determined by their cumulative daytime asthma symptoms score, were randomized to double-blind treatment with zafirlukast or placebo (Table 1). Each day throughout the double-blind treatment, patients recorded PEF, asthma symptoms scores, and 2 -agonist use on diary cards as previously described by Fish et al. 3 FEV 1 was measured at least 6 h after 2 -agonist use at each clinic visit, and the best of three forced expiratory maneuvers was recorded. Patients were withdrawn from treatment at their request or because of a significant adverse event, concurrent illness unrelated to asthma, treatment failure, or administrative reason. Efficacy Assessments The primary efficacy measure for this population was the change from baseline to end point in the measures of pulmonary CHEST / 115 / 2/ FEBRUARY, 1999 337

Table 1 Summary of Entrance Criteria for Patients Enrolled in 13-Week, Double-Blind, Placebo-Controlled Trials* Trial Features Trial 1 Trial 2 Trial 3 Trial 4 Age, yr 18 12 12 12 Entry therapy 2 -agonist alone 2 -agonist alone 2 -agonist alone* 2 -agonist alone Screening FEV 1, % predicted 55%, no upper limit 55%, no upper limit 55%, no upper limit 45% to 80% Daytime asthma symptoms score 8 8 8 10 criteria for randomization Randomization, Z P 1 1 2 1 1 1 1 1 Patients randomized with 15 (17) 88 (12) 26 (14) 132 (29) FEV 1 60% predicted, n (%) Patients randomized to Z or P, n 7, 8 57, 31 14, 12 71, 61 *Theophylline withdrawn during screening or placebo run-in. Cumulative daytime asthma symptoms score (range, 0 to 3/day; maximum, 21/week) from last 7 consecutive days of placebo run-in period. Scale: no symptoms (0); mild symptoms that did not interfere with activities (1); moderate symptoms that interfered with some activities (2); and severe symptoms that interfered with many activities (3). Z zafirlukast; P placebo. function (FEV 1 and morning PEF); secondary end points were the change from baseline to end point in daytime asthma symptoms score, nighttime awakenings, mornings with asthma symptoms, and 2 -agonist use. Response to treatment was evaluated as the percent change in FEV 1 from baseline to end point. A 10% or greater increase from baseline was considered a favorable response to treatment. In the individual trials, the primary efficacy measure was the change from baseline to end point in the daytime asthma symptoms score. Statistical Analysis The patient population analyzed comprised a subgroup of patients pooled from intention-to-treat analyses in four 13-week, placebo-controlled trials. These patients were treated previously with as-needed 2 -agonist alone, received a 20-mg dose of zafirlukast or placebo twice daily, and had a baseline FEV 1 of 60% of predicted. In individual trial protocols, the primary evaluation point was end point, which included a last value carried forward observation for patients who did not complete 13 weeks of treatment. Consequently, we used end-point data for all formal subgroup analyses. In an article that identified various statistical problems encountered when conducting and reporting clinical research, Pocock and colleagues 4 recommended using statistical tests of interaction to analyze and report treatment differences between subgroups. Therefore, we used an analysis of covariance with a trial effect, the assessment baseline as a covariate, strata effect(s), treatment effect, and strata-by-treatment effect(s). For the purpose of this analysis, each center was treated as a random component of the overall variability and not incorporated into the model. Interactions were considered significant if they were associated with an F-test p value of 0.05 and approached significance when associated with an F-test p value of 0.10. p Values for comparison between zafirlukast and placebo were reported using the analysis of covariance model pairwise comparisons. A Bonferroni adjustment for multiple pairwise comparisons was used, and a p value of 0.006 was considered statistically significant. Efficacy evaluations were based on the original planned primary analysis as described by Fish et al. 3 Briefly, analysis of the primary and secondary efficacy assessments was based on endpoint data. End point is defined as the 13-week assessment or the last observed value carried forward for patients withdrawing during the trial treatment period. Safety Assessments The safety of trial medications was evaluated from the results of physical examinations, ECGs, clinical laboratory tests (including clinical chemistry, hematology, and urinalysis), and subjective symptoms interviews. Results Two hundred sixty-one patients with severe persistent asthma (zafirlukast, n 149; placebo, n 112) were included in the subgroup analyses. More than half of this patient population consisted of men who were 30 years of age and who had moderate pulmonary obstruction as demonstrated by an FEV 1 /FVC ratio of 0.70 (Table 2). Baseline asthma characteristics were similar between the zafirlukast and placebo groups. On average, patients had a percent predicted FEV 1 of approximately 53%, between four and five nighttime awakenings per week, and used approximately six puffs of 2 -agonist per day (Table 3). Pulmonary Function Analyses of the subset of patients within each trial for FEV 1 and morning PEF achieved statistical significance only for trial 4, which contributed approximately half of the patients (Table 4). The direction of response was similar across three of the four trials for both measures. No trial was excluded because of the response noted in that trial. As shown in Table 5, the analysis across trials indicated that pulmonary function improved after zafirlukast therapy, with significant (p 0.05) in- 338 Clinical Investigations

Table 2 Distribution of Patients With Severe Persistent Asthma by Demographic and Baseline Characteristics Characteristic Trial 1, n (%) Trial 2, n (%) Trial 3, n (%) Trial 4, n (%) Combined, n (%) Sex Men 9 (60) 59 (67) 19 (73) 62 (47) 149 (57) Women 6 (40) 29 (33) 7 (27) 70 (53) 112 (43) Race White 14 (93) 74 (84) 22 (85) 104 (80) 214 (82) Black 1 (7) 6 (7) 1 (4) 17 (13) 25 (9) Other 0 (0) 8 (9) 3 (12) 11 (8) 22 (8) Age, yr 30 4 (27) 40 (45) 12 (46) 58 (44) 114 (44) 30 11 (73) 48 (55) 14 (54) 74 (56) 147 (56) FEV 1 /FVC ratio 0.70 12 (80) 75 (85) 22 (85) 97 (73) 206 (79) 0.70 3 (20) 13 (15) 4 (15) 35 (27) 55 (21) PEF variability 20% 11 (79) 51 (58) 17 (65) 84 (64) 163 (63) 20% 3 (21) 37 (42) 9 (35) 47 (36) 96 (37) Mean FEV 1, % predicted, screening 60.7 62.1 59.7 59.2 60.3 Range 52 72 39 93.7 52 88 43.9 80.4 39 93.7 Mean FEV 1, % predicted, baseline 51.0 54.0 53.3 52.7 53.1 Range 35 57 41 59.7 44 59.9 40 59.9 35 59.9 creases from baseline to end point, as well as in comparison with placebo, in FEV 1 and morning and evening PEF. Sixty-four percent of patients had an increase of 10% in FEV 1 after zafirlukast treatment compared with 50% of patients who received placebo (Fig 1). Deterioration in pulmonary function was observed for 21% of patients in the zafirlukast group and 35% of patients in the placebo group (Fig 1). When patients were grouped according to PEF variability ( 20% and 20%), no interactions were observed between treatments and the degree of variability. However, for patients with PEF variability indicative of inadequately controlled asthma, ie, those with 20% variability at baseline, morning and evening PEF variability was reduced to 20% Table 3 Baseline Asthma Characteristics for Patients With Severe Persistent Asthma* Characteristic Zafirlukast (n 149) Placebo (n 112) FEV 1, L 1.96 0.44 1.91 0.47 Percent of predicted FEV 1 53.3 4.9 52.9 4.0 Morning PEF, L/min 327.1 81.4 327.6 95.4 Evening PEF, L/min 374.5 88.7 376.2 104.3 PEF variability 17.8 0.9 17.6 1.1 Daytime asthma symptoms score 12.7 2.8 12.8 2.8 Nighttime awakenings, n/wk 4.8 4.9 4.4 4.7 Mornings with asthma, n/wk 5.7 1.9 5.1 2.3 2 -Agonist use, puffs/d 6.1 2.9 5.5 2.8 *Values are given as mean SD. FEV 1 measured 6 h after 2 -agonist and 12 h after theophylline in trial 3 only. Mean daily score multiplied by 7. for significantly more patients in the zafirlukast group than in the placebo group (53% vs 34%, respectively; a 19% difference; p 0.044; 95% confidence interval, 9% to 29%). Also, these patients had an increase from baseline in morning and evening PEF of approximately 40 and 11 L/min, respectively, after zafirlukast treatment (Fig 2). For patients with PEF variability of 20% at baseline, zafirlukast therapy increased morning and evening PEF by 25 and 30 L/min, respectively (Fig 2). Asthma Symptoms After zafirlukast treatment, patients with severe persistent asthma had significant (p 0.01) decreases from baseline to end point compared with placebo in the daytime asthma symptoms score, nighttime awakenings, mornings with asthma, and 2 -agonist use (Table 5). Patients in the zafirlukast group had a 20% reduction from baseline in the daytime asthma symptoms score ( 26%), nighttime awakenings ( 27%), mornings with asthma ( 24%), and 2 -agonist use ( 26%) (Fig 3). When patients were grouped by their FEV 1 /FVC ratio ( 0.70 or 0.70), significant (p 0.05) interaction was observed between treatments and the amount of airflow obstruction for nighttime awakenings and mornings with asthma. Patients who had an FEV 1 /FVC ratio indicative of moderate pulmonary obstruction (ie, 0.70) had fewer nighttime awakenings per week after zafirlukast therapy than did patients who received placebo (Fig 4). Patients with less pulmonary obstruction (FEV 1 /FVC of 0.70) had a similar reduction in awakenings regardless of CHEST / 115 / 2/ FEBRUARY, 1999 339

Table 4 Comparison of Pulmonary Measures at End Point for Patients With Severe Persistent Asthma by Trial Characteristic Trial No. (Nz, Np)* Baseline Zafirlukast End Point Placebo End Point p Value FEV 1, L 1 (7, 7) 1.89 0.10 2.19 0.20 2.04 0.20 0.64 2 (57, 30) 1.98 0.05 2.36 0.06 2.35 0.08 0.95 3 (13, 11) 2.03 0.08 2.29 0.17 2.09 0.18 0.41 4 (68, 57) 1.91 0.04 2.32 0.07 2.12 0.08 0.045 Morning PEF, L/min 1 (7, 7) 344.0 19.2 334.0 7.5 367.4 7.5 0.71 2 (58, 32) 334.5 10.0 361.9 7.2 341.5 9.8 0.099 3 (14, 12) 329.7 11.9 351.1 13.3 320.7 14.4 0.13 4 (70, 60) 320.1 7.7 356.6 6.3 329.4 6.8 0.0042 *Nz number of patients receiving zafirlukast; Np number of patients receiving placebo. End-point values reported are least square means from analysis of covariance model, adjusted for baseline value, so common baseline mean is given in table. Values given as mean SE. treatment (Fig 4). Compared with placebo, patients in both FEV 1 /FVC ratio subgroups reported fewer mornings with asthma after zafirlukast therapy, with the largest reduction occurring for patients with less pulmonary obstruction (FEV 1 /FVC ratio of 0.70) (Fig 4). Safety Seventy-nine percent of patients in the zafirlukast group (n 117) and 71% of patients in the placebo group (n 80) completed the trial. Eighteen patients were withdrawn during the trial because of adverse events, 6 in the zafirlukast group and 12 in the placebo group. Patients were withdrawn from zafirlukast therapy because of asthma exacerbation (n 5) and hypertonia (n 1); none of these events was considered serious. In the placebo group, patients were withdrawn because of asthma exacerbation (n 9), increased levels of aspartate transaminase (n 1), myalgia (n 1), and dizziness and abdominal pain (n 1). Two patients in the placebo group were hospitalized for serious adverse events (n 1 for asthma exacerbation and n 1 for kidney calculus). Asthma exacerbation was defined as a worsening of asthma, requiring the withdrawal of the patient from the trial. In one trial, asthma exacerbation was defined a worsening of asthma resulting in patient withdrawal or requiring a burst of oral corticosteroids. Comparison of the effectiveness of treatments on asthma exacerbation rates favored the zafirlukast group over the placebo group, but the difference was not statistically significant (7 vs 11 or 4.7% vs 10.3%, respectively; p 0.088). No significant treatment differences were observed in the incidence of any adverse event (Table 6). by the finding that 261 patients from the zafirlukast clinical trial program who were diagnosed as having mild-to-moderate asthma on the basis of their treatment history met the NAEPP classification for severe persistent asthma. Although the ideal approach to therapy for these patients involves treatment with an optimized dose of inhaled corticosteroid therapy or combination therapy with inhaled corticosteroids and a long-acting bronchodilator, the subgroup analysis provided an opportunity to evaluate zafirlukast as monotherapy in patients with severe persistent asthma who were undertreated according to the NAEPP guidelines. Overall, 18% of patients who were diagnosed as having mild-to-moderate asthma on the basis of their treatment history (ie, therapy with 2 -agonist only) had severe persistent asthma according to their baseline pulmonary function (ie, anfev 1 of 60% of predicted). Sixty-one patients with an FEV 1 60% of predicted had two or more awakenings per week at baseline. Of these patients, 66% had five or more awakenings per week because of asthma. Our analysis also revealed that within this subgroup of patients with severe persistent asthma, 79% had Discussion This retrospective analysis in steroid-naive patients with severe persistent asthma was prompted Figure 1. Percentage of patients who had an increase or deterioration in FEV 1 after 13 weeks treatment with zafirlukast or placebo. 340 Clinical Investigations

Table 5 Comparison of Asthma Characteristics at End Point for Patients With Severe Persistent Asthma Characteristic Zafirlukast Placebo 95% CI of Difference Between Treatments p Value* Mean % Change From Baseline Zafirlukast Placebo FEV 1, L 2.28 0.05 2.14 0.06 [0.01,0.27] 0.039 21 13 Percent of predicted FEV 1 62.6 1.5 59.1 1.6 [ 0.1,7.0] 0.060 21 14 Morning PEF, L/min 352.1 5.5 329.9 5.9 [8.3,34.1] 0.001 10 3 Evening PEF, L/min 396.2 5.7 377.0 6.1 [5.9,32.6] 0.005 8 2 PEF variability 15.4 1.1 17.7 1.2 [ 4.7,0.7] 0.088 3.4 0.8 Daytime asthma symptoms score 9.8 0.4 11.4 0.4 [ 2.6, 0.7] 0.001 26 12 Nighttime awakenings, No./wk 3.4 0.3 4.6 0.4 [ 2.0, 0.4] 0.002 27 19 Mornings with asthma, No./wk 4.3 0.3 5.1 0.3 [ 1.3, 0.2] 0.007 24 2 2 -Agonist use, puffs/d 4.7 0.3 5.9 0.3 [ 1.8, 0.6] 0.001 26 4 *p 0.006 for statistical significance. FEV 1 measured 6 h after 2 -agonist use. Values are given as mean SE. Mean daily score multiplied by 7. moderate airflow obstruction on the basis of their FEV 1 /FVC ratio ( 0.7), and 37% had poorly controlled asthma according to PEF variability ( 20%). The recommended dosage of zafirlukast (20 mg bid) benefited this group of patients compared with placebo (p 0.05) by inproving pulmonary function and reducing daytime symptoms, nighttime awakenings, and 2 -agonist use. As might be expected, patients with severe disease responded comparably and, in some cases, better to zafirlukast therapy than did patients with milder disease. 3 When reviewing end-point data, we found that patients with severe disease had greater increases in FEV 1 (21% vs 6%, respectively) and morning PEF (10% vs 7%, respectively) and greater reductions in nighttime awakenings (27% vs 20%, respectively) and 2 -agonist use (26% vs 22%, respectively) than did patients with mild-to-moderate disease. The results for FEV 1 and morning PEF for patients who received zafirlukast therapy are consistent with those reported for inhaled corticosteroid therapy in patients with moderate or severe asthma, as well as in patients with chronic asthma who had baseline values similar to those in our analysis. For example, the change from baseline in FEV 1 of 0.32 L with zafirlukast therapy is comparable to the change of 0.28 L reported by Engel et al 5 and 0.27 L observed by Busse et al 6 in trials evaluating 1,600 g budesonide daily, and is approximately 70% of the change of 0.44 L reported by Pearlman et al 7 in a trial evaluating 500 g fluticasone propionate daily. Similarly, the morning FEV 1 change of 25 L/min with zafirlukast was within the range of 12 to 51 Figure 2. Change from baseline to end point in morning and evening PEF for the zafirlukast and placebo groups according to PEF variability. Among patients with 20% PEF variability at baseline, morning and evening PEF variability was reduced to 20% for significantly (p 0.05) more patients in the zafirlukast group than in the placebo group. Figure 3. Percentage change from baseline for morning and evening PEF, FEV 1, daytime asthma symptoms score, nighttime awakenings, mornings with asthma, and 2 -agonist use in the zafirlukast and placebo groups. A significant (p 0.05) difference between the zafirlukast and placebo groups was observed for each assessment. CHEST / 115 / 2/ FEBRUARY, 1999 341

Table 6 Adverse Events Reported in > 2% of Patients Treated With Zafirlukast Zafirlukast, n 149 Placebo, n 112 Adverse Event n % n % Pharyngitis 34 22.8 27 24.1 Headache 17 11.4 17 15.2 Flu syndrome 11 7.4 7 6.3 Sinusitis 6 4.0 3 2.7 Asthma exacerbation 5 3.4 9 8.0 Diarrhea 4 2.7 7 6.3 Accidental injury 4 2.7 3 2.7 Lung disorder* 4 2.7 1 0.9 Rhinitis 4 2.7 0 0 *Diagnosed as chest or pulmonary congestion. Figure 4. Change from baseline to end point in the number of nighttime awakenings per week and in the number of mornings with asthma per week for patients grouped according to treatment and FEV 1 /FVC ratio (f zafirlukast; placebo). A significant (p 0.01) interaction was observed between treatments and the amount of airflow obstruction for both nighttime awakenings and mornings with asthma. L/min reported with 1,600 g budesonide daily 5,6,8,9 and 21 to 27 L/min with fluticasone. 7,8 The effectiveness of an asthma therapy can also be determined by the reduction in the proportion of patients requiring treatment for asthma exacerbation. In our analysis of patients with severe persistent asthma, the number of patients with exacerbations during zafirlukast therapy was almost half that of the placebo group (4.7% vs 10.3%). Similarly, in a study of patients with milder disease, exacerbation rates during the 13 weeks of the trial were half as high in the zafirlukast group as those in the placebo group (3.1% vs 6.5%). 3 The adverse event profile in patients with severe asthma was not different from that observed previously by Fish et al 3 in patients with mild-to-moderate disease. No significant differences were observed between zafirlukast and placebo in the occurrence or severity of adverse events. Hence, 13 weeks of treatment with zafirlukast did not place patients with severe persistent asthma at undue risk. In conclusion, patients with severe persistent asthma who received zafirlukast as monotherapy had clinically significant improvements across all efficacy measures compared with placebo and significant reductions in peak flow variability. Additional studies are required to determine whether zafirlukast may confer additional benefits when combined with an optimized dose of inhaled corticosteroid therapy for the long-term disease management of patients with severe persistent asthma. ACKNOWLEDGMENT: The authors thank Mary Jo Psomas, MS, and Gary Dorrell, MS, ELS, for editorial assistance. References 1 National Heart Lung, and Blood Institute. Executive summary: guidelines for the diagnosis and management of asthma. Bethesda, MD: US Department of Health and Human Services, Public Health Service, National Institutes of Health; 1991. Publication No. 91 3042A 2 National Heart Lung, and Blood Institute. Executive summary: guidelines for the diagnosis and management of asthma. Bethesda, MD: US Department of Health and Human Services, Public Health Service, National Institutes of Health; 1997. Publication No. 97 4051 3 Fish JE, Kemp JP, Lockey RF, et al. Zafirlukast for symptomatic mild-to-moderate asthma: a 13-week multicenter study. Clin Ther 1997; 19:675 690 4 Pocock SJ, Hughes MD, Lee RJ. Statistical problems in the reporting of clinical trials. N Engl J Med 1989; 317:426 432 5 Engel T, Dirksen A, Heinig JH, et al. Single-dose inhaled budesonide in subjects with chronic asthma. Allergy 1991; 46:547 553 6 Busse WW, Chervisnky P, Condemi J, et al. Budesonide delivered by Turbohaler is effective in a dose-dependent fashion when used in the treatment of adult patients with chronic asthma. J Allergy Clin Immunol 1998; 101:457 463 7 Pearlman DS, Noonan MJ, Tashkin DP, et al. Comparative efficacy and safety of twice daily fluticasone propionate powder versus placebo in the treatment of moderate asthma. Ann Allergy Asthma Immunol 1997; 78:356 362 8 Ringdal N, Swinburn P, Backamn R, et al. A blinded comparison of fluticasone propionate with budesonide via powder devices in adult patients with moderate-to-severe asthma: a clinical evaluation. Mediat Inflamm 1996; 5:382 389 9 Johansson SA, Dahl R. A double-blind dose response study of budesonide by inhalation in patients with bronchial asthma. Allergy 1987; 43:173 178 342 Clinical Investigations