i SUMMARY ZENECA PHARMACEUTICALS FINISHED PRODUCT: ACTIVE INGREDIENT: ACCOLATE zafirlukast (ZD9188) Trial title (number): A Dose-ranging, Safety and Efficacy Trial with Zafirlukast (ACCOLATE ) in the Treatment of Pediatric Subjects with Mild-to-moderate Asthma; Up to a 52-week Open-label Safety Extension (9188IL/0139) Clinical phase: III First patient recruited: 28 February 1997 Last patient completed: 23 February 1998 (double-blind phase only) Zeneca approval date: 3 November 1998 Principal investigator and location (center number): James W Baker MD, 545 NW 47th Avenue #310, Portland, OR 97213 USA (0038) Publications: None OBJECTIVES: The objectives of this trial were to assess the effect of zafirlukast compared with placebo on asthma physiology as reflected by changes in 1-second forced expiratory volume (FEV 1 ) and daily measurements of peak expiratory flow rate (PEFR); to assess the efficacy of zafirlukast compared with placebo as reflected by changes in daytime and nighttime asthma symptoms and changes in daily β 2 -agonist usage; to determine plasma concentrations of zafirlukast following twice-daily oral dosing; to assess the safety of zafirlukast compared with placebo; to determine whether the response to zafirlukast treatment increases with increasing doses across the dose range studied; and to assess the safety of zafirlukast 20 mg bid in pediatric subjects aged 5 through 11 years in a 52-week open-label extension; to assess humanistic outcomes (quality of life) and explore health economic consequences of the use of zafirlukast in a pediatric population. ACCOLATE is a trademark, the property of Zeneca Limited.
ii This summary presents the methods, results, and conclusions from the 6-week double-blind period only and not the open-label extension (OLE) period. Data from the OLE period will be summarized in a separate report after the trial has ended. METHODS Design: This was a randomized, double-blind, placebo-controlled, dose-ranging, parallel-group, multicenter, trial in which zafirlukast was given twice daily to pediatric patients with mild-to-moderate asthma. The trial included the following 4 parts: a 1-week observation and screening period; a 7- to 14-day single-blind placebo run-in period; a 6 week double-blind treatment period to assess efficacy and safety, and a 52-week open-label safety-extension period. Patients were also allowed to enter the open-label portion of the trial directly without having participated in the double-blind portion of the trial. Population: Four hundred thirteen pediatric patients aged 5 through 11 years with mild-to-moderate asthma (A total of 379 patients completed the trial.) Key inclusion criteria: During the screening period, patients: (a) were boys or girls aged 5 through 11 years; (b) were nonsmokers; (c) had a documented clinical history of asthma and 1 of the following: demonstrated reversible airway disease shown by at least a 12% increase in FEV 1 after inhaled β 2 -agonist (within 4 weeks of screening) or demonstrated (within 6 months of screening) nonspecific bronchial hyperreactivity to methacholine or histamine challenge; (d) had asthma for which they required therapy with β 2 -agonist; (e) demonstrated an FEV 1 between 50% and 85% predicted FEV 1 without medication (ie, 6 hours after inhaled β 2 -agonist or 8 hours after oral β 2 -agonist, or 48 hours after salmeterol [SEREVENT, Glaxo Wellcome]); (f) performed 3 acceptable forced expiratory maneuvers, with 1 reproducible FEV 1 within 10% of the largest FEV 1 ; (g) had a weight-to-stature ratio within the normal range; (h) demonstrated the ability to swallow tablets. During the double-blind randomization period, patients: (a) demonstrated an FEV 1 of at least 50% and no greater than 85% of predicted FEV 1 without medication (at least 6 hours after inhaled β 2 -agonist or 8 hours after oral β 2 -agonist) on the day of randomization, Visit 3; (b) demonstrated clinically mild-to-moderate asthma, as defined by an asthma-episode score totaling at least 8 during the last 7 consecutive days of the 7- to 14-day single-blind placebo run-in period; (c) performed 3 acceptable forced expiratory maneuvers, with 1 reproducible FEV 1 within 10% of the largest FEV 1. Key exclusion criteria: Patients who (a) were placed at undue risk by a temporary postponement of initiating long-term asthma therapy; (b) who had any clinically significant deviation from the reference range in laboratory results, except for abnormalities related to asthma or allergy; (c) had a history of any illness that might confound the results of the trial or place the patient at undue risk; (d) used any disallowed concomitant medications within a specified time period before screening; and (e) had acute illness within 1 week of screening were excluded from the trial. Dosage: Zeneca Pharmaceuticals, Zeneca Inc (Zeneca, also known as the sponsor) supplied the following trial medications (Formulation number followed by lot numbers): 5-mg (F7213; N63131, T53036), 10-mg (F7212; T63137, T63122), 20-mg (F7211; T63138, T73046, T63132, T63129, and N63128) zafirlukast tablets and matching placebo tablets (F7216; N53039 and N53033). Patients were given either 10, 20, or 40 mg of zafirlukast or matching placebo twice daily for 6 weeks. All patients were issued albuterol inhalers (VENTOLIN, Allen &
iii Hanburys; F10000; ZP0318, ZP0527, and ZPA024) and were instructed to use them in accordance with package labeling. Key assessments: Efficacy: Primary assessments were spirometry (changes in FEV 1 ), daily measures of morning and evening PEFR, asthma-episode scores, total number of nights awakened by asthma, daily inhaled bronchodilator usage, and peak flow variability (PFV). Secondary assessments were school absenteeism for asthma, doctor or hospital contacts for asthma, treatment failures, Juniper Pediatric Asthma Quality-of-life Questionnaire, health economics questionnaire, and Child Health Questionnaire - Parent Form 50 (CHQ-PF50). Pharmacokinetics: Patients had blood samples collected to determine the plasma concentrations of zafirlukast at Visits 1 or 2 and Visit 5. Safety: Safety of zafirlukast was assessed on the basis of adverse event monitoring, subjective symptomatology, clinical laboratory measurements, vital signs measurements, electrocardiography, and physical examinations. Statistical analyses: Because no data were available from trials conducted in pediatric patients treated with zafirlukast on which to base estimation of the sample size for this trial, sample size calculations were based on variability observed in a dose-ranging clinical trial (Trial 0015) conducted with similar asthma-related inclusion criteria in adult asthmatics treated with zafirlukast. One hundred patients per treatment group (400 patients total) were sufficient to detect a 0.20-liter difference in FEV 1 (SD=0.50 L) for pairwise comparisons between treatments at α=0.05, with 80% power. One hundred patients per treatment group also allowed detection of a 0.25-unit difference in asthma symptom score (SD=0.60 units) and a 15-L/min difference in morning PEFR (SD=48 L/min) for pairwise comparisons between treatments with 83% and 56% power, respectively. Intention-to-treat (ITT) and per-protocol (PP) analyses were performed on each primary efficacy assessment (asthma physiology, daytime and nighttime asthma symptoms, concomitant β 2 -agonist use, and PFV). The primary focus of efficacy for this trial was end-point measurements after 6 weeks of trial treatment with the last value carried forward. In consideration of the multiplicity of the statistical tests performed, the end point analysis of FEV 1 expressed as a percent of predicted (in the ITT population) was designated as the primary efficacy analysis in this trial. The group treated with the zafirlukast 10-mg dose was compared with the group treated with placebo at the 0.05 significance level as the primary contrast to confirm the conclusions from Trial 0079 regarding the efficacy of this dose. In addition, the differences between the least square means (lsmeans) of the zafirlukast 20-mg dose and placebo, and the zafirlukast 40-mg dose and placebo were examined to determine the dose-response relationship of this range of doses. Dunnett s 95% confidence intervals were provided for the differences in treatment lsmeans between each of the 3 zafirlukast doses and placebo within the Analysis of Covariance (ANCOVA) framework. Additionally, an assessment of dose response was made by using linear and quadratic contrasts within the ANCOVA framework and by using linear regression techniques, and a contrast for all zafirlukast doses combined compared with placebo was performed. In the absence of strong evidence for a dose response, this contrast provided a good indication of the general efficacy of zafirlukast compared with placebo.
iv The primary method of nonparametric analysis was the van Elteren test, which was the multicenter extension of the Wilcoxon rank sum test, applied to change-from-baseline values. Although the van Elteren test was specified in the statistical analysis plan as the primary analysis for number of nights awakened, the results presented in this CTR are those from the ANCOVA analysis in order to maintain consistency with the methods of analysis for other primary parameters; both methods provided similar conclusions. Nonparametric analysis was also performed to supplement and confirm the ANCOVA analyses, even when the underlying assumptions of the ANCOVA were satisfied. RESULTS Demography: A total of 413 pediatric patients with asthma from 41 research centers in the United States were randomized and entered the double-blind portion of this trial. A total of 379 (91.8%) of 413 patients completed the trial, and 34 patients were withdrawn from the trial. Demographic characteristics were similar among the 4 treatment groups. The mean age of patients was 8.6 years (range 5 through 11 years). Sixty percent of patients in this trial were boys, and 40% were girls; the predominant race was white (67% of patients). The overall mean percent of predicted FEV 1 at baseline (Visit 1) was 75.4%, and the mean baseline total asthma-episode score was 10.8 for all patients. Efficacy: For the primary end point of FEV 1 percent predicted in the ITT population of patients, only the 20-mg and combined zafirlukast treatment groups approached statistically significant differences from placebo at end point. For FEV 1 (L), statistically significant differences from placebo-treated patients were achieved at end point for the 20-mg (p=0.025) and combined (p=0.023) zafirlukast treatment groups, and approached significance at end point for the 40-mg zafirlukast treatment group. Improvements from baseline pulmonary function for the 10-mg zafirlukast treatment group at end point were similar to those observed for the same dose in a previous 4-week dose-ranging trial in a similar population of children (Trial 0079). Although effect size at Week 6 for the 10-mg zafirlukast treatment group appeared diminished compared with Week 3 and end point results, it is important to note that 6 patients who contributed pulmonary function data to end point measurements did not do so for measurements at Week 6. Available pulmonary function data for these patients collected outside the predetermined time frame for Week 6 were instead included in end point analyses, and these results (percent of predicted FEV 1 ) alone were considerably higher than the overall end point mean for the zafirlukast 10-mg treatment group (91.3% and 82.6%, respectively) thereby raising the overall mean value at end point. Hence, percent of predicted FEV 1 end point values were higher than Week 6 values for the 10-mg zafirlukast treatment group. The mean difference between the Week 3 and Week 6 observations for patients who had both Weeks 3 and 6 observations was not statistically significant (p=0.9892). For peak flow assessments of pulmonary function, treatment with the 10-mg and combined zafirlukast groups resulted in statistically significant improvements in both morning (p=0.001 and p=0.034, respectively) and evening (p=0.003 and p=0.009, respectively) PEFR compared with placebo at end point. There were no statistically significant differences from placebo at end point for any of the zafirlukast treatment groups or for the combined treatment group in total number of nights
v awakened for asthma or asthma-episode scores. Total β 2 -agonist use in patients with more than 2 puffs per day of use at baseline was significantly reduced compared with placebo for the 10-mg (p=0.018) and combined (p=0.032) zafirlukast treatment groups, and differences from placebo approached significance for the 40-mg zafirlukast treatment group. Similarly, β 2 -agonist use for asthma symptoms or low PEFR was significantly reduced compared with placebo for the 10-mg zafirlukast (p=0.034) and combined zafirlukast (p=0.053) treatment groups. Improvements in PFV compared with placebo approached statistical significance at end point for the 20-mg zafirlukast treatment group only. In general, most of the effect (compared with placebo) of zafirlukast treatment was achieved during the first 3 weeks of trial treatment. Statistically significant differences were seen more frequently for all zafirlukast treatment groups at the early time points than were seen at end point. Repeated-measures analyses, which incorporated data from all 6 weeks of the trial, showed results that were generally more positive than were the results of the end point analyses. In the case of a consistent effect across time, this analysis allowed for an increased precision of measurement of treatment effect for each patient. The repeated-measures analyses for the 10-mg zafirlukast treatment group showed statistically significant improvements compared with the placebo treatment group in morning PEFR (p=0.006), evening PEFR (p=0.022), total daily β 2 -agonist use (p=0.013), and β 2 -agonist use for asthma symptoms or low PEFR (p=0.009), and differences approached statistical significance for FEV 1 and asthma-episode score. Similarly, repeated-measures analyses for the 20-mg and 40-mg zafirlukast treatment groups also showed statistical significance across multiple efficacy assessments, but neither of these higher treatment groups showed statistical significance for a larger number of efficacy measures than was shown for the 10-mg zafirlukast treatment group. For the repeated-measures analysis of the combined zafirlukast doses, statistically significant differences from placebo favoring zafirlukast were achieved for FEV 1 (p=0.018), evening PEFR (p=0.050), total daily β 2 -agonist use (p=0.002), β 2 -agonist use for asthma symptoms or low PEFR (p=0.008), and asthma-episode score (p=0.028). Stratification analyses by weight, age, weight-adjusted dose, and baseline asthma severity measures showed that no class of demographic variable was identified where zafirlukast treatment was consistently ineffective. This suggests that there was no apparent consistent additional efficacy benefit for the 20- and 40-mg doses of zafirlukast in the overall population. Generally, there was an overall trend towards improvement in QOL and humanistic outcomes for zafirlukast treatment groups. Safety: No patients died during the trial. One hundred sixty-seven patients (46 [43.81%] placebo-treated, 40 [38.46%] 10-mg zafirlukast-treated, 42 [40.0%] 20-mg zafirlukast-treated, and 39 [39.39%] 40-mg zafirlukast-treated) reported a total of 279 adverse events: 76, 79, 59, and 65 events for placebo-, 10-mg zafirlukast-, 20-mg zafirlukast-, and 40-mg zafirlukast-treated patients, respectively. No patient had a serious adverse event in this trial. Fourteen patients (six placebo-treated, three 10-mg, three 20-mg, and two 40-mg zafirlukast-treated patients) had nonserious adverse events that led to withdrawal. Approximately half as many zafirlukast-treated patients in each zafirlukast treatment group were withdrawn from the trial due to worsening of asthma than were placebo-treated patients (2.3% compared with 4.8%). Generally, the proportion of patients with adverse events after treatment with 10, 20, or 40 mg of
vi zafirlukast or placebo was similar. The most frequently reported adverse event after randomization was pharyngitis (Coding Symbols for Thesaurus of Adverse Reaction Terms [COSTART] term that includes cold, cold symptoms, and upper respiratory tract infections), which occurred in 16 (15.2%), 9 (8.7%), 16 (15.2%), and 11 (11.1%) patients treated with placebo, 10 mg, 20 mg, and 40 mg of zafirlukast, respectively. No dose response in safety outcomes occurred (ie, 10-, 20-, and 40-mg doses appeared to be equally well tolerated in this trial). CONCLUSIONS: The results of this trial confirm the efficacy and safety of the 10-mg bid dosage of zafirlukast in children aged 5 through 11 years with asthma. Although statistical significance was not achieved for this dosage for the primary end point (percent of predicted FEV 1 ), it was achieved for morning PEFR, evening PEFR, total daily β 2 -agonist use for patients with a baseline greater than 2 puffs per day, and β 2 -agonist use for asthma symptoms or low PEFR. Changes from baseline across end points were similar to those seen in patients treated with 10 mg bid of zafirlukast in Trial 0079, a previous 4-week dose ranging trial. Approximately half as many zafirlukast-treated patients were withdrawn from the trial due to worsening of asthma than were placebo-treated patients, similar to the effect observed in clinical trial results in adults and adolescents. There was no consistent additional efficacy benefit for the 20- and 40-mg doses of zafirlukast in the overall population. All dosages of zafirlukast (10, 20, and 40 mg bid) were generally well tolerated in children aged 5 through 11 years and were not clinically different from placebo treatment with respect to safety.