Reduction of Eosinophilic Inflammation in the Airways: Data Analysis

Reduction of Eosinophilic Inflammation in the Airways: Data AnalysisBecause this study used a crossover design, the effects of carry-over and period were evaluated. The homogeneity of background factors between the two groups was examined by x2 test (significance level, two-tailed, 15%). The changes from baseline in eosinophil percentages in sputum and peripheral blood were analyzed using an analysis of variance model including the factors of carry-over, period and treatment effects (significance levels: carry-over effect, two-tailed, 10%; period effect, two-tailed, 5%; and treatment effect, one-tailed, 2.5%). Sputum eosinophil percentage, airway responsiveness to histamine (PC20), pulmonary function, and asthma symptoms were analyzed using the same model. All data except PC20 are expressed as mean ± SD. PC20 is expressed as the geometric mean. For safety information, the incidence of total adverse events was compared between periods using Fisher’s Exact Test (significance level, two-tailed, 5%).
Twenty-six patients completed the study, and 3 patients dropped out fully asthma inhaler. The baseline characteristics of the 26 patients are presented in Table 1. One patient was not included in the study because he had received steroid therapy within 4 weeks before enrollment, and two patients were excluded because their percentage of sputum eosinophils at the beginning of period 1 was < 10%. Changes in asthma symptoms and peripheral blood eosinophils were analyzed in 21 eligible patients; 5 patients were excluded because of a lack of data (n = 3), exacerbation of asthma caused by respiratory infection (n = 1), and lack of washout between periods 1 and 2 (n = 1). Induced sputum, PEF, and airway responsiveness to histamine were analyzed in 20 eligible patients; 5 patients were excluded by the reasons described above, and 1 more patient was excluded because of a lack of data.
The symptom score decreased from baseline after 2 weeks and 4 weeks of treatment with montelukast, but these decreases did not achieve statistical significance (8.2 ± 12.1 at baseline, 6.8 ± 10.2 at 2 weeks, 6.6 ± 10.3 at 4 weeks). Placebo administration was not associated with a decrease in asthma symptoms (6.4 ± 7.4 at baseline, 6.5 ± 8.4 at 2 weeks, 6.4 ± 9.0 at 4 weeks).
Table 1—Patient Characteristics (n = 26)

Characteristics Data
Age, yr 36.8 ± 14.5
Duration of asthma, yr 12.3 ± 13.5
Gender, % of patients
Male 13
Female 13
Severity of asthma, % of patients
Mild 19
Moderate 7
Type of asthma, % of patients
Atopic 22
Nonatopic 4
Concomitant therapy, No. of patients
Slow-release theophylline 7
Slow-release theophylline and oral |32 9
Pulmonary function
FVC, L 3.39 ± 1.05
FEVj, L 2.67 ± 0.83
FEVj, % predicted 82.8 ± 16.3
FEVj/FVC, % 78.3 ± 12.2