To determine whether macrolide antibiotics improve pulmonary function and decrease airway inflammation in cystic fibrosis (CF), we treated 10 patients (females; aged 19-26 years, all colonized with P. aeruginosa, none with atypical Mycobacteria) with 3 weeks of placebo, followed by 6 weeks of clarithromycin (500 mg BID) in a single-blind prospective study. We also determined the safety of sputum induction and the reproducibility of assessing inflammatory markers in induced sputum. Subjects performed spirometry and underwent sputum induction (12-min inhalation of 3% saline) at 3-week intervals. We found that sputum induction was well-tolerated. We also found that the reproducibility was high for neutrophil (PMN) number (R = 0.87, P = 0.009), interleukin (IL)-8 (R = 0.73, P < 0.05, free neutrophil elastase (NE) (R = 0.82, P < 0.05), and myeloperoxidase (MPO) levels (R = 0.86, P < 0.05), but was less so for tumor necrosis factor (TNF)-alpha (R = -0.15, P = 0.7). We found no significant difference in pulmonary function after 6 weeks of treatment with clarithromycin (FEV(1) (% predicted) (mean +/- SEM), 2.2 +/- 0.9 (60 +/- 24%) vs. 2.3 +/- 1 (61 +/- 29%)), and no significant differences in any of the inflammatory indices measured. The median (and range) values before and after treatment for indices of airway inflammation in the induced sputum samples were: for PMNs, 8 (1-326) and 21 (0.2 -175) x 10(6) cells/mL sputum; for IL-8, 156 (24-656) and 202 (16-680) ng/mL; for free NE, 260 (31-1,264) and 237 (49-1,048) microg/mL; for TNF-alpha, 20 (7-128) and 35 (17-87) pg/mL; and for MPO, 169 (13-960) and 195 (14-816) microg/mL. We conclude that clarithromycin is not uniformly effective in improving airway obstruction or in decreasing airway inflammation in patients with CF.