Sputum induction and bronchoscopy for assessment of ozone-induced airway inflammation in asthma

Background: Neutrophilic airway inflammation, as defined by cell counts in respiratory tract lining fluid (RTLF), is a key end point in many studies of respiratory toxicity in both healthy and asthmatic subjects. BAIL and sputum induction (SI) are the most common methods of sampling RTLF in such studies. However, the comparability of these methods (BAL and SI) after experimental treatment has not been investigated in a head-to-head controlled trial. Methods: To determine whether BAIL and SI are comparable and can be used in place of each other in the assessment of neutrophilic airway inflammation after ozone (0,I) exposure, we exposed 13 asthmatic subjects to either 0.2 ppm of O-3 or filtered air (FA) followed by either BAL or SI. Subjects then underwent the alternate (O-3 or FA) exposure followed by the same method of RTLF sampling. Next, subjects repeated the same exposure protocol with the alternate method of RTLF sampling. Differences in inflammatory indexes including the percentage of polymorphonuclear neutrophils (%PMNs) between the exposures were then correlated by regression analysis. Results: The %PMNs in sputum was poorly correlated with that in BAL fluid (R = 0.12). The correlation between the %PMNs in sputum and in the bronchial fraction of BAL (BFx) fluid, however, was somewhat higher (R = 0.50). Furthermore, the uncertainty of the estimate of %PMN values in BFx fluid and BAL fluid based on those of sputum values, using regression models, was almost as great as the magnitude of the O-3 effect itself (ie, 9.7% and 5.5% estimate errors for O-3 effects of 17.0% and 7.5%, respectively). Conclusion: We concluded that SI and BAIL indexes are not directly interchangeable in the assessment of O-2-induced airway inflammation in asthmatic subjects.