Sputum Induction as a Method of Analyzing Pulmonary Cells: Reproducibility and Acceptability

Sputum induction has been proposed as a noninvasive method of sampling airway cells for assessing airway inflammation in asthma. Although useful in the research setting, the applicability of this technique to a respiratory clinic, where it might prove useful for clinical management of anti-inflammatory therapy, has not been assessed. We therefore studied the effect of sputum induction in terms of patient acceptability, effect upon airway caliber, and reproducibility of the total cell and differential cell count at 2 weeks in 20 asthmatic subjects first attending an asthma clinic. We compared such patients with normal controls. Thirty-seven subjects underwent sputum induction on two occasions (20 asthmatics and 17 normal subjects) separated by a 2-week interval, using a standardized protocol. Acceptability was assessed by questionnaire. Airway caliber was measured by serial spirometry, using albuterol premedication for asthmatic subjects. Sputum was induced by inhalation of 3.5% saline over 12 min. Total cell and differential counts on induced sputum were assessed and correlated with bronchial hyperresponsiveness, as well as reproducibility of the technique in the clinical setting. All subjects found the process acceptable, although mild side effects occurred in more than 90% of subjects. No differences in acceptability were found between asthmatic and normal subjects. Sputum induction was associated with a significant decrease in forced expiratory volume in 1 sec (FEV1) in normal subjects (from 4.1 +/- 0.17 to 4.02 +/- 0.19 L [p < 0.01] on visit 1 and on visit 2 from 4.01 +/- 0.15 to 3.90 +/- 0.16 L [p < 0.01]), but not in asthmatic subjects after albuterol premedication. Total pulmonary cell yield on the first and second sputum induction days was 1.97 +/- 0.06 x 10(6) cells/mL of sputum and 2.01 +/- 0.05 x 10(6) cells/mL of sputum, respectively, giving a reliability coefficient of 0.77. Less agreement was seen between individual cell differential counts within subjects, but most fell within the expected range on Bland-Altman plots. Sputum induction appears to be safe and acceptable in both normal and asthmatic subjects. A small decrease in FEV1 occurs in normal subjects, which is prevented by albuterol premedication in asthmatics. Reproducibility at 2 weeks yielded similar total numbers of pulmonary cells, but in a clinic population some variability was seen in the number of inflammatory cells, probably because of the small numbers of these cells. This technique may be less valuable in a clinical than a research setting.