Abstract
Measurements of exhaled nitric oxide (eNO) are non-invasive and easily measured. Levels have been postulated as providing an indication of underlying airway inflammation. This series of studies was conducted to investigate the role of eNO in measuring airway inflammation in patients with asthma and COPD (chronic obstructive pulmonary disease).
78 asthma patients controlled on maintenance inhaled corticosteroid were assessed for 2-4 weeks. Measurements of eNO demonstrated good repeatability both within-sitting (coefficient of variation (c.v.) 4.1 %) and between weekly visits (c.v. 10.5%; 95% reference range of -38 to +61% of individual patient means). Patients then underwent cessation of inhaled corticosteroid therapy until loss of control occurred or for a maximum of six weeks. Comparisons were made between eNO, symptoms, lung function, sputum eosinophils and airway hyperresponsiveness to hypertonic saline (4.5%) in predicting and diagnosing loss of control. Sixty patients (77.9%) developed loss of control. Exhaled NO increased to a greater
degree in patients who developed loss of control (2.16 versus 1.44-fold increase, p=0.004). Both single measurements and changes in eNO (10ppb, 15ppb or an increase of >60% over baseline) had positive predictive values that ranged from 80-90% for predicting and diagnosing loss of control. This compared well with other more invasive markers. In a further extension of this design but with a total of 87 patients participating, 65 developed loss of control and were entered into a randomised placebo-controlled, double-blind dose-response study of 50, 100, 200 and 500μg of beclomethasone/day for eight weeks. Linear dose-response relationships between dose and changes in eNO and FEV 1 existed at one-week (p=0.002 and p=0.043 respectively) and at the end of treatment (p=0.015 and p=0.006 respectively). A similar linear dose-response relationship was seen with sputum eosinophils (p=0.037) but not with airway hyperresponsiveness. In differentiating between treatment groups eNO was superior to both FEV 1 and eosinophils. Throughout all of these study phases eNO was shown to correlate significantly with other markers of asthma control. This correlation was particularly strong with sputum eosinophils, where significant correlations were seen whether assessed as point-in-time measurements (range of r=0.46 to 0.62, p<0.002) or as changes-over-time during both deteriorating (r=0.44, p<0.001) and improving asthma control (r=0.40, p=0.002). Finally, the effect of a course of oral prednisone on eNO, lung function and 6-minute walk in 30 COPD patients was assessed. A significant decrease in eNO (-3.31ppb; 95%C.I.: -1.45 to -5.16) was seen in association with an improvement in 6-minute walk (34.36m; 95%C.I.: 22.64 to 46. 08). An eNO of ≤10ppb at baseline was associated with a negative predictive value of 82.2% in predicting a significant improvement in 6-minute walk, but the positive predictive value was only 43.7%. No improvement was seen in lung function.
These results indicate that eNO measurements are reliable and provide important clinical information in the control of asthma, correlating to the degree of sputum eosinophilia. Levels can be used to predict and diagnose poor asthma control as well as demonstrating a dose-response to inhaled corticosteroid therapy. However its usefulness is limited in assessing steroid responsiveness in COPD, perhaps a reflection of the different inflammatory processes occurring. Further longitudinal studies are required to investigate whether adding serial eNO measurements into treatment algorithms leads to improved clinical outcome in both asthma and COPD.