At baseline, having both ENO levels of at least 15 ppb or 20 ppb and bronchoconstriction in response to AMP was a clear predictor for failure of ICS reduction, whereas having either AMP broncho-constriction or increased ENO levels alone was not a predictor. Furthermore, a doubling concentration decrease in PC20 2 weeks after the dose of ICS was halved seems to have some predictive value for failure in ICS reduction, whereas the increase in ENO levels did not have predictive value for exacerbations following ICS reduction.
In 27% (10 of 37 patients) in the current study, an exacerbation of asthma was detected after reduction of ICS; therefore, from these results it is evident that in most patients with asthma with disease well controlled with a moderately high ICS dose, the dose of the drug can be reduced without clinical deterioration. In contrast, previous studies showed that reducing the dose of ICS produced a relapse of asthma in 47 to 78% of patients. Differences in results might be explained, at least in part, by differences in patient characteristics. The overall asthma control in our patients was probably better at study entry than that in other studies, and this was undoubtedly due to our strict inclusion criteria. For example, these differences in patient characteristics are reflected in the as-needed (32-agonist consumption at baseline, which was between zero inhalations and four inhalations per day in the study of Price, but only between zero inhalations and two inhalations per week in this study. The reason for selecting patients with stable asthma in good control is that this group would most likely be considered for ICS reduction. Indeed, Juniper examined the effect of ICS reduction on airway responsiveness to methacholine after 1 year of ICS therapy; although a 50% reduction in ICS dose was not associated with a change in airway responsiveness over a 3-month period, five subjects (35%) experienced a deterioration in symptoms and a decrease in FEVl. Another study observed that the long-term remission during a 2-year treatment period with high doses of budes-onide was maintained for well more than 1 year in patients treated with a dose that was reduced to one third of the previous dosage; however, four patients (21%) deteriorated after reduction of inhaled budes-onide. More recently, Prieto reported that in most patients with asthma well controlled with a moderately high dose of budesonide (800 pg/d), a 75% reduction in ICS dose is possible without loss of asthma control; however, 11 subjects (24%) deteriorated after reduction of ICS. These findings are consistent with the present results.