Background: Inhaled corticosteroid (ICS) therapy has been demonstrated to reduce the risk of COPD exacerbations. It should only be prescribed to COPD patients who are not adequately controlled by dual long-acting bronchodilator therapy and who have ≥2 exacerbations per year and a blood eosinophil count ≥300cells/µL. ICS therapy is widely prescribed outside guidelines to COPD patients, making ICS withdrawal an important consideration. This systematic review aims to provide an up-to-date analysis of the effect of ICS withdrawal on exacerbation frequency, change in lung function (FEV1) and to determine the proportion of COPD patients who resume ICS therapy following withdrawal.
Methods: Randomised controlled trials (RCTs) and observational studies which compared ICS withdrawal with ICS continuation treatment were included. Cochrane Central, Web of Science, CINHAL, Embase and OVID Medline were searched. Risk of bias was assessed using the Cochrane RoB2 tool and the Newcastle-Ottawa Scale. Quality assessment of RCTs was conducted using GRADE. Meta-analysis of post-hoc analyses of RCTs of ICS withdrawal, stratified by blood eosinophil count (BEC), was undertaken.
Results: Ten RCTs (6642 patients randomised) and 6 observational studies (160,029 patients) were included in the results. When ICS was withdrawn and long-acting bronchodilator therapy was maintained, there was no consistent difference in exacerbation frequency or lung function change between the ICS withdrawal and continuation trial arms. The evidence for these effects was of moderate quality. There was insufficient evidence to draw a firm conclusion on the proportion of patients who resumed ICS therapy following withdrawal (estimated range 12-93% of the participants).
Discussion: Withdrawal of ICS therapy from patients with COPD is safe and feasible but should be accompanied by maintenance of bronchodilation therapy for optimal outcomes.
Keywords: COPD; drug withdrawal; exacerbations; inhaled corticosteroid; observational studies; randomised controlled trials.
© 2024 Georgiou et al.