Introduction: Randomized controlled trials (RCTs) in COPD suggest inhaled corticosteroids (ICS) reduce exacerbations in patients with blood eosinophilia. ICS are used in patients with bronchiectasis despite no large RCTs.

Objective: Generate evidence on the association of ICS with bronchiectasis exacerbation using the EMBARC registry. 

Methods: Registry data including CT-confirmed bronchiectasis patients from 31 countries. Patients with history of Asthma, COPD and/or ABPA were excluded. Patients with eosinophil counts >400cells/µl were deemed eosinophilic. Exacerbations captured during annual follow-up were analysed using negative binomial modelling with time in study as an offset adjusted for age, sex, geographical region, BMI, diabetes,?Pseudomonas?infection and FEV1. 

Results: 9730 patients were included; 3184 (32.7%) were ICS users. ICS users appeared more severe with higher Bronchiectasis Severity Index, lower FEV1 and increased?exacerbation frequency?(p<0.001). Those with PCD and TB were more likely to receive ICS (p<0.001). Of the 4269 with blood eosinophil data, 302 were eosinophilic (6.8%). Compared to non-ICS users with normal eosinophil counts, ICS users with eosinophilia had reduced exacerbations during follow-up (RR 0.70 95%CI0.59-0.84, p<0.001) while eosinophilia was associated with increased exacerbations in patients not using ICS (RR 1.17 95%CI1.00-1.38, p=0.053). Hospitalisations were also reduced among eosinophilic ICS users (RR 0.56 95%CI0.35-0.90, p=0.016). No relationship with mortality observed. 

Conclusion: Bronchiectasis patients receiving ICS have different outcomes depending on presence of eosinophilia. RCTs of ICS in bronchiectasis are needed.