Optimising Azithromycin Prevention Treatment in COPD to Reduce Exacerbations

Background: Prophylactic azithromycin is recommended as a treatment to reduce the risk of chronic obstructive pulmonary disease (COPD) exacerbations in people with COPD at high risk of exacerbations. In clinical practice, there is much uncertainty in how to optimally use this valuable treatment in managing COPD. It is unknown whether azithromycin is effective beyond one-year of treatment; what happens when azithromycin is discontinued following a period of use; or temporarily discontinued over the summer when there are fewer exacerbations. Whether there are differences in treatment responsiveness between subgroups of people with COPD is also uncertain. Aim: To evaluate the benefits and risks of complete or seasonal discontinuation of azithromycin chemoprophylaxis vs continued treatment in people with stable COPD at high risk of exacerbations and assess effects in participant subgroups. Methods: Design: A randomised double-blinded, non-inferiority, adaptive-design pragmatic trial of 3 parallel arms (complete discontinuation, vs seasonal discontinuation vs continued azithromycin as standard of care), to test the strategy of discontinuation of prophylactic azithromycin in participants with stable COPD at high risk of exacerbations. Internal pilot to evaluate recruitment will run for 9 months from first participant first visit (FPFV). Randomisation allocation will be 1:1:1. Adaptive design means a treatment arm can be dropped if futile at interim analysis, but remaining arms continue. Setting: General Practitioner (GP) practices, specialist community clinics, hospitals. Target population: Stable COPD participants prescribed azithromycin ≥3 months to reduce risk of COPD exacerbations. Interventions assessed: Complete discontinuation of azithromycin (matched placebo), seasonal discontinuation (azithromycin October-March, matched placebo April-September), continued azithromycin (standard of care). Trial duration and procedures: Median follow up will be 24 months. Participants will have up to 3 visits - baseline, 12 months, 24 months/end of trial, which may coincide with standard of care visits and be in-person or remote depending on participant's individual preference. Telephone follow-up will be conducted at 1 week, 3 months, 6 months, and 18 months. All participants will have active follow up until study end, even if primary endpoint met. If participants have 3 or more exacerbations/year they will be advised to stop their trial medication. Participants may restart regular azithromycin prescription after stopping trial medication if advised by their GP/specialist. Secondary outcomes will be collected over the entire trial period and therefore may include evaluation both on and off trial medication. Outcome measures: Primary endpoint: Time to first exacerbation (TTFE) necessitating additional treatment with antibiotics and/or corticosteroids. Key secondary endpoints (collected over entire trial) 1. Number/rate and severity of exacerbations, length of exacerbation-free status 2. Health related quality of life measured by change in the EuroQol-5 dimension 3. Symptoms (COPD assessment tool (CAT) score and cough) 4. Side effects 5. Mortality 6. Cost effectiveness from National Health Service (NHS) perspective. Sample size: 1311 participants (437 per arm). Assuming a median TTFE of 150 days and non-inferiority threshold of 30 days shorter, equates to the threshold on the hazard ratio scale of 1.25. Sample size is based on 90 percent power for two non-inferiority comparisons (seasonal and placebo compared with continuous as standard treatment), at 2.7 percent significance using a Cox proportional hazards model. Pre-specified factors for subgroup analysis include exacerbation history, forced expiratory volume at one second (FEV1 percentage) predicted, current/ex-smoking status, COPD Assessment Tool (CAT) score, age, blood eosinophils. Conclusion: This pragmatic, real world trial aims to answer the uncertainties regarding prophylactic azithromycin use in COPD.