A randomised pragmatic trial of corticosteroid optimization in severe asthma using a composite biomarker algorithm to adjust corticosteroid dose versus standard care: study protocol for a randomised trial

Background: Patients with difficult-to-control asthma consume 50-60% of healthcare costs attributed to asthma and cost approximately five-times more than patients with mild stable disease. Recent evidence demonstrates that not all patients with asthma have a typical type 2 (T2)-driven eosinophilic inflammation. These asthmatics have been called 'T2-low asthma' and have a minimal response to corticosteroid therapy. Adjustment of corticosteroid treatment using sputum eosinophil counts from induced sputum has demonstrated reduced severe exacerbation rates and optimized corticosteroid dose. However, it has been challenging to move induced sputum into the clinical setting. There is therefore a need to examine novel algorithms to target appropriate levels of corticosteroid treatment in difficult asthma, particularly in T2-low asthmatics. This study examines whether a composite non-invasive biomarker algorithm predicts exacerbation risk in patients with asthma on high-dose inhaled corticosteroids (ICS) (+/- long-acting beta agonist) treatment, and evaluates the utility of this composite score to facilitate personalized biomarker-specific titration of corticosteroid therapy. Methods/design: Patients recruited to this pragmatic, multi-centre, single-blinded randomised controlled trial are randomly allocated into either a biomarker controlled treatment advisory algorithm or usual care group in a ratio of 4: 1. The primary outcome measure is the proportion of patients with any reduction in ICS or oral corticosteroid dose from baseline to week 48. Secondary outcomes include the rate of protocol-defined severe exacerbations per patient per year, time to first severe exacerbation from randomisation, dose of inhaled steroid at the end of the study, cumulative dose of inhaled corticosteroid during the study, proportion of patients on oral corticosteroids at the end of the study, proportion of patients who decline to progress to oral corticosteroids despite composite biomarker score of 2, frequency of hospital admission for asthma, change in the 7-item Asthma Control Questionnaire (ACQ-7), Asthma Quality of Life Questionnaire (AQLQ), forced expiratory volume in 1 s (FEV1), exhaled nitric oxide, blood eosinophil count, and periostin levels from baseline to week 48. Blood will also be taken for whole blood gene expression; serum, plasma, and urine will be stored for validation of additional biomarkers. Discussion: Multi-centre trials present numerous logistical issues that have been addressed to ensure minimal bias and robustness of study conduct.