Late Breaking Abstract - Corticosteroid optimisation in severe asthma using composite biomarkers to adjust dose versus a symptom/risk-based algorithm

Abstract

Introduction: Increasing corticosteroids (CS) to control asthma symptoms and exacerbations has potential for inappropriately high and difficult to down-titrate CS doses.

Aims and Objectives: We tested whether a T2 biomarker-based strategy (BS) to guide CS reduction (factoring fractional exhaled nitric oxide (FeNO), blood eosinophils, serum periostin) vs. a symptom-based control would see fewer exacerbations and better asthma control and lung function.

Methods: In this randomised (4:1/BS:control), single-blind parallel group trial, patients (pts) with severe asthma and FeNO <45 ppb were given group specific instruction on CS adjustment every 8 wks. The primary outcome was proportion of pts achieving CS reduction at Wk48. Secondary outcomes included exacerbation rate, lung function and asthma control.

Results: In the intention-to-treat population (240 BS vs. 61 control pts), the proportion of BS pts on lower CS dose at Wk48 was 28.4% vs. 18.5% in the controls (adjusted OR: 1.71; 95%CI 0.80, 3.63; p=0.165). In the per protocol (PP) population (n=121), significantly more pts had lower CS doses at Wk48 in the BS vs. the control group (30.7% vs. 5.0% (OR: 11.48; 95%CI 1.35, 97.83; p=0.026). Failure to follow treatment advice was the main reason for ITT and PP population differences. There was no difference in secondary outcomes between study arms.

Conclusion: Biomarker based CS adjustment did not result in a greater proportion of pts reducing CS vs. control, probably because many pts did not follow treatment advice. The biomarker strategy seemed beneficial in pts where symptoms and T2 biomarker profile were discordant.