Phenotypic predictors of response to oral glucocorticosteroids in severe asthma.

ABSTRACT

BACKGROUND: Systemic glucocorticosteroids are side-effect prone but often necessary for the treatment of severe asthma (SA). Our goal was to assess the usefulness of medical history, physiological variables and biomarkers as predictors of response to oral steroids. METHODS: After 4 weeks of treatment optimization, 84 patients with SA and 62 with mild-to-moderate asthma (MA) underwent a 2 week double-blind placebo-controlled oral prednisolone intervention (0.5 mg/kg BW daily) (NCT00555607). RESULTS: Responders had a lower FEV1% (73.7 vs. 88.0), lower FEV1/FVC ratio (0.65 vs. 0.73), lower quality of life (SGRQ score 39.1 vs. 31.4), lower total sputum cell number (1.0 vs. 4.5×10(6)) and higher number of sputum eosinophils (16.8% vs. 6.3%) (p<0.05). For all asthmatics, the degree of improvement in FEV1 correlated with sputum eosinophils, level of asthma control, FeNO, quality of life, age of asthma onset and blood eosinophils. In SA, sputum eosinophils≥3% (OR 9.91), FEV1≤60% (OR 3.7), and SGRQ>42.2 (OR 3.25) were associated with a good response to oral prednisolone. The highest sensitivity and specificity to predict more than 12% increase in FEV1 in SA after oral prednisolone was found for sputum eosinophils≥3% and FeNO>45 ppb. CONCLUSIONS: Sputum eosinophils and FeNO were the best predictors of favorable response to oral prednisolone in severe asthmatics. A guided approach to glucocorticosteroid treatment should be recommended as it favors better control of the disease and presumably a lower rate of adverse events. The study has been registered at the site clinicaltrials.gov with number NCT00555607.