Utility of eosinophil peroxidase as a biomarker of eosinophilic inflammation in asthma

Background The relative utility of eosinophil peroxidase (EPX) and blood and sputum eosinophil counts as disease biomarkers in asthma is uncertain. Objective To determine the utility of EPX as a biomarker of systemic and airway eosinophilic inflammation in asthma. Methods EPX protein was measured by immunoassay in serum and sputum in 110 healthy controls to establish a normal reference range and in repeated samples of serum and sputum collected during three years of observation in 480 participants in the Severe Asthma Research Program (SARP)-3. Results Over three years, EPX levels in asthma patients were higher than normal in 27-31% of serum samples and 36-53% of sputum samples. Eosinophils and EPX correlated better in blood than in sputum (rs values of 0.74 and 0.43, respectively), and high sputum EPX levels occurred in 27% of participants with blood eosinophil counts < 150 cells/uL and 42% of participants with blood eosinophil counts 150-299 cells/uL. Patients with persistently high sputum EPX values for three years were characterized by severe airflow obstruction, frequent exacerbations, and high mucus plug scores. In 59 asthma patients who started mepolizumab during observation, serum EPX levels normalized in 96% but sputum EPX normalized in only 49%. Lung function remained abnormal even when sputum EPX normalized. Conclusion Serum EPX is a valid protein biomarker of systemic eosinophilic inflammation in asthma, and sputum EPX levels are a more sensitive biomarker of airway eosinophilic inflammation than sputum eosinophil counts. Eosinophil measures in blood frequently miss airway eosinophilic inflammation, and mepolizumab frequently fails to normalize airway eosinophilic inflammation even though it invariably normalizes systemic eosinophilic inflammation.