Evaluation of serum (S) and urinary (U) eosinophil cationic protein (ECP) and eosinophil-derived neurotoxin (EDN) as biomarkers of eosinophilic (Eo) acute exacerbation of chronic obstructive pulmonary disease (AE).

<bold>Background:</bold> Biomarker-directed treatment strategies using blood eosinophil count (BEC) to guide oral corticosteroid therapy (OCS) in AE have been proposed, but BEC is sensitive to OCS and does not reflect Eo activation. EDN and ECP are markers of Eo inflammation, have slower decreasing kinetics under OCS and therefore are potential biomarkers of Eo AE. In patients with AE requiring hospitalisation, we assessed the levels of S and U EDN and ECP and 1) the differences between patients with an Eo AE and non-Eo AE 2) the correlation with BEC 3) the kinetics of changes under OCS, compared to BEC. <bold>Methods:</bold> U and S EDN and ECP and BEC were determined at admission (T0), 24 and 72 hours later. Based on the BEC at T0, patients were divided in an Eo AE group (BEC >150/µL) or a non-Eo AE group. <bold>Results:</bold> Among 24 patients, 5 presented an Eo AE, 11 a non-Eo AE and 8 received OCS prior to their admission. Analyses performed on patients without OCS prior to T0 showed that there were no differences in the S EDN (24,6± 13,0 vs 23,3± 11,8 µg/L), the S ECP (9,3± 13,7 vs 13,9± 32,5 µg/L), the U EDN (597,9± 388,0 vs 669,8± 445,5 ng/mg of U creatinine) and the U ECP levels (1,3± 0,6 vs 0,7± 1,2 ng/mg of U creatinine) between the Eo AE and the non-Eo AE group at T0. There was no correlation between BEC and S or U, EDN or ECP at T0. While BEC tended to decrease, EDN and ECP remained stable 24h and 72h after initiation of OCS. <bold>Conclusion:</bold> This exploratory study does not support S or U EDN and ECP as biomarkers of Eo AE. In contrast to BEC, EDN and ECP are not significantly affected by 72h of OCS.