Decrease in Eosinophil-Derived-Neurotoxin With Mepolizumab Is Associated With Asthma Improvement

Abstract

Over the past decade, mepolizumab, an anti-IL-5 monoclonal antibody, has been introduced to target eosinophils in severe asthma. Despite its well-documented efficacy [1, 2], recent studies have highlighted a heterogeneous clinical response among patients [3], underscoring the need to identify biomarkers to monitor response to mepolizumab. Currently, eosinophilic inflammation is monitored through eosinophil cell counts, which provide only a limited evaluation of eosinophil activity. Indeed, asthma is characterized not only by an increased number of eosinophils but also by enhanced airway eosinophil degranulation. Although a high blood eosinophil count is associated with treatment response, its extreme variability limits its utility as a biomarker. Similarly, sputum eosinophil counts, while informative, are not routinely available in clinical practice. Given these limitations, recent studies have focused on eosinophil-derived neurotoxin (EDN), a promising stable biomarker of eosinophil degranulation, for asthma monitoring [4-6]. However, the evolution of EDN levels in response to anti-IL-5 therapies remains poorly documented. In this study, we compared EDN levels and eosinophil counts in both blood and bronchoalveolar lavage fluid (BALF) as markers of response to mepolizumab in patients with severe asthma (NCT03797404). EDN was measured at baseline (D0), at 6 months (M6), and at 12 months (M12) using a fluorescence enzyme immunoassay. Regression models were employed to assess associations between the absolute variation in EDN levels (M6-D0) in blood and BALF and concurrent changes in asthma-related characteristics, including disease control, pulmonary function, and exacerbation frequency. Twenty patients with complete D0-M6 EDN kinetics in both BALF and blood were included in the longitudinal analyses (Figure S1). Detailed methods are presented in Appendix S1 and baseline clinical characteristics in Table S1. Additionally, results from a subgroup of 14 patients who underwent an additional bronchoscopy at 12 months are presented in Table S2.

Baseline EDN levels did not differ significantly between BALF and serum both in the 6 months longitudinal population (median = 46 ng/mL vs 146/mL; p = 0.9) and in the 12 months longitudinal population (median = 37 ng/mL vs 118 ng/mL; p = 0.8). At baseline, blood and BALF EDN levels were correlated, as were blood and alveolar eosinophil counts (r = 0.522, p = 0.02 and r = 0.716, p = 0.0006, respectively) (Figure 1A,B). Additionally, EDN levels and eosinophil counts were correlated within both compartments (r = 0.643, p = 0.002 and r = 0.715, p = 0.0006, respectively) (Figure 1C,D).

As expected, mepolizumab led to a significant improvement in asthma control and a reduction in the exacerbation rate (Tables S2 and S3).

EDN level dramatically decreased 6 months after mepolizumab initiation in both blood and BALF (p < 0.0001) and remained low at 12 months (p > 0.99 and p = 0.77 respectively for M6/M12 comparison) (Figure 1E,F, Tables S2 and S3). At M6, BALF EDN levels had decreased by 93% [73%–97%] from baseline, a reduction significantly greater than that observed in serum (70% [20%–82%], p = 0.02). This trend mirrored the reduction observed in blood and BALF eosinophil counts, which decreased by 83% [65%–93%] and 98% [77%–99.9%], respectively, at M6 (Figure 1E,F, Table S3).

The reduction in BALF and serum EDN levels at M6 was associated with a decrease in exacerbations (p = 0.03 and p = 0.0002, respectively) and an improvement in pulmonary function (from p = 0.03 to p = 0.002), independently of OCS intake (Table S4). There was also an association between blood and BALF eosinophil decrease and the reduction of exacerbation (p = 0.02) Additionally, the reduction in BALF EDN levels, but not in eosinophil counts, was associated with better asthma control at M6 (p = 0.03) (Table 1), highlighting the potential interest of EDN as a sensitive marker of local inflammation.

Overall, the associations between reductions in blood eosinophil counts and asthma outcomes were comparable to those observed with decreases in blood EDN levels, showing that EDN performs similarly to eosinophil count in monitoring response to mepolizumab. However, as our cohort did not include poor responders, we could not assess the potential added value of EDN in distinguishing between good and poor responders, an aspect that warrants investigation in larger populations. Notably, only the reduction in BALF EDN levels was significantly associated with improved ACT score, underscoring the potential relevance of local airway inflammation in evaluating clinical response. Future studies should explore EDN measurement in induced sputum, a less invasive and more feasible alternative to BALF. Additionally, urine may represent a promising biological matrix for EDN quantification, as urinary EDN levels are unaffected by circulating eosinophil levels or corticosteroid use [6]. Altogether, our findings support interest for EDN as a biomarker of eosinophilic inflammation in asthma, while highlighting the need to investigate minimally invasive sampling methods in future research.

The authors' contributions to the study were as follows: C.T. was involved in clinical data acquisition; V.G. was involved in EDN measurement and interpretation and wrote the manuscript; V.G., R.N., and L.O. analyzed the data and performed statistical analyses; LdC, S.C.M., and C.T. contributed to designing the study and critically revised the manuscript. All authors approved the final version of the manuscript.

Camille Taillé has received lecture or advisory board fees and grants from AstraZeneca, Sanofi, GSK, Chiesi, Stallergenes Greer, and Novartis. The other authors have no conflicts to disclose.