Mepolizumab does not alter the blood basophil count in severe asthma

Additional supporting information may be found online in the 
Supporting Information section at the end of the article.


S U P P O R T I N G I N F O R M AT I O N
Additional supporting information may be found online in the Supporting Information section at the end of the article.

Mepolizumab does not alter the blood basophil count in severe asthma
To the Editor: Mepolizumab (anti-IL-5) depletes blood and airway eosinophils, and, clinically , allows down-titration of oral corticosteroid and a reduction in the frequency of eosinophil-dependent exacerbations. 1 Basophils also express IL-5Rα, participate in T2-mediated inflammatory pathways 2 and have been associated with exacerbation frequency. 3 Whilst basophil progenitors are unlikely to depend on IL-5 for development, 4 blood basophil counts measured in routine clinical laboratories suggest they decrease following mepolizumab treatment. [5][6][7][8] Our primary objective was to determine whether anti-IL-5 monoclonal antibody treatment reduces blood basophil levels as an additional potential efficacy mechanism. To achieve this, we measured blood basophils, eosinophils and other type 2 inflammatory cells, before and after 16 weeks of mepolizumab ("Nucala," GlaxoSmithKline) by flow cytometry. Patient eligibility criteria are in the online supplement and the study schedule in Figure S1 .
Blood samples were obtained from 26 severe asthma subjects, attending a difficult asthma clinic at a single UK centre, at baseline and following a median (IQR) of 16 (16-17) weeks of mepolizumab, administered as a 100 mg subcutaneous injection every 4 weeks. In 2 cases, it was not possible to obtain post-treatment samples (n = 1, withdrew consent; n = 1, discontinued), totalling 24 ( Figure S1 and Table S1 ). We also recruited 15 nonasthmatic healthy controls (Table   S1 ) to obtain samples at parallel time points but without an intervention ( Figure S1 ). Flow cytometric measurements were compared with data derived through the routine pathology service, which utilizes an ADVIA 2120/2120i analyser (Siemens, UK). A detailed description of the methodology for both approaches is described in the online supplement, and for flow cytometry, the gating strategy is shown in Figure S2 . Our criteria for identifying cell subsets were as follows: eosinophils (CD45 + CD3 -CD193 + CD294 + SSC hi CD123 −/+ ), basophils (CD45 + CD3 -CD193 + CD294 + SSC lo CD123 + ), cTH2 and peTH2 (both and ILC2s (Lineage − CD294 + CD161 + ).
In contrast to eosinophils, basophil concentration and frequency at baseline were similar to that following 16 weeks of  (Figures 1 B and S4 ). These data suggest that basophil concentration and frequency, alongside other T2 inflammatory cells ( Figure S6 ), are likely to be IL-5/mepolizumab-independent in severe asthma. However, our real-world study was not sufficiently powered to detect small differences in relation to basophil concentration or frequency. A strength of our flow cytometric approach is that we have measured cell concentration as well as frequency, and also reported recently identified T2 cell subsets (eg peTH2 cells). Our data suggest there were also no indirect effects of mepolizumab on type-2 polarised T cell or group 2 ILC concentration over this 16-week time frame as indicated by others. 5 Clinically, we observed a significant change in ACQ6 symptom score from a baseline of 2.9 ± 1.6 to 1.  Figure S2 and cumulative in S7 ). This represents a % decrease of −54 ± 25% in asthma compared with +9 ± 55% in the healthy group (mean ± SD, P < 0.0001, Mann-Whitney test). We also noted that eosinophil expression of CRTH2 (CD294) was increased following mepolizumab ( Figure S7 ); however, there was no relationship between the reduction in eosinophil IL-3Rα expression and CRTH2 expression. Importantly, there was no relationship between changes in eosinophil IL-3Ra or CRTH2 expression with ΔACQ6 in these patients.
Furthermore, there were no mepolizumab-dependent effects on eosinophil/basophil Siglec-8, CD69 or IL-5Rα expression (not shown), consistent with others 8 or basophil IL-3Rα (Figure S7 ), . and thus, these parameters were not examined for a relationship with ΔACQ6.
In summary, our flow cytometric data do not support a direct inhibitory effect of mepolizumab on basophil levels, and therefore, clinical benefit is likely to be independent of basophils. Our data suggest that the specificity and sensitivity of basophil detection on routine clinical analysers should be validated prior to reporting/interpreting basophil data in the context of an intervention. Our data do support others 8 that mepolizumab reduces eosinophil, but not basophil, IL-3Rα expression and, importantly, extends the applicability of this phenomenon to the "real-world" scenario. However, neither changes in eosinophil levels nor changes in IL-3Rα expression were associated with clinical efficacy determined by change in asthma control in this study, and thus, biological correlates of response to treatment require further study. Thus, the use of disease biomarkers might help the clinical practice.
MicroRNAs (miRNAs) are small noncoding RNAs associated with disease mechanisms and have been described as good biomarkers.
We have previously described that differential miRNAs expression obtained from eosinophils of asthmatics compared with healthy individuals was able to cluster asthmatics and healthy subjects. 4 Five of these miRNAs (miR-1246, miR-144-5p, miR-320a, miR-185-5p and miR-21-5p) measured in serum were able to differentiate asthmatics from healthy subjects by statistical models. 4 We hypothesized that some of these miRNAs could be specific to asthma, and they may differentiate asthma from other respiratory diseases. Therefore, the aim of the study was to compare these miRNAs between asthmatics, COPD and ACO and to analyse whether they could be used as asthma biomarkers.
Rodrigo-Muñoz and Rial equally contributed to this publication.