Eosinophil subtypes in asthmatic patients treated with mepolizumab, omalizumab and dupilumab

Abstract

Introduction

Two eosinophil subtypes, resident (rEos) and inflammatory (iEos), have been identified, first in mice and later in asthmatic patients. The effects of different biologics on these subpopulations remain unclear. This study evaluates the impact of mepolizumab, dupilumab and omalizumab on the proportion of rEos and iEos in asthmatic patients over 52 weeks.

Methods

We conducted a prospective study with 1-year follow-up of 82 patients with severe asthma: 28 were not receiving biologic therapy, 24 were treated with mepolizumab, 20 with dupilumab and 10 with omalizumab. Clinical variables included Asthma Control Test, exacerbation history and lung function. Blood samples underwent flow cytometry to classify eosinophils into (Siglec-8⁺CD62L^lo+CD11b^hi) for iEos and (Siglec-8⁺CD62L^hi+CD11b^lo) for rEos. Interleukin (IL)-5 receptor expression and levels of free IL-5 and eotaxin I/II were also assessed.

Results

Mepolizumab significantly reduced eosinophil concentrations compared with other groups (p<0.001). Patients without biologics had a high iEos proportion (23.4±13.7%), comparable with the omalizumab and dupilumab groups (31.3±12.8% and 27.6±10.1%). Mepolizumab-treated patients had markedly lower iEos proportions (2.4±4.0%, p<0.001). Stratifying iEos by quartiles, there was a significant relation between the highest quartile and lower FEV₁ % predicted (−15.6%, 95% CI 5.5–25.7) and a higher exacerbation rate (OR 2.24, 95% CI 1.02–4.91). After 1 year, iEos proportions correlated with higher systemic steroid use in mepolizumab and dupilumab groups. At 1-year follow-up, IL-5 levels were higher in omalizumab and dupilumab groups but lower in mepolizumab (p<0.01).

Conclusion

Biological therapies differentially affect eosinophil subtypes, correlating with lung function and steroid use. Monitoring iEos and rEos may serve as a marker of asthma control.

Shareable abstract

Biological therapies affect different eosinophil subpopulations (inflammatory and resident eosinophils). Mepolizumab reduced iEos markedly while dupilumab and omalizumab did not change subpopulations compared with naive patients. https://bit.ly/49BGQdS

Introduction

Asthma is a highly prevalent disease in which eosinophils play a key pathobiological role. They contribute to the inflammation and to the redundancy of the mechanisms leading to the release of inflammatory cytokines in the bronchi by other inflammatory cells [1]. Severe and poorly controlled asthma, which is seen in approximately 5% of all asthmatic patients, is characterised by a higher count of blood eosinophils in over 80% of those cases [2–4].

Asthma is currently divided into type 2 or non-type 2, with the circulating blood eosinophil count used as a clinical marker of patients likely to have underlying type 2 mechanisms driving their disease expression [4], even though there are several studies showing a low correlation in the eosinophil count in blood compared with lung tissue [5–7].

In recent years, eosinophils have been the target of several novel biological therapies. Mepolizumab, an anti-IL-5 monoclonal antibody that steeply decreases blood eosinophil count by blocking IL-5, has been shown to reduce exacerbations and the need of systemic corticosteroids [8, 9]. Dupilumab, an anti-IL-4/IL-13 receptor monoclonal antibody that does not allow eosinophils to enter lung tissue (causing transient elevation of these cells with a progressive reduction in their number over time), was also effective in decreasing the number of asthma exacerbations and the need for systemic corticosteroids [10, 11]. Benralizumab, an anti-IL-5 receptor, has shown similar efficacy in eosinophilic asthma by eliminating all eosinophils due to a receptor–antibody reaction mediated by natural killer cells and macrophages [12, 13]. Omalizumab, an anti-IgE monoclonal antibody, has also been shown to significantly reduce eosinophilic blood count and to reduce asthma exacerbations, improving its control in allergic asthma by preventing mast cells from degranulating and releasing their cytokine content and by causing a downregulation of several inflammatory receptors [14].

There is growing experimental and clinical evidence that there are at least two subtypes of eosinophils in animal models and in asthmatic patients [15–17]. These two subtypes have been labelled as inflammatory eosinophils (iEos) and resident eosinophils (rEos) depending on the molecular surface protein pattern (Siglec-8⁺CD62L^lo IL-3R^hi for iEos and Siglec-8⁺CD62L⁺IL-3R^lo for rEos). This panel has been mostly used in mouse lung tissue. Previous publications have used (Siglec-8⁺CD62L^lo+CD11b^hi) for iEos and (Siglec-8⁺CD62L^hi+CD11b^lo) for rEos in the bloodstream of COPD, asthmatic patients, smokers without COPD and healthy subjects successfully, showing how iEos has also significantly more expression of IL-5R and CD101 in these subsets. iEos are thought to have pathological effects, while the rEos population relates to homoeostatic functions [18–22]. This differentiation is useful for clinical purposes, but it is unclear whether these subsets correspond to perfectly differentiated cell subtypes or represent a continuum of expression of cell-surface markers.

We hypothesised that the currently used biological therapies could result in differences in the proportion of circulating eosinophil subtypes and in the clinical expression of disease severity in the patients on those therapies. Thus, in this study, we explored the alterations in the rEos and iEos caused by mepolizumab, dupilumab and omalizumab in asthmatic patients and compared them with those observed in matched patients not receiving any of those medications over 1 year.

Methods

Patient selections and variables

Four groups of consecutive asthmatic patients diagnosed following the Global Initiative for Asthma diagnostic criteria were recruited for the study between September 2021 and October 2024 [3]: 1) 28 bio-naive moderate-to-severe asthmatic patients; 2) 24 patients treated with mepolizumab; 3) 10 patients on omalizumab; and 4) 20 patients receiving dupilumab. Data on eosinophils before the inclusion of the biological treatment were also obtained. Patients were followed for 1 year to see the changes in the eosinophil pattern through time and its relationship with clinical outcomes, being the baseline the time of the first measure on biologic therapy. The clinical variables collected were the Asthma Control Test (ACT) questionnaire, comorbidities, pulmonary function test, medication received and exacerbations. Exacerbations were defined as visits to the emergency room that led to a course of systemic steroids (CSS) for at least 3 days. Bursts of CSS following domiciliary treatment plans were considered differently. All patients on biological therapy had been treated for at least 3 months, with the medications chosen by their treating specialist according to the accepted criteria for their appropriate use. All patients had less than 10 pack-years. The patients had to be in stable condition, with at least 3 months since the previous exacerbation, and without systemic steroids. The exclusion criteria were any cancer in the last 5 years, eosinophilic diseases different from asthma (EGPA, eosinophilic esophagitis, eosinophilic cystitis or hypereosinophilic syndrome), or being unable to perform spirometry. This study was approved by the ethics committee of the Hospital Universitario de Gran Canaria Dr Negrín (code: 2022-244-1).

Eosinophil phenotypes identification

The methods used in this study have been previously published [17]. In summary, using flow cytometry, the eosinophils were labelled following the previously published criteria in two groups: (Siglec-8⁺CD62L^lo+CD11b^hi) for iEos and (Siglec-8⁺CD62L^hi+CD11b^lo) for rEos (figure 1) [17]. We also measured other proteins not included in the Mesnil pattern as CD125 and CD101. We used a fluorescence-activated cell sorting (FACS) Canto (BD Biosciences, USA). Moreover, eosinophils from peripheral blood were sorted from the granulocyte suspension by FACS Melody Cell Sorter (BD Biosciences, Franklin Lakes, NJ, --USA) based on forward scatter and side scatter characteristics, CD11b⁺CD16⁻ and Siglec-8^hi expression. We also used as a second validation method confocal microscopy with a LSM800 Airyscan microscopy (Zeiss, Germany). Eotaxins and IL-5 were measured by ELISA.

FIGURE 1.

Gating strategy for delineating human blood eosinophils. a) Exclusion of doublets and debris. FSC: forward scatter; SSC: side scatter. b) Cells were gated and analysed for CD11b⁺ SSC^hi, CD16⁻ and Siglec-8^hi and selection of events with loss of CD62L (inflammatory eosinophils; iEos). Figure from Cabrera Lopez et al. [17].

Statistical analysis

Numerical variables are shown by mean±sd or median (interquartile range), depending on their Gaussian distribution, while qualitative variables are shown by frequency and percentage. The chi-squared test or ANOVA, and Kruskal–Wallis test are used for the comparison of qualitative or quantitative variables, respectively. Association was performed using the Pearson or Spearman correlation coefficient. A p-value<0.05 was considered statistically significant.

The software used for the analysis was R Core Team 2019, v.3.6.2 or greater (R Foundation for Statistical Computing, Vienna, Austria) [23].

Results

The clinical characteristics of the study participants are shown in table 1. They were primarily middle-aged patients (55.6 years), with 54% being women. The mean body mass index (BMI) of the whole group was 28.2 kg·m⁻² with a mean ACT score of 21.5. Patients receiving omalizumab had higher IgE levels and were more likely to be receiving inhaled long-acting muscarinic antagonists, whereas patients on dupilumab had more nasal polyps and aspirin-exacerbated respiratory disease (AERD).

TABLE 1.

Baseline characteristics stratified by biologic therapy of the subjects included in the study

	Treatment				p-value	Total
	Non-biologics	Mepolizumab	Omalizumab	Dupilumab		(n=82)
	(n=28)	(n=24)	(n=10)	(n=20)
Age, years	56±12	59±14	55±10	51±11	0.220	55.6±12.4
Sex, women	12 (43)	16 (67)	6 (60)	10 (50)	0.359	44 (54)
BMI, kg·m−2	27.4±5.6	27.9±3.9	29.4±4.2	28.9±5.6	0.633	28.2±5.0
Tobacco					0.643
Pack-years	0 (0–10)	0 (0–0.75)	0 (0–3.7)	0 (0–7.5)	0.441	0 (0–8.5)
ACT	N/A	21.6±5.1	20.4±3.9	22.0±3.5	0.636	21.5±4.3
AERD	6 (21)	5 (21)	N/A	9 (45)	0.043	20 (24)
Rhinitis	15 (54)	15 (63)	5 (50)	14 (70)	0.617	49 (60)
Nasal polyps	13 (46)	11 (46)	2 (20)	15 (75)	0.031	41 (50)
GERD	N/A	6 (25)	4 (40)	2 (10)	0.160	12 (22)
Atopic dermatitis	2 (7)	4 (17)	1 (10)	6 (30)	0.182	13 (16)
IgE, UI·mL−1	100.6 (34.5–195.4)	33.2 (17.4–152.7)	209.5 (135.5–395.0)	46.3 (19.3–88.5)	0.007	65.8 (22.6–171.1)
Exacerbations previous year					0.013
0	18 (64)	20 (83)	5 (50)	15 (75)		58 (71)
1	2 (7)	4 (17)	N/A	3 (15)		8 (11)
≥2	8 (29)	N/A	5 (50)	2 (10)		15 (18)
ICS LABA	23 (82)	22 (92)	10 (100)	19 (95)	0.291	74 (90)
LAMA	9 (32)	11 (46)	8 (80)	4 (20)	0.011	32 (39)
Nasal steroids	15 (54)	1 (4)	2 (20)	9 (45)	<0.001	27 (33)
Antihistamine	7 (25)	1 (4)	2 (20)	3 (15)	0.226	13 (16)
FEV1, L	2.21±0.77	2.23±0.80	1.76±0.60	2.62±0.93	0.056	2.26±0.83
FEV1, % predicted	77±21	79±22	66±25	87±17	0.091	78±21

Data are presented as n (%), mean±sd or median (interquartile range). BMI: body mass index; ACT: Asthma Control Test; AERD: aspirin-exacerbated respiratory disease; GERD: gastro-oesophageal reflux disease; N/A: not available; ICS: inhaled corticosteroid; LABA: long-acting β-agonist; LAMA: long-acting muscarinic antagonist; FEV₁: forced expiratory volume in 1 s.

The eosinophil count at different thresholds for all the groups is shown in table 2. There were differences in the number of eosinophils before treatment in the patients in the different biological therapies. Patients started on dupilumab had lower blood eosinophil count (BEC) than patients selected to receive omalizumab and mepolizumab or those not on biological therapy. The BEC decreased significantly in patients on mepolizumab and omalizumab, with a more important decrease in those receiving mepolizumab. Patients on dupilumab had an increase in the number of blood eosinophils. At 1 year, patients on mepolizumab continued with the lowest proportion of eosinophils, followed by omalizumab and dupilumab showed an increased number of eosinophils in blood.

TABLE 2.

Eosinophils by biological treatment at the beginning of the study

	Treatment				p-value	Total
	Non-biologics	Mepolizumab	Omalizumab	Dupilumab		(n=82)
	(n=28)	(n=24)	(n=10)	(n=20)
Eosinophils before treatment	546 (130–1680)	665 (550–1018)	454 (263–598)	370 (80–620)	0.001#	550 (370–820)
Eosinophils before treatment					0.014
≤300	13 (46)	1 (4)	3 (30)	8 (40)		12 (22)
>300	15 (54)	23 (96)	7 (70)	12 (60)		42 (78)
Eosinophils after treatment	375 (240–675)	95 (55–145)	255 (140–650)	620 (285–800)	<0.001*	265 (120–650)
Eosinophils after treatment					<0.001
<150	2 (7)	20 (83)	3 (30)	4 (20)		29 (35)
150–300	11 (39)	3 (13)	2 (20)	2 (10)		18 (22)
>300–500	6 (21)	1 (4)	2 (20)	2 (10)		11 (14)
>500	9 (33)	N/A	3 (30)	12 (60)		24 (29)
Eosinophils after treatment					<0.001
≤300	13 (46)	23 (96)	5 (50)	6 (30)		47 (57)
>300	15 (54)	1 (4)	5 (50)	14 (70)		35 (43)

Data are presented as n (%) or median (interquartile range). Eosinophil counts are in cells·µL⁻¹. N/A: not available. ^#: p=0.001 between dupilumab versus mepolizumab and between dupilumab versus subjects without biological treatment. *: p<0.05 between eosinophils before and after treatment for all groups.

At the beginning of the study (time of the first collection of the blood sample), the proportion of the two eosinophil subtypes in the four groups differed as shown in table 3 and figure 2. The proportion of iEos (% and sd) in the asthmatic population without biological therapy was 23±14, similar to those previously published [17]. The proportion of iEos was 94% lower in patients on mepolizumab, while patients on omalizumab or dupilumab maintained a proportion that was similar to that of asthmatic patients not receiving biological therapy. These findings were confirmed by using confocal microscopy (supplementary figure S1). Those patients at extreme values of iEos (the upper quartile of iEos for mepolizumab and omalizumab, or the lower quartile of iEos for dupilumab) had a significantly lower FEV₁ % predicted (−15,6%, 95% CI 5.5–25.7; p=0.003) and a higher rate of exacerbations (OR 2.24, 95% CI 1.02–4.91; p=0.04) than those with less iEos (figure 3).

TABLE 3.

Percentage, total number of eosinophil subtypes and differences in interleukin-5 and eotaxins I and II by biologic treatment

	Treatment				p-value
	Non-biologic	Mepolizumab	Omalizumab	Dupilumab
	(n=28)	(n=24)	(n=10)	(n=20)
Blood rEos %, 1 year	76.6±13.9	97.6±4.0	72.5±10.6	72.0±9.7	<0.001
Blood rEos %, 2 years	74.9±14.3	97.2±3.06	89.8±7.4	87±12.8	<0.001
Blood iEos %, 1 year	23.4±13.7	2.4±4.0	31.3±12.8	27.6±10.1	<0.001
Blood iEos %, 2 years	26.9±15.2	2.63±3.1	10.16±7.4	12.9±13.4	<0.001
IL-5, pg·mL−1, 1 year	10.6	5.8	14.1	13.4	0.001
	(7.3–12.0)	(4.7–9.4)	(9.1–15.5)	(9.9–17.1)
Eotaxin I, pg·mL−1 at baseline	94	87	105	208	0.152
	(68–169)	(66–125)	(47–338)	(91–1000)
Eotaxin II, pg·mL−1 at baseline	345	864	720	741	<0.001
	(266–448)	(657–1114)	(565–841)	(425–1083)

Data are presented as median (interquartile range) or mean±sd. rEos: resident eosinophils; iEos: inflammatory eosinophils.

FIGURE 2.

Proportion of patients with inflammatory eosinophils (iEos) at varying thresholds in treatment-naive patients compared with those receiving biological therapies (mepolizumab, omalizumab and dupilumab).

FIGURE 3.

a) Percentage of patients with one or more exacerbations (measured by mean±sd) by extreme and non-extreme proportions of inflammatory eosinophils (iEos) at the beginning of the study and at 1 year. b) Forced expiratory volume in 1 s (FEV₁) % predicted (measured by mean±sd) by extreme and non-extreme proportions of iEos at the beginning of the study and at 1 year. The orange line represents the patients with a proportion (%) of iEos at extreme values (highest quartile for mepolizumab, omalizumab and naive patients and lowest quartile for dupilumab: naive>33%, mepolizumab>4.15%, omalizumab>36.6% and dupilumab<20.2%), while the blue line represents the patients with a proportion (%) of iEos not at extreme values. There is a higher rate of patients with one or more exacerbations, and a lower FEV₁ % predicted at iEos extreme values.

At 1 year, mepolizumab did not change the proportion of iEos, whereas omalizumab and dupilumab decreased this subpopulation significantly, although remaining higher than mepolizumab (supplementary figure S2). Compared with the control group, free IL-5 was reduced as expected in the mepolizumab group, while it was increased in the omalizumab and dupilumab groups (due to the lower sample. size, this analysis was performed by a Holm-adjusted p-value for patients without biologic therapy and patients with dupilumab and omalizumab merged in one group). There were no differences in the eotaxin I, but a trend to higher levels was observed in the dupilumab group, whereas there were significant differences in eotaxin II levels between naive patients and those having biologic therapy (table 3). Differences in CD11b and CD125 expression were observed across groups and between eosinophil subsets (supplementary table S1 and supplementary figure S5). The proportion of iEos at 1 year was again inversely related to the FEV₁ % predicted in naive patients as well as those in mepolizumab and omalizumab (p<0.05) but not in the dupilumab group, although a trend was seen, and it was significantly associated when we used the iEos lowest quartile (figure 4 and supplementary figure S3). There was also a trend relating a higher proportion of iEos to a lower forced vital capacity (FVC) and FEV₁/FVC for all groups, although not significant (supplementary figure S4). ACT and exacerbations were not related in any treatment with the proportion of iEos, but CSS cycles (counting those taken on exacerbations plus those following a treatment plan) were related to a high proportion of iEos in mepolizumab and to a low proportion of iEos in dupilumab (figure 5).

FIGURE 4.

Relationship between forced expiratory volume in 1 s (FEV₁) % predicted and the proportion (%) of inflammatory eosinophils (iEos) in the different groups.

FIGURE 5.

Relationship between the proportion (%) of inflammatory eosinophils (iEos) and steroid cycles in patients treated with mepolizumab, omalizumab and dupilumab

Discussion

To our knowledge, this is the first study that shows the differences in the proportion of circulating inflammatory and resident eosinophil subtypes in asthmatic patients being treated with several biological therapies compared with controls not on biologicals and their longitudinal changes. These results provide some insight as to the effect of these therapies on eosinophils subtypes, their relationship with clinical variables and overall asthma pathophysiology.

Baseline differences were observed across treatment groups. As expected, patients receiving omalizumab exhibited higher IgE levels, whereas those on dupilumab showed a greater prevalence of nasal polyps and AERD, followed by mepolizumab and omalizumab. This distribution likely reflects the efficacy of dupilumab and mepolizumab in nasal polyposis, with affected patients being preferentially treated with these agents. The increased prevalence of AERD in the dupilumab group is probably secondary to the higher burden of nasal polyps.

Compared with the baseline, there was a large decrease in the absolute BEC in patients receiving mepolizumab, which was not observed in the group receiving omalizumab, and this contrasted with the increase that occurred in the dupilumab group. This mirrored the very low proportion of inflammatory circulating eosinophils seen in the mepolizumab-treated patients compared with the other two groups, which have close to the same proportion of iEos as the control people with asthma on no biologic therapy. The decrease in the total number of eosinophils, as well as in the proportion of iEos with mepolizumab in our study, is similar to those published previously by Vultaggio et al. [24] and Fricker et al. [25], and different from those shown by Rimkunas et al. [26], who found a greater proportion of iEos in patients with severe asthma without biologic treatment with a lower reduction in the iEos population when start using mepolizumab. This could be due to differences in the gating strategy using just CD62L versus the whole panel used in this research. In our study, over 80% of the patients on mepolizumab reported no exacerbations, and no patient had two or more exacerbations in the previous year, a finding that is similar to those of Fricker et al. [25] and Vultaggio et al. [24] who also described a relationship between the number of exacerbations and asthma control with the iEos population in patients on mepolizumab. In our study, we observed a relationship between the proportion of iEos and the occurrence of exacerbations in the previous year. We also identified an association between a lower percentage of iEos and a reduced number of CSS cycles during the follow-up period in patients treated with mepolizumab. This finding can be attributed to the distinction between exacerbations and CSS bursts in our study. Notably, patients receiving biologic therapy are typically equipped with written action plans and are trained to manage worsening symptoms independently, thereby reducing their reliance on emergency room visits.

The effect of mepolizumab in iEos is probably related to the blockage of IL-5, as this cytokine is the main driver in the proliferation, maturation and activation of eosinophils [27]. Probably, those patients on mepolizumab with a higher proportion of iEos reflect an insufficient effect of mepolizumab, and therefore, more eosinophils reaching the bronchi. Omalizumab also decreased the total number of blood eosinophils by 46% in our study, a decrease that is higher than the 20% reported by Massanari et al. [14], who attributed this reduction to the levels of free IgE. The reduction in free IgE induced by omalizumab prevents mast cells from degranulation and activation of dendritic cells due to the reduction of their high-affinity receptors [28]. Consequently, IL-5 usually decreases moderately, and so do eosinophils. Importantly, although the proportion of inflammatory eosinophils remained unchanged in patients receiving omalizumab compared with those not on biologic therapy, the absolute number of these cells in circulation was significantly reduced, nearly halved, due to the overall decrease in total eosinophil count. This reduction in absolute eosinophil burden may contribute to the clinical benefits observed with omalizumab therapy [29, 30].

Dupilumab, on the other hand, did not decrease eosinophils in blood, but it is known to drastically lower their ability to enter the lungs [31]. In fact, dupilumab can increase the total eosinophil count in blood after its administration, as indeed happened in our study [10]. This could be explained by the effect of IL-5 on the bone marrow, where it drives eosinophil generation that reaches and remains in the bloodstream as they are not able to enter tissue due to the inhibiting effect of blocking IL-4/IL-13, causing a decreased expression of eotaxins, VCAM in endothelial cells and periostin [32–34]. In our study, we observed that patients on dupilumab had the same proportion of iEos as people with asthma without monoclonal antibodies, but due to the increase in the total number of eosinophils, they actually had an increased number of absolute iEos compared with the controls. This suggests that the benefit reported by dupilumab in patients with severe asthma is not driven by a decrease in the number of inflammatory eosinophils in blood, but by other mechanisms, most likely due to not being able to reach the airways. Supporting this perspective, our findings reveal a contrasting relationship between the percentage of iEos and systemic CSS cycles in patients treated with dupilumab compared with those on mepolizumab. In the dupilumab group, higher iEos percentages were associated with fewer exacerbations and CSS bursts, a pattern opposite to that observed with the anti-IL-5 therapy. This discrepancy may be explained by the inability of iEos to migrate to tissues in patients on dupilumab, leading to their accumulation in the bloodstream. Conversely, the lower proportion of iEos in the blood observed in this group may indicate increased migration by escape mechanisms of these cells to tissues, potentially exacerbating airway inflammation and contributing to a loss of disease control. This perspective is further supported by the observed association between dupilumab-treated patients in the lowest quartile of inflammatory eosinophils and both reduced FEV₁ % predicted and a higher rate of exacerbations (supplementary figure S4).

Another interesting finding of our study is how IL-5 levels change with the different treatments. Patients with omalizumab and dupilumab show a higher level of IL-5 than patients not in biological therapy and those on mepolizumab. This could suggest an early upregulation in the blood of this cytokine, leading to the observed uprising in the iEos subpopulation in those groups. It is also interesting how the proportion of iEos lowers in the dupilumab and omalizumab groups one year later, suggesting a downregulation of these inflammatory cytokines over time when patients are on these biologic drugs. However, the variation over time should be explored specifically in the future.

The findings in this study expand our knowledge about the effects of different biologics on eosinophil subtypes but leave many areas in need of research. Interestingly, the three biologics studied resulted in clinical benefits as shown by the similar ACT scores as well as a relatively low number of reported exacerbations. However, in our study, iEos were not related to ACT scores, opposite to the findings in pulmonary function and systemic corticosteroids use.

The analysis of eotaxins revealed significant differences between the groups. Eotaxins are classified into three types: eotaxin I, II and III. There was a trend indicating a higher concentration of eotaxin I in patients treated with dupilumab. This phenomenon could be attributed to the elevated eosinophil blood count observed in the dupilumab group, as eotaxin I is released by eosinophils, among other cell types [35]. In contrast, patients not undergoing biologic therapy exhibited significantly lower levels of eotaxin II. This disparity may be explained by the difference in asthma severity and/or by an upregulation mechanism aimed at recruiting more eosinophils from the microcirculation and mobilising them from the bone marrow into the tissue [36].

Our study has several weaknesses. First, while baseline eosinophil counts were readily available from medical records, we were unable to obtain measurements of eosinophil subtypes before the initiation of biologic therapy. Second, our study focused solely on circulating eosinophils, without assessing eosinophils in sputum or airway samples. Nevertheless, circulating BECs remain the most widely utilised marker of type 2 inflammation and are instrumental in identifying patients likely to benefit from biologic treatments. Third, the sample size, particularly in the omalizumab group, may have influenced our findings, as this group started with 10 patients but only five remained by the end of the study. Finally, eotaxins and IL-5 levels were measured exclusively at baseline, limiting our ability to assess their longitudinal patterns and their association with eosinophil subtypes over time. This investigation has several strengths, nonetheless. We used a complete pattern of markers for eosinophil differentiation and asthma-related cytokines which is more accurate than using CD62L alone. In addition, we had a well-phenotyped control group that can help determine the magnitude and relative changes of the BEC and subtypes in the biologically treated patients. We also followed patients for 1 year, relating clinical variables as well as the effects of these biological therapies to the changes in the eosinophils’ subpopulations.

In summary, we show how the concentration of iEos is high in asthmatic patients and differs between subjects when they are on different biologic treatments. Mepolizumab clearly reduced this eosinophil subpopulation, whereas omalizumab preserved the proportion of inflammatory eosinophils and dupilumab led to an increase in both the total eosinophil count and the inflammatory subset. Finally, iEos are related to pulmonary function tests and CSS bursts in mepolizumab and dupilumab. These findings could help to explain the beneficial effect of biologics in asthmatic patients.

Supplementary material

Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.

Supplementary material

00763-2025.SUPPLEMENT