Exploring the risk of infection events in patients with asthma receiving anti-IL-5 monoclonal antibodies: A rapid systematic review and a meta-analysis

Introduction Benralizumab, mepolizumab, and reslizumab are novel monoclonal antibodies approved for asthma, targeting eosinophilic inflammation. Benralizumab is directed against IL-5 receptor (IL-5R), while mepolizumab and reslizumab are directed against IL-5. The three drugs cause a reduction in eosinophils, but benralizumab also causes a cytotoxic effect on eosinophils and basophils. Recently, it has been reported that suboptimal responders to benralizumab presented exacerbations associated with concomitant infections and sputum neutrophilia and the incidence of infections was greater in patients receiving benralizumab compared to mepolizumab and reslizumab. For this reason, we wanted to explore potential differences in terms of infectious adverse events between the three different anti-IL-5 antibodies. Methods We performed a rapid systematic review on PubMed up to April 28, 2022. We included randomized controlled trials (RCTs) evaluating benralizumab, mepolizumab, or reslizumab in patients with asthma. Included outcomes were the reporting of any respiratory tract infection and any emergency department (ED) or hospital admission for infection or asthma exacerbation. A Mantel-Haenszel meta-analysis was performed with Cochrane RevMan 5.4 to estimate pooled odds ratios (OR) with 95 % confidence intervals (CI). A subgroup analysis for the different active treatments was performed. Results From 163 references we included 21 studies reporting the results of 23 different RCTs for a total population of 9156 patients. All studies compared anti-IL-5 antibodies against placebo. Anti-IL-5 treatment resulted in non-significant differences compared to placebo in the odds for nasopharyngitis (OR = 0.90; 95 % CI from 0.76 to 1.07), pharyngitis (OR = 1.45; 95 % CI from 0.92 to 2.28), upper respiratory tract infection (URTI) (OR = 0.97; 95 % CI from 0.82 to 1.15), rhinitis (OR = 1.01; 95 % CI from 0.71 to 1.44), pneumonia (OR = 0.56; 95 % CI from 0.10 to 2.01), and influenza (OR = 0.84; 95 % CI from 0.65 to 1.09). We observed significant reductions in the reporting of sinusitis (OR = 0.75; 95 % CI from 0.53 to 1.06), bronchitis (OR = 0.71; 95 % CI from 0.59 to 0.86), and ED or hospital admission due to asthma exacerbation for overall anti-IL-5 antibodies compared to placebo (OR = 0.59; 95 % CI from 0.40 to 0.88). We were not able to discriminate whether exacerbations were associated with infections or to increased sputum eosinophilia. From the subgroup analysis, we observed differences in directions and magnitudes of the effect size in the reporting of some events. Benralizumab was associated with increased odds of pharyngitis (OR = 1.56; 95 % CI from 0.97 to 2.52) and a similar trend was observed for mepolizumab in the reporting of rhinitis (OR = 1.85; 95 % CI from 0.72 to 4.78), both non-statistically significant. In terms of effect size, benralizumab also showed higher odds for bronchitis and pneumonia in comparison to mepolizumab and reslizumab (OR = 0.76, OR = 0.69, and OR = 0.60 for bronchitis and OR = 0.80, OR = 0.20, and OR = 0.45, respectively, all non-significant). Conclusion Anti-IL-5 treatments might have different effects on the reporting of some infection events in patients with asthma. However, the evidence is limited by sample size and far than conclusive and suggest the need of future studies to evaluate the risk of infections in patients with asthma receiving anti-IL-5 treatments.

Introduction: Benralizumab, mepolizumab, and reslizumab are novel monoclonal antibodies approved for asthma, targeting eosinophilic inflammation.Benralizumab is directed against IL-5 receptor (IL-5R), while mepolizumab and reslizumab are directed against IL-5.The three drugs cause a reduction in eosinophils, but benralizumab also causes a cytotoxic effect on eosinophils and basophils.Recently, it has been reported that suboptimal responders to benralizumab presented exacerbations associated with concomitant infections and sputum neutrophilia and the incidence of infections was greater in patients receiving benralizumab compared to mepolizumab and reslizumab.For this reason, we wanted to explore potential differences in terms of infectious adverse events between the three different anti-IL-5 antibodies.Methods: We performed a rapid systematic review on PubMed up to April 28, 2022.We included randomized controlled trials (RCTs) evaluating benralizumab, mepolizumab, or reslizumab in patients with asthma.Included outcomes were the reporting of any respiratory tract infection and any emergency department (ED) or hospital admission for infection or asthma exacerbation.A Mantel-Haenszel meta-analysis was performed with Cochrane RevMan 5.4 to estimate pooled odds ratios (OR) with 95 % confidence intervals (CI).A subgroup analysis for the different active treatments was performed.Results: From 163 references we included 21 studies reporting the results of 23 different RCTs for a total population of 9156 patients.All studies compared anti-IL-5 antibodies against placebo.Anti-IL-5 treatment resulted in non-significant differences compared to placebo in the odds for nasopharyngitis (OR = 0.90; 95 % CI from 0.76 to 1.07), pharyngitis (OR = 1.45; 95 % CI from 0.92 to 2.28), upper respiratory tract infection (URTI) (OR = 0.97; 95 % CI from 0.82 to 1.15), rhinitis (OR = 1.01; 95 % CI from 0.71 to 1.44), pneumonia (OR = 0.56; 95 % CI from 0.10 to 2.01), and influenza (OR = 0.84; 95 % CI from 0.65 to 1.09).We observed significant reductions in the reporting of sinusitis (OR = 0.75; 95 % CI from 0.53 to 1.06), bronchitis (OR = 0.71; 95 % CI from 0.59 to 0.86), and ED or hospital admission due to asthma exacerbation for overall anti-IL-5 antibodies compared to placebo (OR = 0.59; 95 % CI from 0.40 to 0.88).We were not able to discriminate whether exacerbations were associated with infections or to increased sputum eosinophilia.From the subgroup analysis, we observed differences in directions and magnitudes

Introduction
Novel treatments for asthma are directed against eosinophilic inflammation by targeting IL-5.IL-5 is a homodimeric cytokine which acts as the primary modulator of eosinophils.In Europe, three monoclonal antibodies directed against IL-5 have been approved.Mepolizumab and reslizumab bind directly to circulating IL-5 and reduce eosinophil counts by inhibiting IL-5 signalling.Benralizumab, on the other hand, is a monoclonal antibody directed against the IL-5 receptor (IL-5R) alpha chain with a cytotoxic effect capable of completely depleting the eosinophilic as well as basophilic population [1].
The role of eosinophils in relation to the pathophysiology of asthma has been extensively discussed in the literature [2].A high eosinophil count is associated with more disease severity [3].Eosinophils accumulate in the lungs in the inflammatory setting causing tissue damage and promoting Th2-mediated inflammatory signalling [4].
On the other hand, the role of eosinophils in protecting the body from infection remains debated.Traditionally, the function of eosinophils has traditionally been associated with protection against parasitic infections.However, studies have also shown they are involved in fungal [5,6] and viral infections [7].For example, it has recently been reported that eosinopenia is linked with acute respiratory deterioration during SARS-CoV2 infection [8,9].Other studies associated eosinopenia with poor outcomes in patients with acute exacerbations of chronic obstructive pulmonary disease (COPD) [10] and identified eosinopenia as a possible marker of severe infection and sepsis [11].
A recent prospective cohort study reported a suboptimal response, namely the presence of exacerbations or failure to reduce prednisone by at least 50 %, in about 27 % of patients receiving benralizumab.Authors found that only a minority of exacerbations in this group of patients were associated with sputum eosinophilia.Conversely, the majority of them were associated with concomitant infections and sputum neutrophilia.Also, the incidence of respiratory infections increased in the assessed population while receiving benralizumab and was significantly greater compared to that observed in a group of patients receiving mepolizumab or reslizumab [12].
Given these observations, we decided to conduct a rapid systematic review and a meta-analysis including randomized controlled trials (RCTs) evaluating anti-IL-5 monoclonal antibodies for asthma to explore potential differences in terms of the reporting of infectious adverse events.

Search strategy and selection criteria
We performed a literature search on PubMed from the beginning up to August 28, 2023 to include RCTs fulfilling our inclusion criteria.The search strategy included "benralizumab", "mepolizumab", "reslizumab", "asthma", and "RCTs" as keywords used as MeSH and free terms and combined with Boolean operators (Supplementary Material S1).We included full-text English articles reporting data of RCTs, in patients with asthma of any age, evaluating benralizumab, mepolizumab, or reslizumab against placebo or any active comparator.Observational studies, case series, case reports, reviews, editorials, commentaries, congress abstracts, and study with no outcomes were excluded.Two study authors (R.G. and A.P.) independently screened literature citations for inclusion and discrepancies were resolved by collegial discussion.

Assessed outcomes
Included outcomes of our systematic review were the reporting of any respiratory tract infection coded according to the MedDRA dictionary.Infection events were collected separately for each available category from included RCTs.Additional outcomes were emergency department (ED) or hospital admission for infection and ED or hospital admission asthma exacerbation.Respiratory tract infections were defined as upper respiratory tract infections (URTI), including nasopharyngitis, pharyngitis, sinusitis, acute sinusitis, and rhinitis and lower respiratory tract infections (LRTI), such as bronchitis, pneumonia, and influenza.R. Giossi et al.

Data extraction
Two study authors (R.G. and A.P.) independently extracted included study data on an Excel spreadsheet and discrepancies were resolved by discussion.Extracted data consisted in: first author, year, period, study duration, patients age and sex, RCTs main inclusion and exclusion criteria, total patient included, the different treatment and anti-IL-5 dose regimens, and prespecified included outcomes.

Statistical analysis
Events from outcomes of interest were collected from included studies.Whenever the number of events was not completely reported, we estimated the number of adverse events from available rates or odds ratios (ORs), confidence intervals (CI), or p-values.For all comparisons, we used a random-effect Mantel-Haenszel meta-analysis performed with Cochrane RevMan 5.4 software.Heterogeneity was evaluated with the I-squared statistic.Pooled estimates were reported as ORs and 95%CI.A subgroup analysis was  Abbreviations: CS = corticosteroids; EO = eosinophilic; ICS = inhalatory corticosteroids; LABA = long-acting beta agonists; SABA = short-acting beta agonists.

R. Giossi et al.
performed based on the different anti-IL-5 to allow comparisons in-between treatments.

Role of the funding source and ethical considerations
The study was not sponsored and did not receive any funding.Due to the nature of the study on already published data without the involvement of new human participants or animals an IRB approval was not necessary.

Lower respiratory tract infections and influenza
From included RCTs, we managed to extract data on the following lower respiratory tract infections: bronchitis, pneumonia, and influenza.Anti-IL-5 treatment was associated with a significant reduction of bronchitis (OR = 0.71; 95 % CI from 0.59 to 0.86).Similar results were obtained in the subgroup analysis, with, in order of effect size from the lower to higher, a significant reduction in the odds for benralizumab (OR = 0.76; 95 % CI from 0.60 to 0.96), a non-significant reduction in the odds of mepolizumab (OR = 0.69; 95 % CI from 0.40 to 1.20), and a significant reduction in the odds of reslizumab (OR = 0.60; 95 % CI from 0.40 to 0.90), compared to placebo (Fig. 8).Conversely, anti-IL-5 treatment resulted in non-significant reduction of pneumonia (OR = 0.56; 95 % CI from 0.10 to 2.01).In terms of effect size, among the three treatments, mepolizumab showed the lowest odds for pneumonia (OR = 0.20; 95 % CI from 0.02 to 1.80), followed by reslizumab (OR = 0.45; 95 % CI from 0.10 to 2.01), and benralizumab (OR = 0.80; 95 % CI from 0.27 to 2.36), all non-significantly different from placebo (Fig. 9).Anti-IL-5 treatment resulted in non-significantly reduced odds for influenza (OR = 0.84; 95 % CI from 0.65 to 1.09).Influenza was not reported by patients included in RCTs on mepolizumab (OR = not estimable), while benralizumab (OR = 0.81; 95 % CI from 0.59 to 1.11) and reslizumab (OR = 0.92; 95 % CI from 0.59 to 1.43) resulted in nonsignificantly reduced odds for influenza, respectively (Fig. 10).

ED or hospital admission
In included studies, ED or hospital admission was reported as an outcome related to asthma exacerbation.On the contrary, ED or hospital admission for infection was not reported.Neither was it possible to extract data on whether an infection was present at the time of ED or hospital admission due to asthma exacerbation.This outcome was reported by 12 included RCTs [16][17][18][19]23,25,26,[28][29][30]33].Anti-IL-5 antibodies were associated to reduced odds of ED or hospital admission for asthma exacerbation (OR = 0.59; 95 % CI from 0.40 to 0.88).In terms of effect sizes, benralizumab showed the lowest odds for ED or hospital admission due to asthma exacerbation (OR = 0.47; 95 % CI from 0.20 to 1.12), followed by mepolizumab (OR = 0.53; 95 % CI from 0.20 to 1.42) and reslizumab (OR = 0.74; 95 % CI from 0.47 to 1.17), all statistically non-significant (Fig. 11).

Discussion
We conducted a systematic review and a meta-analysis to explore the association of anti-IL-5 antibodies administration with the reporting of infective events in patients with asthma.From our analysis emerged a significant reduction in the reporting of sinusitis and bronchitis and a reduction in the odds for ED or hospital admission due to asthma exacerbation for overall anti-IL-5 antibodies compared to placebo.These observations are likely correlated to the mechanism of action of anti-IL-5 antibodies and reflect their clinical efficacy on eosinophilic inflammation.Indeed, anti-IL-5s are currently being evaluated also for chronic rhinosinusitis and nasal polyposis [34][35][36].These conditions share physio-pathological mechanisms with asthma and are frequently reported in patients with asthma.Thus, the significant reduction of sinusitis observed in our study with anti-IL-5 antibodies may be related to the concurrent reduced inflammation of the nasal mucosa while the reduction of bronchitis likely reflects an improvement in the bronchial system's status.
The reporting of other included infection events (nasopharyngitis, pharyngitis, URTI, acute sinusitis, rhinitis, pneumonia, and influenza) was non-significantly different.However, we observed differences both in directions and magnitudes of some effect sizes.In particular, benralizumab was associated with increased odds of pharyngitis, although at limits of statistical non-significance; a similar trend was observed also for mepolizumab on the reporting of rhinitis.Similar to bronchitis, anti-IL-5 treatment was associated with a non-significant reduction in the odds for pneumonia.However, for both infections and more importantly pneumonia the ORs of benralizumab were closer to the line of non-significance compared to those of reslizumab and mepolizumab in terms of effect size, (i.e., benralizumab reduced less the occurrence of bronchitis and especially pneumonia).This trend, albeit limited by the sample size and the low number of reported patients with the event, was similar to that observed in the aforementioned prospective cohort study [12].
Anti-IL-5 treatment was associated with a significant reduction of ED or hospital admission for asthma exacerbation.In terms of effect sizes, the greatest reduction in ED or hospital admission was observed with benralizumab, followed by mepolizumab and reslizumab.This reduction was non-significant in all the three drug subgroups; however, this could be explained by the relatively low frequency of the observed outcome, leading to large confidence intervals.Of note, the cumulative effect sizes and their directions were consistent across the three cumulative subgroups.In this context, the comparison of the trend of the clinical efficacy (i.e., the reduction of ED or hospital admission for exacerbation) and the trend of the reporting of some infection events is conflicting.In terms of effect size and direction, on one side, drugs such as benralizumab and mepolizumab reduced exacerbations and anti-IL-5 treatment reduced  sinusitis and bronchitis; on the other hand, benralizumab was less effective in reducing pneumonia and even seemed to increase pharyngitis.Also, we were not able to evaluate ED or hospital admission for infections and neither to discriminate if the cases of exacerbation leading to ED or hospital admission were pure asthmatic re-accruals or were associated with infections, neutrophilia, or other signs of infection.
Altogether, our findings led us to speculate whether benralizumab might increase the risk of some infections in asthmatic patients due to its mechanism of action which targets IL-5 receptors instead of circulating IL-5 such as mepolizumab and reslizumab.While benralizumab inhibits the binding of IL-5 to its receptor and induces antibody-dependent cell cytotoxicity on IL-5R + cells [1], mepolizumab action could be exerted through reducing the mobilization of eosinophils from bone marrow, reducing the maturation of eosinophils from progenitors, having a different selectivity on different subset of eosinophils, and/or a return of eosinophils to peripheral blood from the tissues [37].

Study limitations
Our study presents some limitations and caveats.Events presented in RCTs are generally presented as "patient with event" over the total number of patients and are divided by different event categories.However, the same patient could be included in multiple, also similar, events.For this reason, we could not cumulate all respiratory tract infection events to avoid the possible artefactual duplication of included patients even when some events were similar in the classification (e.g., nasopharyngitis, pharyngitis, sinusitis, acute sinusitis).Also, as discussed before, we could not distinguish exacerbations associated with markers of eosinophilia (i.e., anti-IL-5 inefficacy) from exacerbations possibly associated to infections that may have elicited asthma worsening (i.e., a possible effect of anti-IL-5 in increasing the risk of some infections in a subset of individuals).

Conclusion
Our results suggest that anti-IL-5 treatments might have different effects on the reporting of some infection events in patients with asthma.The almost complete depletion of eosinophils by benralizumab might lead to an increased risk of some infections.However, the evidence is limited and these results are far from conclusive and strongly suggest the need of future studies to evaluate the risk of infections in patients with asthma receiving anti-IL-5 treatments.

Data availability statement
The data supporting the findings of this study are available within the article.No further data were used for the realization of this study.

Table 1
Included studies characteristics.
M and F, 12 to 75 yo, with asthma, on medium dose ICS/LABA, two or more exacerbations in the last year requiring systemic CS or temporary ioncrease of usual oral CS M and F, 12 to 75 yo, with asthma, on high dose ICS/LABA, two or more exacerbations in the