Measuring inflammation in paediatric severe asthma: biomarkers in clinical practice

Eosinophils

Eosinophils are central drivers of type 2 asthma, both in allergic and nonallergic forms, and represent the predominant inflammatory cell types in the airways of children with severe asthma [17]. Upon activation, eosinophils contribute to airway damage and chronic inflammation and promote remodelling, subepithelial membrane thickening and goblet cell metaplasia [17].

So far, sputum eosinophilia is the best-characterised and most useful biomarker. The normal range for sputum eosinophilia is defined as 1–2%, with female sex and atopic status associated with higher sputum eosinophil levels [18]. Some recent longitudinal studies have reported controversial results on temporal variability in sputum eosinophil counts in adult patients with severe asthma: while some authors have shown that 70% of patients with eosinophilic asthma (sputum eosinophils ≥2%) retained their sputum phenotype over 5 years [19], others revealed that 60% of the patients changed their phenotype from eosinophilic (sputum eosinophils ≥2%) to non-eosinophilic and vice versa [20]. A similar variability in sputum inflammatory cells over time was observed in 61% of children (including those with severe asthma) followed over 1 year in a longitudinal study by Fleming et al. [21]. Specifically, they classified sputum samples into eosinophilic (>2.5% eosinophils), neutrophilic (>54% neutrophils), mixed granulocytic (>2.5% eosinophils, >54% neutrophils) or paucigranulocytic (≤2.5% eosinophils, ≤54% neutrophils). Raised levels of inflammatory cells (up to 78%) were reported in at least one sputum sample of these children, regardless of asthma severity; also, the lack of temporal phenotype stability was not related to changes in asthma control or ICS therapy [21].

Sputum eosinophil count is increased in symptomatic patients with asthma, in patients with asthma during exacerbations, and in corticosteroid-naïve patients [22–24]. Increased eosinophil number in sputum has also been associated with airway hyperresponsiveness (AHR) and obstruction, and inversely correlated with forced expiratory volume in 1 s (FEV₁) [25].

In asthma, corticosteroids are effective at suppressing eosinophilic inflammation [26]. In general, sputum eosinophilia is associated with steroid responsiveness [26]; however, the persistence of elevated sputum eosinophilia may suggest either non-adherence or inadequate corticosteroid dosing [21]. In such cases, titrating ICS doses based on sputum eosinophilia may reduce the likelihood of asthma exacerbations [25]. Adjusting treatment based on sputum eosinophilia has been performed with various sputum eosinophil cut-offs, ranging from 2% to 8%; however, the most commonly used cut-off is sputum eosinophilia ≥3% [18]. Unfortunately, there are insufficient data available to extend the clinical application of induced sputum eosinophilia in tailoring asthma medication in children [23].

The clinical utility of peripheral blood eosinophilia in the assessment of children with severe asthmahas long been explored. Although there is no standardised cut-off, a blood eosinophil count of 300 cells·μL⁻¹ has commonly been usedasathreshold to indicate eosinophilic inflammation [27].

Despite being influenced by some confounding factors (allergen exposure, parasitic infections and systemic corticosteroid treatment), blood eosinophils have shown the highest accuracy in the identification of sputum eosinophilia, compared with other biomarkers such as F_ENO and serum periostin [28]. Nevertheless, peripheral blood eosinophilia does not always reflect pulmonary (airway or mucosal) eosinophilia in children with severe asthma [29].

The correlation between blood eosinophil count, asthma severity and AHR has been demonstrated in several studies conducted in children [30, 31]. Analysing the 12-month longitudinal dataset of a cohort of children and adolescents with severe asthma exacerbations, Shah et al. [32] recently found that elevated blood eosinophils were associated with acute asthma exacerbations and hospitalisations, but optimal cut-off points for blood eosinophils differed by age (≥150 versus ≥300 cells·μL⁻¹ for children versus adolescents, respectively). However, the addition of a second biomarker of type 2 inflammation, such as serum IgE and/or F_ENO, was useful in predicting asthma exacerbations only in adolescents [32].

In the Individualized Therapy for Asthma in Toddlers (INFANT) trial, the presence of blood eosinophilia (≥300 cells·µL⁻¹), in combination with aeroallergen sensitisation, has been identified as a significant predictor of response to ICS in children with asthma [33]. Thus, blood eosinophil count could be useful to identify patients for ICS prescription and patients at high risk of exacerbations. Instead, higher eosinophil count has been assessed to be a sensitive and practical predictive biomarker for biological therapies targeting allergic and/or eosinophilic pathways in patients with severe asthma [7]. A high blood eosinophil count (>300 cells·μL⁻¹) has been reported as a potential biomarker to predict successful omalizumab treatment effects in children with severe allergic asthma [34]. In combination with F_ENO and periostin, blood eosinophils were shown to identify patient subgroups that may achieve more significant benefit from omalizumab therapy [34]. Blood eosinophil count has also evolved as a predictive biomarker for the efficacy of treatment with anti-interleukin (IL)-5 (mainly mepolizumab) in patients with severe eosinophilic asthma [35]; in particular, a baseline blood eosinophil threshold of ≥150 cells·μL⁻¹ or a historical blood eosinophil threshold of ≥300 cells·μL⁻¹ allow selection of patients with severe eosinophilic asthma who are most likely to achieve clinically significant reductions in the rate of exacerbations with mepolizumab treatment [36].

Sputum eosinophilia also adequately predicts response to biological drugs. Patients with refractory asthma are more likely to respond to anti-IL-5 or anti-IL-4/IL-13 treatment if they have sputum eosinophils of >3%. A significant decrease in exacerbation rates, an improved quality of life and corticosteroid-sparing effects have been observed in adults treated with mepolizumab [35]. Hence, the newest Global Initiative for Asthma (GINA) document suggests assessment of sputum eosinophil count in adults with moderate or severe asthma who are managed in experienced centres [37]. No similar recommendations are available for children.

Serum IgE

IgE plays a critical pathophysiological role in allergy and asthma, influencing the functioning of several immune and structural cells of the bronchial wall [38–40]. By binding to its high- and low-affinity receptors, IgE exerts several biological functions: degranulation and mediator release by mast cells/basophils; promotion of mast cell maturation, survival and cytokine production; increase in allergen presentation by dendritic cells and B-cells; activation of eosinophils and T-helper type 2 (Th2) cells; activation, proliferation and cytokine release by smooth muscle cells; inhibition of type I interferon production by dendritic cells; and other indirect effects on innate immune cells. As a result, IgE is primarily responsible for both the acute and chronic phases of inflammation characteristic of allergic asthma [41].

Total IgE and allergen-specific IgE have been strongly associated with asthma in >80% of children [42]. In infants with viral-induced wheezing, a raised level of total IgE is considered a risk factor for asthma development [43], while high levels of allergen-specific IgE correlate well with asthma severity, mainly in children [42, 44]. Consequently, IgE measurements should be routinely performed in patients with asthma, and total IgE should be checked in every child diagnosed with severe asthma, for the consideration of add-on therapies such as omalizumab [42].

Periostin

Periostin is an extracellular matrix protein secreted by airway epithelial cells, endothelial and smooth muscle cells, and lung fibroblasts, in response to IL-4 and IL-13. Following its release, periostin acts on keratinocytes, promoting the release of pro-inflammatory cytokines such as thymic stromal lymphopoietin (TSLP) and IL-24 and regulating epithelial–mesenchymal interactions. In this context, periostin actively participates in airway remodelling, eosinophil recruitment and mucus production from goblet cells [45, 46]. Thus, it has been associated with the type 2 asthma endotype.

Serum periostin levels ≥10 ng·mL⁻¹ identify an inflammatory status. However, age-dependent changes in periostin values commonly occur, probably due to bone turnover and growth, especially in children. Moreover, the presence of comorbidities, including rhinitis, rhinosinusitis with or without nasal polyposis, and atopic dermatitis, can affect its baseline values [47]. High levels of periostin have also been found in sputum or EBC, reflecting severe airway inflammation [47].

A meta-analysis reported high serum periostin levels in patients with asthma compared with healthy controls, despite a wide heterogeneity in periostin values among patients with asthma [47]. Particularly high serum periostin levels were found in those with corticosteroid-resistant asthma [48]. There appears to be a moderate or weak correlation between periostin and eosinophil count, F_ENO and serum total IgE [47]. Clinically, serum periostin levels have been inversely correlated with predicted FEV₁, probably reflecting the periostin-mediated fibrogenic effects [47]. In adults with severe asthma, periostin seems to be able to predict clinical response to monoclonal antibody therapies, including omalizumab and lebrikizumab [47].

In contrast to adult asthma, the usefulness of periostin in childhood asthma is still debated and controversial. Some studies on periostin have shown significantly higher values in children with asthma compared with healthy controls, a significant correlation between higher levels of serum periostin and induced AHR, and a moderate relationship between blood eosinophils and IgE in children with asthma [49, 50]. In contrast, Konradsen et al. [51] and Inoue et al. [52] did not detect correlation of serum periostin with blood eosinophils or F_ENO in childhood asthma patients. In a real-world setting, we recently demonstrated that serum periostin does not reflect asthma control grade in children with asthma [53]. It remains unclear if periostin has a predictive value for identifying severe asthma in children. Finally, serum periostin measurements cannot be readily assessed in most clinical laboratories.

Exhaled biomarkers

Urine metabolites

Urine metabolites, such as bromotyrosine and leukotriene E₄, could potentially represent new clinical biomarkers to diagnose asthma. Some studies have found increased urinary bromotyrosine and leukotriene E₄ levels in patients with airflow limitation and inadequately controlled asthma. Moreover, both these urine metabolites could predict risk of asthma exacerbations and response to treatments including ICS, leukotriene receptor antagonists and long-acting β-agonists [71–73].

Recently, the eosinophil granule-derived proteins, such as eosinophil protein X (EPX) (also known as eosinophil-derived neurotoxin), have been investigated in the urine samples of patients with asthma. Reflecting the eosinophil activation, these proteins have proved to be reliable for assessing airway inflammation. Over 70% of included studies in a recent meta-analysis reported higher EPX levels in children with asthma compared to controls [74]. Nevertheless, due to the variability of metabolic pathways associated with paediatric asthma and their weak correlation with other biomarkers, such as sputum cell count and F_ENO level, the utility of urine metabolites needs to be further investigated [75].

Neutrophils

Despite neutrophils being the most commonly detected cell type in induced sputum samples, they do not yet represent established markers to define the non-type 2 asthma endotype [76]. The cut-off values used to define sputum neutrophilia are different between cohorts of patients with asthma, ranging from 40% to 76%. In healthy subjects, the detected median neutrophil percentage is about 37% (10th and 90th percentiles of 11% and 64%). Tobacco smoke, pollution exposure and acute/chronic infections can also contribute to increased sputum neutrophil counts. Generally, a sputum neutrophil count ≥61% allows the diagnosis of neutrophilic asthma [77]. A cut-off of 5% in BAL neutrophils has been adopted to classify as “neutrophil-high” or “neutrophil-low” children with severe asthma who underwent bronchoscopy [78]. Currently, none of the proposed cut-offs have been approved [77].

Clinically, Shaw et al. [79] reported evidence that severity of asthma, as well as severe exacerbations, could be objectively assessed by the sputum neutrophil count. However, the prevalence of neutrophilia in severe asthma remains somewhat uncertain, especially in childhood asthma, in which neutrophilic airway inflammation has been recently associated with better lung function [80].

Conversely, Teague et al. [81] reported that children with pauci-granulocytic asthma (with no evidence of increased numbers of eosinophils or neutrophils in sputum or blood and in whom anti-inflammatory therapies are ineffective at controlling symptoms) experienced a lower post-bronchodilator airflow limitation, less blood eosinophilia, fewer doses of prednisone, and fewer infectious species than children with mixed granulocytic BAL [81]. Moreover, a strong relationship between sputum and blood neutrophils, as well as with asthma severity, was also reported, suggesting neutrophilic lung inflammation as a risk factor for poorly controlled asthma [82, 83].

Sputum neutrophil count can also reflect response to treatment. Cluster analyses from the Severe Asthma Research Program (SARP) found that the highest sputum neutrophil count was correlated with the worst lung function despite maximal bronchodilator and ICS treatment [84].

Given the limited accuracy and ability of available blood biomarkers to reflect sputum neutrophilia, as well as the unavailability of targeted therapy for neutrophilic asthma, sputum neutrophilia currently lacks a clinical application [85].

Other cells

Natural killer T (NKT) cells, a heterogeneous T-cell population, are involved in both the autoimmunity and adaptive immune responses [86]. Recently, their functional role has been investigated in asthma. By using flow cytometric analysis, Hamzaoui et al.[87] reported a significant increase in the number of NKT cells in the sputum of patients with severe asthma compared with mild asthmatic and healthy control groups (p<0.05). The invariant natural killer T (iNKT) cells, which feature the expression of T-cell receptors recognising lipid antigens, do not seem to differ significantly between children and adults with asthma compared to healthy subjects[88]. The mucosal-associated invariant T (MAIT) cells, which recognise microbial-derived vitamin B2 (riboflavin) metabolites, are significantly lower in the peripheral blood and sputum of patients with moderate to severe asthma, thus negatively correlate with clinical severity [89]. Mast cells can be detected in both eosinophilic and non-eosinophilic asthma. However, recently, Bergamo de Araujo-Paulino et al. [90] provided evidence that mast cell subtypes and their related gene expression profiles are associated exclusively with airway eosinophilia and could predict clinical asthma outcomes as well as corticosteroid response.

Further investigations on the number and functional properties of lung iNKT, MAIT and mast cells will hopefully provide more information on asthma pathogenesis. To date, their clinical utility is still under debate.