**5. Non-invasive biomarkers limitations: can more invasive sampling do better?**

Asthma syndromes are characterised for being dynamic, with varying changes in symptoms pattern, lung function, inflammation and remodelling throughout time. In this setting, noninvasive direct airway sampling, such as exhaled breath analysis, seems especially appealing, allowing easy and repeatable measures over time. However, low molecular concentrations and variable sample dilution lead to difficulties in methods sensitivity and validation, with consequent issues in replication of biomarker findings (**Table 2**). In comparison, bronchoscopy allows direct visual examination of the airways and direct collection of fluid (bronchoalveolar lavage, bronchial washing) and tissue (brushing, biopsy). These techniques are mostly impractical because they are invasive, require specialised equipment and qualified personnel, have contraindications and carry potential risks / complications. Therefore, ethical issues preclude bronchoscopic sampling broad use in asthma, even less when repeated samplings are needed, thus being mainly reserved for selected severe patients and for research purposes. Apart from practical issues, standard bronchoscopy techniques hold several other limitations, including lack of reproducibility and sample dilution effect, despite recently proposed improvements (Table 2) [75]. In between invasive and non-invasive airway samplings, semiinvasive induced-sputum analysis may also reflect the airways and is easier to perform. Moreover, although indirect, blood sampling is minimally invasive and is a known relevant biomarker source in asthma.

Interestingly, the metabolomics analysis of urine also discriminated healthy individuals from those with asthma [69], and could distinguish patients with stable asthma from those with acute exacerbations based on profiles [69, 70]. Metabolomics analysis of urine samples has also been recently suggested as a useful clinical tool to differentiate asthma from COPD [71].

Pinkerton et al. [72] demonstrated for the first time that differences between healthy controls and asthma patients could be detected via micro-RNA (miRNA) expression in EBC, and suggest that different types of inflammation may have unique miRNA signatures. These small non-coding RNAs are known to be important in the post-transcriptional regulation of inflam‐

Proteomics has also recently been applied to EBC. Liquid chromatography (LC)-MS has been used to separate and detect proteolytic peptides present in EBC with differentiating profiles based on asthma status [73]. However, this preliminary study faced several problems such as insufficient sample volume, possible salivary contamination and difficulties in peptides

Besides allowing an overview of molecular signatures, the "omics" approach may potentially lead to new knowledge regarding asthma pathophysiology, due to its untargeted, hypothesisgenerating approach. All biomedical researchers are facing not only the opportunities but also the challenges in accessing, managing, analysing and integrating diverse data sets that are larger, more diverse and more complex than ever before, and that exceed the abilities of current management and analysis approaches [60, 74]. Composite biomarkers research such as that coming from molecular profiling assays including various "omics" is a live example that needs to be critically interpreted and cautiously validated to yield truly significant advances in

**5. Non-invasive biomarkers limitations: can more invasive sampling do**

Asthma syndromes are characterised for being dynamic, with varying changes in symptoms pattern, lung function, inflammation and remodelling throughout time. In this setting, noninvasive direct airway sampling, such as exhaled breath analysis, seems especially appealing, allowing easy and repeatable measures over time. However, low molecular concentrations and variable sample dilution lead to difficulties in methods sensitivity and validation, with consequent issues in replication of biomarker findings (**Table 2**). In comparison, bronchoscopy allows direct visual examination of the airways and direct collection of fluid (bronchoalveolar lavage, bronchial washing) and tissue (brushing, biopsy). These techniques are mostly impractical because they are invasive, require specialised equipment and qualified personnel, have contraindications and carry potential risks / complications. Therefore, ethical issues preclude bronchoscopic sampling broad use in asthma, even less when repeated samplings are needed, thus being mainly reserved for selected severe patients and for research purposes. Apart from practical issues, standard bronchoscopy techniques hold several other limitations, including lack of reproducibility and sample dilution effect, despite recently proposed

mation, thus opening a new research field using non-invasive direct air sampling.

identification due to their low concentration.

26 Asthma - From Childhood Asthma to ACOS Phenotypes

personalised medicine.

**better?**


BAL: bronchoalveolar lavage; EBC: exhaled breath condensate; EBT: exhaled breath temperature; eNose: electronic nose; FeNO: fractional exhaled nitric oxide.

**Table 2.** Pros and cons of main biomarker sample sources in asthma.

In this section, we will discuss these sampling methods and related current main biomarkers for asthma management.

#### **5.1. Sputum biomarkers**

Induced sputum is a validated sampling method of the more central airways. Sputum is collected after inhalations of hypertonic saline. Although relatively safe, induced-sputum requires specialised training, equipment and laboratory processing. Monitoring lung function during the induction procedure reduces the risk of excessive bronchoconstriction. Patient's active cooperation is needed for collection, making this technique unsuitable for some patients, especially for children below the age of 7 years [76].

Induced-sputum provides a rich source of soluble and cellular biomarkers and has exception‐ ally allowed a successful single biomarker-based clinical management approach in asthma. This is the case with sputum eosinophil percentage, which identifies patients who have eosinophilic and non-eosinophilic asthma phenotypes and can be predictive of poor asthma outcome and targeted treatment response, with demonstrated treatment-guided superior efficacy in reducing asthma exacerbations in adults [2, 27, 77, 78]. Thus, sputum eosinophil percentage acts as a key marker and correlates with severe exacerbations and AHR. It has also been useful in a panel of biomarkers to select patients who may benefit from IL-5 targeted therapies, including mepolizumab (anti-IL-5), reslizumab (anti-IL-5) and benralizumab (anti-IL-5R). In contrast with adults [77, 78], eosinophil sputum-guided therapy was not associated with decreased asthma exacerbations or improved asthma control in school-aged children and adolescents [79]. Sputum inflammatory phenotype was shown to be unstable in children with asthma, and this was not related to treatment or disease control [80].

Besides eosinophils, other sputum biomarkers are currently in research. Sputum neutrophils are often related to severe non-eosinophilic asthma with fixed airway obstruction. Soluble sputum biomarkers have been associated with asthma severity (e.g. eosinophilic cationic protein, LT, IL-4, IL-5, IL-13, IL-6, IL-12, tumour necrosis factor-α, granulocyte-macrophage colony-stimulating factor), exacerbations (e.g. IL-8, neurokinin A) or remodelling (procollagen synthesis peptides, tissue inhibitors of metalloproteinase or transforming growth factor-β) [10]. Many biomarkers can be measured, but most require highly sensitive detection methods and results may be affected by sputum processing or variable dilutions. These factors need to be taken into account to select and validate useful biomarkers in sputum.

Induced sputum may also be an interesting source for composite biomarkers. Unsupervised clustering of induced-sputum gene expression profiles identified three transcriptional asthma phenotypes that related to clinical and inflammatory parameters (resembling eosinophilic, neutrophilic and paucigranulocytic asthma) [81]. Differentially expressed genes were related to immune and inflammatory responses, proving a framework to investigate asthma endo‐ types.

In summary, logistic and practical difficulties have precluded the wide use of induced sputum in clinical practice, but sputum eosinophil percentage is recommended as a supplemental measure in future asthma clinical research studies to identify specific cellular profiles and to predict or to monitor a treatment response in adult patients [27]. It is important to highlight that sputum eosinophils and FeNO are not duplicative outcome measures, even though low sputum eosinophil and low FeNO are strongly linked [27].

### **5.2. Blood biomarkers**

BAL: bronchoalveolar lavage; EBC: exhaled breath condensate; EBT: exhaled breath temperature; eNose: electronic

In this section, we will discuss these sampling methods and related current main biomarkers

Induced sputum is a validated sampling method of the more central airways. Sputum is collected after inhalations of hypertonic saline. Although relatively safe, induced-sputum requires specialised training, equipment and laboratory processing. Monitoring lung function during the induction procedure reduces the risk of excessive bronchoconstriction. Patient's active cooperation is needed for collection, making this technique unsuitable for some patients,

Induced-sputum provides a rich source of soluble and cellular biomarkers and has exception‐ ally allowed a successful single biomarker-based clinical management approach in asthma. This is the case with sputum eosinophil percentage, which identifies patients who have eosinophilic and non-eosinophilic asthma phenotypes and can be predictive of poor asthma outcome and targeted treatment response, with demonstrated treatment-guided superior efficacy in reducing asthma exacerbations in adults [2, 27, 77, 78]. Thus, sputum eosinophil percentage acts as a key marker and correlates with severe exacerbations and AHR. It has also been useful in a panel of biomarkers to select patients who may benefit from IL-5 targeted therapies, including mepolizumab (anti-IL-5), reslizumab (anti-IL-5) and benralizumab (anti-IL-5R). In contrast with adults [77, 78], eosinophil sputum-guided therapy was not associated with decreased asthma exacerbations or improved asthma control in school-aged children and adolescents [79]. Sputum inflammatory phenotype was shown to be unstable in children with

Besides eosinophils, other sputum biomarkers are currently in research. Sputum neutrophils are often related to severe non-eosinophilic asthma with fixed airway obstruction. Soluble sputum biomarkers have been associated with asthma severity (e.g. eosinophilic cationic protein, LT, IL-4, IL-5, IL-13, IL-6, IL-12, tumour necrosis factor-α, granulocyte-macrophage colony-stimulating factor), exacerbations (e.g. IL-8, neurokinin A) or remodelling (procollagen synthesis peptides, tissue inhibitors of metalloproteinase or transforming growth factor-β) [10]. Many biomarkers can be measured, but most require highly sensitive detection methods and results may be affected by sputum processing or variable dilutions. These factors need to

Induced sputum may also be an interesting source for composite biomarkers. Unsupervised clustering of induced-sputum gene expression profiles identified three transcriptional asthma phenotypes that related to clinical and inflammatory parameters (resembling eosinophilic, neutrophilic and paucigranulocytic asthma) [81]. Differentially expressed genes were related

nose; FeNO: fractional exhaled nitric oxide.

28 Asthma - From Childhood Asthma to ACOS Phenotypes

for asthma management.

**5.1. Sputum biomarkers**

**Table 2.** Pros and cons of main biomarker sample sources in asthma.

especially for children below the age of 7 years [76].

asthma, and this was not related to treatment or disease control [80].

be taken into account to select and validate useful biomarkers in sputum.

Peripheral blood can be collected across all age groups, with minimal risk. Some biomarkers are routinely standardised in medical institutions and therefore readily available, such as eosinophils, total serum IgE and allergen-specific IgE. The latter are used to define atopy, which can be accurately, easily and more readily detected by skin prick test. Atopy modestly increases the probability of asthma, but is not essential for diagnosis. Though it is useful to characterise patients, atopy itself is recognised to be heterogeneous, including both "Th2 high" and "Th2-low" phenotypes [5]. Specific sensitisations are useful in clinical practice to suggest clinically relevant allergen avoidance and consider allergen-specific immunotherapy. However, total IgE or allergen-specific IgE quantification cannot predict the response to treatment and are otherwise weak biomarkers in asthma.

Blood eosinophil absolute count has long been associated with asthma and remains a recom‐ mended supplemental asthma biomarker [27]. Although it may not reflect the airways and be unspecific, blood eosinophilia supports asthma diagnosis and is an independent risk factor for exacerbations and fixed airflow limitation. Blood eosinophil counts are useful to subphenotype asthma and to monitor systemic biologic effects of pharmacologic interventions in patients with asthma, including (inhaled) corticosteroids, anti-IgE, LT antagonists and 5-lipoxygenase inhibitors [27]. Furthermore, blood eosinophil counts emerged as predictive biomarkers of clinical benefit from IL-5- and IL-13-targeted therapies, being associated with a "Th2 bronchial signature" [82].

Another promising "Th2-high" serum biomarker is the extracellular matrix protein periostin. The expression of periostin is upregulated by IL-13 in bronchial epithelial cells and, unlike IL-13, is abundant and readily detectable in peripheral blood [82]. Interestingly, a multi-centre study collecting matched sputum, bronchoscopy and peripheral blood samples from patients with asthma showed that serum periostin was the best single predictor of airway eosinophilia, with a further advantage of lower intrasubject variability over time than FeNO or blood eosinophilia [82]. However, conflicting results have recently been reported [83, 84]. Neverthe‐ less, periostin levels have been associated with asthma severity and its levels have also been shown to be important to predict lebrikizumab (anti-IL-13) clinical benefit, with greater reduction in severe exacerbations and greater improvement in lung function in the "periostinhigh" patients [85]. A greater decrease in exacerbations with anti-IgE therapy has also been reported in "periostin-high" patients. Healthy subjects and lebrikizumab-treated patients still have measurable levels of serum periostin, thus other systemic sources of periostin than IL-13 need to be explored [82].

Overall, blood eosinophils, serum periostin and FeNO reflect "type 2" airway inflammation in different ways and are only weakly correlated; therefore, combinations of these biomarkers obtained with minimally or non-invasive samplings may further enable optimisation of treatment benefit [82, 86, 87].

Recently, application of "omics" technologies to peripheral blood and invasive sampling with unsupervised clustering are yielding crucial data to capture the complexity of various asthma phenotypes and add new insights on asthma endotypes and treatment response. Given its maturity, transcriptomics analysis using microarrays is the current state-of-the-art method for asthma signature discovery [60]. For instance, gene expression profiling of bronchial epithe‐ lium identified distinct subtypes of patients with asthma with "Th2-high" or "Th2-low" phenotype [5], supported the involvement of endotoxin and macrophage activation in corticosteroid resistance, and suggested that corticosteroids also exert their beneficial effects through activity on bronchial smooth muscle [60]. "Omics" technologies developments, with data comparison and validation, will lead to the integration of composite signature biomarkers in phenotyping asthma and improvements in our understanding of asthma. Ultimately, breakthroughs in asthma treatment may be reached through the development of innovative targeted therapies [12, 60].

Non-invasive procedures for biomarker analysis form the backbone for day-to-day clinical asthma management. However, invasive tests may provide important information to pheno‐ type and direct therapy in patients with severe refractory asthma [88]. These techniques bring significant additional knowledge in asthma research that needs to be integrated with noninvasive procedures outcomes to allow truly innovative steps in biomarker discovery for asthma management.
