Molecular profiling of exhaled breath condensate in respiratory diseases

Abstract

Background

Respiratory disorders, , continue to pose a major global health burden. Their complexity and heterogeneity challenge accurate diagnosis, effective monitoring, and therapeutic decision-making. Exhaled breath condensate (EBC) provides a reliable, non-invasive means of sampling the molecular environment of the airways.

Aim

This review presents the state-of-the-art in EBC-based omics approaches—particularly metabolomics and proteomics—to characterize molecular signatures associated with chronic respiratory (e.g. asthma, chronic obstructive pulmonary disease, and rhinitis) and infectious diseases (e.g. COVID-19).

Results

We critically examine findings from studies applying nuclear magnetic resonance (NMR), mass spectrometry (MS), and sensor-based technologies to analyze EBC across various respiratory conditions. NMR, valued for its reproducibility and minimal sample preparation, consistently discriminates among disease phenotypes, identifies distinct metabotypes, and monitors treatment response over time. MS-based approaches afford enhanced sensitivity and specificity, enabling detailed profiling of inflammatory mediators, such as lipid-derived eicosanoids and amino acid derivatives. Proteomic studies reveal protein-level alterations associated with inflammation and tissue remodeling. In COVID-19 and long COVID, metabolomic and volatile compound profiling distinguishes affected individuals from healthy controls suggesting clinical potential. However, inconsistent sample processing and lack of analytical standardization remain limiting factors.

Conclusions

EBC profiling shows clear promise for improving diagnosis, monitoring, and stratification in respiratory medicine. Yet, translation into clinical practice is hindered by limited standardization and validation. Broader, longitudinal studies will be essential to establish robust molecular signatures across disease states. This review underscores the timely need to implement breathomics investigations to gain mechanistic insight into the underlying biology of respiratory diseases.

Introduction

Chronic respiratory diseases

Given their substantial global burden and progressive nature, chronic respiratory diseases (CRDs) represent a critical and complex area of investigation in public health and clinical medicine. This umbrella term comprises a heterogeneous group of long-term pulmonary conditions characterized by persistent inflammation, airflow limitation, and structural alterations of the airways and lung parenchyma. Among the most prevalent CRDs are asthma, chronic obstructive pulmonary disease (COPD), cystic fibrosis, bronchiectasis, occupational lung diseases (e.g. asbestosis), and pulmonary hypertension. Clinically, these non-communicable diseases often present with chronic cough, dyspnea, wheezing, and chest tightness, reflecting the underlying pathophysiological processes that compromise ventilatory efficiency and gas exchange over time [1].

The etiology of CRDs is multifactorial, resulting from the interplay between genetic predisposition, environmental exposures, and lifestyle behaviors. Cigarette smoking is the most prominent and extensively documented risk factor, exerting its deleterious effects through both active smoking and passive (secondhand) exposure. The toxic chemicals of tobacco smoke not only induce DNA damage in pulmonary tissues, generating oxidative stress, but they also disrupt immune homeostasis, leading to persistent inflammation, impaired adaptive T cell responses, and compromised host defenses against pathogens. Together, these mechanisms contribute to airway remodeling and chronic tissue damage, which are particularly evident in conditions such as chronic obstructive pulmonary disease (COPD) [2–4].

Along with cigarette smoke, prolonged exposure to environmental and occupational pollutants (e.g. airborne particulate matter, chemical irritants, and dust) represents a significant risk factor for CRD development. Likewise, a history of recurrent lower respiratory infections, particularly during childhood, further increases long-term susceptibility to CRDs. Other important contributors include genetic predisposition and age-related decline in respiratory function, both of which further exacerbate disease risk and progression [5]. The convergence of these various etiological factors underscores the multifactorial nature of CRD pathophysiology and highlights the need for integrated prevention and management strategies.

As defined by the World Health Organization (WHO) (Chronic respiratory diseases (who.int)), CRDs remain incurable. Nevertheless, in the absence of a definitive cure, various therapeutic strategies have been developed to alleviate symptoms, optimize pulmonary function, and improve CRD patients’ quality of life (QoL). Current interventions primarily focus on relieving airflow obstruction, reducing dyspnea, and effectively managing persistent symptoms, thereby restoring a degree of functional stability and day-to-day well-being.

Preventive measures, particularly smoking cessation and regular physical activity, continue to play a key role in slowing down CRD progression while ameliorating respiratory function and overall cardiovascular health. From a pharmacological perspective, the combination of inhaled corticosteroids and bronchodilator—including beta-adrenergic agonists and anticholinergic agents—has shown efficacy in achieving symptom control, providing both anti-inflammatory and bronchodilator effects. However, the long-term administration of these medications has been associated with a growing incidence of severe side effects, underscoring the need to minimize off-target effects by developing novel pathway-specific inhibition strategies acting more precisely on specific biological pathways involved in CRDs [6]. Accordingly, the identification of modifiable risk factors and the design of nuanced, patient-specific therapeutic strategies have become central to effective CRD management. A multifaceted care mode—integrating lifestyle modifications, optimized pharmacological regimens, and patient-centered education—could significantly influence disease progression and improve long-term clinical outcomes. In parallel, sustained efforts in translational research and the implementation of proactive public health policies are necessary to reduce disease incidence and alleviate the burden of CRDs on individuals and healthcare infrastructures worldwide.

Advancing respiratory research through mass spectrometry and nuclear magnetic resonance

The respiratory system plays a central role in maintaining homeostasis by facilitating the exchange of oxygen and carbon dioxide, both essential for cellular metabolism. However, diseases that impair its function, such as the aforementioned CRDs and infectious respiratory illnesses, require a closer look into the biological pathways that drive disease onset, progression, and heterogeneity. Bridging this gap is an essential step to refine disease classification, which in turn enables a more accurate patient stratification and the identification of novel actionable therapeutic targets.

Metabolites and proteins within biological systems are shaped by anabolic and catabolic processes and, as such, constitute blueprints of cellular biochemical activity, presenting a more straightforward correlation with phenotypes than genes [7]. In the context of respiratory disorders, such as asthma [8], COPD [9], and allergic rhinitis [10], metabolomics has proven valuable in defining phenotypes and uncovering molecular endotypes. In particular, metabolomics has provided critical insights into SARS-CoV-2 infection, revealing distinct metabolic signatures linked to coronavirus disease 2019 (COVID-19). These signatures have enabled patient differentiation based on disease severity, evaluation of host responses to pharmacological and vaccine-based treatments, and longitudinal assessment of metabolic changes spanning from acute infection to either full recovery or the insurgence of long COVID [11].

In this scenario, mass spectrometry (MS) has emerged as a sensitive, specific and rapid approach able to characterize the molecular mechanisms associated with respiratory health and disease due to its sensitivity, specificity, and speed. Its broad applicability extends across numerous aspects of respiratory research, enabling the precise measurement of thousands of proteins and metabolites in various biological samples [12,13]. Several MS-based approaches, such as gas chromatography (GC), liquid chromatography (LC), inductively coupled plasma (ICP), and matrix-assisted laser desorption/ionization (MALDI), are used in respiratory diseases, ranging from omics-based research and biomarker discovery to qualitative and quantitative analyses [14].

Of particular relevance is the integration of MS-based proteomics with computational data analysis, which allows thorough profiling of specimens such as blood [15–17], bronchoalveolar lavage fluid [18,19], nasal lavage fluid [20], sputum [21], lung tissue biopsies [22], and most notably exhaled breath condensate (EBC)—the primary focus of this review. Despite their efficacy, the inherent sensitivity of MS-based omics approaches presents several challenges, including the complex composition of the metabolome and proteome, their wide concentration ranges, and tissue specificity. Additional complexity is due to the interconnected nature of metabolites, requiring specialized analytical methods for accurate interpretation. To address these challenges and ensure robust study design, researchers need to implement prospective planning, standardized procedures, and meticulous data recording [23]. Nevertheless, MS-based omics have made significant strides in dissecting the molecular mechanisms of respiratory diseases, providing a foundation for significant advancements in the field.

Unlike MS, which measures mass-to-charge ratios of ionized fragments to identify molecular weight and substructures, nuclear magnetic resonance (NMR) analyzes the magnetic properties of atomic nuclei to provide detailed information on molecular structure and atom connectivity. Although less sensitive than MS, NMR spectroscopy has been widely used to characterize the biochemical profiles of metabolites in EBC samples, particularly small and volatile organic compounds (VOCs). In particular, it has been successfully applied to discriminate patients with COPD [24,25], asthma [26,27], or other respiratory disorders from healthy individuals. The method is non-destructive, typically requires minimal or no sample preparation, and allows rapid spectral acquisition (10–15 min). Moreover, NMR-derived spectra are rich in information, supporting both software-assisted clustering and direct metabolite identification, with peak intensities proportional to nuclei abundance. However, its two main limitations include the scarce sensitivity, especially for low abundant metabolites, and its inability to perform targeted analysis of specific, known compounds.

To improve sensitivity and strengthen the analytical potential of these so called ‘breathomics’, different analytical techniques, such as NMR and MS, are often combined. This integrated approach has enabled the molecular characterization of respiratory diseases such as asthma, COPD, allergic rhinitis, and SARS-CoV-2 infection (Figure 1). Despite the wealth of information derived from omics-centric studies, a major limitation still remains the difficulty in translating these discoveries into validated clinical biomarkers.

Figure 1.

Molecular profiling of exhaled breath condensate in respiratory diseases is mainly performed using MS-based techniques and NMR. A wide range of biomolecules can be analyzed using these methods, focusing on both chronic and infectious diseases.

The path from biomarker discovery to clinical implementation is indeed quite challenging, requiring rigorous validation processes and stringent reproducibility. Bridging this translational gap calls for more effective integration and interpretation of large-scale omics datasets to ensure both biological relevance and clinical utility. Enhancing the specificity and sensitivity of disease-related molecular detection is essential, and standardization across laboratories is equally important. Ultimately, the key challenge is to translate rich metabolomics insights into clinically actionable tools that can improve patient care and outcomes [28].

Exhaled breath condensate

EBC, obtained by condensing exhaled aerosol from spontaneous tidal breathing, represents a non-invasive biological matrix for the detection of biomarkers originating from the lower respiratory tract. Its collection is simple, well-tolerated, and feasible across various patient populations, such as children and individuals on mechanical ventilation. Several commercially available devices, such as Ecoscreen^®, TURBO-DECCS^®, RTube^®, and Anacon glass condenser are available for standardized EBC collection.

EBC is composed of three main components: aerosolized non-volatile droplets from the airway lining fluid, water-soluble molecules condensed after aerosolization, and the water vapor from the airways. It contains a wide array of biomarkers derived from the airway lining fluid (e.g. VOCs), which makes it a suitable medium for airway inflammation assessment, non-volatile organic mediators (e.g. adenosine, isoprostanes, leukotrienes, peptides, and cytokines), and inorganic compounds (e.g. nitric oxide, ammonia, and carbon monoxide). More recently, larger macromolecular complexes, such as extracellular vesicles carrying microRNAs, have been identified in EBC, likely derived from lung epithelial cells. The concentration of these mediators can be modulated by disease activity or therapeutic interventions, suggesting their diagnostic and/or prognostic value.

To ensure the reliability of EBC measurements, several pre-analytical issues must be addressed. Salivary contamination is a major concern that can be minimized through techniques such as periodic swallowing [29], the use of a nose clip to prevent nasal contamination [30], and the application of saliva traps or swan-neck collection tubes. Mouth rinsing and pre-sampling with sodium bicarbonate (4.5%) can reduce oropharyngeal microbial activity and nitrogen oxide levels [31]. Environmental pollutants also pose a risk of sample contamination [32]. As spontaneous breathing patterns may also influence EBC composition, subjects should refrain from exercise for at least one hour prior to EBC collection [33]. Quiet tidal breathing is recommended, as high dead-space ventilation combined with low tidal volume often yield samples from conducting airways rather than peripheral lung regions [34]. Lastly, several physiological conditions such as sex, age, circadian rhythm, infections, and water excess should also be taken into account as potential confounders [35].

In light of these characteristics, EBC analysis offers a valuable non-invasive approach for investigating CRDs such as asthma, COPD, and rhinitis. Integration with MS-based approaches (e.g. metabolomics and proteomics) has allowed thorough profiling of low-molecular-weight metabolites. This analytical strategy has not only contributed to the understanding of the complex metabolic and signaling pathways characterizing respiratory diseases, but it has also supported current precision medicine paradigms focusing on molecular phenotyping [36].

Overall, by providing valuable insights into disease onset, prognosis, and progression, EBC-based metabolomics is set to uncover novel pathophysiological mechanisms, reinforcing the importance of multi-biomarker profiles in defining and characterizing distinct respiratory disease phenotypes.

To enhance detection sensitivity and minimize ionization suppression, MS is often coupled with separation techniques, such as gas chromatography (GC), high-performance liquid chromatography (HPLC), and capillary electrophoresis (CE) [37].

The exploration of the metabolome offers a dual strategy, allowing for either an untargeted or targeted approach based on specific research goals. Untargeted metabolomics allows for a broad survey of the molecular landscape, facilitating the discovery of novel biomarkers and complex biochemical patterns. In contrast, targeted approaches focus on specific metabolites or classes, characterizing precise biological mechanisms that underlie pathological processes [38].

Gas chromatography-MS (GC-MS) is particularly effective in profiling volatile and semi-volatile thermally stable compounds in EBC, and together with chemical derivatization, it also allows analyzing non-volatile compounds [37,38]. VOC identification in EBC typically relies on GC-MS or GC coupled with tandem MS (GC/MS/MS) [39]. However, for unknown mass peaks, structural validation requires cross-referencing against mass spectral databases and libraries. Conversely, liquid chromatography-tandem MS (LC-MS/MS) allows for the simultaneous identification and quantification of targeted metabolites with high sensitivity and specificity. In fact, this technique has proven effective in detecting 8-isoprostaglandin F2α in human EBC [40] and is widely employed to assess inflammatory biomarkers, such as leukotrienes [41], isoprostanes [42], and eicosanoids [43], in both healthy individuals and those affected by respiratory diseases. In this regard, Cruickshank-Quinn et al. demonstrated the usefulness of LC-MS in investigating lung diseases and smoking effects among 133 asthmatic individuals across three different cohorts. Their study compared EBC, saliva, and saliva-contaminated EBC, measuring amino acids and eicosanoids using targeted LC-MS. They also applied untargeted metabolomics to profile EBC constituents and conducted proteomic analyses using both human and bacterial databases. The study provides a comprehensive overview of EBC constituents, serving as a foundation for researchers to explore more specialized techniques for detecting additional or less abundant biomarkers [44].

MS-based proteomics and peptidomics analyses have proven instrumental in identifying disease-specific proteins—whether in their intact or digested form—and endogenous peptides. These approaches not only enhance early detection, prognosis, and therapeutic monitoring in respiratory diseases but also yield mechanistic insights into disease processes. By examining changes in protein expression, degradation patterns, post-translational modifications, and protein-protein interactions, MS-based proteomics and peptidomics provide mechanistic insights into the pathophysiology of respiratory disorders, fostering biomarker discovery.

As mentioned earlier, NMR spectroscopy has similarly gained traction as a robust analytical platform for the study of EBC. It permits the non-destructive identification and quantification of metabolites in samples from patients with conditions such as asthma, COPD, and lung cancer. Moreover, NMR has facilitated the discovery of biomarkers linked to inflammation, oxidative stress, or other pathophysiological processes within the pulmonary system [45,46] (Figure 2).

Figure 2.

Exhaled breath condensate (EBC) molecular characterization: advantages, limitations and challenges of mass spectroscopy (MS)- and nuclear magnetic resonance (NMR)-based approaches.

Taken together, current literature demonstrates substantial progress in applying metabolomic and proteomic methodologies to EBC for disease characterization and monitoring. However, most studies to date focus on individual conditions, which limits the detection of shared molecular features and hinders the comparative analysis across disease types.

This review addresses this gap by synthesizing EBC-derived molecular data across a spectrum of chronic and acute respiratory diseases. Through the integration of metabolomic and proteomic insights, we aim to identify both common and disease-specific biomarkers linked to asthma, COPD, rhinitis, and respiratory infections. In doing so, this review contributes to a more refined understanding of disease mechanisms and supports the development of targeted diagnostic and therapeutic strategies.

Metabolomic and proteomic studies on asthma EBC

Asthma is a heterogeneous, chronic respiratory condition primarily affecting the lower respiratory tract. Its inherent heterogeneity manifests in diverse clinical presentations, treatment responses, and disease trajectories across the lifespan. Characterized by recurrent episodes of breathlessness, wheezing, chest tightness, and coughing, whose intensity and occurrence vary over time, asthma also presents with variable expiratory airflow limitation and heightened airway responsiveness to stimuli, such as exercise and inhaled irritants. Bronchial asthma has been widely recognized as an inflammatory disease, and substantial efforts have been made to identify the underlying biochemical and inflammatory phenotypes to guide targeted and personalized therapies [47,48].

At the population level, a subset of individuals—particularly those with late-onset asthma—experience accelerated lung function decline, which may end up in fixed airflow obstruction in severe cases. The frequency and severity of symptoms are modulated by several factors, including exposure to known and unknown allergens, irritants, respiratory infections, physical exercise, and cold weather [49]. These diverse triggers can exacerbate symptoms, underscoring the complexity of asthma and the need for a comprehensive understanding of its pathobiology.

Both genetic predisposition and environmental exposures contribute to asthma pathogenesis. While onset frequently occurs in childhood, adult-onset asthma is often non-allergic and may present in association with conditions such as nonsteroidal anti-inflammatory drug (NSAID) intolerance, chronic rhinosinusitis, and nasal polyposis. Asthma, as a multifaceted condition, frequently coexists with a spectrum of co-morbidities, including both allergic and non-allergic systemic conditions [50,51]. Adding further complexity, some forms of asthma can undergo spontaneous remission, particularly during late childhood and adolescence [52–54].

While traditional classification methods categorize asthma based on symptom severity or disease control, a personalized or stratified approach is gaining traction [55]. This emerging framework considers disease heterogeneity, underlying causal mechanisms, and individual variability in response to treatment. Recognizing asthma as a heterogeneous condition has enabled the identification of different asthma phenotypes, each reflecting distinct symptom patterns, treatment responses, and molecular signatures. However, establishing a direct and consistent link between specific pathological features and clinical patterns or treatment responses remains a challenge. This complexity continues to hinder precise endotyping, as highlighted in the 2024 GINA Main Report (2024 GINA Main Report - Global Initiative for Asthma - GINA (ginasthma.org).

Researchers have increasingly focused on identifying key metabolites that may help diagnose, monitor, and treat asthma. Since EBC composition reflects the biochemical makeup of the airway lining fluid, it has been frequently used to assess potential biomarkers in the EBC from asthmatic patients. Among them, one easily detectable exhaled surrogate of airway inflammation is fractional exhaled nitric oxide (FeNO), which is associated with eosinophilic airway inflammation. FeNO plays a significant role in diagnosing asthma, defining different inflammatory patterns, and assessing disease control, adherence to therapy, and response to corticosteroids [56]. In addition to FeNO, nitrogen oxides (NOx) formed by the reaction of NO with other reactive oxygen species (e.g. H₂O₂) and oxidative stress biomarkers such as isoprostanes and malondialdehyde—resulting from lipid peroxidation—have shown correlation with asthma [57–61]. Other promising biomarkers include EBC pH, which tends to be lower in asthma and is significantly associated with disease presence [62,63]. Eicosanoids, due to their central role in airway inflammation, are also frequently found in altered concentrations in asthmatic patients [64]. Finally, inflammatory epithelial-derived proteins, such as periostin, which is increased [65], and ezrin which is reduced [66], demonstrate differential expression, underscoring their potential as biomarkers for diagnosis, endotype discrimination [67], and monitoring therapeutic efficacy [68].

Metabolomic profiling in both children and adults has proven effective in distinguishing asthmatic individuals from healthy controls (HCs) and identifying relevant biological pathways actionable for therapeutic intervention. EBC proteomic and metabolomic investigations continue to reveal novel candidate biomarkers and provide mechanistic insights into asthma onset and progression.

Among these, the proteomic study by Bloemen et al. [69] is particularly interesting. These researchers analyzed EBC samples from 40 asthmatic children aged 6 to 12 years—classified according to GINA guidelines as having well-controlled (n = 26), partially controlled (n = 13), or uncontrolled (n = 1) asthma—and 30 HCs. While single peptides showed limited discriminative power, specific peptide patterns derived from exhaled proteins were able to differentiate partially controlled and uncontrolled asthmatic patients from HCs. In addition, a unique peptide signature differentiated well-controlled asthmatic patients from healthy states. Cytokeratins were among the most abundant proteins in EBC, although not all proteins present in the observed patterns could be identified or correlated with known biomarkers [69].

The utility of EBC metabolomics, or ‘breathomics,’ in asthma research was demonstrated by Carraro et al. [70], who analyzed samples from 42 asthmatic children, 31 with non-severe asthma, treated with inhaled corticosteroids [ICS] or not, 11 with severe asthma, and 15 HCs. Using Orbitrap LC-MS, the researchers demonstrated that metabolomic profiling clearly distinguished the three groups [70]. Furthermore, Ferraro et al.⁷¹applied an untargeted LC-MS analysis to EBC and a targeted MS analysis to urine samples from 26 asthmatic children and 16 age-matched HCs. Based on the ensuing metabolomic profiles, asthmatic children were distinguishable from HCs at baseline and after a three-week course of beclomethasone dipropionate (BDP) treatment. Key metabolites included omega-amino fatty acids, prostaglandins, N-acyl amides, and mono glycerophospholipids. Interestingly, despite significant clinical and functional improvements, the metabolomic profile did not substantially change, and neither BDP nor its metabolites were detected in urine. The patients’ endogenous steroid metabolism also appeared unaffected by the treatment [71].

NMR-based metabolomics is a valuable tool in asthma studies as well. Carraro et al. [27] performed NMR analysis on EBC samples from 25 children with allergic asthma (8 with intermittent asthma without ICS treatment, 17 with persistent asthma and ICS treatment) and 11 age-matched HCs. They identified acetylated and oxidized compound profiles that effectively distinguished asthmatic from HCs, indicating elevated oxidative stress markers in EBC from asthmatic subjects and the presence of asthma-related metabolic pathways [27]. More specifically, retinoic acid and deoxyadenosine—both associated with airway remodeling and inflammation—were identified as key metabolites in severe asthma, while the active vitamin D metabolite ercalcitriol was observed in non-severe asthma forms. This finding supports an inverse relationship between vitamin D levels and asthma severity [72].

Further demonstrating the potential of breathomics in asthma studies, Sinha et al. [73] combined NMR spectroscopy with machine learning to examine 89 asthmatic children and 20 HCs. Even though they did not quantify specific metabolites, their model could distinguish asthma patients from HCs based on NMR spectra patterns. Furthermore, they were able to cluster the EBC spectra from asthma patients into three groups with different clinical and histological features, suggesting that EBC-NMR can discriminate among asthma endotypes [73].

Aiming to identify relevant metabolites, Chiang et al. [74] also used NMR-based metabolomic profiles of EBC samples to derive a four-metabolite signature—lactate, formate, butyric acid, and isobutyrate—which effectively differentiated children with asthma from HCs [74]. Along the same lines, Ntontsi et al. combined NMR and UHPLC-MS to discriminate severe from mild-to-moderate asthma based on EBC metabolomics, identifying five discriminatory metabolites, with lysine being the most distinctive [26]. Similarly, Maniscalco et al. [75] investigated the existence of an asthma-obesity metabotype via NMR analysis of EBC from 25 obese asthmatics compared to 30 obese non-asthmatic and 30 lean asthmatic individuals. This study was designed based on prior evidence suggesting that obesity may be considered a risk factor for asthma and may influence its severity, control, and therapeutic response [75]. Remarkably, NMR profiling successfully distinguished obese asthmatic patients; key differentiators included increased levels of glucose, butyrate, and acetoin levels and decreased levels of formate, tyrosine, ethanol, ethylene glycol, methanol, n-valerate, acetate, saturated fatty acids, and propionate. The differentiation between obese and lean asthmatics featured a similar profile, with increases in glucose, n-valerate, acetoin, isovalerate, and 1,2-propanediol levels and decreases in formate, ethanol, methanol, acetone, propionate, acetate, lactate, and saturated fatty acid levels [76].

Focusing on lipidomics, Sanak et al. [77] analyzed EBC samples from 115 adult asthmatic subjects (including 62 with aspirin intolerance) and 38 HCs to quantitatively assess 19 eicosanoids using complementary methods including HPLC and gas chromatography-MS (GC-MS). Palmitic acid concentrations were used to normalize for sample dilution. This highly sensitive eicosanoid profiling of EBC revealed characteristic alterations in asthma, particularly an upregulation of cyclooxygenase (COX) pathway activity, lipoxygenase-derived arachidonate products, and elevated levels of cysteinyl leukotrienes. A distinct phenotype associated with hypersensitivity to aspirin and other NSAIDS showed marked increases in prostaglandin D₂ and E₂ metabolites, as well as higher concentrations of 5- and 15-hydroxyeicosateraenoic acid compared to aspirin-tolerant subjects. Thus, this study not only differentiated asthmatic from HCs but also discriminated between aspirin-sensitive and aspirin-tolerant asthma phenotypes [77]. On Table 1 are summarized metabolomic and proteomic studies on EBC from asthmatic patients.

Table 1.

Summary of metabolomic and proteomic studies on EBC from asthmatic patients.

Author, Year [Ref]	Aim	Study Subjects	Conclusion
Carraro et al. 2007 [27]	To evaluate the feasibility of NMR–based metabolomics applied to EBC	Healthy and asthmatic children	NMR-based metabolomics distinguishes asthma patients with profiles suggestive of acetylated and oxidized compounds
Bloemen et al. 2010 [69]	To identify disease-specific proteolytic peptides or protein patterns in asthmatic patients via proteomics	Healthy and asthmatic children	EBC contains proteins useful for non-invasive asthma diagnosis and monitoring
Sanak et al. 2011 [77]	To assess eicosanoids in asthma phenotypes through MS	Asthmatic patients and HCs	Eicosanoid profiling to identify asthma-related changes, especially an increase in cyclooxygenase and lipoxygenase pathways and a distinct profile for hypersensitivity to NSAIDs
Carraro et al. 2013 [70]	To detect disease-specific metabolite patterns in asthmatic patients using LC-MS	Healthy and asthmatic children of varying severity	EBC metabolomics allows non-invasive asthma diagnosis and stratification of severe forms. ▲retinoic acid, deoxyadenosine, and ercalcitriol.
Sinha et al. 2017 [73]	To investigate the potential diagnostic or endotype discovery utility of global NMR spectrum patterns combined with machine learning	Asthmatic patient and HCs (< 18 years old)	NMR spectra of EBC differentiate patients from controls and form three clinical endotype clusters
Ferraro et al. 2020 [71]	To apply LC-MS untargeted analysis to assess EBC gathered from asthmatic children	Healthy and steroid-naive asthmatic children before and after treatment	Breathomics distinguishes asthmatic from healthy children but not treatment effect. Putative markers: prostaglandins, fatty acids, and glycerophospholipids
Ntontsi et al. 2020 [26]	To discriminate severe asthma from mild-to-moderate asthma using NMR and identify key metabolites by LC-MS	Severe and mild-to-moderate asthmatic subjects	Five discriminatory metabolites; lysine identified as key differentiator
Chang-Chien et al. 2021 [74]	To diagnose pediatric asthma using NMR	Asthmatic patients and HCs	EBC metabolomic signature: ▲lactate, formate, butyrate, and isobutyrate

NMR: nuclear magnetic resonance; LC-MS: liquid chromatography–mass spectrometry; EBC: exhaled breath condensate; HCs: healthy controls; NSAIDs: nonsteroidal anti-inflammatory drugs.

Metabolomic studies on COPD EBC

COPD is defined by the Global Initiative for Chronic Obstructive Lung Disease (GOLD) as a heterogeneous lung condition characterized by chronic respiratory symptoms including dyspnea, cough, sputum production, and acute exacerbations. This complex disorder arises from abnormalities in the airways (e.g. bronchitis and bronchiolitis) and alveoli (e.g. emphysema), leading to persistent, non-reversible and often progressive airflow obstruction [78].

The development of COPD is shaped by lifetime interactions between genetic predisposition and environmental exposures. Among external risk factors, tobacco smoke and environmental pollution represent the most significant contributors. On the genetic side, mutations in the SERPINA1 gene that lead to α-1 antitrypsin deficiency (AATD) are recognized as a major hereditary risk factor [79].

The diagnosis of COPD is established through spirometry, which confirms the presence of non-fully or not-reversible airflow obstruction. The key diagnostic criterion is a post-bronchodilator forced expiratory ratio (FER) of forced expiratory volume in one second/forced vital capacity (FEV1/FVC) < 0.7. Disease severity is further stratified based on the FEV1% predicted: mild (≥ 80%), moderate (50-80%), severe (30-49%), and very severe (< 30%).

This condition typically manifests with symptoms such as dyspnea, limitations in physical activity, and cough, with or without sputum production. Additionally, individuals may also experience acute respiratory events marked by a notable escalation in respiratory symptoms, commonly referred to as exacerbations. Moreover, COPD patients often experience the coexistence of various comorbidities (e.g. cardiovascular diseases, metabolic conditions, anxiety, and depression), adding complexity to their overall health management. The interplay of these comorbidities can significantly impact the prognosis, exacerbation frequency, and QoL of individuals living with COPD [80].

Even though COPD is a prevalent, preventable, and treatable respiratory condition, delayed or incorrect diagnoses remain a widespread issue, often resulting in patients receiving inappropriate therapies or no treatment at all. Indeed, underdiagnosis continues to represent a major challenge, hindering timely intervention and effective disease management. Improving awareness and implementing comprehensive diagnostic approaches are imperative to address this issue, ensuring accurate and timely diagnoses. This is essential to preserve lung function, improve patients’ QoL, and provide targeted care.

Several studies have shown that EBC could be promising source of biological macromolecules (e.g. protein, lipids, and nucleic acids) for COPD studies [81]. These investigations have indeed provided potential biomarkers for disease diagnosis and stratification, while also shedding light on disease mechanisms through the assessment of specific biomarker profiles, the identification of disease-characterizing metabolites, and the discrimination of relevant metabolic pathways involved in phenotypization. For instance, analyses of EBC from COPD patients have revealed altered levels of various growth factors and inflammatory biomarkers, including platelet derived growth factor (PDGF), granzyme B, IL-6, IL-8 [82], extracellular ATP [83], soluble HLA-I and HLA-II molecules [84], and various eicosanoids, such as 8-isoprostane and 15-isoprostane, PGE2 and PGF2α, and LTB4 [85–87].

Evidence of ongoing tissue remodeling is also supported by increased concentrations of MMP-9, MMP-12, TIMP-1, TIMP-4 [88], and intriguingly alpha-1 antitrypsin [89]. Moreover, changes in markers of cellular stress have been reported, including elevated malondialdehyde (MDA) and decreased lactate levels [90].

Despite these findings, large biobank and registry-based studies have shown that current exhalative biomarkers, such as FeNO, PGE2, and 8-isoprostane, lack sufficient specificity to distinguish from idiopathic pulmonary fibrosis or HCs, underscoring the urgent need for the identification of novel biomarkers [91]. To address this gap, a number of studies have examined the omic profiles of EBC in COPD patients. Noticeably, almost all published reports to date indicate that selected protein levels in smokers EBC are higher than those observed in non-smokers, though the clinical and biological significance of these observations remains unclear [92]. However, only a few studies focusing specifically on the proteomic profiles of EBC among COPD patients have been carried out. In a seminal study, Fumagalli et al. compared non-smokers, healthy smokers, COPD patients without emphysema, and individuals with pulmonary emphysema associated with AATD, identifying distinct proteomic fingerprints: 44 proteins in non-smokers/healthy smokers, 17 in COPD patients, and 15 in AATD subjects. Using SELDI-MS and Western blot analysis, the authors detected and validated several inflammatory cytokines (i.e. IL-1, IL-2; IL-12, IL-15, INF-α, INF-γ, and TNF-α), as well as signaling and regulatory proteins, including cytokeratins, surfactant protein A isoforms, calgranulin and α1-antitrypsin. Of note, cytokeratins I and II were predominant (34%) in EBC of HCs, both smokers and non-smokers. In contrast, cytokines such as IL-1α, IL-1β, IL-2, IL-12α, IL-12β, and IL-1, were identified as the dominant protein type (62%) in COPD and AATD patients, alongside INF-α, INF-γ, TNF-α, and complement C3. Noteworthy, comparative EBC proteomic analysis between COPD and AATD patients revealed significant differences. In AATD patients, α ₁-antitrypsin was not detected, whereas IL-1α and lysozyme-C—undetectable in the other COPD patients—were identified. Although these differences in proteomic profile might represent a valuable tool to better understand the pathogenetic mechanisms involved in COPD, form their data it remains unclear whether the observed differences in protein number between nonsmokers, healthy smokers, and the patient groups reflect true disease-specific alterations or are simply the result of variations in protein concentration and sample handling procedures. This uncertainty underscores the importance of corroborating EBC proteomic findings with broader evidence from the literature. In this regard, many of the identified markers are consistent with previous reports on lung inflammation [93]. For example, Sun et al. [94] employed tandem mass tag (TMT)-based quantitative MS to analyze EBC samples from 19 COPD patients and 19 HCs. A total of 257 proteins were identified. Twenty-four proteins were differentially expressed in COPD patients compared to the control group. Gene ontology (GO) analysis indicated that these were predominantly cytoplasmic, while Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis showed they were involved in inflammatory responses [94].

The metabolomic profiling of COPD patients has primarily been performed using NMR. De Laurentiis et al. [95] employed (1)H-NMR to analyze 36 paired samples of EBC and saliva obtained from healthy subjects, laryngectomized patients, and individuals with COPD. The study aimed to validate the clinical feasibility of NMR-based metabonomic analysis of EBC in adults while accounting for pre-analytical variables, such as saliva and disinfectant contamination. Their findings demonstrated that EBC metabolomics using NMR is a reliable approach, capable of distinguishing COPD even in the presence of natural and artificial contaminants. Key discriminatory metabolites included increased acetate and pyruvate, alongside reduced or absent succinate, glutamine, choline, and phosphorylcholine [95]. In a subsequent study, the same group analyzed 50 duplicate EBC samples from healthy smokers, smokers with COPD, and individuals with pulmonary Langerhans cell histiocytosis (PLCH). NMR spectra revealed distinct metabolic profiles for each group. COPD patients were characterized by increased levels of acetate, propionate, and 2-propanol, with decreased 1-methylimidazole and isobutyrate, whereas PLCH patients showed low 2-propanol and high isobutyrate. These findings suggest that COPD and PLCH display disease-specific biomarker signatures that is independent of smoking habits [96]. Along the same lines, Bertini et al. [25] conducted an (1)H-NMR-based metabolomics study using EBC from 37 stable COPD patients and 25 non-obstructed controls. Their metabolomics analysis showed significantly lower levels of acetone, valine, and lysine in COPD patients, along with significantly higher levels of lactate, acetate, propionate, serine, proline, and tyrosine. Unsupervised analysis of NMR spectra stratified the COPD group into three distinct metabolic clusters (metabotypes), although no obvious differences in clinical data were observed between them [25]. This concept of distinct metabotypes was questioned by Kilk et al. who conducted a targeted MS-based analysis of EBC and serum metabolites, focusing on amino acids and lipid species. They also carried out untargeted profiling of both biofluids in COPD patients and HCs with different clinical characteristics. The study identified sphingomyelins as the most effective markers for differentiating COPD patients from HCs. Furthermore, the untargeted analysis unveiled group differences in unsaturated fatty acid-containing lipids, ornithine metabolism, and plasma protein signatures, which aligned with GOLD classifications but ruled out the existence of discrete metabotypes [97].

Investigating the diagnostic potential of metabolomics, Maniscalco et al. [45] used NMR spectroscopy to discriminate between COPD and asthma patients. Their results demonstrated increased ethanol, methanol and reduced formate and acetone/acetoin in COPD, distinguishing it metabolically from asthma [45]. In good keeping, Ghosh et al. [98] in 2021 performed NMR profiling in male smokers with asthma-COPD overlap—a subgroup exhibiting mixed features of both diseases—identifying eight perturbed metabolites, including increased propionate, isopropanol, and acetone and decreased valine, which allowed them to distinguish asthma–COPD overlap (ACO) from both asthma and COPD patients [98]. Similarly, Choudhury et al. [99] used NMR to compare serum and EBC metabolomic profiles in 57 healthy donors, 59 COPD patients, and 60 COPD patients with pulmonary hypertension. They found fifteen serum and 9 EBC metabolites allowing patients classification. Notably, downregulation of alanine and pyruvate coupled to upregulation of glycerol and lactate were found in both serum and EBC of COPD patients with pulmonary hypertension, indicating conserved trends across biofluids and suggesting metabolic reprogramming toward glycolysis and lipolysis [99]. In contrast, Zabek et al. [100] examined serum, urine, and EBC to differentiate COPD from obstructive sleep apnea syndrome (OSAS) and found that while metabolite panels from serum and urine successfully classified the two conditions, incorporating EBC data did not improve diagnostic accuracy, indicating that exhaled metabolomics may not be effective for distinguishing COPD from OSAS [100].

The application of breathomics for follow-up was demonstrated by Montuschi et al. [46] who used carbon polymer sensor e-nose, quartz crystal sensor e-nose, and NMR-based metabolomics to track metabolic changes following inhaled corticosteroid (ICS) treatment and withdrawal among COPD patients. Each technique yielded consistent results and complemented the spirometry data, providing a multidimensional perspective analysis for evaluating the anti-inflammatory effects of corticosteroids. Focusing on NMR-based analysis of EBC, sputum, serum, and urine, the same group demonstrated that different corticosteroid regimens yield distinguishable metabolomic signatures. Importantly, each biofluid contributed orthogonal information: for instance, serum formate increased over time, whereas EBC formate decreased, highlighting the value of multi-compartment analysis [24]. Similarly, Maniscalco et al. [101] monitored COPD patients undergoing pulmonary rehabilitation, reporting a progressive remodeling of the metabolome, with reductions in acetate, methanol, 3-hydroxyisovalerate, isobutyrate, and acetone levels together with an increase in isopropanol, 1-propanol, lactate, and fatty acids levels among rehabilitated COPD patients. Intriguingly, the progressive decrease of methanol paralleled the reduction in dyspnea and fatigue and the increase in walking distance, suggesting a close correlation between metabolic inflammation-related metabolites and clinical features [101]. A summary of omics studies on EBC from COPD patients is reported on Table 2.

Table 2.

Summary of omics studies on EBC from COPD patients.

Author, Year [Ref]	Aim	Study subjects	Conclusion
de Laurentiis et al. 2008 [95]	To validate the NMR metabonomic analysis of EBC in adults	HCs, laryngectomized patients, and COPD patients	EBC NMR fingerprint shows: ▲ acetate and pyruvate, ▼succinate, glutamine, choline, and phosphatidylcholine
Fumagalli et al. 2012 [93]	To obtain proteomic fingerprints for future research	Non-smokers, healthy smokers, COPD patients, and AATD-associated emphysema patients	Inflammatory cytokines, cytokeratins, surfactant proteins, calgranulins, and other proteins may be indicators for distinct COPD variants
de Laurentiis et al. 2013 [96]	To use NMR-to differentiate patients with shared risk factors	Current smokers without COPD, COPD smokers, and PLCH subjects	The EBC NMR profile of COPD is different from that of healthy smokers (▲ acetate, propionate, 2 propanol and ▼ 1metyl-imidazole, isobutyrate) and PLCH subjects (▼ 2-propanol and ▲isobutyrate)
Bertini et al. 2014 [25]	To identify COPD-specific EBC metabolite profiles	Stable COPD patients and non-obstructed controls	COPD: ▲ lactate, acetate, propionate, serine, proline, tyrosine; ▼ acetone, valine, and lysine. Distinct metabotypes observed
Zabek et al. 2015 [100]	To differentiate COPD from OSAS via biofluid analysis	COPD and OSAS patients	Serum and urine markers can identify the two groups; EBC metabolites do not aid diagnosis
Kilk et al. 2018 [97]	To characterize COPD with MS-based metabolomics of both peripheral blood and EBC	COPD patients with different clinical characteristics and HCs	MS based metabolomic approaches succeed in identifying biomarkers and elucidating disease mechanisms, but no clear co-variate-independent metabotypes were found
Maniscalco et al. 2018 [45]	To distinguish COPD from asthma by NMR	Newly diagnosed COPD and asthma patients	Discriminatory model validated that successfully distinguishes COPD from asthma
Montuschi et al. 2018 [46]	To track metabolic changes associated with ICS using multimodal analysis	COPD patients under ICS treatment and after withdrawal	NMR, e-nose, and classical spirometry yield complementary data for anti-inflammatory treatment assessment
Sun et al. 2019 [94]	To identify differentially expressed proteins in EBC	COPD patients and controls	24 inflammation-related proteins may serve as novel EBC biomarkers for COPD
Vignoli et al. 2020 [24]	To use NMR metabolomics to assess treatment-specific signatures	COPD patients under different pharmacologic regimens	Different biofluids provide orthogonal information useful for model-based classification
Ghosh et al. 2021 [98]	To ascertain the utility of NMR in ACO endotyping and diagnosis	Patients with ACO, COPD, asthma, and controls	Eight discriminatory metabolites identified for ACO; predictive model validated
Choudhury et al. 2021 [99]	To define metabolic signature of COPD with pulmonary hypertension	COPD patients with or without pulmonary hypertension and HCs	Found that metabolic features in blood and EBC can not only efficiently distinguish COPD with pulmonary hypertension from HCs but also from other COPD patients
Maniscalco et al. 2022 [101]	To correlate metabolic changes with clinical outcomes	COPD patients undergoing pulmonary rehabilitation	Methanol and other biomarkers correlated with symptom improvement; potentially useful for follow-up

COPD: chronic obstructive pulmonary disease; AATD: α-1 antitrypsin deficiency; NMR: nuclear magnetic resonance; EBC: exhaled breath condensate; ACO: asthma–COPD overlap; OSAS: obstructive sleep apnea syndrome; HCs: healthy controls.

Metabolomic studies on rhinitis EBC

Rhinitis is a broad term encompassing various nasal symptoms, including nasal congestion or obstruction, rhinorrhea, sneezing, and pruritus, resulting from inflammation and/or dysfunction of the nasal mucosa. As one of the most prevalent medical conditions worldwide, rhinitis contributes substantially to global morbidity and imposes a significant economic burden. Importantly, its clinical relevance extends beyond localized nasal discomfort, as it is recognized as a risk factor for asthma and is tightly linked to chronic conditions like rhinosinusitis [102].

Rhinitis, a complex and prevalent condition, is due to various etiologies, such as infections, allergic responses, environmental irritants, medications, hormonal imbalances, and neuronal dysfunction. Traditionally, rhinitis is categorized into three major clinical phenotypes: i) allergic; ii) non-infectious, non-allergic; and iii) infectious. However, mixed phenotypes are also observed in clinical practice, underscoring the multifactorial nature of this condition [103].

Allergic rhinitis (AR) is an immunoglobulin E (IgE)-mediated atopic condition causing inflammation of the nasal mucosa. AR results from a type I hypersensitivity response triggered by exposure to common allergens, like dust mite fecal particles, cockroach residues, animal dander, molds, and pollens, which initiates the recruitment of inflammatory cells, such as mast cells, CD4-positive T cells, B cells, macrophages, and eosinophils, into the nasal lining. A hallmark of AR is the prevalence of T helper 2 (Th2) cells in the nasal mucosa, which release cytokines such as IL-3, IL-4, IL-5, and IL-13, thereby promoting IgE production by plasma cells. Upon subsequent allergen exposure, crosslinking of IgE on mast cells triggers the release of histamine and leukotrienes, which contributes to the clinical manifestations of AR, such as arteriolar dilation, increased vascular permeability, itching, rhinorrhea, mucus secretion, and in the lungs, bronchoconstriction. The ensuing late-phase inflammatory response, occurring 4-8 h later, further amplifies these effects, leading to persistent symptoms, particularly chronic nasal congestion [104].

Non-infectious, non-allergic rhinitis (NAR) is marked by chronic inflammatory symptoms of the nasal mucosa, featuring a minimum of two nasal symptoms, such as nasal obstruction, rhinorrhea, sneezing, and/or an itchy nose. Diagnosis involves a thorough past medical history assessment and the exclusion of endonasal infectious etiologies and allergic sensitization. Within the spectrum of NAR, at least six subphenotypes have been identified: drug-induced rhinitis, gustatory rhinitis, hormone-induced rhinitis, rhinitis of the elderly, atrophic rhinitis, and idiopathic rhinitis [105].

Infectious rhinitis is primarily attributed to viral pathogens rather than bacterial agents. Common viruses implicated include rhinovirus, coronavirus, adenovirus, influenza virus, parainfluenza virus, respiratory syncytial virus, and enterovirus. These viruses compromise epithelial integrity by disrupting tight junctions and cell membranes, invading epithelial cells, and manipulating host cell metabolic processes to support viral replication—culminating in cell death. Symptoms of infectious rhinitis generally follow a self-limiting course and often resolve without medical intervention [106].

Rhinitis is therefore a complex condition characterized by various clinical presentations and underlying pathophysiological mechanisms. Because of this heterogeneity, it can be difficult for clinicians to accurately identify the specific type of rhinitis a patient has. Identifying these overlapping entities is thus essential for developing more precise diagnostic tools and tailored, effective treatments.

To date, only a single study has comprehensively investigated the omic profile of EBC samples associated with rhinitis, highlighting a scarcity of research in this domain [107]. A limited number of studies have focused on individual markers such as FeNO [108] and adenosine [109]. Cap et al. [110] investigated the presence of eicosanoid in non-asthmatic individuals with seasonal allergic rhinitis and found increased levels of leukotrienes B4 and E4 during pollen season. Likewise, Kowal et al. [111] who reported increased levels of 5-oxo-ETE, leukotriene D4, and 8-iso-PGE2 in patients reactive to dust mites. An additional study using an electronic nose showed how allergic and asthmatic patients could be discriminated based on their volatilome profiles [112]. Despite these promising findings, a comprehensive characterization of rhinitis using omic approaches is lacking.

Celik et al. [107] investigated oxidative stress levels in children with asthma, AR, or both, seeking to determine whether their coexistence amplifies oxidant stress within a unified airway concept. The study used malondialdehyde (MDA) and reduced glutathione (GSH) as biomarkers to compare nasal and oral EBC levels among children with asthma, AR, both asthma and AR, and HCs. All patient groups showed increased MDA and reduced GSH levels compared to HCs. Interestingly, oral MDA was lower in patients with concomitant asthma and AR than in those with asthma alone. These findings suggest that oxidative stress is elevated in both asthma and AR, and that their co-occurrence does not intensify oxidative stress in pediatric airways [107] Table 3.

Table 3.

Omics studies on EBC from rhinitis patients.

Author, Year [Ref]	Aim	Study subjects	Conclusion
Celik et al. 2012 [112]	To evaluate oxidative stress levels in children with asthma and/or AR using nasal and oral EBC	Children with asthma, AR, asthma and AR, and HCs	Asthma and AR in children are individually associated with elevated oxidative stress in the airways, but their coexistence does not further increase it

AR: allergic rhinitis; EBC: Exhaled breath condensate; HCs: healthy controls.

EBC studies in infectious respiratory diseases: the example of COVID-19

Infectious respiratory diseases are caused by microorganisms, like bacteria, viruses, or fungi, and often manifest with acute symptoms, rapid transmission, and dynamic epidemiological patterns. The worldwide impact of the COVID-19 pandemic, caused by the novel coronavirus SARS-CoV-2, has drawn unparalleled global attention and reshaped the approach to infectious diseases, influencing strategies for control, surveillance, and treatment to safeguard public health.

These diseases encompass infections of both the upper or lower respiratory tract. Upper respiratory tract infections (URTIs) present a spectrum of conditions, including the common cold, laryngitis, pharyngitis/tonsillitis, sinusitis, and acute otitis media. Lower respiratory tract infections (LRTIs) comprise acute bronchitis, bronchiolitis, pneumonia, and tracheitis [113]. URTIs affect the nose, sinuses, pharynx, larynx, and large airways, predominantly stemming from viral sources, except in cases of epiglottitis and laryngotracheitis, where Haemophilus influenzae type b may be responsible for severe instances, and bacterial pharyngitis, which is primarily attributed to Streptococcus pyogenes. URTIs are defined as self-limited irritative and inflammatory diseases of the upper airways, typically accompanied by cough, in the absence of pneumonia, chronic bronchitis, emphysema, or COPD. Common viral pathogens include rhinovirus, influenza virus, adenovirus, enterovirus, and respiratory syncytial virus.

URTIs typically result from the direct invasion of the upper airway mucosa by pathogens, most often transmitted through inhalation of infected respiratory droplets. Symptomatic treatment is recommended for viral infections, while antibacterials are usually administered to patients with streptococcal pharyngitis or H. influenzae-caused epiglottitis [114]. LRTIs involve various causative agents, with viruses predominantly responsible for bronchitis and bronchiolitis, while bacteria, particularly Streptococcus pneumoniae, are commonly associated with community-acquired pneumonias. On the other hand, atypical cases of pneumonia can be attributed to pathogens like Mycoplasma pneumoniae, Chlamydia spp, Legionella, Coxiella burnetii, and viruses. These infections reach the lower respiratory tract through inhalation, aspiration, or hematogenous spread, leading to inflammation, increased mucus secretion, and impaired mucociliary clearance. In severe cases, such as bronchiolitis, inflammation, and epithelial necrosis, small airway obstruction may occur. LRTI patients typically present with cough, fever, chest pain, tachypnea, and sputum production. Pneumonia may also cause non-respiratory manifestations like confusion, headache, myalgia, abdominal pain, nausea, vomiting, and diarrhea. The bacterial predominance and potential for systemic involvement distinguish LRTIs from URTIs.

In the context of respiratory viruses, COVID-19 is a distinct pathogen that has garnered significant attention owing to its causative agent, SARS-CoV-2. COVID-19 manifests through a diverse array of symptoms, ranging from fever, cough, and shortness of breath to fatigue, muscle aches, headache, sore throat, and loss of taste or smell. In severe cases, it can lead to complications like pneumonia, acute respiratory distress syndrome (ARDS), and COVID-associated-coagulopathy, which results from an interplay between thrombosis and inflammation and can be fatal [115].

SARS-CoV-2 primarily enters the body through the respiratory tract, often through inhalation of infected droplets. Once inside the host, SARS-CoV-2 targets epithelial cells of both the upper and lower respiratory tract by binding to angiotensin-converting enzyme 2 (ACE2) receptors, facilitating viral entry and replication. This infection can initiate a cascade of respiratory symptoms, from mild cough to severe ARDS, driven by an immune response that causes widespread lung inflammation, tissue damage, and impaired oxygen exchange [116].

Beyond the acute phase, significant subset of individuals experiences post-acute sequelae, commonly referred to as long COVID. This syndrome is characterized by persistent symptoms affecting various organ systems, extending well beyond the respiratory tract. While ongoing respiratory symptoms, such as chronic cough and breathlessness, are frequent, other manifestations include fatigue and cognitive impairment [117]. This prolonged impact underscores the complex pathophysiology of COVID-19 and its long-term impact on health.

In this regard, our group conducted an untargeted metabolomics study using two-dimensional gas chromatography mass spectrometer (GCxGC-TOFMS) on EBC from 26 COVID-19 patients, 11 cardiopulmonary edema patients, and 20 HCs. The objective was to identify novel biomarkers for noninvasive diagnosis and disease monitoring. Interestingly, among 26 differentially abundant molecules, the most discriminating where lipid or their derivates, including several monoacylglycerols with a putative protective effect. These findings suggest that EBC metabolomic profiling can differentiate COVID-19 not only from HCs but also from other pathologies [118].

Using a targeted LC-MS approach focused on oxidated polyunsaturated fatty acids (oxylipins) Borras et al. [119] observed increased levels of cytochrome P450, COX and 15-LOX products in EBC from COVID-19 patients. These included PGA2, PGD2, LXA4, 5-HETE, 12-HETE, 15-HETE, 5-HEPE, 9-HODE, 13-oxoODE, and 19(20)-EpDPA. Fittingly, all these molecules are linked to inflammatory and thrombotic pathways, corroborating their potential relevance to COVID-19 pathophysiology [119].

Berna et al. [120] identified candidate biomarkers for COVID-19 in pediatric patients using GCxGC-TOFMS to analyze two cohorts of patients comprising 22 SARS-COV-2 positive and 27 negative controls. Their targeted analysis of odorant VOCs unveiled a panel of 6 upregulated molecules, including straight-chain aldehydes (i.e. octanal, nonanal, and heptanal), acids (i.e. decane and tridecane), and 2-pentyl furan. This VOC panel successfully distinguished infected individuals in an independent validation cohort [120].

In a pilot study, Grassin-Delyle et al. [121] used TOFMS to discriminate between adults with COVID-19-associated acute respiratory syndrome (ARSD) and those with non-COVID ARDS. Specific VOCs—methylpent-2-enal, 2,4-octadiene 1-chloroheptane, and nonanal—allowed for efficient discrimination of COVID-19-positive patients, supporting the potential diagnostic value of breathomics [121].

Ruszkiewicz et al. [122] adopted a point of care approach using gas chromatography-ion mobility spectrometry (GC-IMS) to distinguish COVID-19 from other respiratory diseases, such as COPD, asthma, and bacterial pneumonia. Once again aldehydes (i.e. ethanal, octanal), ketones (i.e. acetone, butanone), and methanol discriminated COVD patients [122].

More recently, Paris et al. [123] aimed to define EBC biomarkers of the post-COVID-19 clinical state. Using NMR, they analyzed EBC from 38 patients experiencing persistent symptoms following viral clearance, compared to 38 matched controls. Their results revealed a distinctive spectral profile characterized by increased ethanol, lactate, and acetoin, alongside reduced levels of acetate, acetone, fatty acids, isocaproate, isovalerate, methanol, and valerate. Remarkably, these metabolic alterations correlated with clinical biochemical parameters in some patients. In addition, the observed metabotype was associated with specific changes in EBC microRNAs, suggesting a potential mechanistic link to post-COVID symptomatology. A summary of omics studies on EBC from COVID-19 patients is reported on Table 4.

Table 4.

Omics studies on EBC from COVID-19 patients.

Author,Year [Ref]	Aim	Study subjects	Conclusion
Barberis et al. 2021 [118]	To identify COVID-19 biomarkers by MS and differentiate it from other diseases	26 COVID-19 adult patients, 20 HCs, and 11 cardiopulmonary edema patients	25 modulated molecules in COVID-19; among lipid markers, monoacylglycerols support the differential diagnosis
Berna et al. 2021 [120]	To detect COVID-19 VOC biomarkers in children by MS	22 SARS-COV-2 positive and 27 negative children	Discrimination possible and validated in a second court through the use of a 6 molecule panel (▲ octanal, nonanal, heptanal, decane, tridecane and 2-pentyl furan)
Grassin-Delyle et al. 2021 [121]	To identify COVID-19 positive patients among ARDS patients	A cohort of ARDS patients with and without COVID-19	VOCs can discriminate COVID-19 patients, implying translational diagnostic utility
Ruszkiewicz et al. 2021 [122]	To develop point-of-care MS diagnosis for respiratory diseases	Patients with COVID-19 and several other respiratory diseases	Aldehydes, ketones, and methanol enable discrimination of COVID-19 from other respiratory diseases
Borras et al. 2023 [119]	To characterize alteration in oxylipins during COVID-19 by LC-MS	COVID-19 patients and HCs	Multiple upregulated products of cytochrome P450, COX and 15-LOX
Paris et al. 2023 [123]	To identify long COVID biomarkers and correlate them with microRNA profiles via NMR	Symptomatic post-COVID-19 patients and matched controls	NMR-based profile shows ▲ethanol, lactate, and acetoin ▼ acetate, acetone, fatty acids, isocaproate, isovalerate, methanol, and valerate

COVID-19: coronavirus disease 2019; MS: mass spectrometry; VOC: volatile organic compound; ARDS: acute respiratory distress syndrome; NMR: nuclear magnetic resonance; LC-MS: liquid chromatography–mass spectrometry.

Discussion

In chronic respiratory research, there is a pressing need to establish robust biomarker profiles capable of distinguishing among various respiratory diseases. These biomarkers would not only enhance diagnostic accuracy but also serve prognostic purposes by providing cues on disease evolution and treatment effectiveness. From the reviewed studies, it is clear that a more comprehensive identification of specific metabolites that can accurately characterize conditions such as asthma, COPD, and rhinitis, could lead to the identification of endotypes defined by metabolic features, also referred to as ‘metabotypes’. These unique molecular signatures are essential for advancing both diagnostic precision and our understanding of the complex metabolic pathways involved in respiratory disease pathophysiology and progression.

In this scenario, the application of NMR, MS, or electronic nose-based metabolomic technologies represents a powerful investigative approach in respiratory diseases. In some cases, these tools are progressing toward clinical implementation, as exemplified by portable devices such as Inflammacheck^®, which measures H₂O₂ in EBC, and other point-of-care diagnostic efforts [122].

Although each method provides different and often orthogonal information, certain patterns of specificity are beginning to emerge. Both MS and NMR have demonstrated the ability to identify disease-specific metabolic profiles and metabotypes. NMR-based approaches, while generally less labor-intensive, are particularly useful for distinguishing between patient subgroups, although they tend to rely on a smaller set of volatile, low-molecular-weight metabolites. Conversely, MS-based techniques offer high sensitivity and specificity, especially when integrated with targeted workflows and specialized enrichment/separation techniques. The strength of MS-based analysis lies not only in biomarker identification but also in its ability to shed light on the pathophysiological mechanisms associated with chronic and infectious respiratory conditions. This is especially true for EBC proteomics, which enables noninvasive investigations of lung biology—providing a valid alternative to biopsy—though it remains limited by its analytical complexity.

When we consider the classes of molecules identified as discriminatory and the functionally enriched pathways, several trends are recognizable. In COPD patients, NMR-based metabolomic studies frequently detect increased levels of monocarboxylic acids, such as acetate, pyruvate/lactate, and propionate, and decreased levels of acetone/acetoin [95,96]. Similarly, COVID-19 patients show increased ethanol, lactate, acetoin, and decreased acetate, acetone, and fatty acids [123]. In contrast, the discriminating metabolites in asthma include increased acetoin and decreased methanol, ethanol, and formate [45,76]. Taken together, these findings suggest that enriched biochemical pathways are primarily associated with fatty acid catabolism and glycolysis, functionally linked to metabolic and oxidative stress [74,98].

When we evaluate the results of MS-based approaches, we find compelling evidence for the involvement of lipids and fatty acids, not only as metabolic fuels and sources of smaller molecules, but more importantly as precursors of oxidized compounds driving signaling pathways involved in inflammation and thrombosis. Indeed, several studies have employed targeted MS approaches to quantify prostaglandins, thromboxane, leukotrienes, and lipoxins, which often require enrichment procedures for accurate measurement [77,111,119]. Untargeted MS studies have expanded the molecular landscape further, identifying compounds involved in amino acid metabolism and other pathways, depending on the extraction an enrichment procedure used [120,124], reflecting both the complex nature of metabolic alterations in respiratory diseases and the current lack of methodological standardization across studies.

Looking ahead, the rapid advancement of omics technologies holds promise for the discovery of novel biomarkers, the identification of therapeutic targets, and the development of personalized treatment strategies in respiratory medicine. However, realizing this potential will require substantial efforts in the areas of clinical validation and methodological standardization. The integration of these approaches into clinical practice could signal the onset of a precision medicine era for respiratory care.

Future investigations should also extend beyond the diseases currently represented in the literature, broadening the scope to include a wider range of respiratory conditions and molecular profiles. Incorporating larger, more diverse cohorts—along with longitudinal designs—will improve the robustness and generalizability of findings.

Funding Statement

The authors received no financial support for the research, authorship, and/or publication of this article.

Ethics approval and consent to participate

Non applicable

Consent for publication

Not applicable.

Disclosure statement

All the authors have no relevant affiliations of financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript.

Availability of data and materials

Data sharing is not applicable to this article as no data were created or analyzed in this study.