Exhaled Metabolite Patterns to Identify Recent Asthma Exacerbations
Aim: To evaluate whether eNose could identify patients that recently had asthma exacerbations.
Take home message: Exhaled breath analysis using eNose technology can reliably identify recent asthma exacerbations by detecting distinct breath patterns. This non-invasive approach shows promise as a tool for monitoring asthma stability and could support timely interventions in asthma management.
Introduction
Asthma is a prevalent chronic respiratory disease, characterized by recurrent symptoms such as wheezing and shortness of breath. Exacerbations, which are acute episodes of worsening symptoms, significantly impact patients’ quality of life and increase healthcare costs. Predicting these exacerbations is crucial for effective asthma management, but current biomarkers, such as sputum eosinophils, are often invasive and not predictive. The study investigates the potential of using an electronic nose (eNose) to detect exhaled metabolite patterns that indicate recent asthma exacerbations. The hypothesis is that patients with a recent exacerbation display distinct exhaled breath profiles detectable by eNose technology.
Methods
The study was a cross-sectional analysis involving 361 adult asthma patients. Data were collected from the BreathCloud database, which includes clinical and exhaled breath data. Exclusion criteria included a diagnosis of COPD or lung cancer and a smoking history of ≥10 pack-years. Exacerbations were defined as acute worsening of symptoms requiring oral corticosteroids (OCS) within the last three months. Breath samples were analyzed using the SpiroNose®, an eNose device with metal oxide semiconductor sensors that detect volatile organic compounds (VOCs). Sensor data were pre-processed and analyzed using principal component analysis (PCA) and linear discriminant analysis (LDA). The dataset was divided into a training (70%) and validation (30%) set to assess the model’s performance.
Results
The study included 252 patients in the training set and 109 in the validation set. Exacerbations were reported in 16.3% of the training set and 10.1% of the validation set. Patients with exacerbations had higher antibiotic use and increased maintenance inhaled corticosteroid (ICS) usage compared to those without exacerbations. LDA analysis of the eNose signals successfully discriminated between patients with recent exacerbations and those without, achieving an area under the curve (AUC) of 0.76 (95% CI: 0.69–0.82) in the training set and 0.76 (95% CI: 0.64–0.87) in the validation set. Sensitivity analysis confirmed the robustness of the results, even when excluding patients based on recent antibiotic use, smoking status, or ICS usage.
Discussion
The findings demonstrate that eNose technology can accurately identify asthma patients who have recently experienced exacerbations, suggesting that these patients exhibit a unique exhaled “breathprint.” This study is the first to validate the use of exhaled breath analysis for detecting recent exacerbations in adult asthma patients. The accuracy, although slightly lower than that observed in similar studies on COPD patients, indicates that distinct VOC patterns may exist for different respiratory conditions. Factors such as antibiotic use and ICS treatment could influence VOC profiles, warranting further research to differentiate exacerbation signals from medication effects. The study’s strengths include its large, diverse asthma population and the use of a validated discriminant model. However, limitations include the reliance on patient-reported exacerbation data and the lack of precise timing for exacerbation events. Future longitudinal studies are needed to explore the dynamics of VOC changes during asthma exacerbations.
In conclusion, eNose analysis offers a non-invasive, promising approach for monitoring asthma control and detecting exacerbations, potentially enabling timely intervention and improved disease management.
