Exhaled breath analysis by use of eNose technology: a novel diagnostic tool for interstitial lung disease
Aim: To assess the accuracy of exhaled breath analysis using eNose technology to discriminate between ILD patients and healthy controls, and to distinguish ILD subgroups.
Take home message: Electronic nose (eNose) technology can reliably distinguish ILD patients from healthy controls and effectively differentiate between specific ILD subtypes such as IPF, CTD-ILD, and CHP. This highlights the potential of exhaled breath analysis as a novel, noninvasive diagnostic tool for early and accurate diagnosis of ILDs, which could significantly reduce diagnostic delays and improve patient management in clinical practice.
Introduction
Interstitial lung diseases (ILDs) include over 200 disorders that significantly impact patient morbidity and mortality. Accurate diagnosis of ILDs is challenging due to nonspecific symptoms like cough and dyspnea, often leading to misdiagnosis and delays in treatment. Current diagnostic approaches, such as multidisciplinary team (MDT) discussions, are considered the gold standard but remain invasive and time-consuming. Given the expanding treatment options, there is a critical need for noninvasive diagnostic tools. Exhaled breath analysis using eNose technology, which detects volatile organic compounds (VOCs), is a promising, noninvasive approach to diagnose and monitor ILDs.
Methods
This single-center, cross-sectional study included 322 patients diagnosed with ILD and 48 healthy controls at the Erasmus Medical Center. Diagnosis was confirmed through MDT discussions, following established criteria. Breath analysis was performed using the SpiroNose®, a validated cloud-connected eNose. Participants provided exhaled breath samples through a standardized protocol. Data analysis included partial least square discriminant analysis (PLS-DA) and receiver operating characteristic (ROC) analysis to evaluate the discriminatory ability of the eNose between different patient groups. A 2:1 split was used to create training and validation sets for larger subgroups, and statistical analyses were performed using R software.
Results
The study demonstrated that eNose technology could accurately differentiate between ILD patients and healthy controls, achieving an area under the curve (AUC) of 1.00 for both training and validation sets, with 100% sensitivity, specificity, and accuracy. When distinguishing idiopathic pulmonary fibrosis (IPF) from other ILD subtypes, the model showed an AUC of 0.91 (95% CI: 0.85–0.96) in the training set and 0.87 (95% CI: 0.77–0.96) in the validation set. High diagnostic accuracy was also observed in differentiating individual ILDs, with AUC values ranging from 0.85 to 0.99. This suggests that eNose technology can effectively discriminate between ILD subtypes, including IPF, chronic hypersensitivity pneumonitis (CHP), and connective tissue disease-associated ILD (CTD-ILD).
Discussion
The findings indicate that exhaled breath analysis using eNose technology has strong potential as a noninvasive diagnostic tool for ILD. The eNose achieved complete discrimination between ILD patients and healthy controls and demonstrated high accuracy in distinguishing between different ILD subtypes. Compared to previous studies, the larger sample size in this study likely contributed to improved model performance. Despite demographic differences between healthy controls and ILD patients, additional analyses confirmed that these differences did not affect the breathprint outcomes. Limitations include the lack of a definitive “gold standard” for ILD diagnosis and the cross-sectional nature of the study, which precluded evaluation of the impact of medication use on breathprint results. Future research should focus on larger, multicenter cohorts and include longitudinal data to validate these findings. If confirmed, eNose technology could facilitate early, accurate diagnosis of ILDs, potentially guiding treatment decisions and reducing diagnostic delays.
This study supports the potential of eNose technology as an innovative tool for ILD diagnosis, which could significantly enhance clinical practice by providing a rapid, noninvasive, and reliable diagnostic method.
