Differentiating interstitial lung diseases from other respiratory diseases using electronic nose technology
Aim: To investigate whether exhaled breath analysis using an eNose has potential as application for early detection of ILD amongst patients with respiratory symptoms.
Take home message: eNose technology can accurately differentiate interstitial lung disease from other respiratory conditions non-invasively, showing promise as a quick and reliable diagnostic tool that could enable earlier detection and treatment.
Introduction
This study investigates the potential of electronic nose (eNose) technology to accurately differentiate interstitial lung disease (ILD) from other respiratory conditions, including asthma, chronic obstructive pulmonary disease (COPD), and lung cancer. ILD often presents with non-specific respiratory symptoms, making diagnosis challenging and frequently delayed. eNose technology analyzes volatile organic compounds (VOCs) in exhaled breath, offering a non-invasive tool for early disease identification.
Methods
The study was a cross-sectional, multicenter trial conducted at Erasmus University Medical Center and Franciscus Gasthuis & Vlietland in the Netherlands. It included 322 patients: 161 with ILD, 65 with asthma, 50 with COPD, and 46 with lung cancer. Exhaled breath samples were collected using the SpiroNose®, and data were analyzed using partial least squares discriminant analysis (PLS-DA). Breath profiles were compared across all patient groups to assess the accuracy of eNose in distinguishing ILD from other respiratory diseases.
Results
The eNose demonstrated high accuracy in differentiating ILD from other respiratory conditions, achieving an area under the curve (AUC) of 0.99 in the test set. ILD was distinguished from asthma (AUC 1.00), COPD (AUC 0.96), and lung cancer (AUC 0.98). Subgroup analyses showed consistent results across different smoking statuses and patient demographics. Additionally, the technology could differentiate idiopathic pulmonary fibrosis (IPF) from COPD and lung cancer with AUCs of 0.93.
Discussion
The findings confirm that eNose technology can reliably identify patients with ILD among those with various respiratory symptoms, offering a non-invasive, rapid, and accessible diagnostic method. The ability to distinguish ILD from other respiratory conditions, including lung cancer and COPD, highlights its potential for early screening and referral. Future studies should focus on validating these findings in larger, diverse cohorts, particularly in primary care settings where early diagnosis is critical.
This research supports the integration of eNose technology as a point-of-care diagnostic tool, improving the diagnostic pathway for ILD and potentially reducing delays in treatment initiation.
