Validation of electronic nose technology as diagnostic tool for different fibrotic interstitial lung diseases

<bold>Introduction:</bold> Diagnosing fibrotic interstitial lung disease (ILD) requires chest CT-scan, (non)invasive investigations, and multidisciplinary teams (MDT). Diagnosis and treatment delays are common. Previous single-center studies showed that profiling of exhaled volatile organic compounds using electronic nose (eNose) sensor technology has potential to identify ILD quickly and noninvasively. We aim to validate eNose technology for diagnosing different fibrotic ILDs in an international multicenter cohort. <bold>Methods:</bold> Patients with 6 fibrotic ILD MDT diagnoses that often cause diagnostic dilemmas, were included in a prospective longitudinal study in 3 European ILD expert centers. The SpiroNose was used for exhaled breath analysis. Baseline breath data, split in training and test sets, were analysed with partial least squares discriminant and ROC analyses. <bold>Results:</bold> 372 patients (36.3% female) were included of whom 40.3% (n=150) had IPF. Differentiating breath profiles of patients with IPF and other fibrotic ILDs resulted in AUC of 0.94 (95%CI 0.91-0.97) in the training and 0.92 (0.87-0.98) in the test set. Moreover, individual ILDs could be discriminated with AUCs ranging 0.89-0.95 (Figure 1). <bold>Conclusions:</bold> This international study shows that eNose technology can differentiate breath profiles from patients with various fibrotic ILDs. eNose could help reducing diagnostic delay in patients with pulmonary fibrosis. <fig><object-id>erj;64/suppl_68/OA2862/F1</object-id><object-id>F1</object-id><object-id>F1</object-id><graphic></graphic></fig>