Exhaled breath analysis for identifying eosinophilic and neutrophilic inflammation in a mixed population of patients with asthma or COPD

Rationale: Asthma and COPD are complex and overlapping diseases including various inflammatory phenotypes with different responses to treatment [Agusti ERJ 2016]. Volatile organic compounds in exhaled air are associated with inflammatory profiles in asthma and COPD. Aim: To determine diagnostic accuracy of exhaled breath analysis by electronic nose for neutrophilic and/or eosinophilic inflammation in a mixed patient population. Methods: This was a cross-sectional study in asthma and COPD patients with eosinophilic (countu003e0.3x10 9 /L) and/or neutrophilic (proportionu003e61%) inflammation in peripheral blood. As part of spirometry breathprints were collected in duplicate by SpiroNose [De Vries JBR 2015]. Data-analysis involved signal processing, ambient correction and statistics based on principal component analysis (PCA) followed by discriminant and ROC analysis. Results: Exhaled breath data of 70 patients (41 asthma/29 COPD) with 1. neutrophilia (n=23) 2. eosinophilia (n=20) 3. neutrophilia and eosinophilia (n=6) and 4. none (n=21) were available. PC2 showed a significant difference (p 75%) and eosinophilia (countu003e0.4 x10 9 /L) ROC curves raised to 0.88±0.06 and 0.87±0.07. Conclusion: Breath analysis by SpiroNose is able to identify neutrophilia and/or eosinophilia amongst patients with chronic airways disease. Higher cut-off values for inflammation increased diagnostic accuracy demonstrating a dose-effect relationship for the expired inflammatory signal.