Detection of exhaled breath olfactory signature of pah with artificial nose: From proof-of-concept to clinical validation - 20/12/14

Novel tools for the non-invasive detection of pulmonary arterial hypertension (PAH) are needed in order to facilitate early diagnosis, particularly in high-risk groups. Volatile organic compounds (VOCs) in the exhaled breath are potential specific biomarkers of disease pathophysiology. We hypothesized that PAH can be detected by exhaled VOCs analysis using a novel artificial nose based on sensor arrays. In the proof-of-concept (PoC) study, we performed a case-control study in 25 PAH patients and 25 healthy controls to assess the diagnostic performance of the artificial nose. Exhaled breath were collected and analysed using an “offline” method. Predictive models were developed using discriminant factor analysis (DFA). Canonical score derived from DFA model was able to discriminate PAH from controls with high discriminative power with receiver operator curve analysis AUC=0.97, P<0.05. Furthermore, the artificial nose separated BMPR2+ heritable PAH versus idiopathic PAH with AUC=0.91, P<0.05. Based on the results of PoC, we have now initiated a validation study aiming to enrol 100 healthy controls, 150 PAH subjects, 50 subjects with chronic thromboembolic pulmonary hypertension, and 40 asymptomatic BMPR2 mutation carriers at high-risk for PAH. Exhaled breath VOCs is also analysed using an “online” method, with real time analysis while the patient exhales, in order to facilitate translation to clinic environment. Gas chromatography-mass spectrometry is also used to identify individual VOCs that compose the olfactory signature to identify biomarkers and signalling pathways.