(535) Detection of Staphylococcus aureus in exhaled breath of children with cystic fibrosis

Background: Staphylococcus aureus (SA) is an early pathogen in the respiratory tract of children with cystic fibrosis (CF), identified in up to 80% of patients between the age of 11 to 17 years. However, obtaining lower airway samples for microbiology testing can be challenging. There is an unfilled need for a non-invasive tool that can be used to screen for early CF respiratory infections at the clinical point-of-care. The electronic nose (eNose) is a technology that can detect lung diseases and infection with pathogens such as Pseudomonas aeruginosa (PA) in exhaled breath. We aimed to investigate whether SA infections can be detected in children with CF by eNose breath analysis.

Methods: In this cross-sectional observational study, breath profiles were collected from pediatric CF patients attending routine CF clinic visits. A cloud-connected eNose, the SpiroNose (de Vries et al. 2018 ERJ) was used for breath analysis, with advanced signal processing and ambient correction. Sub-group analyses were performed according to pathogens identified by routine airway microbiology testing performed in sputum or throat swabs, as per clinic routine. Statistical analysis was based on linear discriminant discriminant scores used for receiver operating characteristics (ROC) analysis and determining the area under the curves (AUC-ROC).

Results: Breath was analysed from 100 clinically stable children with CF (median age 12.0, range 5–18 years, ppFEV₁ 91, range 43-126,) and 25 age matched healthy controls (HC). The eNose distinguished between CF and HC with high accuracy of 96.0% and AUC-ROC 0.985 (95% CI 0.966-1). Within the CF group, eNose distinguished between CF patients with positive airway cultures for any CF pathogen and no CF pathogens (no growth or usual respiratory flora) with an accuracy of 79.0% and AUC-ROC 0.791 (95% CI: 0.669-0.913) and between patients positive for SA only and no CF pathogen with an accuracy of 74.1% and AUC-ROC 0.797 (95% CI: 0.698-0.896). Similar differences were seen for PA infection vs no CF pathogens (87.1%, AUC-ROC 0.945, 95% CI: 0.869-1). Interestingly, SA- and PA-specific signatures were driven by different sensors in the SpiroNose. There were no differences in breath profiles comparing patients with SA positive cultures obtained by either throat swab (n=15) or sputum (n=29).

Conclusions: In this prospective observational study, breath profiles of children with CF obtained by the SpiroNose were distinguishable from those of age-matched HC with high accuracy. Differences between HC and CF breath profiles were not explained by airway infection with CF pathogens alone, as similar differences were also seen between HC and CF children with negative airway cultures. Our findings further suggest that airways infection with Staphylococcus aureus (SA) can be detected by eNose in breath of children with CF at the point-of-care, and that SA can be distinguished from Pseudomonas aeruginosa (PA) infection. Future studies are needed to identify the VOCs responsible for the observed differences to provide further insight into the role of SA in CF lung disease.

Acknowledgements:

References: