Background: Airway cultures are an important part of monitoring CF lung disease. Bacteria such as Pseudomonas aeruginosa (PA) and Staphylococcus aureus (SA) frequently cause infections. Yet, collecting reliable samples from young children and individuals on elexacaftor/tezacaftor/ivacaftor therapy poses challenges. Exhaled breath testing, utilizing volatile organic compounds (VOCs) analysis, offers a non-invasive and culture-independent diagnostic strategy, as shown earlier for PA.
 
Aim: To identify exhaled VOCs that enables detecting SA in the airways of people with CF.
 
Methods: The study included 91 participants from two CF cohorts: I-BALL 2 (ages 0-6, n=19) and TERRIFIC-MILE (ages >12, with and without ETI therapy, n=72). Breath samples were analyzed by gas chromatography?mass spectrometry, focusing on VOCs associated with the presence of Staphylococcus aureus, identified through a literature review. SA presence was confirmed by cultures from bronchoalveolar lavage (n=17), sputum (n=50), or cough swabs (n=24). A Sparse Partial Least Squares Discriminant Analysis (sPLS-DA) model of the targeted VOCs, validated by K-fold cross-validation, was used to predict the presence of SA.

Results: The sPLS-DA model showed an accuracy of 70.65% (95% CI: [0.60, 0.80]), with a sensitivity of 92.31% and a specificity of 42.50% to predict the presence of SA. This was statistically significant (P-value: <0.01). Key VOCs used in the model were Benzene acetaldehyde, Butane, 2-methyl-propanal, and dimethyl disulfide.

Conclusion: GC-MS breath analysis showed that VOCs could predict a positive airway culture for SA, offering a promising non-invasive alternative to airway cultures despite its lower specificity.