Developing breath biomarkers of S. aureus cystic fibrosis lung infections through the secondary analysis of IMPACT-Breath clinical study data

About This Grant

Staphylococcus aureus is the most common cause of lung infections in the US cystic fibrosis (CF) population, with nearly half (48.3%) of persons with CF testing positive for S. aureus in 2023. Testing for CF lung infections heavily relies upon the collection of sputum, but as CF care improves through the use of highly effective cystic fibrosis transmembrane conductance regulator therapy, the production of sputum has declined, though the risk of lung infections persists. Therefore, novel methods for diagnosing lung infections without sputum are urgently needed. Our goal is to develop highly sensitive and specific, non-invasive breath tests for the detection of S. aureus and other lung infections. The overall objective of this proposal is to develop biomarkers for S. aureus lung infections through the secondary analysis of breath volatilomics data and clinical metadata collected for the IMproving P. Aeruginosa deteCTion using Breath (IMPACT-Breath) clinical study. We hypothesize that breath biomarkers can detect S. aureus lung infections and discriminate them from other mono-infections and co-infections with > 80% accuracy. We will test this hypothesis via three Specific Aims: Aim 1: Identify biomarkers that differentiate breath samples from pwCF who are Staphylococcus aureus positive vs. negative. Using machine learning algorithms, we will build classification models that discriminate breath samples from subjects with new or chronic S. aureus infections (Sa+) vs. S. aureus negative (Sa-). Aim 2: Identify breath biomarkers that differentiate S. aureus vs. P. aeruginosa lung infections. P. aeruginosa is the second most common lung pathogen in the US CF population and in the IMPACT-Breath study cohort. Using the P. aeruginosa infection status from the IMPACT-Breath study, we will build classification models that discriminate breath samples from subjects infected with S. aureus vs. P. aeruginosa (Sa+ vs. Pa+). Aim 3: Identify breath biomarkers that differentiate S. aureus mono-infections from S. aureus-P. aeruginosa co-infections. Using IMPACT-Breath data, we will build classification models that discriminate breath samples from subjects with S. aureus mono-infections vs. S. aureus–P. aeruginosa co-infections (Sa+ vs. Sa+Pa+). The expected outcomes of this project are putative breath biomarkers for detecting S. aureus lung infections and estimates of their sensitivity and specificity, which will provide us with the necessary preliminary data to design a prospective clinical study to validate S. aureus biomarkers in the breath of pwCF.