A NOVEL breath test that identifies bacterial infections by detecting unique metabolic signatures may offer clinicians a rapid, non-invasive diagnostic and monitoring tool, according to findings presented at ESCMID Global 2025.
Researchers from the University of California, San Francisco (UCSF) and St. Jude Children’s Research Hospital evaluated the test using a laser-based platform known as integrated cavity output spectroscopy (ICOS) in preclinical models. The test involves administering 13C-labelled sugar compounds, which are selectively metabolised by bacteria but not mammalian cells, leading to the production of [13C]CO2 detectable in exhaled breath. The team assessed five bacterial pathogens, including Staphylococcus aureus and Escherichia coli, and administered 13C-maltose and 13C-mannitol to infected mice, using various infection models such as pneumonia, osteomyelitis, and myositis.
Results demonstrated that [13C]CO2 was reliably detected only in infected animals, with healthy mice producing no signal following administration of the tested tracers. 13C-maltose and 13C-mannitol showed high specificity for bacterial metabolism, unlike 13C-glucose or 13C-sorbitol, which were also metabolised by mammalian cells. In a further validation step, E. coli-infected mice treated with ceftriaxone exhibited a marked decline in [13C]CO2 levels after 24 hours, mirroring the drop in bacterial load.
These findings support the test’s potential not only to diagnose infections but also to monitor treatment response in real time. Although this study did not quantify the test’s sensitivity, previous research suggests that the ICOS platform offers practical advantages over traditional isotope ratio mass spectrometry, including portability and reduced cost (estimated at under 100,000 USD per unit).
The authors emphasised that confirming the test’s specificity in healthy humans will be essential before clinical use, but they are optimistic about its broad future applications. If validated, this approach could significantly improve point-of-care diagnostics in emergency and intensive care settings.
Reference
Lopez-Alvarez M et al. [13C]CO2 breath testing for detecting and monitoring bacterial infection. ESCMID Global 2025, 11-15 April, 2025.
