Rapid and Accurate Diagnosis of Bacterial Infections via Profiling of Metabolic Volatile Organic Compounds Using Proton Transfer Reaction Mass Spectrometry

Bacterial infections have become the second leading cause of death worldwide, and antimicrobial resistance (AMR) is a major factor contributing to treatment failure. Rapid and accurate bacterial identification (ID) and antibiotic susceptibility testing (AST) are critical for improving patient outcomes and controlling the spread of resistant strains. In this study, we developed a novel, rapid, and accurate diagnostic method for bacterial infections based on metabolic volatile organic compounds (VOCs). First, we performed untargeted profiling of metabolic VOCs from 9 clinically prevalent pathogens (Escherichia coli, Klebsiella pneumoniae, Enterobacter cloacae, Pseudomonas aeruginosa, Acinetobacter baumannii, Staphylococcus aureus, Staphylococcus epidermidis, Enterococcus faecalis, and Enterococcus faecium) using proton transfer reaction mass spectrometry (PTR-MS). Notably, all 10 species (including the culture medium control) could be successfully distinguished using only four VOCs: isovaleraldehyde, methanethiol, acetaldehyde, and dimethyl disulfide. The entire ID workflow only takes 3 h, achieving 98.0% accuracy. And then, we established a phenotypic AST method leveraging bacterial activity-related characteristic VOCs (ARC-VOCs). Testing the susceptibility of these 9 pathogens to different antibiotics achieved 99.1% accuracy (455 bacterial-antibiotic combination), with a turnaround time of 3–4 h. In summary, this metabolic VOC-based method enables bacterial infection diagnosis (ID + AST) within 6–7 h. It cuts the diagnostic time of the widely used clinical platform VITEK 2 by more than 50%, while maintaining high accuracy. This approach holds promise for guiding clinicians to develop timely personalized treatment regimens and curbing the transmission of drug-resistant bacterial strains.