# Rapid, Breath Volatile Metabolite-Based Diagnostic for In Vivo Identification and Antibiotic Resistance Profiling of Bacterial Pathogens in Ventilator-Associated Pneumonia

> **NIH NIH R01** · BRIGHAM AND WOMEN'S HOSPITAL · 2022 · $1,125,660

## Abstract

Project Summary/Abstract:
The lack of diagnostics that rapidly and accurately identify bacterial infections drives empiric antibiotic
prescribing in patients with pneumonia – ultimately, 37-50% of these antibiotics are unnecessary. These issues
are amplified in the intensive care unit (ICU), where antimicrobial resistance is common, the risk of imminent
clinical deterioration and death is high, and clinicians are under pressure to make rapid treatment decisions.
Ventilator-associated pneumonia (VAP) is the most common ICU hospital-acquired infection, responsible for
approximately half of all ICU antibiotic prescribing. Time to effective antibiotic treatment is a critical
determinant of outcome, but many patients with VAP receive inadequate empiric treatment due to the high
prevalence of resistant organisms in VAP. Clinical findings in VAP are highly nonspecific, and 30-60% of
antibiotics prescribed for suspected VAP are ultimately unnecessary. Despite a high pulmonary bacterial load
in patients with VAP, the lung has traditionally been a particularly inaccessible space without the use of
invasive diagnostic procedures. We have established proof of concept in murine VAP models that there are
bacterial species-specific breath volatile metabolite signatures in VAP caused by Staphylococcus aureus,
Pseudomonas aeruginosa, Escherichia coli and Klebsiella pneumoniae, and that microbial breath volatile
metabolites have markedly different responses to antibiotic exposure within a few hours in phenotypically
susceptible (S) vs. non-susceptible (NS) organisms. In close collaboration with industry partners and a team of
experts in antimicrobial resistance, microbiology, VAP, advanced statistical methods, and regulatory matters,
we propose further development of an advanced, miniaturized gas chromatography-differential mobility
spectrometry (GC-DMS) diagnostic platform for the rapid, noninvasive, breath-based diagnosis of VAP and its
most common causative pathogens, S. aureus, P. aeruginosa, K. pneumoniae, E. coli, Enterobacter cloacae,
and Acinetobacter baumannii, exploiting differential volatile metabolite responses to effective and ineffective
antibiotic therapy to obtain in vivo phenotypic information about antibiotic susceptibility. Using thermal
desorption-GC-tandem mass spectrometry and in parallel, a rapid GC-DMS diagnostic device, we will
systematically characterize these species-specific breath signatures and early responses to antibiotic therapy in
S vs. NS organisms in murine VAP models and in patients with suspected VAP, defining and validating breath
signatures that (a) identify VAP, distinguishing it from other ventilator-associated conditions and respiratory
tract colonization, (b) identify its underlying microbial etiology, and (c) determine whether the microbe is S or
NS by examining its early response to antibiotics, and create GC-DMS algorithms that identify these signatures
in breath data automatically, in preparation for a 510(k) clearan...

## Key facts

- **NIH application ID:** 10196932
- **Project number:** 5R01AI138999-04
- **Recipient organization:** BRIGHAM AND WOMEN'S HOSPITAL
- **Principal Investigator:** Sophia Koo
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2022
- **Award amount:** $1,125,660
- **Award type:** 5
- **Project period:** 2018-06-20 → 2024-05-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10196932

## Citation

> US National Institutes of Health, RePORTER application 10196932, Rapid, Breath Volatile Metabolite-Based Diagnostic for In Vivo Identification and Antibiotic Resistance Profiling of Bacterial Pathogens in Ventilator-Associated Pneumonia (5R01AI138999-04). Retrieved via AI Analytics 2026-05-25 from https://api.ai-analytics.org/grant/nih/10196932. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*