Breath as a Diagnostic Fluid: Is It Finally Ready for Prime Time?

Jane E. Hill (1,2)
(1) University of British Columbia (2) Shirley Diagnostics, Inc

Jane Hill, PhD (Presenter) University of British Columbia

Presenter Bio: Jane Hill, PhD, is an Associate Professor at the University of British Columbia (UBC) and a leading researcher in breath science and technology, specifically using mass spectrometry (MS) to detect respiratory infections. She specializes in metabolomics and volatolomics, utilizing secondary electrospray ionization-mass spectrometry (SESI-MS) and GC-MS to identify bacterial lung infections and develop noninvasive diagnostics.

Relevant Financial Disclosures (within past 24 months, reported on Apr 20, 2026) Other Potential Conflicts Shirley Diagnostics, Inc. Discussing Products? NO, presenter will NOT present educational content related to products/services of the ineligible company(ies) listed above. Conflict mitigation NOT REQUIRED.

Dogs can smell cancer. A Scottish nurse can smell the musky odor of Parkinson’s. The sweet, fruity smell of diabetes can be detected in the breath of people with untreated diabetes. Using the breath to detect a disease has long been a blue-sky goal. The utilization of an abundant, information rich, and non-invasive sample seems ripe for clinical transformations, and yet, we do not see proliferation of these tests in the medical system. Why?

In this talk, I will review the state of the breath science field and highlight the advances, challenges, and remaining roadblocks for this promising technology. The presentation will highlight clinical research and commercial activity in the breath field, including applications in infectious disease screening, metabolic disease tracking, and cancer detection as well as some of the enabling technologies needed for clinical practice to become a reality.

I will share two stories that screen for two diseases, tuberculosis (the largest infectious disease killer globally) and MRSA detection. TB screening results are based on a study involving the recruitment of 1,000 subjects across four countries and with a goal to reach the WHO’s target product profile of a sensitivity of 90% and a specificity of 80% with fewer than ten breath molecules. After reaching this ambitious goal, there are still some critical innovations needed to translate these biomarkers into an FDA-approved device, which will be discussed. In addition, I will share about the use of volatile molecules to track drug resistance phenotypes in Staphylococcus aureus. The MRSA project will highlight the process by which many breath researchers go about entering a disease sphere with this modality. In this case, a progression from bacterial culture, to mouse work, through to human lung specimen testing (including breath), will be shared as a case study of the accumulation of proof-of-concept typically conducted for a clinical application.

The detection of heart failure using the acetone in breath is another case study that will be highlighted. Breath acetone is a major exhaled metabolite whose levels change with metabolic substrate use, for example during fasting or physical activity. In heart failure, it is hypothesized that increasing breath acetone signals progression toward exacerbation, supported by our findings of elevated levels at hospital admission in decompensated patients, normalization at discharge, and increased levels during cardiopulmonary exercise testing, suggesting a metabolic shift from fatty acid to ketone body usage. These insights can aid in understanding heart failure physiology as well as provide a potential tool to monitor physical and pharmacological therapies during recovery.

Metabolic changes are omnipresent in many diseases, including insulin resistance. I will share the development of a non-invasive diagnostic tool for insulin resistance detection via ten breath biomarkers in a cohort adolescents. The breath model shows an accuracy of 77.8%, with a sensitivity of 73.1% (60–83% within cross-validation) and a specificity of 81% (70–89% within cross-validation). The resulting model showed a high correlation (R = 0.95, p < 0.001) with current gold-standard blood measures, suggesting that breath analysis could replace invasive screening tools for detecting early-stage metabolic disease.

Tracking drugs using breath aka, a subgenre pharmacometabolomics, is a noninvasive way to monitor drug exposure and treatment-induced metabolic responses. I will share about the profiling of antiseizure therapy, salbutamol, propofol, and lidocaine, supporting therapeutic drug monitoring and clinically relevant phenotyping in a combined human evidence-base of n = 163 across pediatric and adult cohorts in a European clinical and regulatory environment.

Detecting and tracking cancer progression is the final area to be covered in this presentation. The focus here is on using breath for the triage of patients with non-specific symptoms to specialized investigations for gastrointestinal cancers in adult patients. In studies of over 1,000 patients for esophageal, gastric, colorectal, pancreatic, and liver cancer, biomarkers for each have been proposed and are currently undergoing rigorous clinical and chemical validation in the United Kingdom.

Taken together, the impact of breath science in the clinical world, could be, literally, breath-taking with respect to providing screens that would improve patient morbidity and mortality outcomes as well as the economic bottom line. In the context of these maturing scientific stories, I will also share common pitfalls in the field as well as highlight some of the key steps needed to progress breath science into the clinical sphere.