Breath Volatile Flux Reveals Age-Dependent Metabolic Markers for Breast Cancer Detection

Breast cancer remains a dominant health risk for women globally with early detection a primary indicator for successful treatment and survival. Current approaches for diagnosis are invasive, costly, and accuracy can be improved. Breath testing using volatile organic compounds (VOCs) as biomarkers presents an exciting avenue for non-invasive cancer diagnosis. However, breath sampling for cancer diagnosis has not yet delivered reliable biomarkers or technology. This is, in-part due to methodology, which introduces confounding factors when attempting to scale to diverse patient populations. We utilize a novel approach to human clinical breath sampling, in which we quantify breath volatile flux. Breath volatile flux considers inhaled air alongside exhaled air to generate a dynamic breath profile which enhances our ability to identify compounds metabolized in humans. We present a novel breath collection platform into which breast cancer clinic patients, in a pilot study of 60 women, provided 5 mins duration breath samples for analysis by gas chromatography mass spectrometry (GC-MS). GC methods targeted a suit of 11 previously identified compounds alongside non-targeted scans. When examining the data across the entire cohort, butanone was the only significantly altered (increased) breath volatile and was able to separate benign tumour and cancer patients from normal patients. However, patient age was observed as a primary confounding factor and reduced the accuracy of butanones diagnostic potential. When age was controlled for, chloroform, styrene and isopropyl alcohol acted as indicators of breast cancer health status. Furthermore, once age was accounted for and cancer patients were identified, grade of cancer was indicated by chloroform and DMS fluxes. Using the top 5 discriminate compounds and receiver operator curves we were able to identify cancer from normal patients with area under the curve of 93.4%, grade 2/3 cancers from normal patients with 97.6% AUC, and benign from normal patients with 90.5% AUC. This study suggests that volatile flux measurements from breath allows successful identification, and separation of, cancer, benign tumour and healthy patients.