Breath analysis as a method for breast cancer early detection

Abstract

1522 Background: Breath markers from exhaled breath specimens have recently been utilized to detect lung cancer. Phillips et al has applied this technology to breast cancer (BC), identifying methylated-alkane patterns to predict the presence or absence of BC in 3 groups of women: those with and without BC on breast biopsy and age-matched healthy controls. The purpose of our study was to build on this work by collecting breath samples with a simple portable device, unique from prior work, to determine whether BC patients have breath volatile organic compound (VOC) patterns distinct from those in healthy volunteers. Methods: A prospective IRB-approved pilot study enrolled 19 healthy volunteers >40 years of age, BC-free within 5 years and a negative mammogram within one year of breath collection and 8 newly diagnosed stage II-IV BC patients. After fasting for 2 hours, the study subjects deeply exhaled into a Markes Bio VOC Breath Sampler 5 times at 5-minute intervals trapping alveolar breath VOCs. Potential interferents were identified through a subject questionnaire completed during the breath collection session. The exhaled breath samples were analyzed by thermal desorption/gas chromatography/mass spectrometry. Results: In conformance with other researchers, our results indicate that individual compound identification is not appropriate for determining the presence or absence of disease, but aggregate low-dimensional summaries and compound quantities result in specific (unique?) patterns that can confirm disease. Therefore, principal components analysis (PCA) was used to construct informative linear combinations of individual VOCs with high discriminatory power. A definite clustering of the presence of BC from cancer-free points was found; e.g. for C33, 18 of 24 controls clustered below 0 while all 8 cancer patients clustered above 0.1 (overall range -0.4 - 0.3). Conclusions: Our data show promising evidence that BC patients can be differentiated from healthy volunteers through distinct breath VOCs. We continue to accrue to this pilot study with the intent of developing portable collection and sensing devices that could have implications for women without access to standard screening; interval screening for patients at high risk for BC; and disease monitoring for BC patients after completion of treatment. No significant financial relationships to disclose.