Using Exhaled Breath Analysis as a Biomonitoring Tool – Determining Smoking Status Using Breath Biomarkers
Using Exhaled Breath Analysis as a Biomonitoring Tool – Determining Smoking Status Using Breath Biomarkers
S. Patassini, M. O’Neill, D. Layton, E. Krizek, G. De Palo, M. Williams, Y. Chen, T. W. Wilson, W. Murch and M. Allsworth
Owlstone Medical Ltd., Unit 183, Cambridge Science Park, Milton Road, Cambridge, CB4 0GJ
Full Poster:
Abstract:
Background:
This study aimed to discover and validate breath-based biomarkers for the discrimination of smokers and non-smokers, and to test the feasibility of breath analysis as a tool for biomonitoring and exposure research.
Methods:
Exhaled breath samples from 73 subjects were captured and analysed using the Breath Biopsy Platform. Each breath sample was collected over a 10-minute period using the ReCIVA Breath Sampler. Volatile organic compounds (VOCs) from breath were pre-concentrated onto sorbent tubes. ReCIVA uses pumps, pressure sensors and software to selectively sample breath fraction and volume. A CASPER Portable Air Supply minimized contamination from environmental sources. Samples were shipped to Owlstone Medical and analysed using TD-GC-TOF-MS.
Results:
475 molecular features (MFs) with distinct mass spectrums and retention times were identified across multiple breath samples. Analysis revealed 26 MFs that were significantly different between the groups. The statistically significant features were analysed by quantifying how well the two classes separated from each other in the dimension of the feature across all samples – yielding ROC-AUCs ranging between 0.72 and 0.96. Combinations of MFs, analysed using linear discriminant analysis and random forest, discriminated between groups with an accuracy of 96% and 97% respectively, using 10-fold cross-validation. Tentative identification of MFs using the NIST library indicated that many are combustion related compounds including benzene, toluene and ethylbenzene.
Verification of the validity of our observations was performed using breath samples collected from 136 individuals. A subset of the tentatively identified MFs were quantified at the parts per billion level using pure synthetic standards as VOCs surrogates. This quantitative analysis confirms BTEX compounds (benzene, toluene, ethylbenzene, xylenes) measured using Breath Biopsy can discriminate smokers from non-smokers.
Short discussion/conclusions:
This study shows that VOCs in breath can be used to discriminate between smokers and non-smokers and supports breath analysis as a novel technique for exposure related biomarker discovery and biomonitoring research.
Comments
Please sign in to leave a comment.