Improving the interpretation of exhaled breath volatile data for clinical applications through mathematical and statistical methods
Chris A. Mayhew
Institute for Breath Research, Faculty for Chemistry and Pharmacy, University of Innsbruck,
Innrain 80-82, A-6020 Innsbruck, Austria
E-mail: christopher.mayhew@uibk.ac.at
Nearly 1500 trace volatile organic compounds have been found in exhaled breath [1], with detected volume mixing ratios ranging from sub-parts per trillion by volume up to hundreds of parts per million by volume. These volatiles may arise endogenously – reflecting normal or abnormal metabolic processes and pathological states – or exogenously, through ingestion, inhalation, or dermal absorption of chemicals. In many cases, both endogenous and exogenous sources contribute simultaneously. This talk will present the advantages and limitations of exhaled breath volatiles for applications in the health sciences, with a focus on the potential development of diagnostically useful clinical breath test devices. After decades of research, no spectroscopic or spectrometric investigation has identified with any level of confidence any endogenous volatile organic compounds that provide biomarkers or fingerprints to diagnose a disease [2, 3]. This mainly results from a lack of standardization and from a number of confounding factors that are often overlooked in the analysis of endogenous volatiles contained in exhaled breath, which if ignored will significantly limit the interpretation of exhaled breath volatile data and thus prevent meaningful outcomes. Crucial confounding factors that are often overlooked are those that influence the alveolar volatile concentrations according to the Farhi equation, namely cardiac output, alveolar ventilation, blood:air partition coefficients and mixed-venous blood concentrations. Another potential confounding factor is associated with the contributions of volatiles produced in the oral cavity through microbial activity [4]. Microbially produced oral volatiles could be misattributed to systemic processes, particularly in untargeted discovery studies where the metabolic origin of detected compounds is not rigorously established. These concerns are becoming more and more relevant in the context of the increasingly frequent machine learning-driven breathomics [5], for which algorithms could blindly take the volatile fingerprints in untargeted discovery studies and hence include oral microbial volatiles as discriminatory features without understanding their non-diagnostic origin. The concentration of an exhaled endogenous breath volatile will in addition be affected if that volatile is also present in the ambient air [6]. This talk will highlight the importance of combining mathematical modeling and statistical analysis with experimental results to enable a far more informed interpretation of the exhaled breath volatile data. Together, they directly address the primary crucial aspects of breath research so that exhaled volatiles can begin to fulfill their promise as transformative, non-invasive diagnostic biomarkers. Finally, we will discuss how exogenous, rather than endogenous, breath volatiles have possibly a greater potential of being exploited as useful metabolic probes to the human body, capable of providing unique biomarkers for use in diagnostically useful tests that could be clinically acceptable.
References
[1] Drabinska, C. Flynn, N. Ratcliffe et al, A literature survey of all volatiles from healthy human breath and bodily fluids: the human volatilome, J. Breath Res. 15 (2021) 034001 doi: 10.1088/1752-7163/abf1d
[2] Ruzsanyi, F. Lochmann, S. Jürschik et al, The Origin and Emission of Volatile Biomarkers, Chapter 1 of Volatile Biomarkers for Human Health: From Nature to Artificial Senses, First Edition, Editor Hossam Haick (RCS, 2022) ISBN 978-1-83916-430-9
[3] S. Modak, Why have only a handful of breath tests made the transition from R&D to clinical practice?, J. Breath Res. 18 (2024) 012001 doi: 10.1088/1752-7163/acff7d
[4] S. Petralia et al, The oral microbiome and its effect on exhaled breath volatile analysis – the elephant in the room, J. Breath Res. 19 (2025) 046004 doi: 10.1088/1752-7163/adf505
[5] Smolinska et al, Current breathomics – a review on data pre-processing techniques and machine learning in metabolomics breath analysis, J. Breath Res. 8 (2014) 027105 doi: 10.1088/1752-7155/8/2/027105
[6] Beauchamp, Inhaled today, not gone tomorrow: pharmacokinetics and environmental exposure of volatiles in exhaled breath, J Breath Res. 5 (2011) 037103 doi: 10.1088/1752-7155/5/3/037103.
