Targeted breath analysis: exogenous volatile organic compounds (EVOC) as metabolic pathway-specific probes
Edoardo Gaude,Morad K Nakhleh,Stefano Patassini,Jasper Boschmans,Max Allsworth,Billy Boyle,Marc P van der Schee
DOI: https://doi.org/10.1088/1752-7163/ab1789
2019-05-17
Journal of Breath Research
Abstract:Breath research has almost invariably focussed on the identification of endogenous volatile organic compounds (VOCs) as disease biomarkers. After five decades, a very limited number of breath tests measuring endogenous VOCs is applied to the clinic. In this perspective article, we explore some of the factors that may have contributed to the current lack of clinical applications of breath endogenous VOCs. We discuss potential pitfalls of experimental design, analytical challenges, as well as considerations regarding the biochemical pathways that may impinge on the application of endogenous VOCs as specific disease biomarkers. We point towards several lines of evidence showing that breath analysis based on administration of exogenous compounds has been a more successful strategy, with several tests currently applied to the clinic, compared to measurement of endogenous VOCs. Finally, we propose a novel approach, based on the use of exogenous VOC (EVOC) probes as potential strategy to measure the activity of metabolic enzymes <i>in vivo</i>, as well as the function of organs, through breath analysis. We present longitudinal data showing the potential of EVOC probe strategies in breath analysis. We also gathered important data showing that administration of EVOC probes induces significant changes compared to previous exposures to the same compounds. EVOC strategies could herald a new wave of substrate-based breath tests, potentially bridging the gap between research tools and clinical applications.Exhaled human breath contains hundreds of volatile organic compounds (VOCs) [<a href="#jbrab1789bib1">1</a>]—low molecular weight metabolites that are excreted in breath as the result of metabolic processes in the body—that can be of endogenous and/or exogenous origin [<a href="#jbrab1789bib2">2</a>]. Endogenous VOCs are produced from internally available metabolic substrates and are part of physiological and pathological mechanisms. Exogenous VOCs are introduced into the body from an external source such as diet, environmental exposure, medication, etc, and are expressed in exhaled breath after circulation and/or internal metabolism etc [<a href="#jbrab1789bib3">3</a>].Several endogenous and exogenous breath VOCs have been investigated, in the last decades, as potentially useful and non-invasive biomarkers for various diseases with applications ranging as wide as lung cancer, cardiovascular diseases, asthma, cystic fibrosis, infectious diseases, and chronic inflammatory diseases [<a href="#jbrab1789bib4">4</a>–<a href="#jbrab1789bib6">6</a>]. The general approach that has been applied to the discovery of mainly endogenous disease-related biomarkers entails the comparison of breath VOC profiles between groups of supposedly healthy subjects, and/or cohorts of patients with a documented disease. Application of statistical techniques helps with discovery of VOCs that are differentially expressed in control and diseased groups, leading to identification of candidate biomarker(s). Although this approach has been trialled and tested in several hundreds of experiments, none of these candidate endogenous or exogenous exhaled VOCs has paved its way into routine diagnostic use. In this article we will explore potential reasons for the current lack of VOC tests in the clinic, and present some data pointing towards a different experimental strategy which could complement current unsupervised discovery research.Historically, breath research has deployed the standard model of 'omics' research described in the previous paragraph: (small-size) cross-sectional case-control studies, screening as many potential biomarkers as possible, using complex mathematical algorithms to distil signal from noise, and to assess their statistical power when compared to control groups [<a href="#jbrab1789bib6">6</a>, <a href="#jbrab1789bib7">7</a>]. Upon observation of potentially discriminating repeatable spectral features, these non-targeted approaches rely on identification of compounds through comparison with database of chemical suspects, for which an exhaustive framework has been proposed [<a href="#jbrab1789bib8">8</a>]. Although many technical and methodological aspects are constantly being addressed and improved, two main challenges are however fundamentally linked to the scientific method applied in untargeted biomarker discovery studies and are shared with other 'omics' approaches such as transcriptomics and proteomics.Untargeted study designs typically have a very small ratio of subjects to tested variables. This creates a considerable challenge to find relevant biomarkers in a very complex highly dimensional dataset. Especially, if a biomarker is expected to only be relevant in a minority of patients there are two 'evils' which need to be carefully balanced; the risk of overfitting a model resulting in falsely identified 'voodoo' correlations and the risk of obscuring true biomarkers [<a<p>-Abstract Truncated-</p>
biochemical research methods,respiratory system