Analysis of Volatile Organic Compounds in Expired Air in Healthy Volunteers: Comparison of Three Mass Spectrometry Techniques for the Characterization of Volatolome in Clinical Studies

Precision medicine, or personalized medicine or individualized medicine, represents an important source of hope for alleviating the social and economic burden of severe pathologies. There is no commonly accepted definition of the notion of "personalized medicine". However, according to the European Council, personalized medicine is a medical model that relies on the characterization of people's phenotypes and genotypes (e.g. through molecular profiling, medical imaging, lifestyle information) to offer the right therapeutic strategy to the right person at the right time and/or to establish the existence of a predisposition to a disease and/or to ensure targeted and timely prevention. Personalized medicine is linked to the broader notion of "patient-centred care", which takes into account the general need for health systems to better meet the needs of patients. Advances in research have made it possible to develop "omic" signatures of therapeutic responses to certain cancers or rare diseases. However, if a few omic or cellular biomarkers have been developed, they remain today insufficient and too complex to foresee their generalization for the individualized treatment of patients. Thus, today there is significant potential for the development of precision medicine for severe pathologies. A major obstacle in precision medicine is the unavailability of biomarkers that are easy to access, non-invasive, measurable with high-performance techniques, fast, easy to use, reproducible, inexpensive and easily deployable on a large scale.

The analysis of exhaled air (volatolomics) is derived from the latest "omics" technology, metabolomics, devoted to the analysis of small molecules in the body, and allows real-time detection, sick bed, volatile organic compounds (VOCs) eliminated through the lungs. Thousands of VOCs have been identified in exhaled air following infectious, inflammatory or pathological events with examples in the field of infectious diseases for the diagnosis of tuberculosis, invasive fungal infections, bacterial colonization of the airways or ventilator-associated pneumonia in intensive care patients . For viral infections, animal models of influenza infections and clinical studies in patients with chronic obstructive pulmonary disease also suggest the benefit of VOC analysis . In this infectious context, the "respiratory fingerprint" detected reflects a mixture of metabolites of microbial origin, direct biomarkers of the presence of pathogenic agents, and metabolites generated by the host in response to the infection. Thus, the analysis of exhaled air, the main advantages of which are totally non-invasive sampling and the instantaneous analysis allowed by certain technologies (result in a few minutes) could be used for diagnosis, large-scale screening, surveillance of infections and prediction of response to treatment. The technological challenges for its realization are linked to the great chemical diversity of the VOCs to be studied and the particularly low abundance of many of them.

The reference technology for the analysis of VOCs is mass spectrometry (MS), which uses instruments consisting of - an ionization source whose function is to ionize the VOCs contained in exhaled air, - an analyzer which sorts the ions formed according to their mass to charge ratio (m/z) and - a detector which allows the quantitative analysis of the m/z signals of a sample . Several types of mass spectrometers can be used, and, in the absence of a consensual and standardized method, have practical methods for carrying out different analyzes which also lead to the generation of signals specific to each of them, the nature, the complexity and the completeness of the information contained being heterogeneous. Each type of instrument has advantages and disadvantages in terms of ease of sampling, speed of analysis, completeness of information and technical and analytical constraints for carrying out the analyses. For example, some instruments such as proton transfer reaction mass spectrometers (PTR-MS) or those using soft ionization by chemical transfer reaction (SICRIT) are relevant for perform measurements online and in real time, without storing a sample of exhaled air but vary by their mode of ionization and the resolution of the associated detectors. However, their level of information generated does not generally allow VOCs to be identified (soft ionization, absence of chromatographic separation, etc.). Indeed, once a signature of VOCs (characterized by their m/z) is discovered, their formal chemical identification is then the critical and obligatory step to improve knowledge on the physiology and regulatory processes of VOCs as well as to set up and clinically validate specific quantitative measurement methods based on portable technologies (sensors, etc.). Two-dimensional gas chromatography coupled with mass spectrometry (GCxGC-MS) is the most advanced technology at present for this purpose, combining two chromatographic columns with complementary stationary phases and ionization of VOCs by electron impact before MS detection. Compounds coeluting in conventional gas chromatography can be separated by GCxGC, and several teams have published proof-of-concept studies using GCxGC-MS for breath biomarker discovery for lung cancer diagnosis , tuberculosis or severe asthma phenotyping . One of the main challenges consists in establishing the concordance of the signals generated by the different technological approaches, some employing a preliminary chromatographic separation, others not, and some employing soft ionization methods whereas those of the others are on the contrary hard. Thus, the availability of datasets obtained on the same population with these complementary technological approaches will allow significant progress for the identification of the COVs of interest in clinical studies, beyond the simple comparison of the analytical performances of the different methods.