Merging the multi-measurement approach to Breathomics

Volatile organic compounds (VOCs) found in exhaled breath (EB) have the potential to transform the diagnosis of a range of diseases, notably cardiovascular diseases such as asthma, lung cancer, pneumonia and heart failure by providing a rapid, non-invasive and repeatable sampling methodology. The use of near patient techniques which could provide data in near real time, such as proton transfer reaction – time of flight – mass spectrometry (PTR-ToF-MS offers the ability to gain an insight into human metabolism. Challenges such as a lack of standardised data processing methodologies and a lack of clinically relevant studies have restricted the uptake of EB analysis in routine clinical care.

This thesis describes a novel methodology for handling real time mass spectrometry data, leveraging the information gained from the direct sampling methods possible mass spectrometry techniques such as PTR-ToF-MS. The method described demonstrates a substantial increase in the ability to accurately predict groups of individuals in a statistical model of participants based upon their smoking history. This thesis also describes the integration of EB analysis using PTR-ToF-MS into a large clinical study with the recruitment of over 300 participants, which is uncommon in the breath analysis literature. The use of a large-scale study revealed challenges in multiclass, such as the shortfalls in using identifying individual VOCs which have sufficient discriminatory power to be of clinical interest. Finally, this thesis describes the use of EB analysis in cutting edge clinical research, with two major clinical trials investigating improvements in patient care. The outcomes from the two clinical trials demonstrate that overall breath patterns can discriminate between groups of individuals with clinical relevance but falls short of accurately identifying chemical features which significantly discriminate between groups. Despite challenges noted throughout this these, the use of PTR-ToF-MS as a method for the analysis of EB still holds potential for future research.