Cross Platform Validation of Mass Spectrometry for the Diagnosis of Colorectal Cancer Using a Breath Test Georgia Woodfield (Presenter) Imperial College London Presenter Bio: I am a gastroenterologist working in North West London, part way through my registrar training. I am currently taking 3 years out of my medical training to do a PhD in "Breath Testing for Colorectal Cancer" with Professor Hanna And Professor Atkin at Imperial College London. This mainly involves running two large studies based in London looking at diagnostic accuracy and feasibility of breath testing for colorectal cancer. I use mass spectrometry (GC-MS and PTR-MS) to analyse volatile organic compounds present in breath. I am particularly interested in early cancer diagnosis and endoscopic screening. Given that I am new to the mass spectrometry field, I am excited to meet colleagues working in this area and find out about other ongoing and related projects.

Authors: Georgia Woodfield (1,2), Ilaria Belluomo (1), Gengping Lin (1), Piers Boshier (1), Andrea Romano (1), John Martin (2), Chris Groves (3), Brian Saunders (1,4), Wendy Atkin (1), George B Hanna (1,2)
1) Imperial College London 2) Imperial College Healthcare NHS Trust 3) St Georges Healthcare NHS Trust 4) St Mark’s Hospital, London

A breath test for colorectal cancer (CRC) could be a valuable non-invasive diagnostic tool. The Colorectal BReath Analysis (COBRA) prospective study aims to determine the accuracy of breath volatile organic compounds (VOCs) for detecting CRC, building on previous literature. This crucially depends on accurate breath profiling. We performed cross platform validation of breath VOCs from the first 406 patients using gas chromatography mass spectrometry vs. proton transfer reaction mass spectrometry. This proved to be an accurate, reliable method for identifying and quantifying VOCs, with good inter-instrument agreement. This widespread use of a breath test for CRC seems technically possible.

Introduction

Colorectal cancer (CRC) is the 2nd most common cause of cancer death in the UK, often due to late diagnosis as a result of non-specific presenting symptoms (1,2). A breath test could be a valuable non-invasive tool to aid early diagnosis, whilst potentially reducing the number of unnecessary colonoscopies for well patients. Prior studies have shown promising results using exhaled propanal for detecting CRC (96% sensitivity/76% specificity) (3). The Colorectal BReath Analysis (COBRA) study aims to determine the diagnostic accuracy of using breath volatile organic compounds (VOCs) for detecting CRC in a prospective study of 2000 patients. However, before implementing a new diagnostic test, a crucial requirement is accurate and reproducible quantification and identification of VOCs in human breath with fast and efficient analysis of samples.

Mass spectrometry, particularly gas chromatography mass spectrometry (GC-MS) is an increasingly recognised method of measuring breath VOCs, with detection of trace compounds in the order of parts per billion (ppb). Proton transfer reaction mass spectrometry (PTR-MS) is less widely used in the literature, however it lends itself to widespread clinical use with its ability to process 100 samples per day of continuous unsupervised analysis. In order to justify the use of PTR-MS for high throughput analysis for the COBRA study, we aimed to validate the results using GC-MS, and use the two different but complementary platforms to aid accurate compound identification.

Methods

Exhaled breath (500mls) was collected using the ReCIVA™ breath sampling device, onto thermal desorption tubes from patients over 4 London hospitals. All were nil by mouth. We used both PTR-MS and GC-MS, coupled to thermal desorption units, for breath analysis. This allowed cross platform validation of results for each patient.

Results

406 patients were recruited between July to December 2017 as part of the COBRA study (on-going). These patients had a wide range of colorectal pathology including CRC. Comparison between the concentrations of known abundant compounds showed excellent correlation between the two mass spectrometers. We also analysed potential cancer compounds with very good results, where the good correlation of PTR to GC data greatly aided the identification of unknown m/z ratios.

Conclusions & Discussion

Simultaneous analysis of breath by both GC-MS and PTR-MS is a good method for identifying and quantifying VOCs, with good agreement between instruments. This paves the way for widespread clinical use of a breath test for diagnosis of CRC.

1) De Angelis R, Sant M, Coleman MP, Francisci S, Baili P, Pierannunzio D, et al. Cancer survival in Europe 1999–2007 by country and age: results of EUROCARE-5—a population-based study. The Lancet Oncology.15(1):23-34.

2) England PH. National Cancer Intelligence Network: Routes to diagnosis 2015 [Available from: http://www.ncin.org.uk/publications/routes_to_diagnosis.

3) Markar S, Chin S, Romano A, Wiggins T, Antonowicz S, Paraskeva P, Ziprin P, Darzi, A, Hanna G. Annals of Surgery; 2017; volume published ahead of print.