Early oesophageal cancer typically has non-specific symptoms that are commonly ascribed to benign conditions. Current guidelines recommend referral for endoscopy, only in the setting of ‘red flag’ symptoms that are frequently associated with inoperable disease. Consequently, 7 out of 10 new cases of oesophageal cancer are diagnosed at an advanced stage, meaning less than a third of patients are eligible for potentially curative therapy, and less than 1 in 5 patients survive beyond 5 years.

Previous studies have established the presence of an altered profile of volatile organic compounds (VOCs) in oesophageal cancer. These studies have all assessed exhaled breath VOCs at a single point in time, using a cross-sectional study design. Longitudinal variation in exhaled breath VOCs and urinary VOCs, as well as the contribution of the upper gastrointestinal microbiome and the effect of therapeutic intervention in oesophageal cancer patients remain unknown. Understanding this variation is critically important for interpreting intra-subject variability, and the likely mechanisms of VOC production and release. Furthermore, the expected decline in target VOCs following chemoradiotherapy and/or tumour resection would both verify the link between the tumour and altered VOC production, as well as offering an opportunity to utilise VOCs in a longitudinal setting to assess treatment response and disease recurrence.

Participating centres:

Aims:

1. Determine longitudinal variations in exhaled breath VOC concentrations during intended curative therapy for oesophageal cancer.

2. Validate the previously published model for exhaled VOC breath test of oesophageal cancer.

3. Correlate the exhaled VOCs of oesophageal cancer patients to important clinical parameters.

4. Determine the relationship between exhaled VOCs and response to neoadjuvant chemoradiotherapy.

5. Investigate exhaled VOCs following tumour resection, and their association with disease recurrence.

6. Correlate the exhaled and urinary VOC concentrations in oesophageal cancer patients.

7. Investigate the longitudinal variations in the upper gastrointestinal microbiome and its relationship to exhaled and urinary VOC concentrations.

Methods:

SEATTLE is a longitudinal, single-centre cohort study with a sample size of 50 patients with oesophageal cancer recruited at a single, high-volume centre (Virginia Mason Medical Center, Seattle). The study involves four time points for breath and biofluid (urine and saliva) collection:

1. Before commencing treatment for oesophageal cancer.

2. Following conclusion of neoadjuvant therapy (at the time of a routine outpatient appointment or on the day of surgery).

3. Following surgical resection (on the day of hospital discharge).

4. At the time of routine postoperative follow-up at Virginia Mason Medical Center (between six months to one-year following surgery).

Breath was collected using standardised procedures and transported using thermal desorption (TD) tubes to the Imperial College London VOC Laboratory. Biofluids (urine and saliva) were collected and transported in cryovials on dry ice at -80^oC. The VOCs from the breath, urine and saliva samples are being analysed using mid-polar, polar, and two-dimensional gas chromatography-time of flight mass-chromatography (GC-MS-TOF).

Bioinformatics analysis will link identified microbiota and their functional pathways with their identified end metabolites and long-term GI symptoms. This will be performed using in-house bioinformatic pipelines that exist for integrative analysis of multi-omics data, existing open access resources (e.g. MetaboAnalyst), or customised R scripts (e.g. MixOmics).