Background: Each year in the United Kingdom there are over 44,000 deaths related to gastrointestinal (gut) cancers that include oesophageal (food pipe), stomach, pancreas, liver and bowel cancers. These cancers are difficult to detect as the symptoms are similar to non-cancer conditions. The delay in suspecting cancer and referral to a camera test (endoscopy) or scans can lead to the growth of cancer. Even with urgent referral only 4 out of every 100 patients are found to have cancer. If diagnosed early, the majority of people can survive gastrointestinal cancer. However, as the disease progresses, the number of people who survive falls significantly. Early diagnosis is therefore key.

The proposed solution is a simple breath test for patients with relevant symptoms. Breath testing is quick, acceptable and can be offered directly by a GP. The breath samples are analysed for small molecules called volatile organic compounds. These molecules provide early indicators of gastrointestinal cancer, to identify those who would benefit most from investigations. Picking up gastrointestinal cancers earlier will increase the chances of curing the disease.

Aims: breath molecules that are specific to gastrointestinal cancers have been previously identified. The aim is to confirm these molecules and examine the accuracy of the breath test. How to implement the test in primary care will also be studied.

Design and Methods: Breath testing will be performed on at least 8,000 patients referred by GPs on suspected gastrointestinal cancer pathways. Patients will be followed up to determine whether they have gastrointestinal cancer or not.

The breath test will take place in national NHS hospitals before patients undergo camera tests or scans. The breath test involves them breathing out into a bag allowing the breath to be collected into special tubes. The breath samples in the tubes will be analysed in the local laboratory. Which molecules in patients' breath samples correctly identify cancer will be confirmed.

The key members will be engaged with to implement the breath test in primary care and NHS and assess the affordability of the breath test.

Patient and public involvement and engagement (PPIE): A diverse PPIE group of 6 members affected by gastrointestinal cancer have identified clear benefits to patients and GPs and provided ways to improve patient recruitment. Work was with charities and appointed a patient representative to the local committee to support dissemination.

The results will be disseminated through scientific papers and reports, conference presentations, public-facing areas, social media and cancer charity newsletters.

Background: In the UK, there are over 75,000 new cases annually of gastrointestinal cancers (oesophageal, gastric, pancreatic, liver and colorectal). The overall 5-year survival for these cancers is poor and even lower in deprived areas. Symptoms related to gastrointestinal cancers are non-specific, therefore the decision to refer for investigations is challenging. Currently, the cancer yield of urgent and non-urgent referral pathways is 4.4-5% and 0.1-1.7% respectively.

Breath analysis offers a simple, quick and non-invasive method to detect volatile organic compounds (VOCs) in breath that are specific to gastrointestinal cancers. The triage test would maximise patient compliance, particularly in certain ethnic groups and deprived areas.

The aim is to validate VOCs in the detection of multi-gastrointestinal cancers and determine how well does a single breath test perform in the detection of multi-gastrointestinal cancers. Also the optimal strategy to implement breath test in primary care and the cost-effectiveness of such a test will be assessed.

Methods:

Multicentre clinical study: At least 8,000 individuals with suspected cancer, referred from primary care along upper and lower gastrointestinal cancer pathways, will be recruited from 40 centres. The aim is to have 256 cancers and 6144 controls with complete dataset (reliable breath data and results of reference test). Breath samples will be collected on thermal desorption tubes which will be analysed using gas chromatography-mass spectrometry (GC-MS). The diagnostic accuracy of breath test to detect cancer and estimate the most likely cancer from upper and lower gastrointestinal referral pathways will be tested. In addition, the views of key stakeholders and co-design breath test delivery pathways in primary care will be captured. Guidance on best practice in breath test delivery for primary care will be developed. Also a cost-effectiveness and budget impact analysis to understand the affordability and financial impact on the NHS will be performed.

Therefore, in summary, a multi-outcome prediction model to identify gastrointestinal cancer with the highest probability, along with a comprehensive plan to integrate a cost-effective breath test within the NHS will be developed.

Anticipated impact and dissemination: The non-invasive breath test will provide direct patient benefit through earlier and accurate detection of gastrointestinal cancers, along with higher patient acceptability. Ensuring timely referral will translate to curative treatment and improved long-term survival. Outcomes will be disseminated via conference presentations, publications, the local public advisory group (PAG), charities, the local research website and social media.

Inclusion Criteria:

• Aged ≥ 18 years old.
• Participants with non-specific symptoms referred from primary care along upper and lower gastrointestinal cancer pathways for the reference test (upper/lower gastrointestinal endoscopy, CT/MRI/ultrasound ± histological/cytological confirmation).
• Willing and able to provide informed written consent to take part in study.

Exclusion Criteria:

• Previous oesophageal, gastric, liver, pancreatic or colonic resection
• History of another cancer within three years
• Co-morbidities preventing breath collection
• Pregnant participants
• Unable or unwilling to provide informed written consent.

Patients seen in a hospital (secondary or tertiary level care) setting