Breath Research Narrow Validation for Gastrointestinal Cancer Detection

In the United Kingdom, approximately 75,000 patients are diagnosed with, and 45,000 patients die from, the five major gastrointestinal cancers (oesophageal, gastric, pancreatic, liver and colorectal) each year. Early-stage disease is commonly associated with non-specific symptoms that mimic benign conditions.

Diagnosing cancer at an advanced stage limits curative treatment options. The overall 5-year survival for oesophageal, gastric, pancreatic, liver and colorectal cancer remains poor (15.9%, 20.7%, 9.0%, 12.6% and 50.7% respectively) and is strongly dependent on the stage at diagnosis. Currently, only 14.1%, 19.1%, 14.8%, 13.6% and 37.3% of oesophageal, gastric, pancreatic, liver and colorectal cancers are diagnosed at an early stage.

Earlier detection improves survival and quality of life by increasing access to curative treatment, enabling minimally invasive surgery, shortening hospital stays and facilitating an earlier return to normal activities. Avoiding unnecessary invasive investigations also reduces patient anxiety and the risk of procedure-related complications. Streamlining the diagnostic pathway has the potential to reduce unnecessary investigations and waiting times for endoscopy and cross-sectional imaging.

To address the challenge of earlier detection, the investigators of this study have developed a series of non-invasive breath tests that use volatile organic compound (VOC) analysis to identify symptomatic patients at risk of gastrointestinal cancers. In the future, a GP seeing a patient with symptoms suggestive of gastrointestinal cancer could request a breath test to help guide onward referral for endoscopy or imaging. Such tests could benefit patients with non-specific symptoms who do not currently meet existing referral criteria.

As part of this research programme, the investigators have conducted several clinical studies to identify VOC biomarkers associated with gastrointestinal cancers and develop VOC-based clinical prediction models (CPM) for the following cancers:

• Oesophageal and gastric adenocarcinoma (AROMA1)
• Oesophageal squamous cell carcinoma (ViSON)
• Pancreatic cancer (VAPOR1)
• Liver cancer (VOCAL1)
• Colorectal cancer (COBRA1)

The aim of the BRAVE study is to externally validate these VOC-based clinical prediction models in an enriched population, providing confidence in their diagnostic performance before progressing to the next phase of blinded validation studies.

Inclusion criteria

Adult participants ≥ 18 years old who meet at least one of the following criteria

Colorectal arm:

• Cancer: Histologically-confirmed CRC*
• Control: Non-specific abdominal symptoms with normal or benign colonoscopy findings

Pancreatic arm:

• Cancer: Histologically-confirmed PDAC*
• Control: Non-specific abdominal symptoms with a radiologically-normal pancreas

Oesophagogastric arm:

• Cancer: Histologically-confirmed OGC*
• Control: Non-specific abdominal symptoms with normal or benign upper gastrointestinal endoscopy findings

Oesophageal squamous arm:

• Cancer: Histologically-confirmed OSCC*
• Control: Non-specific abdominal symptoms with normal or benign upper gastrointestinal endoscopy findings

Liver arm:

• Cancer: Histologically- or radiologically-confirmed HCC or CCC*
• Control: Non-specific abdominal symptoms with a radiologically-normal liver

• Patients with suspected cancer (e.g. based on imaging) but who do not have histological confirmation prior to participating in the study may still be recruited and followed up to determine whether or not a diagnosis of cancer was subsequently confirmed.

Exclusion criteria

• Patients who have already received chemotherapy, radiotherapy or surgery for their cancer
• History of another cancer (other than non-melanoma skin cancers) within three years
• Participants with co-morbidities preventing breath collection
• Unable or unwilling to provide informed consent
• (For Oesophagogastric arm only): Allergies to any of the constituents of the nutrient drink including glucose, glycerol, iron sulphate, Maltodextrin (Corn, Potato), Xanthan Gum, Potassium Chloride, tyrosine, phenylalanine, and glutamic acid
• (For Colorectal arm only): Participants receiving bowel prep in the previous 7 days