Oesophageal Cancer, Oesophageal Squamous Cell Carcinoma
Conditions
Keywords
Breath biomarker, Early detection, Oesophageal cancer, Oesophageal squamous cell cancer, Volatile organic compounds
Brief summary
Oesophageal Squamous Cell Carcinoma (OSCC) is a cancer of the food pipe that affects around 2000 patients in the UK every year. It is often detected at an advanced stage, resulting in poor survival (5-year survival less than 20%). Early detection can improve survival (5-year survival \>70%). Therefore, early detection is vital to improving survival. There are no national screening guidelines, and an endoscopy (A camera test to look at the food pipe) is the only available test to detect OSCC. Early detection of OSCC is challenging for many reasons. Firstly, early disease symptoms are non-specific, which patients often overlook. Secondly, 'Alarm' symptoms such as weight loss, difficulty swallowing or vomiting blood are signs of advanced stage. Lastly, endoscopy is an invasive test with associated risks and significant discomfort. The investigators propose to develop a breath test for patients with non-specific symptoms. Breath testing has the ideal characteristics for a triage test because it is non-invasive, simple to perform, cost-effective and highly acceptable to patients. The test is based on identifying volatile organic compounds (VOCs, small molecules) that are produced by the cancer and released in breath. The breath test will be offered by General Practitioners (GPs) to patients with non-specific symptoms. Those who test positive will be referred for an urgent camera test, and those who test negative can be reassured.
Detailed description
In this prospective multicentre case-control study, the investigators will recruit a total of 518 patients. These will be divided into the following groups: 1. Cancer group (n=259): Patients with treatment naive, histopathology confirmed OSCC. 2. Control group (n=259): Patients who have undergone or are undergoing an upper gastrointestinal (GI) endoscopy as part of their investigation for upper GI symptoms and are found to have either - (i) A normal upper GI tract or (ii) Benign upper GI disease. Eligible and willing participants will be asked to provide two breath samples by exhaling into single-use breath collection bags. Using a custom designed gas sampling pump, the breath VOCs will be transferred onto thermal desorption (TD) tubes at a controlled flow rate. When the participants' breath sampling is complete, room air (Blank) samples will be taken onto additional TD tubes using the same process. Once collected, the TD tubes will be transported to Imperial College London (The Hanna lab), where they will be analysed.
Interventions
Participants will maintain a clear fluid diet for a minimum of 6 hours prior to breath collection. Participants will be asked to provide a breath sample by exhaling into single-use breath collection bags. Using a custom designed gas sampling pump, the breath VOCs will be transferred onto TD tubes at a controlled flow rate. This breath sampling procedure will be repeated once. When the participants' breath sampling is complete, room air (Blank) samples must be taken onto additional TD tubes using the same procedure. The TD tubes will be sealed with long-term storage caps using the CapLok Tool.
Sponsors
Imperial College Healthcare NHS Trust
Hull University Teaching Hospitals NHS Trust
Norfolk and Norwich University Hospitals NHS Foundation Trust
University Hospitals Coventry and Warwickshire NHS Trust
Oxford University Hospitals NHS Trust
Velindre NHS Trust
Cardiff and Vale University Health Board
Newcastle-upon-Tyne Hospitals NHS Trust
Liverpool University Hospitals NHS Foundation Trust
Bedfordshire Hospitals NHS Foundation Trust
University Hospitals, Leicester
The Clatterbridge Cancer Centre NHS Foundation Trust
Portsmouth Hospitals NHS Trust
University Hospital Plymouth NHS Trust
NHS Tayside
NHS Highlands
The Christie NHS Foundation Trust
North Cumbria University Hospitals NHS Trust
Maidstone & Tunbridge Wells NHS Trust
University Hospitals of Derby and Burton NHS Foundation Trust
Bradford Teaching Hospitals NHS Foundation Trust
East and North Hertfordshire NHS Trust
Imperial College London
Study design
Observational model
CASE_CONTROL
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
18 Years to 90 Years
Healthy volunteers
No
Inclusion criteria
Participants with all the following characteristics will be eligible for inclusion in the study: 1. Cancer cohort (n=259): Patients with treatment naïve, histopathology confirmed OSCC. 2. Control cohort (n=259): Patients who have undergone or are undergoing an endoscopy (OGD) as part of their investigation for upper GI symptoms and are found to have either: * A normal upper gastrointestinal tract * Benign upper gastrointestinal disease
Exclusion criteria
Participants with the following characteristics will not be eligible for inclusion in the study: 1. Received some form of treatment (chemotherapy, radiotherapy, immunotherapy, endoscopic resection, or surgery) for OSCC 2. History of another cancer in the last five years 3. Non-squamous cell oesophageal cancer 4. Barrett's oesophagus (with or without dysplasia) 5. Previous oesophageal or gastric resection 6. Unable to provide written consent or lack capacity. 7. Pregnant women
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| To develop a breath test for detection of OSCC | 36 months | The investigators hypothesise that breath VOCs in OSCC are unique and different to those in non-cancer patients. Replicate breath samples will be analysed by two independent assays using different Thermal Desorption - Gas Chromatography - Time of Flight Mass Spectrometry (TD-GC-TOF-MS) instruments. One assay will use a mid-polar, and the other a polar column stationary phase for optimal determination of VOC chemical classes. The samples will then be recollected and analysed by two-dimensional (2D) TD-GC-TOF-MS for robust VOC identification. Confirming the identity of the top VOCs driving the model will provide chemical validation. Data obtained will be used to refine the model using molecular network analysis. Statistical methods will estimate the probability of OSCC as a function of measured VOCs. Calibrations and method validation will be used to assess and correct for discrimination. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| To evaluate the effects of known confounding variables on the target VOCs in OSCC | 36 months | To develop this clinical triage test, the investigators will use logistic regression modelling. They will assess the linearity of relationships between OSCC risk and continuous variables and explore categorical variables and two-way interactions. Inclusion of patient characteristics (e.g., age, gender, alcohol use and smoking), clinical variables (e.g., symptoms) and VOCs in the model will be determined using a combination of statistical significance (Wald p\<0.05), size of the model coefficients and clinical opinion, drawing on contemporary methods for clinical prediction rules. The investigators will use model reduction techniques such as the LASSO (least absolute shrinkage and selection operator) or Harrell's step-down approach to prevent over-fitting. They will also assess optimism. If there is missing data, the investigators will apply multiple imputation techniques. |
Countries
United Kingdom
Contacts
Primary ContactSameera Sharma, MBBS; MRCS
Outcome results
None listed