Lung cancer remains the leading cause of cancer-related death in the UK, with over 35,000 deaths annually and most cases diagnosed at a late stage. Current screening programmes using low-dose CT scans target only high-risk individuals, missing around 30% of lung cancer cases, including many women and never-smokers. There is no simple, non-invasive tool to help triage patients with persistent respiratory symptoms who fall outside formal screening criteria. Inflammacheck® measures hydrogen peroxide and other breath biomarkers linked to airway inflammation and oxidative stress. Preliminary studies (VICTORY and ExPeL) have shown strong diagnostic performance for distinguishing lung cancer from other respiratory conditions. VICTORY 2 aims to validate and refine the AI model supporting Inflammacheck®, enabling accurate, rapid, and affordable triage for suspected lung cancer in NHS settings.
Group A: Lung cancer confirmed by multidisciplinary team (MDT) diagnosis.
Group B: Suspected lung cancer (biopsy negative or awaiting further investigation).
Group C: Pneumonia (radiologically or clinically confirmed).
Group D: Non-malignant airways diseases (asthma or COPD).
1. Active respiratory infection (e.g., tuberculosis, bronchopneumonia) other than pneumonia (e.g., tuberculosis) at the time of breath sampling.
Eligible patients will be identified by members of the clinical or research team during attendance at relevant hospital or community clinics, including: * Suspected lung cancer clinics and respiratory outpatient services * CT imaging lists for patients awaiting percutaneous CT lung biopsy * Inpatient wards with respiratory or pneumonia admissions; and * Community diagnostic hubs or screening services such as Total Lung Health Check.
Other
Other
No intervention
Other interventionThis is a multi-centre, mixed retrospective-prospective observational study with no intervention or randomisation. The study involves data analysis and AI model development based on exhaled breath measurements, demographic data, and clinical diagnosis.
Refine an AI-based diagnostic model
To develop and refine an AI-based diagnostic model using Inflammacheck® breath biomarker data from a diversified cohort to reliably distinguish participants with lung cancer from those with non-malignant respiratory conditions and healthy controls. Primary Endpoint: An AI model with diagnostic performance (Area Under the Receiver Operating Characteristic Curve (AUROC) target value of ≥0.80) that discriminates lung cancer from non-malignant respiratory conditions and healthy controls.
Click on the lead sponsor to view all of their trials.
