A breathalyser-style test for detecting pancreatic cancer
Article metadata
Article Date: 10 December 2025
Article URL: https://www.nature.com/articles/d41586-025-03946-0
Author: Esme Hedley
Image: Portrait of George Hanna
Summary
Surgeon George Hanna and colleagues at Imperial College London are developing a non-invasive breath test to detect pancreatic ductal adenocarcinoma (PDAC) earlier. The test captures volatile organic compounds (VOCs) from an exhaled sample into a sorbent tube; the trapped compounds are analysed by mass spectrometry to identify patterns linked to cancer, the host response and tumour-associated bacteria. The procedure takes about five minutes to collect and aims to deliver laboratory results within 24 hours. Initial multi-centre work involving more than 700 participants across 18 UK centres produced promising early results, and a large validation study (approximately 6,000 people) is running to 2028. The team also plans broader trials of breath tests for other gastrointestinal cancers and hopes to scale to tens of thousands of tests in coming years.
The project started with pilot work in 2007, progressed through multi-centre trials and methodology development, and received recent funding (including £1.1 million from Pancreatic Cancer UK) to advance validation. Key development challenges include preventing loss or contamination of VOCs during collection, transport and storage, and ensuring the trial cohort represents the UK population across socio-economic and ethnic groups. Health-economic modelling suggests a UK-wide screening pathway using the test could save the NHS around £155 million per year if implemented effectively.
Key Points
- Test concept: a breathalyser-style collection of VOCs captured on sorbent tubes and analysed by mass spectrometry to detect PDAC.
- Speed and convenience: breath collection takes ~5 minutes; lab-based results targeted within 24 hours.
- Detection basis: the model uses VOCs from the tumour, the host response and tumour-associated bacteria to distinguish cancer from non-cancer.
- Clinical progress: early multi-centre studies (700+ participants) show promise; a large prospective validation (c. 6,000 people) runs to 2028.
- Scale and ambition: plans to develop breath tests for multiple gastrointestinal cancers and test ~30,000 people over the next few years.
- Main challenges: VOC loss/contamination, transport and storage logistics, and ensuring representative recruitment across socio-economic and ethnic groups.
- Potential impact: modelling estimates around £155 million annual NHS savings if used as a frontline triage test to speed diagnosis and increase curative treatment opportunities.
Why should I read this?
Because pancreatic cancer is normally found too late — and this could change the game. It’s a five-minute, non-invasive test you could do in a clinic or pharmacy that aims to flag trouble fast. If the validation pans out, it’s the sort of simple tech that could mean earlier surgery for more people and big savings for health services. Worth a look — and worth keeping an eye on as the validation study progresses.
Author style
Punchy: this work matters. Early detection for PDAC is currently poor and survival rates are low; a reliable, cheap, quick triage test would be a major clinical step forward. The article’s practical focus and real-world trial data make it essential reading for clinicians, diagnostics developers and health-policy planners.
Context and relevance
Early detection is the critical bottleneck in pancreatic-cancer outcomes. This breath-VOC approach sits alongside other non-invasive diagnostics (liquid biopsies, biomarker panels) and taps into growing interest in metabolomic and microbial signatures as disease indicators. If the sensitivity and specificity reported in validation are clinically useful, the test could be integrated as a frontline triage tool to expand testing beyond strict referral pathways and reach people with vague symptoms. The project also highlights common barriers to deployable diagnostics: sample stability, standardised collection, transport logistics and representative recruitment. Its potential NHS cost savings and scalability ambitions align with health-system priorities for earlier diagnosis and value-based care.
Limitations and caveats
The interview outlines promising early results but does not publish final sensitivity/specificity figures yet. Real-world performance, false-positive burden, cost of mass spectrometry analysis and logistics of national roll-out remain to be demonstrated in the ongoing validation study.