Pancreatic cancer remains one of the most difficult cancers to diagnose early, largely because symptoms often appear late and are usually non-specific. For many patients, signs such as abdominal pain, back pain or indigestion may not immediately suggest pancreatic cancer, delaying diagnosis until the disease is advanced.
Professor Tatjana Crnogorac-Jurcevic, Professor of Molecular Pathology and Biomarkers at Queen Mary University of London’s Barts Cancer Institute, has spent almost three decades studying pancreatic ductal adenocarcinoma, the most common form of pancreatic cancer. Her research focuses on improving early detection through non-invasive biomarker testing, particularly using urine samples.
Her team identified a panel of urinary biomarkers, including LYVE1, REG1B and TFF1, which have shown strong potential for detecting pancreatic cancer earlier. When combined, these biomarkers have demonstrated around 90% accuracy in early detection and have been validated in more than 1,000 retrospectively collected samples. Importantly, the biomarkers were shown to rise up to two years before clinical presentation and remained robust across factors such as gender, race and time of collection.
A key strength of urinary testing is that it is completely non-invasive and could support home sampling. This makes it especially promising for surveillance of high-risk groups, including people with a family history of pancreatic cancer, genetic risk factors such as BRCA or p16 mutations, and other defined high-risk populations.
Professor Crnogorac-Jurcevic also highlighted the importance of distinguishing pancreatic cancer from benign pancreatic and hepatobiliary diseases, such as chronic pancreatitis. Comparing cancer samples only with healthy controls can overstate accuracy, so her team has focused on testing biomarkers against clinically relevant benign disease groups.
Looking ahead, early detection is likely to depend on combining biomarkers with imaging, clinical risk factors and AI-driven tools. Radiomics and AI-based analysis of CT, MRI and other imaging data could work alongside biomarker testing to improve risk stratification and diagnosis.
However, several barriers remain before biomarker tests can become routine in clinical practice. These include the need for large-scale validation, pre-diagnostic testing, prospective clinical studies, regulatory approval, accredited laboratory workflows, cost-benefit analysis and evidence of real clinical utility. Professor Crnogorac-Jurcevic’s team is currently involved in the UroPanc trial, which is prospectively testing biomarkers and clinical data.
Over the next three to five years, the biggest opportunity lies in detecting pancreatic cancer when it is still small and localised. If tumours can be found when they are under two centimetres and confined to the pancreas, survival could improve substantially. For Professor Crnogorac-Jurcevic, the future of progress lies in better surveillance of enriched high-risk groups, well-validated biomarkers, improved imaging and more effective treatment strategies after surgery.
Earlier diagnosis remains the most urgent priority. By bringing together non-invasive testing, risk stratification and advanced imaging, researchers hope to shift pancreatic cancer detection to a stage where treatment can make a far greater difference.
