Published on 13 April 2026
Data from cancer patients across Cheshire and Merseyside has been used in research that could change UK clinical practice in prostate cancer care.
This major national study has shown that artificial intelligence can significantly improve predictions about how a patient’s prostate cancer is likely to behave.
The findings come from biopsies taken in the CHHiP clinical trial from hundreds of men treated for localised prostate cancer at The Clatterbridge Cancer Centre (CCC) between 2002 and 2011, and their health monitored for more than 10 years afterwards.
CCC was the largest recruiting site in the UK for the landmark CHHiP trial and the biopsies from 1,800 men across the country were analysed in the new research, which was led by London’s Institute of Cancer Research.
Prostate cancer cases are expected to rise sharply over the coming years, but not all prostate cancers behave in the same way. Some grow slowly and may never cause harm, while others return or spread despite treatment. Current tools that assess this risk do not always capture these differences, which can lead to over‑treatment or under‑treatment.
The new research used an AI pathology test which analysed the initial biopsy slides and combined this with clinical data and information on patient outcomes over 12 years of follow-up.
The AI test, a multimodal artificial intelligence biomarker, looks more closely at the biology of the cancer using digital scans of standard tissue samples. This allows doctors to see differences between patients that are not visible using standard methods alone.
The study showed that the AI test could divide patients into low‑, medium‑ and high‑risk groups more accurately than current tools. After 10 years around 85% of men in the AI low‑risk group had no signs of their cancer returning. This fell to 73% in the medium‑risk group and to just 45% in the high‑risk group. The AI test was also better at predicting whether cancer would spread to other parts of the body.
As the AI test uses routine pathology samples already collected as part of diagnosis, it could be introduced into clinical practice without extra procedures for patients.
Researchers say the findings – recently presented at the 2026 American Society of Clinical Oncology Genitourinary Cancers Symposium – show how AI can help personalise prostate cancer treatment, ensuring men at higher risk receive the right level of care, while avoiding unnecessary treatment for those with less aggressive disease.
The research is the latest to use data from the CHHiP trial, one of the largest and most important prostate cancer radiotherapy studies ever run in the UK. Patients received high‑quality, modern treatment, and all samples went through strict quality checks, making the results particularly robust. Clatterbridge played a leading role in the trial, recruiting more patients than any other centre and helping to build one of the strongest evidence bases for improving prostate cancer care.
One of the leaders of the CHHiP trial at Clatterbridge, consultant oncologist Prof Isabel Syndikus, said: “It is fantastic that data collected at Clatterbridge has been used in this important study, which shows how artificial intelligence can give us a much clearer picture of which prostate cancers are likely to come back or spread.
“By using information already available from routine biopsy samples from the CHHiP clinical trial, this approach has real potential to help doctors personalise treatment and make better‑informed decisions for patients.”