Completed Cancer Computing & AI

MICA: The exploitation of a novel image-based risk stratification tool in early prostate cancer - the Re-IMAGINE Consortium

In plain English

AI plain-English summary

Every year, one million European men undergo a prostate biopsy—yet many are over-diagnosed, over-treated, or missed entirely. This project tackles a fundamental gap: doctors cannot reliably tell which early prostate cancers will become dangerous and which will not. The Re-IMAGINE Consortium will combine molecular markers from blood, urine, and tissue with high-quality MRI scans to predict each man’s risk—low, medium, or high—without needing a biopsy. The team has already shown that MRI identifies at-risk men twice as accurately as standard biopsy and misses no men with potentially fatal disease. If successful, this could transform prostate cancer care from a system of imprecision and unnecessary harm into one where men receive fewer but better biopsies, appropriate treatment, and fewer life-long side effects. It would also save healthcare resources by avoiding costly, ineffective treatments. The consortium will recruit men from three London centres, including those who have never had a PSA test, to test MRI performance in a lower-risk population. The work is driven by patients and advocacy groups from the outset.

View original technical description
Background: One in eight men will develop prostate cancer and the risk increases with age. Most will not die of it, but some do. Our ability to identify those men whose cancer will decrease either their quality or quantity of life remains poor. Nowhere else in modern medicine are the errors of over-diagnosis, over-treatment, missed-diagnoses and poor risk-stratification more extreme. These four errors - which impact directly on the 1million men who undergo prostate biopsy each year in Europe alone - result in the application of treatments that have little or no benefit, leave men suffering with significant life-long harms and waste valuable healthcare resources. Aims: Our Re-IMAGINE Consortium has been created to undertake discovery that will correct these four key errors. We will, for the first time, be able to combine the underlying molecular changes in the cancer with state-of-the-art imaging. This will allow us to predict prostate cancer status for the individual (low, medium or high risk) without recourse to biopsy. It will also allow us to predict which prostate cancers are likely to progress over time and which are not - something that has largely eluded us so far. By doing so, men will be subject to fewer but better biopsies, improved risk stratification, appropriate treatment allocation, more benefit, less harm and more cost-effective care. Methods: We will achieve this by building on established partnerships between patients, advocacy organisations, clinicians, imaging experts, molecular biologists, methodologists and a broad range of industrial partners. The vision of Re-IMAGINE is built on our recently published work showing that magnetic resonance imaging (MRI) of the prostate was 100% better at identifying men at risk compared to the standard prostate biopsy, and moreover missed no men with potentially fatal disease. MRI is now certain to become the future cornerstone of the risk-stratification process for men at risk of early prostate cancer. However, we know very little about the use of MRI in combination with other markers in the body. Through working with patients and advocacy groups, we have developed several work strands (WS) that will allow us to realise our aims: WS1 will create the first group (cohort) of men who receive a prostate cancer diagnosis by means of an MRI-based pathway. These men will, after appropriate consent, donate some tissue, blood and urine for marker analysis. They will be recruited from 3 major centres in London that have high quality MRI in place. WS2 will test the performance of MRI in a population where cancer is less common - in men who have not had a PSA test. This will allow us to predict the exact proportion of men at risk within the community. These work streams will generate a lot of data (clinical information, imaging data, as well as information from markers in blood, tissue, and urine). We will combine all this data for each patient using a secure data warehouse and analyse it using advanced mathematical techniques. Such a complex project will require high levels of project management and coordination. This will be achieved by building on existing and fruitful collaborations. Due to its ambition, likely impact and potential for positive disruption we are pleased that the early development work was both informed and driven by patients and their families and in some specific instances (WS2) by the public. This patient and public involvement will intensify during the period of data gathering, analysis and dissemination. Impact: Re-IMAGINE will transform the management of early prostate cancer from an era in which imprecision and uncertainty was the norm to one in which any patient can expect - and benefit from - a precise prediction of his own risk and better care as a result.

View the original record at the funder ↗

Researchers

Anthony C.C. (Ton) Coolen (Co-Investigator)Caroline Moore (Co-Investigator)Gerhardt Attard (Co-Investigator)Hashim Ahmed (Co-Investigator)Hayley Whitaker (Co-Investigator)Louise BROWN (Co-Investigator)Mark Emberton (Principal Investigator)Mieke Van Hemelrijck (Co-Investigator)Shonit Punwani (Co-Investigator)

Related Research

Grants with similar aims, by meaning.

BeSpoke: Building and Evaluating a risk Stratified PrOstate pathway for cancer screening, detection, treatment and surveillancE
Smarter identification and management of Early prostate cancer: improving Lives and outcomEs through Clinical Translation of novel magnetic resonance imaging (SELeCT)
Developing microstructural and metabolic magnetic resonance imaging to address the diagnostic and prognostic unmet needs in breast cancer
Integrating MRI with spatial transcriptomics to identify “Radio-Spatial Genomic” features of aggressive prostate cancer using artificial intelligence.
MRI to Histology mapping using Deep Learning

Original classification

Research Grant

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.