THIS FUTURISTIC PROSTATE CANCER TEST COULD PREDICT YOUR TRUE RISK
Imagine a future where the moment you’re born, a virtual avatar is created and steadily filled with personalised information that is used to assess your cancer risk at each stage of life.
When you’re born, this avatar will hold your DNA, sequenced from a blood spot. Then over the course of your life, more layers of data are added, perhaps collected from wearable sensors, or combined with the results of scans and blood tests. This vast information resource, personalised entirely to you, will be known as your “digital twin”.
Doctors can use your digital twin to advise you on how to adapt your lifestyle to reduce your cancer risk, and if ultimately diagnosed, an oncologist will consult it to predict the treatments that will suit you best.
Amid the chaos of the modern-day NHS, such a futuristic vision of healthcare might seem more akin to Blade Runner than the real world. But primitive versions of digital twins are already being pioneered in a few hospitals around the UK, while many believe they will ultimately reshape cancer medicine in the years to come.
“In time, we’ll end up having digital twins for a number of different diseases,” predicts Prof Rakesh Heer, a urology researcher at Imperial College London. “It’ll mean that we can compare between populations of individuals, those that get a disease versus those that don’t. We can do this now with clinical trials, but they cost too much. Digital twins will allow us to get these questions answered quicker, and at scale, without the same cost.”
Let’s take a look at how they’re already being used – and what’s likely to come.
From planes to prostate cancer – how tumours will be studied in the future
According to Mikael Benson, who leads the digital twin research group at the Karolinska Institute in Stockholm, the concept behind digital twins has a long history in engineering, and in particular the aerospace industry. Through creating detailed computer models of planes, all kinds of experiments can be conducted in the virtual world to study everything from lightning strikes to air turbulence, with the results used to make them safer and more resilient. Now, the same idea is being used to study tumours.
To make a formal diagnosis of prostate cancer, for example, oncologists place pieces of prostate tissue on glass slides which are then analysed under a microscope and graded according to a scale called the Gleason score. “Patients with higher Gleason scores do significantly worse,” says Prof Nicholas van As, a consultant oncologist and the chief medical officer at the Royal Marsden NHS Foundation Trust. “But for people in the middle, there’s lots of different treatment options and different ways you can manage their disease, for example with radiation or hormone therapy.”
How is this new technology being used now?
Along with a few other hospitals around the UK, the Royal Marsden’s oncologists have begun making more informed treatment decisions for prostate cancer patients through applying newly developed AI tools to their digital twins. These virtual clones are created using digitised versions of the prostate biopsy slides, along with additional pieces of information unique to that patient such as their age, PSA level and other clinical data, before being fed into a computer algorithm called ArteraAI.
For patients with intermediate-high Gleason scores, ArteraAI is able to use this data to predict whether they can significantly benefit from various additional hormonal therapies in addition to radiotherapy. “It’s really exciting,” says van As. “No one really knows exactly what it’s picking up, because it’s beyond the human eye, but it’s been really well validated. It’s likely to transform how we manage and treat prostate cancer in the next five years.”
But helping to predict response to treatments is merely the beginning of how this technology is set to be increasingly involved in cancer care.
Finding a better screening tool for prostate cancer
Last year, Prostate Cancer UK unveiled the landmark Transform trial, a £42m initiative backed by the NHS, the National Institute for Health and Care Research, the UK government and various other donors. Its main aim is simple – finding a better way of screening men for prostate cancer.
As Heer, one of Transform’s lead scientists, explains, current assessments like the PSA test are limited as they’re incapable of distinguishing between relatively benign or indolent tumours which are unlikely to spread beyond the prostate, and more life-threatening tumours. While widespread use of the PSA test saves lives, it also leads to many men undergoing unnecessarily aggressive treatment.
“Surgery and radiotherapies can have lifelong consequences such as incontinence and loss of erections, that men have to live with,” says Heer.
To improve on this, Transform will recruit several hundred thousand men from around the UK. Digital twins will be created based on data collected by analysing their blood and urine, sequencing their microbiome from faecal samples, taking MRI scans of their prostate and estimating their genetic risk of prostate cancer from saliva swabs. Men who are found to have cancer will undergo biopsies, the results of which will be added to their digital twin.
Through this vast data reserve, Heer is optimistic that completely new ways of diagnosing and assessing prostate cancer risk can be identified. He believes it can identify a range of molecular signals, known as biomarkers, which can help specialists pinpoint the right treatment for a patient, assess their risk of side effects, and monitor men who have gone into remission to track whether their disease is at risk of recurrence.
Recommended
Five things that can help lower your risk of prostate cancer, from diet to exercise
The unique power of having digital twins is that, as with aircraft, it allows virtual experiments to be carried out. Heer points out that numerous companies and research groups have already identified promising prostate cancer biomarkers in small studies, but lack the money to run the kind of expensive, large clinical trial which would be needed to prove their efficacy. But through using Transform’s digital twins, he says that researchers will be able to sift through these various biomarkers and test them on the digital clones of tens of thousands of people at a time.
“If it shows a signal, we can quickly turn that into a test that’s available to patients,” says Heer. “I think this offers a quicker route to getting patients new diagnostics.”
Beyond cancer
Digital twin initiatives are also taking off in other fields of medicine. At Imperial College London, researchers are using data from scans, wearables and implanted heart monitors to create virtual representations of the hearts of patients with pulmonary arterial hypertension, a life-threatening form of cardiovascular disease, as part of a project called CVD-Net. The hope is that this will lead to new ways of forecasting when treatments have stopped working, or when patients will experience fluctuations in their condition.
Benson has also been involved in a major European Union project using digital twins to develop personalised treatment approaches for a range of autoimmune and inflammatory diseases such as rheumatoid arthritis, systemic lupus and ulcerative colitis.
Tackling harder-to-treat cancers
There’s particular optimism that the digital twin approach will point to new ways for tackling some of the hardest to treat cancers. For example, a major challenge for oncologists is to distinguish between benign pancreatic cysts and those which are likely to develop into tumours. “Such cysts are increasingly being picked up by different medical imaging techniques, but most are harmless, and it’s difficult to predict those which aren’t,” says Benson.
Benson has recently been involved in a project using data from the UK Biobank – a vast database of biological information collected from hundreds of thousands of British people – to create digital twins for people with pancreatic cysts. With the database recording which individuals went on to develop pancreatic cancer, he has been able to identify biomarkers which could potentially be used as new screening tools for the deadly disease in future.
“Having biomarkers would be hugely transformative for a very, very difficult cancer to treat,” says van As. “With pancreatic cancer, if you don’t find it early, you’re never going to be able to cure it, and the mortality rates are extremely high.”
The future of cancer treatment
Ultimately, van As feels that the future for all cancers is for digital twins to be created for everyone from birth, providing doctors with the ability to offer ever more personalised advice when it comes to prevention and treatment, based on these reserves of stored biological information.
“Every bit of health data these days is stored electronically, and therefore you could constantly update the digital twin to form this massively rich dataset, which will have predictive potential,” says van As. “You’ll be able to start seeing patterns which show that someone has a higher chance of developing a cancer later in life, and so you may be able to intervene.”
Recommended
NHS to pilot AI prostate cancer tests
2025-10-04T06:30:40Z
