Skip to main content

Together we are beating cancer

Donate now
  • For Researchers

The ‘mad thought’ that changed millions of lives

Henry Scowcroft
by Henry Scowcroft | In depth

17 November 2026

0 comments 0 comments

Clinical trial

The STAMPEDE prostate cancer trial formally wound down earlier this year. Henry Scowcroft looks at its practice-changing legacy, its profound impact on clinical trial design, and its successor, STAMPEDE2

“I’ve always liked playing around with a paper and a pen while thinking mad and ridiculous thoughts,” says statistician Professor Mahesh ‘Max’ Parmar. “If you’re not willing to have mad thoughts, you’re probably not thinking creatively enough.”

That creativity of thought has ultimately reaped dividends: Max’s ‘mad’ ideas have not only transformed prostate cancer treatment: they’ve helped reconceptualise clinical trial design – not just in cancer, but across conditions as diverse as COVID-19 and multiple sclerosis.

In the process, they’ve transformed millions of lives.

It was the late 1990s when inspiration first struck. Max – a maths graduate who’d gravitated towards a career in medical statistics – was running the Medical Research Council’s Cancer Trials Office in Cambridge, a small team that was one of the few running UK-wide oncology trials. Back then, most systemic cancer therapies were chemotherapies or, for some cancers, relatively simple hormone-based treatments.

But thanks to decades of research, a new generation of experimental therapies was arriving. And a question rattling around Max’s head was how to deliver them to the clinic, and to people with cancer, as efficiently as possible.

I started to wonder: why are we testing one drug at a time? Why couldn't we test many things at once, in a single platform?

Max Parmar
Mahesh (Max) Parmar is a Professor of Medical Statistics and Epidemiology and Director of both the MRC Clinical Trials Unit at UCL and the Institute of Clinical Trials and Methodology at University College London.

“Strange as it may sound today, clinical trials weren’t so much of a national priority – so I was thinking about how we could do things more efficiently and faster. And I started to wonder: why are we testing one drug at a time? Why couldn’t we test many things at once, in a single platform?” he recalls. “It wasn’t a particularly complicated idea, and I wondered what was holding us back.”

Prostate cancer – preparing the ground

As is often the case with seeds of inspiration, the issue was the lack of fertile soil. This would eventually arise a few years later, in 2002 in a pub in York, when Max and his trials unit colleague Matthew Sydes pitched their ‘mad idea’ to an eager group of oncologists and urologists who specialised in prostate cancer.

It was a field that was desperate for progress, because treatment for men with advanced prostate cancer had been largely the same for decades, recalls Professor Nick James, now a consultant at the Royal Marsden and a researcher at The Institute of Cancer Research – who was at that critical meeting. “Back then, essentially, you gave men hormone therapy, and they generally got better for a while. But then they progressed and died. And that literally hadn’t changed since 1941.”

But across this barren landscape, green shoots were finally starting to appear: pharmaceutical companies, having recently spent heavily on breast cancer, were now turning their attention to prostate cancer. “They’d realised that there was an equally common cancer killing pretty much the same number of people,” says Nick. As a result, a crop of newer drugs were arriving.

Chief among them was a new chemotherapy drug, docetaxel. Nick, then a relatively junior clinician scientist, had been UK lead for a pivotal commercial docetaxel trial which showed a survival advantage among people with late-stage, hormone-resistant disease – the first time a chemotherapy drug had done so.

It was a field that was desperate for progress, because treatment for men with advanced prostate cancer had been largely the same for decades.

Nick James
Professor Nick James is Professor of Prostate and Bladder Cancer Research at the Institute of Cancer Research and the Royal Marsden Hospital in London.

“There was a bit of a reluctance to give chemo to men with prostate cancer,” says Nick. This was partly due to the relatively limited evidence, but also because most men were seen by urologists, rather than oncologists.

“Urologists would generally prescribe hormone therapy, and if they needed to refer onwards it’d be for palliative radiotherapy.”

And although chemo could palliate symptoms too, its harsh side effects meant that “nobody was that interested in it, and no one really wanted to run the docetaxel trial – so I volunteered.”

The trial’s eventual success – prolonging survival by around 20%, equivalent to several months – got Nick and his colleagues thinking: what if drugs were given much earlier – alongside men’s initial hormone treatment, rather than after their tumours had become resistant?

Beyond docetaxel, there were several other promising ideas kicking around. Bisphosphonates, already used to protect bones against metastasis and preserve quality of life, were showing signs of disease-modifying effect, as was another bone-targeted drug, the radionuclide strontium-89. Hormones like stilboestrol – a synthetic oestrogen analogue – were showing promise, as were anti-inflammatories like celecoxib.

Nick, together with a close-knit group of more senior oncologists and urologists – including Professors Malcom Mason, Noel Clarke and David Dearnley – had begun to discuss how to trial these agents in the earlier, hormone-sensitive setting.

Of course, there was a challenge: men in this situation lived longer, so any effect would take much longer to show. But on the plus side, there was a relatively large population who might participate. “We reasoned that there were about 10,000 men a year dying of metastatic prostate cancer, and another 5,000 or so with high-risk locally advanced disease, who essentially die a bit later. And no one was doing trials in this population – so it should be relatively easy to recruit.”

The question occupying their minds as they travelled back from meeting in Max and Matt in York was: which of these treatments should we test first and why?

An iconic ‘failure’ paves the way

As Nick and his colleagues were trying to untangle the promising ideas in prostate cancer, Max had been trying to bring his ‘mad idea’ to life – and by now it had a name: the ‘multi-armed, multi-stage trial’, or MAMS trial.

As well as testing multiple treatments in parallel, the concept had another feature: it enabled a trial to span multiple stages in a single protocol: “At intermediate time points, you could compare each of the research arms against the control and ask, ‘is it worth carrying on with that research arm?’,” says Max. By identifying suitable interim metrics such as progression-free survival (PFS), signs of efficacy could be quickly detected, allowing promising arms to continue from stage 2 to stage 3, while quickly weeding out ineffective treatments.

The statistical challenge was substantial. The design needed to compare multiple treatments against a common control, allow repeated interim looks at the data, and eventually report clear and reliable answers, all within a single trial framework. Many researchers thought it was simply too complicated to work in practice.

However, by 2001, Max and his colleagues had refined the idea sufficiently to launch an ambitious five-arm ovarian cancer trial, ICON5, comparing different chemotherapy combinations. Initially it was a huge success, recruiting “like wildfire” and eventually drawing in 4,200 participants in just a few years. But at the first interim analysis, the wheels came off: none of the combinations proved superior to standard of care. “Nothing worked,” remembers Max.

However, word of ICON5’s radical, innovative design – and rapid recruitment – had spread. Although none of the experimental treatments proved superior, the trial demonstrated something equally important; large numbers of patients and clinicians were willing to participate in a study that tested several questions simultaneously. The treatments may have failed but the methodology had succeeded. And now, prostate cancer clinicians were keen to work with Max’s team to see whether the same thinking could help them chart a way forward.

Considerations for clinical trials. Illustrative

The idea takes root

In York, the group reached broad and enthusiastic agreement to pursue a multi-arm, multi-stage trial in men with either newly diagnosed metastatic prostate cancer, or localised, high-risk disease. Since both disease states were treated with hormone therapy as standard, this would be the control arm, with experimental arms adding one or more agents to it.

Given that the trial would likely take a decade or more, it was decided that Nick would lead it – he was early enough in his career to see it through, and unencumbered by senior responsibilities. “Nick also came with lots of energy and ideas, and was the sort of person who wouldn’t take ‘no’ for an answer,” says Max – characteristics that would prove vital in the years to come.

The basics agreed, the group’s attention then turned to getting their idea off the ground. “In order to do it, we had to engage with UK researchers and clinicians and see what they would think was a reasonable thing to do,” says Nick.

There was also the small matter of which drugs to include.

“After a lot of discussion, we decided zoledronate, a bisphosphonate, had to be in there, because the licensing trials suggested that there was a survival advantage, even though that wasn’t the trials’ primary endpoint,” recalls Nick. The docetaxel data made it an “obvious choice” too, and Nick’s prior connections with docetaxel’s manufacturer, Sanofi, facilitated this.

A third treatment – the anti-inflammatory drug celecoxib – was also selected based on epidemiological evidence suggesting improved survival, and a plausible biological mechanism. This established a principle which would last the entire length of the trial: testing pre-existing drugs not primarily licenced for cancer – ‘repurposing’ – as well as early testing of novel cancer drugs.

In the end, they settled on six arms: a control arm; a docetaxel arm; a zoledronate arm; a celecoxib arm; and two additional arms – one combining zoledronate with docetaxel, and another combining it with celecoxib.

And finally, there was the matter of what to call their radical idea.

Eventually, Matt Sydes proposed the name: “Systemic Therapy in Advancing or Metastatic Prostate cancer: Evaluation of Drug Efficacy” – STAMPEDE for short.

“And then, of course, we needed to find someone to pay for it,” says Nick.

Growing the trial

With the design almost finalised, the team approached Cancer Research UK for core funding, with additional funding to come from Max’s MRC Cancer Trials Group.

But when it came to presenting it to Cancer Research UK’s funding committee, “they initially thought it was mad,” remembers Max. “Some people said it was too complex – ‘doctors won’t like it; patients won’t understand it’.

“Two things got it through,” he says. “The main thing was the patient voice on the committee – which was relatively new in those days, but so important.”

The other, Max says, was the backing of the committee’s influential chair Professor Stan Kaye. An ovarian cancer clinician, Kaye had heard how well the ICON5 trial had recruited, “and that probably clinched it,” says Max: “sometimes you need a bit of luck and coincidence to get you through.”

In 2004, Cancer Research UK’s clinical research committee approved funding to support the trial, which would initially open at 6 centres around the UK, to iron out any on-the-ground issues, with plans to scale up over time.

We spent a lot of time doing roadshows to rooms full of people, explaining what we were doing, getting the momentum to build.

A year later, in October 2005, STAMPEDE randomised its first participant.

Even then, it wasn’t plain sailing, says Nick. “There was a fair degree of scepticism”, especially from urologists with regards to chemotherapy. “Care of these patients largely sat in urology clinics, so for the trial to recruit, we had to change the whole referral pattern,” he says. “We spent a lot of time doing roadshows to rooms full of people, explaining what we were doing, getting the momentum to build.”

And build it did. Far from being ‘too complex’, as word got out recruitment became a runaway success – “it just flew,” Max says – and this was due, in large part, to a then-under-appreciated aspect of MAMS trials: with multiple intervention arms, participants had a high chance of being randomised to an experimental treatment. This, of course, is not a reason to do these types of trials, but a direct consequence of their design. But it is seen by many doctors and patients as an attractive component of such trials.

Danny Garrett
Danny Garrett in an Oncology Research Nurse in uro-oncology clinical trials, University College London Hospitals

As a research nurse at UCL Hospital, Danny Garrett helped enrol scores of men to STAMPEDE, and seen men’s positivity first-hand. “Since participants know they’re getting a good standard of care anyway, it makes them much more amenable,” he says. “They’ll either continue with the ‘gold standard’, or they’ll be adding in this extra option that wouldn’t normally be available.”

Over the next two decades, STAMPEDE would go on to recruit just under 12,000 men with prostate cancer, and branch out to more than 120 sites across the UK and Switzerland.

Adding an arm

By 2008, STAMPEDE was in full swing and recruiting well – and wider momentum was building in the prostate cancer field. In particular, a new drug was making waves: the hormone blocker abiraterone (sold commercially as Zytiga).

Developed by researchers at The Institute of Cancer Research, and commercialised through Cancer Research UK’s translational arm, Cancer Research Horizons (then known as Cancer Research Technology), it was showing dramatic results in early phase trials – and Nick and other clinicians could see the impact it was already having in men with later-stage disease.

At this point Nick started to wonder, could they bring abiraterone into STAMPEDE? And it really was a radical idea.

“We’d talked a lot about how to stop arms in these trials,” Nick says, “but then we started to wonder, how do you add an arm?

This would necessitate the development of a MAMS platform trial – an ongoing, perpetual “umbrella” trial that can run indefinitely, spanning multiple generations of experimental therapies under a single master protocol.

It was clear that funders and regulators would take some convincing, he recalls, “but ultimately it was a very simple logic: we could do one of three things. One, ignore new treatments like abiraterone. Two, we could start a second, competing trial in the same population – which is madness, why would you compete with your own trial?”

The third option was to set up a new arm within the existing trial. “No one had done this before. It seemed mad… but the other options were even madder,” Max says.

The argument was swayed further by trial’s stellar track record in recruitment and, in 2010, STAMPEDE opened a seventh arm, adding abiraterone to standard hormone therapy – with manufacturer Johnson & Johnson covering the drug’s cost. “When we amended the protocol, we were very worried it might get strangled in NHS R&D bureaucracy,” recalls Nick. “We put a lot of effort into preparing the ground and put a deadline on NHS Trusts approving it. We were very nervous they might not – and we could kill our own trial.”

In the event, the opposite happened and the speed of approval was breathtaking, with 80 sites recruiting to the new comparison within weeks. Typically it takes a year or more to get that many sites recruiting to a new trial.

“We probably brought about £50-60m of abiraterone into the NHS through STAMPEDE,” Nick says. “And J&J got a very cheap phase 3 trial – it was very much a win-win situation for the trial, for the company and above all for the thousands of men with prostate cancer who took part and received the drug.”

With the principle established, the abiraterone arm was soon joined by an eighth, radiotherapy, arm. And when another new hormone-blocking agent, enzalutamide, arrived, the team added a ninth arm, trialling it in combination with abiraterone. A tenth arm opened in 2016, to test a diabetes drug, metformin; a final eleventh arm substituted one of the ‘standard’ hormone therapy drugs with an oestrogen patch.

“The operational complexity of the trial became substantial,” says Max. The underpinning statistics also grew even more challenging, requiring constant dialogue with regulators. “At one point, we had 40 people working on it in our trials unit.” says Max, who by then had relocated his unit from Cambridge to UCL and become Director.

The first results

By 2013, it became clear that the celecoxib arms weren’t showing sufficient PFS benefit – they closed to new recruitment, although the participants would continue to be followed up. The combined zoledronic acid and celecoxib arm also failed so show clear benefit, and also closed.

One of the strengths of a MAMS platform trial is not simply finding effective treatments quickly, but identifying ineffective approaches early.

These closures were an important success in their own right. One of the strengths of a MAMS platform trial is not simply finding effective treatments quickly, but identifying ineffective approaches early. STAMPEDE was generating quicker answers whichever way the results fell.

In other arms, docetaxel and the combined docetaxel/zoledronic acid arm, the story was looking much more promising, and in 2015, ten years after the trial initially launched, “we were all called to a meeting at the Clinical Trials Unit,” remembers Nick. STAMPEDE had reached its first significant landmark: the docetaxel results were ready.

The unveiling of the data was, he says, “the statistical equivalent of a striptease: ‘these are the patients that are in the trial’, ‘these are their age characteristics’, ‘this is the…’ And everyone’s like, come on, get to the results!”

When they eventually came, they exceeded all expectations: adding docetaxel to hormone therapy dramatically improved men’s overall survival by around 25%. It would change clinical practice worldwide.

More was to follow two years later, with the results of the abiraterone arm: another huge survival gain of 35% – and another practice-changing option for men with advanced disease. Nick recalls presenting the results at the annual American Society of Clinical Oncology meeting in Chicago. “It was standing-room only”, he recalls. “It was just massive – and super-exciting because now we’d shifted the dial on survival, a lot, twice.”

But STAMPEDE wasn’t done there. The following year, the radiotherapy arm showed a survival advantage among men with a lower metastatic burden – and once again, guidelines changed to capitalise on the discovery. And as abiraterone and enzalutamide became more widely available, STAMPEDE showed that there was no additional advantage in combining them – a ‘negative’ result that still helped guide clinical practice the world over.

Sitting alongside these results came a slew of less eye-catching but important, practice-guiding findings, looking at prognostic signatures, genomic alterations and other biological features of prostate cancer, giving ever-deeper insight into the disease’s underlying biology.

A good thing comes to an end?

In theory, STAMPEDE could have gone on adding arms, recruiting men, and cranking out results. But after two decades, the world had changed – and questions were starting to be raised about the trial’s long-term future. “It had become a beast – and every now and again, you’ve got to make some big decisions,” Max says.

For one thing, the changing regulatory environment kept throwing curveballs. “Every time we were getting reviewed, we were inspected on new regulations, even though the set-up was based on decisions made more than 15 years ago,” he says. This was adding a significant administrative burden to an already stretched trial team.

And then there were the now-antiquated databases and processes through which the trial ran: on the front line, the paper-based case report forms (CRFs) used to manually collect participant data were a ‘nightmare’ compared to more modern trial set-ups, says research nurse Danny Garrett. “We had to physically post the paper CRFs to the MRC Trials Unit – we could scan them individually and send them by secure email – but that was a faff.”

After much agonising, the decision was made to end STAMPEDE in its current form. Recruitment stopped in 2024, and in April 2026, the last centre formally closed (although participants are still followed up passively, via their NHS records).

But that’s far from the end of the story. “For the majority of the patients in the trial, we have their baseline pathology, we have all of their imaging, blood samples, genomics… the translational impact has already been colossal,” says Nick, pointing both to a charity-funded translational programme, STRATOSPHere, and commercial partnerships with companies like Veracyte.

These partnerships and collaborations underpin another key feature of STAMPEDE: the sheer number of people involved in making it tick. For example, across the participating hospitals, the trial delegation logs feature more than 3,000 names. And critically, academic credit is shared across the team, with designated comparison leads presenting crucial outcome data at major conferences; research fellows fronting up key translational work; and patient representatives presenting patient-reported outcomes. “This gives the whole team a sense of ownership and distributes the academic credit widely, rather than focussing it on a small central core,” says Nick.

And next, rising from the ashes is STAMPEDE’s successor trial: STAMPEDE2 – effectively a ‘clean reboot’: “We’ve reset all the procedures, the processes, the databases – the lot,” says Max. The trial is now open with two arms: one testing a form of advanced radiotherapy called SABR, and another testing a radioisotope-based treatment called PSMA-lutetium.

More are being considered, looking at agents as diverse as T-cell-engager immunotherapies, novel targeted molecular therapies, and – capitalising on recent data on the benefits of physical activity in cancer – structured exercise regimens. Of particular interest, says Nick, is using ctDNA blood tests to test novel approaches in patients who seem at particularly high-risk of relapse after hormone therapy.

Also in the works is STAMPEDE3, another MAMS platform trial looking specifically at non-metastatic disease.

The MAMS that changed the world

It’s hard to think of a trial that’s made more waves than STAMPEDE. Beyond its impact on prostate cancer, the underpinning MAMS platform concept has transformed the wider fabric of clinical trial design and practice. There have now been more than 80 randomised MAMS trials around the world tackling challenges such as progressive multiple sclerosis, Covid-19 and post-surgical infections, as well as a whole range of different cancer types. They are helping shift clinical trials from one-off comparisons towards enduring research infrastructures.

I did a quick back-of-envelope calculation, and I reckon it’s reasonable to say that the use docetaxel and abiraterone in newly diagnosed men, as a result of trials like STAMPEDE, has led to more than a million life-years gained globally – and growing.

Dozens more are being designed and both Nick and Max now regularly speak to people all around the world to help develop these trials.

“STAMPEDE’s most visible success has been the identification of treatments that have improved and prolonged the lives of millions of men with prostate cancer,” says Max. “But its deeper legacy lies in creating a framework that could adapt as scientific understanding advanced, continually evaluating new opportunities for progress.”

For Nick, the impact on men with prostate cancer underpins STAMPEDE’s importance. When it launched in 2005, there were few trials for advanced, hormone-sensitive prostate cancer. “By doing something at such scale so early, I feel we’ve redefined that space,” says Nick. “Now it’s incredibly crowded,” and, as a result, there’s been a measurable increase in survival showing up in real-world healthcare data. “I did a quick back-of-envelope calculation,” says Nick, “and I reckon it’s reasonable to say that the use docetaxel and abiraterone in newly diagnosed men, as a result of trials like STAMPEDE, has led to more than a million life-years gained globally – and growing.”

And that translates into impact on individual lives, says Danny: “To see patients that were metastatic and still alive years later, that took a while to get my head round. And to see these drugs being used day-to-day in the patients you see, or read about it in the press, it makes you think, ‘I worked on that trial’. It makes all the hours spent in getting questionnaires and doing the data entry worth it.”

Reflecting on what STAMPEDE’s stellar success means for Nick personally, he’s struck by a particular moment: “Standing in front of a packed conference auditorium with thousands of people – that’s pretty special feeling – like you’re doing some sort of stadium rock… but instead of Oasis giving everyone a good sing-along at Wembley, you’re actually telling them how to save people’s lives.”

STAMPEDE in numbers

20 years

12,000 patients

3,000+ site staff

22 clinical fellows

4 practice-changing results

20% improvement in prostate cancer survival

~1,000,000 life-years added globally

Henry Scowcroft

Author

Henry Scowcroft

Henry is an award-winning science writer and author of Cross Everything, an acclaimed memoir about cancer. He worked at Cancer Research UK between 2003 and 2022, and is now a freelance communications consultant, writer and editor. 

Tell us what you think

Leave a Reply

Your email address will not be published. Required fields are marked *

Read our comment policy.

Tell us what you think

Leave a Reply

Your email address will not be published. Required fields are marked *

Read our comment policy.