Every little helps – how we’re transforming breast cancer survival
Every year, more than 50,000 people in the UK are diagnosed with breast cancer. Thanks mainly to advances in treatment, death rates have halved since the 1970s and survival is continuing to improve year on year, particularly for cancers diagnosed at an early stage. Much of the credit for this impressive achievement can be traced back to a team of dedicated statisticians working in the Oxford Clinical Trial Service Unit (CTSU, now part of the NDPH).
By pulling together results from hundreds of research groups all around the world over three decades, they’ve pieced together vital information about how best to treat patients with early breast cancer. Although each improvement may be small, it all adds up to many tens of thousands of lives saved over the years, with still more benefits to come in the future.
The start of the story
Back in the 1980s, when Professors Sir Richard Peto and Sir Rory Collins set up the CTSU, it wasn’t clear how best to treat breast cancer in its early stages, before there’s any evidence of the disease spreading around the body. The first option was surgery to remove the tumour, but whether subsequent treatments – such as chemotherapy, hormone therapy (principally the drug tamoxifen) or radiotherapy – could help prevent the cancer from coming back and improve survival wasn’t known.
At the time, the CTSU was running a clinical trial testing whether tamoxifen or chemotherapy made a difference to survival, and the tamoxifen results seemed very promising. But, in the words of Professor Richard Gray who was working on the trial at the time, they seemed “too good to be true, so we wanted to check what benefits other trials had shown from the treatment”. To build up a bigger picture, the CTSU team brought together UK and international researchers running trials in early breast cancer, collecting their data and crunching it all together using a technique known as meta-analysis.
“At first people were saying it was a waste of time because none of the other trials of tamoxifen had reported any survival benefit,” explains Gray, “but when you look at all the trials together you realise that most favour tamoxifen and, when the results are combined in a meta-analysis, there’s a small but clear survival benefit. We found the same for chemotherapy. The investigators were really surprised by the results when we put all the numbers together.”
The results of the meta-analysis - presented at a meeting held near Heathrow airport back in 1984 - showed the power of bringing together the results of many smaller trials. It was game-changing. Building on these initial findings, the CTSU team proposed to collect individual patient data from each trial and hold another meeting. This time it would take place in Washington DC, just before the US National Institutes of Health were due to bring out a statement on how best to treat early stage breast cancer.
Race against time
As well as looking at the benefits of tamoxifen and chemotherapy, the plan was to combine data from trials of radiotherapy. But obtaining and analysing data from over 100 trials in just 3 months turned out to be a challenging task, especially as the computers at the time took half an hour just to do one analysis. In the end, it came right down to the wire.
“The night before we flew out for the meeting we were still working on the analysis,” recalls Gray. “We’d only just got the first results and they looked amazing, but it still wasn’t finished. We were even thinking about cancelling our morning flight and going on Concorde later that day! We were working on the plane, working on our presentation all night. We had the programmer back in Oxford reading us results over the phone as we were still doing more analyses.”
It’s not often that a scientific conference leads to high drama, but Gray laughs as he remembers the hectic final moments leading up to the presentation of their findings. “It could have been a film script – our data handouts were being printed across town and we had to hire a car and rush to pick them up, weaving in and out of the traffic with these huge boxes of paper. Just as everyone came out for the coffee break we slammed them down, literally hot off the press!”
Fortunately, the results made all the stress worth it. For example, combining all the studies showed even more clearly that chemotherapy and tamoxifen improved survival, slashing the number of people whose cancers came back within the first year. The CTSU team’s analyses were so convincing that the conference agreed they should form the basis of their recommendations for how best to treat patients with early breast cancer. The full results of the meta-analysis were finally published in the New England Journal of Medicine in 1988, and a decades-long international collaboration – the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) – was firmly established.
Building on early success
Co-ordinated by a 20-strong team here at the NDPH, the EBCTCG continues to gather data from clinical trials in early stage breast cancer all over the world, analysing it every five years. Today, they combine results from more than 500 research groups, encompassing hundreds of thousands of patients.
Many new and often unexpected results have emerged since those first meetings. For example, the collaboration has revealed that taking tamoxifen for five years rather than one or two is better for preventing cancer recurrence. By looking at the trials of longer treatment, the team showed that tamoxifen could benefit women under 50 as well as older women. More effective types of chemotherapy have also been identified. Long-term analysis has shown that the survival benefits from both tamoxifen and chemotherapy are larger than first thought, and that radiotherapy after surgery also helps to improve survival.
By taking such a ‘helicopter view’ of clinical trials, the EBCTCG can identify gaps in the research and suggest new trials to fill them. For example, the ATLAS and aTTom trials, developed by CTSU, showed that 10 years of tamoxifen was even better than 5 years. The project has incorporated trials of new treatments as they’ve become available, including hormone therapies known as aromatase inhibitors, as well as more modern combinations of chemotherapy drugs and precision radiotherapy techniques. They’re looking at newer drugs including Herceptin (trastuzumab), and investigating developments in personalised medicine that might allow tailoring treatment according to the genetic makeup of each patient’s tumour. And they’re also turning their attention to breast cancer screening, as well as identifying the best way to treat ductal carcinoma in situ (DCIS), a very early type of tumour.
Importantly, because so many people are now surviving breast cancer for many years, the EBCTCG researchers are also looking at long-term side effects of these life-saving treatments. For example, some types of hormone therapies can increase the risk of bone thinning and fractures, or raise the chances of developing womb cancer. And because radiotherapy for breast cancer inevitably ends up with some exposure to the chest, it can lead to heart disease. Understanding the true scale of side effects across many clinical trials builds a more accurate picture of the risks and benefits of each type of treatment, helping doctors to make the best decision for each patient.
The long game
Since those early meetings in the 1980s the EBCTCG has identified many treatments that each bring relatively small improvements yet add up to a major difference in survival. And because breast cancer is so common and affects so many people around the world, this translates into many tens of thousands of lives saved each year. Here in the UK, breast cancer death rates for women in middle age have halved since the 1980s. Similar reductions have been seen in other countries, and we expect this progress to continue.
Although the statistics speak to the power of this approach – gathering data from many trials and crunching it all together - there are no quick answers. It can take years to run trials, follow up patients, and gather and analyse the results. It can also be frustrating dealing with researchers or pharmaceutical companies who sometimes are reluctant to give up their data for meta-analysis. But this is the only way we can know for sure how best to treat people.
As Professor Gray says, “It takes a long time and it can be very tough. You’ve got to get your hands dirty in the data but it does make a difference. I was on holiday in Cuba recently and the woman running the guesthouse I was staying in told me she was taking tamoxifen for breast cancer and had recently been told to take it for ten years instead of five. That’s a direct result of our work, and it’s very nice to do something that’s been useful to so many people all round the world.”