This is a study of a multicancer early detection test which is a blood test that is looking for essentially all types of cancer. It’s going to be used in screening, so testing people who don’t have symptoms of cancer, who don’t think they’ve got cancer, to see if we can find cancers early.
In 2020 we first started talking about setting up such a trial and we recruited 142,000 people and we randomised them into two groups. In one group we collected bloods and analysed those bloods and if it was positive those people were referred on for diagnostic investigation. In the other group we collected bloods but we just stored them. People were blinded as to which group they were in, unless they had a positive test having been screened they didn’t know whether they had a negative test or had just been in the control arm of the study.
We wanted to look at whether we could see a reduction in the number of people who got an advanced stage cancer, stage III or stage IV cancer. Traditionally, cancer screening studies have always been done, or virtually always been done, looking for a reduction in cancer mortality. But that takes many years and we felt that if this was working it would be such a paradigm change to the way we do cancer screening that we needed to find out as soon as possible. So we looked at what had happened in the past and almost always if you see a reduction in advanced stage cancers, the trial goes on to show a reduction in mortality and if you don’t, you don’t. So that’s not true for every single trial but if you look at types of cancer screening as a whole it’s always true. So in breast screening there’s always a reduction in advanced stage and a reduction in mortality in the meta-analyses, other forms of screening don’t work and you see that it doesn’t work in either setting. So that was one of the reasons why we wanted to do this, to get the information as quickly as possible about the trial.
Could you outline the methodology?
It’s a randomised controlled trial. We recruited people in mobile units which we took around the country. So we were really taking the screening to the population rather than… And we invited people who were registered, so on a population register, for the National Health Service in England in eight parts of the country, so eight of 20 parts of the country. So really going around the whole country recruiting people from age 50 to 77 and making great efforts to make sure that we included people from all sections of socioeconomic status, educational background, ethnicity, males, females, different age groups within that. We got a little bit of balance. Then everyone had three blood tests, so if they were in the screening arm they were screened three times one year apart, so baseline, 12 months and 24 months.
Then what we’re presenting here is three years after the last person was recruited, we’ve got complete cancer registration, cancer incidence, on that group of people.
What did you find?
The main endpoint was to see a reduction in stage III and stage IV cancer and we didn’t achieve that, there was no reduction. But there were lots of secondary endpoints, all of which were achieved, the ones we can look at now, and a very coherent picture of what’s going on. So there was a reduction in stage IV cancers overall. When we look at what happened in each round of screening, when you screen, particularly when you first screen you expect to see a large increase in screen-detected cancers and if there are stage IV cancers there it’s too late to prevent them, they’re already there at the time you recruit the patients. So we saw an increase in stage IV cancers, stage III cancers, in the first part of the trial.
By the end of the first year there was a substantial increase in stage III and IV cancers combined and then in years 2 and 3 we see a reduction in stage III and stage IV cancers. So the effect of the screening is as we would expect – you find a lot of things and it’s not soon enough to see a reduction in those band events, as it were, within the first year but we start seeing that greater and greater.
One of the exciting findings that wasn’t a specified secondary endpoint is that we saw a reduction of 25% in the number of people who presented with cancer through the Emergency Room. So it’s well known that people who first present and are diagnosed with cancer through Accident & Emergency, Emergency Room, have much worse prognoses than people who have cancers diagnosed through a referral to secondary care in a more calm and without going through a crisis. So a 25% reduction in those people is also great news.
What impact could these findings have?
It suggests to me that multicancer screening using liquid biopsies, and this is looking at the methylation of cell-free DNA within the blood, really does have the prospect of changing the paradigm for how we manage cancer. So even if it’s quite a small reduction in cancer mortality, so it’s possible that this could lead to a 10% reduction in cancer mortality, which doesn’t sound great but that would be across all cancers. So in the population you’d be talking about thousands of cancer deaths per year avoided in a single country. Across Europe and North America that could be 50,000 in ten years, half a million deaths avoided. So it could be transformative. Because the primary endpoint wasn’t met I don’t suppose there’s going to be any big national screening programmes introduced in the next 12 months or so, we need to see further follow-up, what happens to these groups. But the fact that the tests do what they’re intended to do, they find cancers early in people with no symptoms, is very clear. Yes, it could be hugely transformative but it’s not going to happen in the next six months.
