Major NHS trial provides new evidence on multi-cancer blood test

The NHS-Galleri trial, involving more than 140,000 people in England, evaluated whether adding the Galleri multi-cancer early detection (MCED) blood test to usual NHS care could reduce the number of cancers diagnosed at a late stage.

The results, published in the New England Journal of Medicine and Nature Medicine, show that the trial did not meet its primary endpoint of reducing the combined number of stage III and IV cancers. However, fewer stage IV cancers were observed among people offered the blood test, and more cancers were diagnosed at earlier stages. Further follow-up is planned to understand how these findings develop over time. Additionally, over 50% of those testing positive were diagnosed with cancer. That compares with about 6% of people referred with “urgent suspected cancer” symptoms and 26% of women with a positive breast screening test in England.

The trial was run by The Cancer Research UK Cancer Prevention Trials Unit at Queen Mary University of London (CPTU) in partnership with colleagues from King’s College London and Guy’s and St Thomas’ NHS Foundation Trust. Queen Mary Professor Peter Sasieni, was one of the three scientific leads of the trial. Over 40 members of CPTU staff, past and present, have contributed to the delivery of this ambitious project.

Testing a new approach to cancer screening

MCED tests aim to detect a common signal associated with several different types of cancer from a single blood sample. The Galleri test looks for patterns in DNA circulating in the blood and, when it detects a cancer signal, predicts the part of the body where that signal is most likely to have originated.

The NHS-Galleri trial recruited 142,250 people aged 50 to 77, who were randomly assigned either to an intervention group, whose blood was tested using the MCED test, or a control group receiving usual NHS care. Participants were invited to provide blood samples over three annual screening rounds.

The New England Journal of Medicine paper reports the main trial outcome. After three rounds of screening, there was no significant difference between the groups in the incidence of stage III and IV cancers combined across 12 prespecified cancer types.

However, when stage IV cancers were considered separately (a pre-specified secondary outcome), their incidence was lower in the intervention group. This difference increased as the trial progressed because initially there is an excess of stage IV cancer diagnoses in the intervention group from undiagnosed stage IV cancers at entry into the trial. There were also more stage I and II cancers diagnosed among people in the intervention group. Additionally, there were fewer cancers that were diagnosed after presentation in Accident and Emergency. The researchers emphasise that longer follow-up is required to establish how these differences evolve and whether they ultimately translate into improved cancer outcomes.

Professor Peter Sasieni, Professor of Cancer Epidemiology at Queen Mary University of London, said:

“The NHS-Galleri trial is the first randomised controlled trial anywhere to evaluate a multi-cancer early detection blood test at this scale. Although we did not see a reduction in our primary endpoint of stage III and IV cancers combined, we observed fewer stage IV cancers among people offered the test, alongside an increase in cancers diagnosed at earlier stages.

“The companion study also gives us important evidence about how the test performs over repeated rounds of screening. Used in people without a clinical suspicion of cancer, the test detects far more early-stage cancers than routine NHS screening; and a positive test cannot be ignored – about 50% of participants with a positive test had cancer diagnosed. Longer follow-up will be crucial to understanding what these findings could mean for cancer outcomes and the potential future role of multi-cancer early detection tests.”

How well did the blood test perform?

A companion paper in Nature Medicine reports how the MCED test itself performed during the three annual screening rounds.

Around one in every 100 participants tested received a positive cancer-signal result in each round. Across the three rounds, 937 participants with a positive test were subsequently diagnosed with cancer. Depending on the screening round, around 46% to 58% of positive results were associated with a cancer diagnosis.

The test also had high specificity, of around 99.5% to 99.6% across the three screening rounds, meaning that false-positive results were uncommon – fewer than 1 in 200 tests resulted in a false-positive result. When cancer was detected, the test’s prediction of where in the body the cancer signal originated was also accurate in more than 90% of cases when the first two predicted sites were considered.

Taken together, the two papers provide a wealth of high-quality evidence both about the clinical impact of adding MCED testing to NHS care and about the performance of the test when deployed at large scale.

The NHS-Galleri trial was funded by GRAIL and conducted in collaboration with The Cancer Research UK Cancer Prevention Trials Unit at Queen Mary University of London and NHS partners. Data analysis for the main trial publication was conducted independently by statisticians at Queen Mary and verified by GRAIL.

The NHS-Galleri trial recruited over 142,000 people in just over 10 months making it the fastest recruiting clinical trial since the Salk polio vaccine trial in 1950. It is also almost unique in succeeding to recruit such a diverse and representative group of people from across England. If anything, the most deprived sections of society were over recruited. The trials unit ensured that they were available to answer participants questions and that is reflected in the numbers who returned to provide further blood samples at 12 and 24 months. Queen Mary is proud of its pivotal role in this ambitious and inclusive research.

Further analyses are planned, including longer-term follow-up of cancer outcomes. The NHS and other health organisations will consider the evidence from the trial as they assess the potential future role of multi-cancer early detection testing.