Why look for many cancers with one blood test?
In the US, recommended screening covers only a few cancers: breast, colorectal, cervical, and lung cancer in some smokers. According to the authors, these cancers account for less than 25% of US cancer deaths. Many others, like pancreatic cancer, have no screening test at all and are often found late, when cure is less likely.
Tumors release small pieces of DNA into the blood, called cell-free DNA. The test in this study reads chemical tags on that DNA, called methylation, which look different from one cancer to another. A computer algorithm then decides whether there is a cancer signal and names the organ where it most likely started.
What did the researchers do?
Between December 2019 and December 2020, 6,662 adults aged 50 or older with no signs of cancer gave a blood sample at seven US health networks. A little over half had a higher risk: a smoking history, an inherited risk, or a cancer treated more than three years earlier. Results went to their doctor, who chose freely which tests to order after a signal. Everyone was followed for one year.
The study was funded by GRAIL, the company that developed the test and now sells the Galleri test. The version used here was an earlier one: a refined version was built during the study but was not used for medical decisions.
The results: 92 signals, 35 confirmed cancers
Out of 6,621 usable results, the test found a cancer signal in 92 people, or 1.4%. Of those, 35 really had cancer (38%) and 57 did not: these are false positives. When cancer was real, the test named the right organ as its first or second guess in 97% of cases. Nearly three quarters of the cancers found had no recommended screening test, and almost half of the new cancers were caught at an early stage.
After a signal, it took a median of 79 days to reach an answer (the median is the middle value): 57 days when cancer was real and 162 days for false alarms, often with several scans. Most importantly, the test missed many cancers: of 121 people diagnosed during the year, 86 (71%) had no signal.
Key figures
- 1.4%of participants got a "cancer signal detected" result (92 of 6,621)
- 38%of people with a signal actually had cancer (35 of 92)
- 79 daysmedian time to confirm or rule out cancer after a signal
- 71%of people diagnosed with cancer within the year had no signal on the test (86 of 121)
A multi-cancer blood test? Talk it through first, with Mima
This test is not a proven screening tool: it can raise a false alarm or miss a cancer, and it does not replace your regular screenings. If it is offered to you, the most important step is talking it through with your doctor before you take it.
Mima helps you get ready for that visit by gathering your questions, such as: "What would we do if the result came back positive?" She also keeps track of your regular screenings and their dates.
What this means for patients
For now, this test does not replace regular screening. In the study, every participant, even with no signal, was told to keep up with recommended screening. A negative result does not mean "no cancer."
The study also shows what follows a positive result: weeks or even months of tests, and not always a cancer at the end. Before taking this kind of test, a good question for your doctor is: what would we do if the result came back positive?
What are the limits of this study?
This was a feasibility study with no comparison group, so it cannot say whether the test lowers cancer deaths. Participants were mostly white (91.7%), college educated, and already well screened, and the sponsor paid for all follow-up tests. Follow-up lasted only one year, at the height of the COVID-19 pandemic.
The authors also point to the risk of overdiagnosis, meaning finding cancers that would never have caused harm. They are waiting for larger studies, including the UK's NHS-Galleri trial, which randomized 140,000 adults.
Frequently asked questions
What is the Galleri test?
It is a blood test from GRAIL that looks for signs of many cancers at once in cell-free DNA. The PATHFINDER study tested an earlier version of it in adults aged 50 and older.
How accurate is the Galleri multi-cancer blood test?
In PATHFINDER, 38% of positive results turned out to be cancer, and 99.1% of people without cancer correctly tested negative. But the test missed 86 of the 121 cancers diagnosed during the year.
What happens if a multi-cancer blood test is positive?
Your doctor orders follow-up tests, such as blood work and scans, to confirm or rule out cancer. In this study, 62% of signals were false alarms.
Does it replace mammograms or colonoscopy?
No. Participants were told to keep up with regular screening, and some of the cancers the test missed were found by those standard screenings.
The original study
Schrag D, Beer TM, McDonnell CH, et al. “Blood-based tests for multicancer early detection (PATHFINDER): a prospective cohort study” The Lancet, October 1, 2023. DOI : 10.1016/S0140-6736(23)01700-2
Read the full study in The LancetThis article summarizes a scientific study and is not medical advice.