Blood tests promise to detect cancer early. How well do they work?

Washington Post

15-07-2025

For years, it has been one of the most tantalizing ideas in medicine: One day, a simple blood test will detect nascent cancers — allowing doctors to intervene when tumors may be easier to treat.
A study published this month in Cancer Discovery found that three years before a handful of cancer patients were diagnosed, some promising tests could detect trace amounts of cancer genetic material floating in their blood, shed by tumors the people didn't yet know existed.
Start-up companies have developed different versions of these 'multi-cancer early detection' tests designed to flag the errant genetic signatures of more than a dozen cancers. Some doctors and medical practices are already using them, and the field has been propelled forward by innovation and hype.
Turning these innovative technologies into usable medical tools for the broader population has been complex. But a study recruiting subjects this summer, funded by the National Cancer Institute, seeks to begin to do that by answering basic questions about how the tests perform in comparison to more conventional cancer screening.
A screening test given to millions of healthy people has to be accurate to avoid costly and anxiety-inducing follow-ups. And for it to become part of routine medical care, there needs to be evidence that the test can reliably find cancers — perhaps far earlier than routine methods. Most importantly, scientists want to see clear evidence that earlier detection leads to longer lives.
To begin to fill this knowledge gap, scientists at nine sites around the country are recruiting up to 24,000 people between the ages of 45 and 75 for the Vanguard Study. Participants will be randomly assigned to receive either regular screening or one of two multi-cancer detection tests. All of the groups will be told to continue with regular cancer screenings and will be followed over two years.
The blood tests used in the study will screen for a variety of cancers, including many for which there aren't more traditional screening tests: bladder, breast, colorectal, esophageal, stomach, liver, lung, ovarian, pancreatic and prostate.
Researchers, guided by what they learn from this first trial, hope to launch a much larger trial, with about 150,000 people, to definitively establish the effectiveness of these tests in preventing deaths.
Bert Vogelstein has been working on these ideas for 30 years at Johns Hopkins Medicine and is a scientific founder of Exact Sciences, a company working to make blood tests for cancer screenings, but is not involved in the new trial. He said the field is still in its adolescence.
'But adolescents can do a lot of great things, that's clear in athletics and it's clear in science,' Vogelstein said. 'I think what NCI is doing is great. And I applaud those efforts, because tests can do good or harm.'
The benefit of early cancer detection is intuitive. People could roll up their sleeves and get a blood test to flag 20 cancers, allowing physicians to give early treatments when they have no symptoms. That could dramatically reduce the need for people to go through years of illness and advanced cancer treatments.
'That dream, to me, is decades off, but we're starting on it now — which is why it's so exciting,' said Scott Ramsey, a physician and researcher at the Fred Hutchinson Cancer Center and a principal investigator with the Cancer Screening Research Network, which will be running the Vanguard Study. 'We're not talking about treating people who are sick; we're talking about preventing people of dying in cancer in the first place.'
The harms of screening are not always as clear to the public, but they show the trade-offs inherent when providing a medical intervention to healthy people.
Ramsey pointed to the history around prostate specific antigen (PSA) tests, which identified many prostate cancers early — some of which would have never shown symptoms or caused death. That meant that many men were aggressively treated, with surgery and radiation that caused them to be incontinent or impotent for years or decades.
The recommendations on how to test for prostate cancer changed as these trends became clear over decades. Ramsey said that now is the right moment to be subjecting these cancer blood tests to scrutiny. The potential for the tests is enormous, but there are fundamental questions that need to be answered about how and when they should be used — and, perhaps, when they shouldn't.
The recent study in Cancer Discovery illustrated one of the major problems facing the field. To detect cancers early, the tests need a big boost in sensitivity, because they need to find fragments of cancer tumor DNA that are scarce.
That study found that it was possible to detect cancer three or more years than a traditional diagnosis. 'It can be done — and that's an important part of the article, much earlier than had ever been shown before,' said Vogelstein, part of the team behind the study.
But to detect those trace amounts of genetic material, the sensitivity of the test had to be roughly 50 times higher than what it is in tests today, he said — and would likely cost a few thousand dollars.
Ramsey said current tests can be $900 or more. Health insurance carriers typically won't pay for them yet, because it isn't clear what their benefits and harms are.
'The bottom line is we don't know whether when taking one of these tests will give you an opportunity to live longer without cancer — or longer overall,' Ramsey said. 'There's never been a study that established that.'