
In recent years, we’ve gotten much better at fighting cancer — if it’s detected early. But some cancers can sneak up and grow for years without showing any symptoms. This is where the new blood test comes in.
Researchers were surprised to see they could detect signs of cancerous tumors in the bloodstream so much earlier.
“Three years earlier provides time for intervention. The tumors are likely to be much less advanced and more likely to be curable,” says lead study author Yuxuan Wang, an assistant professor of oncology at the Johns Hopkins University School of Medicine.
A blood test for cancer
The research comes from a team at Johns Hopkins, which analyzed blood samples collected as part of a long-running cardiovascular study called ARIC (Atherosclerosis Risk in Communities). This study began in the late 1980s to track heart health in thousands of Americans. But the blood samples it gathered over decades are now revealing insights far beyond the heart.
Using ultra-sensitive DNA sequencing, researchers focused on samples from 26 people who developed cancer within six months of giving blood, and 26 matched individuals who did not. They used a test known as a multicancer early detection (MCED) assay, designed to search for tiny fragments of mutated DNA that tumors release into the bloodstream.
At the first time point — just before diagnosis — the MCED test flagged cancer in 8 of the 26 people who got it. That’s not surprising; it’s pretty much what researchers expected. What stunned researchers came next. For six of those eight people, earlier blood samples were also available, drawn more than three years before the cancer diagnosis. In four of those six cases, the team found the same tumor mutations already present.
“These results demonstrate that it is possible to detect circulating tumor DNA more than three years prior to clinical diagnosis, and provide benchmark sensitivities required for this purpose,” the study authors write.
How the test works
Tumors, even when minuscule, shed DNA into the bloodstream. Scientists call this cell-free tumor DNA, or ctDNA. Detecting it is a bit like looking for a needle in a haystack, only the haystack is the human genome and the needle is a single mutated fragment out of billions.
Thanks to improvements in sequencing technology, the search is becoming more feasible. But the key challenge has always been sensitivity. It’s a big haystack. So, how can we detect cancer when there’s barely any anomalous DNA to find? This is the billion-dollar question.
Multicancer early detection tests are a new generation of diagnostics that’s even more ambitious. Rather than screening for one specific cancer — like a mammogram for breast cancer — they look for markers of dozens of cancers all at once, using just a small blood sample. The rate isn’t flawless, but it’s definitely promising because it scans for multiple types of cancer.
“This study shows the promise of MCED tests in detecting cancers very early, and sets the benchmark sensitivities required for their success,” says Bert Vogelstein, M.D., Clayton Professor of Oncology, co-director of the Ludwig Center at Johns Hopkins and a senior author on the study.
Could this become common practice?
This isn’t the only promising blood test for cancer. A study from April 2025 describes an electric-field molecular fingerprinting (EMF) test that uses laser pulses and AI to analyze blood plasma, reporting over 80% accuracy for lung cancer; another breakthrough comes from the University of Edinburgh, where researchers combined spectroscopy with machine learning to screen for early breast cancer using blood samples. Meanwhile, for pancreatic cancer — a notoriously silent killer — researchers at Oregon Health and Science University developed a rapid, ultra‑low‑cost blood test that can also achieve over 80% accuracy. However, integrating them into large-scale studies that screen for multiple types of cancer remains challenging.
In a best-case future, a person might go for an annual blood test and receive an early alert about a developing cancer — long before pain, fatigue, or visible symptoms drive them to the doctor.
These tests are not yet standard in medical practice. Several companies and research groups are racing to validate them in large clinical trials. The Johns Hopkins study doesn’t prove that these tests are ready for broad use, but it strongly suggests they can work years earlier than expected.
“Detecting cancers years before their clinical diagnosis could help provide management with a more favorable outcome,” adds Nickolas Papadopoulos, Ph.D., professor of oncology, Ludwig Center investigator and senior author of the study. “Of course, we need to determine the appropriate clinical follow-up after a positive test for such cancers.”
Our best new weapon against cancer could be scouting. If confirmed at scale, MCED tests could become part of routine checkups — detecting silent killers years before they strike. All of it from a single tube of blood.
The study was published in Cancer Discovery.