What a multi-cancer blood test can and cannot tell you

For most of medical history, finding cancer early has meant looking for one cancer at a time: a mammogram for breast, a smear for cervical, a stool test for bowel. A multi-cancer early-detection test, or MCED, tries something different. From a single tube of blood, it looks for a signal that could come from any of dozens of cancers at once. The best known is the Galleri test. It is worth understanding what that actually means, because the idea is genuinely clever and the limits are genuinely real.

Cancers shed fragments of DNA into the bloodstream as their cells grow and die. The chemical tags sitting on that DNA, a layer of biology called methylation, look different in cancer cells than in healthy ones. An MCED test reads those patterns and asks two questions: is there a cancer signal here, and if so, where in the body is it most likely coming from. That second question matters, because a signal without a location would send you for every scan under the sun. Galleri predicts the origin correctly around nine times in ten, which turns a vague alarm into a focused search.

The strength of these tests is their reach and their precision about false alarms. They can look for cancers that have no screening programme at all, which is most of them. And when they flag a signal, they are right more often than not: in the largest trial, around half of people with a positive result were found to have cancer. A positive result is uncommon, and it is worth taking seriously.

The limit is on the other side. These tests miss more cancers than they find, especially early ones. In the same large trial, the test picked up roughly a third of the cancers that were diagnosed within a year of the blood draw. That is the number that matters most for how you read a clear result: it is reassuring, but it is not a clean bill of health. A negative MCED test does not mean there is no cancer, and it is not a reason to skip your mammogram, your bowel-screening kit or a visit to the doctor about a symptom that is bothering you.

There is also an open question about whether finding cancer this way helps people live longer. Detecting a signal is not the same as changing an outcome. The largest study did not meet its main target of cutting late-stage diagnoses over three years, and the data on survival is still some way off. The technology clearly detects cancer, sometimes early. Whether that translates into more years of life is not yet proven, and anyone offering the test should say so.

None of this makes an MCED test a bad idea. It makes it a considered one. For the right person, keeping up with standard screening and adding a wider look can be a sensible extra layer. For someone else, it can add anxiety and a cascade of follow-up tests without adding much. The way to tell the difference is a proper conversation before the blood is taken, which is exactly where a doctor comes in.

If you want to talk through whether a multi-cancer test makes sense for you, our doctors will give you the full picture.