Can a Blood Test Really Screen for 50+ Cancers?
A blood test that screens for 50+ cancers sounds incredible. Here’s what the data actually shows.
Prefer to listen to a human conversation?
A patient came in recently having already ordered a multi-cancer blood test through another clinician. She brought in the result — clean, unambiguous, two words: cancer signal not detected.
She was relieved. And honestly, the simplicity of it was understandably appealing. No prep, no procedures, no waiting rooms, no ambiguous imaging findings with follow-up recommended in six months. Just a blood draw, a clear answer, and a sense that something had been done.
This is the promise of multi-cancer early detection blood tests, or MCEDs — a category that includes products like Galleri ($949) and CancerGuard ($689), and one that is expanding fast. You may have seen them advertised online, heard about them from a friend, or been asked about them by a patient who found an urgent care clinic with a sign in the window that just says: “Blood test for 50 cancers. Get it done today.”
Underneath that marketing is a genuinely compelling scientific idea, and underneath that idea are some numbers that deserve a real conversation. So let’s have it.
The Claim
If you spend time on social media — or drive past certain urgent care clinics — you’ve probably encountered this pitch: a single blood draw can now screen for 50 or more cancers simultaneously. These multi-cancer early detection tests promise to find tumors before they cause symptoms, at the stages when they’re most treatable, including cancers for which no standard screening currently exists. The marketing language is evocative. CancerGuard’s website declares: “Your cells have a story. We’re listening.” Results, they promise, will be “negative or positive. That’s it.” No jargon. Just clarity.
The appeal is obvious, especially for anyone who has sat through an indeterminate mammogram result or tried to make sense of what “we’d like to take another look” actually means. The idea that one clean yes-or-no answer could replace all of that is deeply, emotionally attractive.
And these tests are not fringe technology. They are actively being studied by major health systems. One of them is currently the subject of the first randomized controlled trial of its kind, funded by the NHS. They are real. They are scientifically grounded. And they are not ready to be used the way they are being marketed.
Why It’s Going Viral
The pandemic changed how a lot of people relate to their health. There was a heightened awareness of mortality that didn’t fully go away when the acute crisis did. And in the years since, social media feeds have been full of stories about young, seemingly healthy people receiving devastating cancer diagnoses. Celebrities. Friends of friends. The colleague who ran marathons. These stories aren’t wrong to share, but they’re not representative, and they do something particular to our sense of risk: they make the statistically unlikely feel close and personal and inevitable.
At the same time, there’s a growing cultural narrative — one that has real traction for real reasons — that what you receive through the traditional healthcare system is barely the minimum. That standard screenings are a starting point, not a finish line. That proactive people do more. For patients who have felt dismissed, rushed, or under-served, this narrative lands hard. And for anyone who has watched a loved one diagnosed late with a cancer that might have been caught earlier, it is impossible to dismiss.
The companies building these products understand this. “Is CancerGuard right for you?” their website asks, and then lists: “You are proactive about your health. You have a family history of cancer. You have increased risk factors such as smoking, alcohol use, or obesity.” That is not a list of specific criteria. That is a description of most adults who are paying attention to their health at all.
And the contrast with standard cancer screening is striking. A colonoscopy requires a day off work, a prep that no one enjoys, a procedure where you are mildly sedated while someone inserts a camera into your colon, and then a wait for results that are often not perfectly clear. A blood draw takes ten minutes. The emotional math is not hard to do.
What the Science Shows
The science behind these tests is real and interesting. All cells — including tumor cells — shed fragments of their DNA into the bloodstream as they turn over. These fragments are called cell-free DNA, and when they come from a tumor, they’re called circulating tumor DNA. What MCED tests do is look at the methylation patterns on that cell-free DNA — epigenetic modifications that serve as a kind of fingerprint for where the DNA originated and whether it came from a cancer cell. AI models trained on thousands of these patterns can then flag abnormal signals and estimate which tissue or organ the signal may have come from. It is genuinely sophisticated technology.
But here is where the numbers get complicated.
The largest validation study on Galleri, published in Annals of Oncology, enrolled over 5,000 participants with known cancers and without. The overall sensitivity — meaning the test’s ability to correctly identify true cancer cases — was 51.5%. That sounds reasonable until you look at where that number comes from. For stage I cancers (the early ones these tests are most hoping to catch), sensitivity dropped to 16.8%. For stage IV cancers (metastatic, often already clinically apparent), sensitivity was 90.1%. In other words: this test is best at finding the cancers that are already the hardest to miss, and least good at finding the ones it claims to catch.
Specificity was high at 99.5%, suggesting a false positive rate of only 0.5%. But here is where a critical statistical nuance applies. Because cancer is relatively rare in the general population, even a high specificity does not protect against a meaningful number of false positives in absolute terms. The positive predictive value — the probability that a “cancer signal detected” result actually reflects a true cancer — was 44.4%. That means that in the validation population, more than half of people who received a positive result did not have cancer.
The PATHFINDER study, published in The Lancet in 2023, looked at over 6,000 adults 50 and older across seven U.S. health systems. Cancer signal was detected in 1.4% of participants. Of those, only 38% were ultimately diagnosed with a cancer within the 12-month follow-up period. The other 62% had no cancer diagnosis, but what they did have was an average of 79 days of diagnostic workup: labs, imaging, and in 49% of cases, at least one procedure. One patient underwent an inguinal orchiectomy — testicle removal — based on findings that turned out not to be cancer. During the 12-month study period, 122 cancers were diagnosed in the full cohort. Only 35 of those were detected by the MCED test. The rest were found through standard screening, incidental findings, or clinical symptoms.
There is also an NHS-funded randomized controlled trial underway — the first RCT on an MCED test, with 142,000 participants tested annually for three years. Preliminary data, shared in a brief PDF circulated by the company, shows a favorable trend toward fewer late-stage diagnoses but no statistically significant reduction in the combined rate of stage III and IV cancers. The primary endpoint has not been met. No published data yet demonstrates that performing this screening in any population reduces the risk of death from cancer.
Clinical Nuance
None of this means these tests are a scam. The technology is real, the studies are ongoing, and the landscape is evolving rapidly. What the data shows today may look different in 2027 or 2030. We are at an early and genuinely uncertain moment.
But here is what we want patients to understand: this test cannot replace standard cancer screening. The Galleri website says so clearly in its own fine print. A breast cancer sensitivity of 30.5% does not supplement a mammogram’s 70-90% sensitivity — it does not come close. A normal result does not clear you. A positive result does not diagnose you. And the workup that follows a positive result can be extensive, expensive, emotionally taxing, and in some cases, physically risky.
When someone brings this up in a clinical conversation, it is worth slowing down to ask what is really driving the question. Sometimes it is health anxiety. Sometimes it is a family history that hasn’t been fully explored — and that might qualify the patient for genetic testing that could actually change their management. Sometimes it is a sense that they haven’t been getting enough from the healthcare system, which may or may not be true, but is worth exploring. Sometimes patients who are overdue on a colonoscopy or a mammogram are asking about this instead, and that shift in attention deserves a gentle redirect.
There are specific clinical scenarios where this type of testing becomes a more reasonable part of the conversation: a patient with a strong family history, negative genetic testing, elevated pretest probability for a specific cancer, or a situation where conventional workup would involve meaningful procedural risk. In those cases, shared decision-making might include this as one option. But for the general population, the American Cancer Society’s guidance is right: this should involve explicit informed consent about high specificity but limited early-stage sensitivity, the potential for false positives requiring extensive workup, the absence of mortality data, and the out-of-pocket cost.
And if someone brings it up and ultimately decides to move forward after a full conversation, that is a reasonable outcome. People have the right to spend their money on optional tests. What they deserve is an honest account of what those tests can and cannot do.
The Antidote
For Patients
(Gentle reminders and affirmations)
Wanting to know if I have cancer is not irrational. It is human.
The fear I feel when I hear about young, healthy people getting cancer diagnoses makes sense. Those stories are real — and they are not representative of what is most likely to happen to me.
A blood test that screens for cancer is not the same as a blood test that rules out cancer.
If this test says “cancer signal detected,” that does not mean I have cancer. It means more investigation is needed.
If this test says “cancer signal not detected,” that does not mean I am free of cancer. It has significant limitations in detecting early-stage disease.
This test does not replace my mammogram, my colonoscopy, my Pap smear, or other recommended screenings.
I deserve a real conversation about what this test can and cannot do — not just a confident-looking result.
There may already be screenings or genetic testing I qualify for that I haven’t explored yet. That conversation is worth having with my doctor.
For Clinicians
(Language you can borrow, tweak, or make your own)
“I’m really glad you brought this up. These tests are real technology — not fringe — and they’re actively being studied. Let’s talk about what the data shows.”
“The sensitivity for early-stage cancers — which is what we’re most hoping to catch — is about 17% for stage I disease. That’s meaningfully lower than standard screening tools for many cancer types.”
“If this test comes back positive, it doesn’t diagnose you with cancer. It triggers a workup. In studies, more than half of positive results didn’t end up being cancer. But that workup — labs, imaging, sometimes procedures — takes time, money, and often comes with real anxiety.”
“This is a test you can add on top of your standard screening, but it cannot replace it.”
“Before we talk about this, I want to make sure we’ve covered everything you already qualify for — are you up to date on your colonoscopy and mammogram? And based on your family history, I want to ask about whether you might qualify for genetic testing.”
“I’m not opposed to this if you’ve thought it through and understand the limitations. What I want for you is a decision you’ve made with full information.”