Grail’s multicancer detection test could soon become the first blood test of its kind to secure Food and Drug Administration approval.
The Galleri test screens for many types of cancer from one blood draw. It works by analyzing DNA fragments for a shared cancer signal and then predicts its origin. A positive signal needs follow-up testing such as imaging or biopsy to confirm cancer.
An FDA advisory committee reviewed the clinical evidence for Galleri last month and endorsed its effectiveness by a 6-4 margin, reflecting caution among panelists about what the test can and cannot do to improve early detection of cancer. Grail, a publicly traded company that was spun out of Illumina, says Galleri should be used in addition to guideline-recommended single-cancer screenings. The test is intended for people age 50 and older.
FDA approval could expand insurance coverage for the test, which lists for $949. A new law paves the way for Medicare coverage starting in 2028.
As the medical world awaits the FDA’s decision, MedTech Dive asked five cancer specialists about Galleri’s potential to find cancer earlier, the test’s limitations, and their expectations for the use of multicancer testing. Their answers follow.
Responses have been edited for length and clarity.
Dr. Elizabeth O’Donnell, director of the Multi-Cancer Early Detection Clinic at Dana-Farber Cancer Institute, associate professor of medicine at Harvard Medical School and an author of the PATHFINDER 1 clinical trial:
The first question about Grail’s test is, how good is it at identifying abnormal from normal, and when they find an abnormal signal, how often is it actually cancer? That’s what’s called a positive predictive value. In the NHS-Galleri study, the positive predictive value was just over 50%, and the PATHFINDER 2 study was 62%.
What is the sensitivity when you do a clinical trial that has capped at 12 months? How well are you finding cancers, and what are you missing? The episode sensitivity is still relatively low. It’s only about 30% to 40%. The positive predictive value is pretty good at 60%, but the episode sensitivity still suggests that we’re missing cancers.
One of the concerns that people have about introducing a novel screening test is that you’re going to have all of these false positives to work up. It’s important to understand in the clinical trial that 0.9% of those taking the test, less than 1%, had positive results. That’s a very low number.
How many of these are going to be false positives? You want to make sure that the specificity is very high, which it is in these studies. You don’t want to create a lot of false positives, both from a patient anxiety standpoint, but also from a workload standpoint from primary care doctors who are trying to do the due diligence around these.
There’s critique of the design of the NHS-Galleri study, which is unfortunate because at the end of the day, they tried to do what we would do for a drug, a randomized, controlled, longitudinal study. But it’s hard to go first.
The science is iterative. It’s going to get better. It is getting better if you look at the differences between PATHFINDER 1 and PATHFINDER 2, but it takes a lot of work to make things better.
I really see these as complementary to existing screenings. If 70% of cancers don’t have screening tests, then really we should be looking at how well it’s doing there, and that’s what a lot of these studies show. They did pick up a lot more cancers.
The long-term data, in terms of how that affects mortality and patient outcomes, is still pending.
I focus on multi-cancer detection tests, but there are also many single-cancer detection tests that are being studied. This is an entire field in medicine now that’s evolving. This is only going to expand, in my opinion.
Dr. Parijatham Thomas, associate professor of clinical cancer prevention and breast medical oncology and medical director of the Cancer Prevention Center at The University of Texas MD Anderson Cancer Center:
Based off the data we have, this does not replace other cancer screenings.
In June, at the American Society of Clinical Oncology annual conference, they presented the NHS-Galleri study. This is the largest randomized trial with one of these multicancer detection tests that we have currently. They looked to see, could they reduce the incidence of late-stage cancers, the combination of Stage 3 and 4 cancers, after three annual screening rounds. The key thing here is they did not meet their primary endpoint. They did not meet that goal that they were trying to reduce a significant number. They did see there was a trend toward a decrease, but it was not statistically significant.
How well does this test pick up a cancer? I think it is good at picking certain cancers. It may not be good at picking up all cancers. That’s a very important fact to let people know, that it’s not equally good across all cancers.
We could increase our screening for cancers that we don’t have a way to screen for. But it’s not necessarily going to be the best for everything.
For example, its actual sensitivity for breast cancer and prostate cancer is very low. It is not very good at all to detect those cancers. So it’s still very important you go and have your annual screening mammogram or discuss PSA screening with your primary care doctor for men. In NHS-Galleri, they’re talking about reducing Stage 3 and 4 cancers in a pre-specified group of 12 primary cancer types. Of note, prostate and breast are not in those 12.
Our current standard of care screenings have very high sensitivity. Our screening mammogram, our colonoscopy and Pap smears have very high sensitivity. That’s to give a comparison.
We need to continue our screening, and that’s really something I don’t want people to misunderstand.
Dr. Charles Atwood, head of the lung cancer screening program at VA Pittsburgh Healthcare System, associate professor of medicine, University of Pittsburgh, and lead researcher for the REFLECTION study:
If you look at the PATHFINDER 2 study, they found that the positive predictive value had increased from the previous studies that Grail had performed. With PATHFINDER 2, it bumped up. That’s very encouraging. What that says is that if you have a positive Galleri test, the likelihood that it really represents a cancer and not a false positive has gotten better.
The trends toward better, in a sense, accuracy of the test are improving. But they’re not there yet. So people who use the test, doctors who prescribe it or patients who use it, just need to bear that in mind.
The world does not have great cancer early detection. Even if you can’t detect a cancer at its earliest Stage 1 phase, catching it before it’s metastasized and in Stage 4 is a real benefit to the patients who will receive therapy for that cancer. In other words, a Stage 2 or 3 cancer is way better to treat than a Stage 4 cancer.
Every organ’s cancers are unique. I don’t want to overgeneralize, but I do think it will help humanity to be able to detect cancer early. More health systems, now that it’s going to receive FDA approval, will begin recognizing some value in it. The marketplace will sort it out.
It’s a somewhat expensive test, but it’s not that much more expensive than if you got a colonoscopy screening for colon cancer or a chest CT for lung cancer detection, or other types of cancer detection.
Dr. Arthur Burnett, professor of urology and oncology at Johns Hopkins University and director of the Male Consultation Clinic at Johns Hopkins Hospital:
There are a lot of implications here, from scientific to practical.
I do think it’s potentially widely used. If it’s widely used, it may tell us something’s awry with the person’s body or the potential malignancy.
At the same time, I would just offer some cautionary thoughts. It’s got to meet certain rigor with regard to its accuracy. Otherwise, it could lead to potential risk for overdiagnosis, additional cost for a lot of tests that could be both financially concerning as well as create a lot of anxiety if misused.
Much like PSA, which screens for the prostate, it tells us something is not right, but it can’t in and of itself be immediately diagnostic for prostate cancer. That would be an erroneous thought.
To what extent is it accurately driving us clinicians to determine what is a true condition versus something that may be pre-malignancy? Is it picking up things that we’re not sure exactly what their true threat would be? These kinds of things have to be well defined in terms of where it should be used. Should it be done with any other suspicion of health changes that would prompt it? Should it be used widely with everybody of all ages? We’d have to define this.
The other thing is that much like the storyline that we have surrounding inherited genetic susceptibilities and changes, does that put you in a category not only of worry, but also in a category where you might not be insurable? If you get a positive test, then you’ve got to get a workup. If we can’t find something necessarily, is it going in your record? There’s insurance challenges if that goes on the record, just like having a certain kind of gene susceptibility that we’re still trying to figure out the threat of it.
Dr. Karthik Giridhar, oncologist at Mayo Clinic and senior author of the PATHFINDER 2 study:
Three numbers encourage me. First, the false-alarm rate. Only about 1 in 280 people without cancer got a positive result. Specificity was 99.6%, and it held up in the U.K. trial too. Second, when the test was positive, it was right about 60% of the time. That’s far higher than most single-cancer screening tests. Third, it told doctors where to look. It predicted the cancer’s location correctly more than 90% of the time. That kept most follow-up testing to scans and blood work. Only 0.6% of everyone tested ended up with an invasive procedure after a positive result.
The test is best at finding aggressive cancers that shed a lot of DNA into the blood. In our study it caught most liver and bile duct cancers, colorectal cancers, pancreatic and gallbladder cancers, and head and neck cancers, generally 75% to 90% of those diagnosed within the year. It also did well with lymphomas. It’s weakest with slow-growing cancers. It caught fewer than 1 in 10 prostate cancers and fewer than 1 in 4 breast cancers. It’s best at the cancers we’re worst at finding, and worst at the cancers we’re already good at finding. That’s why it works as an addition to screening, not a replacement.
Adoption is already happening. Grail has sold over 595,000 commercial Galleri tests through June 30, 2026. FDA approval would likely speed that up. But broad use depends on insurance coverage, and coverage will depend on stronger evidence.
The biggest benefit is finding deadly cancers like pancreatic and ovarian cancer before symptoms, when more patients can still be treated with curative intent.
The biggest risk is false reassurance. A negative result does not rule out cancer, and we don’t know if earlier detection helps save lives. If people skip their mammograms or colonoscopies because of a negative blood test, we could do real harm.
These tests will grow and be further adopted. The need is for clear communication and ongoing evidence generation.
