And not talk about the GPTs that we use because we're starting. All right. I'm Dave Westenberg. I wrote this with ChatGPT. I cover life science tools and diagnostics here. With us is GRAIL. Very happy to have the President, Josh Ofman, and the CFO, Aaron Freidin. Thank you a lot for being here, both. We'll jump right into Q&A. Grail is the oldest MCED company. Can you talk about some of the advantages of being kind of the first mover with data to market and, you know, fingers crossed for the approval?
Sure. Thank you. First of all, it's great to be here. Thank you for having us. Good to see everybody. So, you know, GRAIL was founded in 2015 with this big idea about really changing the whole paradigm about how we detect cancer. There's really, when you think about how we go about doing that today, there's really nothing acceptable about what we're doing today. Our current single cancer screening tests find about five, you know, 14% of all the cancers using those five single cancer tests. But about 80% of all the deaths are coming from cancers we're not looking for at all. So GRAIL set out to change that picture. And they did it with some very innovative technology that was validated early on and finally validated in 2019. And so we've been way ahead of the field.
We knew from the beginning that this would be a Herculean effort as it related to evidence generation. Screening is unlike diagnosing existing disease. The bar for market access, the bar for evidence is much, much higher. So we set out on a course to develop world-class levels of evidence, rigorous evidence from case-controlled studies, interventional trials, and randomized clinical trials. We wanted to seek FDA approval because we knew that that would be a huge, competitive advantage and set the bar very high for the whole field and also enable commercial payers to be able to provide coverage. We knew that we want to be in broad population screening. So we've conducted these very large studies. And that's given us a sizable lead and advantage. We can use the data we've collected to continually improve our AI, which is a huge part of our test.
It's machine learning-based. It's gotten us the ability to learn how to implement a test like this in actual clinical practice, to develop relationships with clinicians, and understand what patients and clinicians need from a test like this, and then from a regulatory and a policymaker perspective, what it's gonna take to change the whole paradigm to live in a world where we are detecting, you know, most cancers early, and our data really speak for themselves, and I think that's part of the lead. We've also been able to build, in advance of broad market access, so we've built a world-class laboratory. We've made some of those capital investments, in advance of having that broad market access, which we're expecting to attain.
Perfect. What are the potential outcomes for the FDA approval? In addition, you know, obviously there's yes and no. You got it. You didn't get it. But, you know, what are some of the potential labels that could come out of that?
It's a great question. I mean, we are, you know, we've been in a modular PMA submission now for several years with the FDA. We'll complete that submission. We've just narrowed the time range to the first quarter in 2026. And it'll be one of the largest data submissions of a diagnostic test that the FDA has ever seen before. It'll be over 100,000 subjects. It'll be an enormous data set that, by the way, nobody has ever seen before, of this magnitude. How it gets labeled is a very interesting question. I know the FDA's thinking about many, many things. The way our test works is that it detects a signal in blood of apparently methylated DNA and in genomic regions that are known to be informative for cancer. So these are hypermethylated, hypomethylated fragments of DNA around tumor promoter and suppressor genes.
We see this pattern that is almost never seen in people without cancer. That pattern in itself, when we recognize it, has a very high specificity for cancer, and in other words, a very low false positive rate, and generates an extremely high positive predictive value. That's on a cancer signal. Then we localize that cancer signal to different organs and regions of the body. So we might say cancer signal detected, predicted origin, pancreas or ovary or liver or stomach. The FDA has never created a label for a test like that before that looks at a shared signal. We have detected over 50 different types of cancer. We still group them, so it's actually well over 100. We group them. Our CSO categories are predicted categories. There are 19 of them.
So one of the labeling options will be around, you know, all the cancers that we find, you know, which is well over 50. Another option might be to do it around the cancer signal predictions. And another could be that it's just they find a shared cancer signal, and then we just have a list of all the cancers we've ever found, which is in the hundreds. So there are a lot of ways of labeling this. And I know the FDA is thinking hard about what to include in the label, what not to include in the label. There's another option where there's a page one of tumor types that we found so many of them that the FDA feels comfortable labeling, and then a page two of all the other tumor types that have less evidence around them but that we found multiple times.
So, I think the big idea here is that we're gonna be able to return whatever result that we get. And the FDA has assured us of that. And that's really what matters, because we're still finding cancers that we've never even trained our algorithm on, because it's a shared cancer signal.
No, very helpful. As we took a look at PATHFINDER 2, what do you think is the most important data, from both a physician point of view and a guideline point of view?
Yeah. So I think whenever you look at data from these large data sets about Galleri or any other MCED test, you should think about it in kind of two dimensions. There's the public health benefit, and then there's the performance of the test. So from the public health benefit, some of the most important features to look at are what we call the cancer detection rate. So we know that our current paradigm of screening for one cancer at a time only is finding about 14% of all the cancers. What we've shown in our trials is that when you add Galleri to standard of care screening, the cancer detection rate together is increased more than sevenfold. So let me just repeat that. When you add Galleri to standard of care single cancer screening, you can increase the cancer detection rate by sevenfold.
When you add prostate cancer, which is a Grade C recommendation, we increase it by threefold. So we can increase by threefold and sevenfold the number of cancers and only add a very small number of false positives. So that's the big public health benefit, okay? The other public health benefit is that we're finding a lot more cancer early. Of all the cancers we found in PATHFINDER 2, over half of them were in stage one and two. And that's pretty remarkable given what the skeptics said when we first launched this test. Was that, you're only gonna find all these late-stage cancers that are shedding all this DNA into the blood? Well, over half the cancers we found were stage one and two, and 70% of the cancers we found were in stages one through three. So those are enormous public health benefits.
From an individual performance perspective, the key metrics are the false positive rate. Single cancer screening tests, I'll use Cologuard as an example, have false positive rates in the double digits. Mammograms, low-dose CT of the lungs, Cologuard, 10%-15%. The false positive rate for Galleri, which finds over 50 different types of cancers, is 0.4%. So an order of magnitude lower than anything that's ever been seen before. That's a key safety metric. The other key clinical metric is the positive predictive value, or how confident are you when you get a positive result. And the positive predictive value for Galleri is 62%, which is, again, an order of magnitude higher than anything that's ever been seen with screening before. Most positive predictive values are between 4% and 9%.
Again, if you have a positive cancer signal, you're most likely to be sitting there with cancer right now. So it gives the doctor a lot of confidence. The sensitivity for deadly cancers is quite high. It's about 70%, even in interventional trials, not just case-control studies. And the other big advantage we have is we've now studied Galleri in multiple interventional studies in the screening population. No other test developer has done that. And it's surprising to me that they're launching products based simply on case-control data. GRAIL would never have done that. We waited for our interventional study to be completed before we would launch the product. There simply is no way to know about the safety or benefits of your product without doing an interventional study in the intended-use population. There's just no way to know anything about it.
What I'm gonna focus on something. You know, you mentioned the 7X benefit of doing your test on top of traditional screening.
Yep.
Now, you take the people that do the Function Health or the HIMS, etc., and you probably do have a really good compliant market. So as we look at, you know, you're already on that on that system. Are you seeing that kind of double-fold benefit? Because the, the users that tend to order you also tend to be hypervigilant on, that kind of screening. And, you know, I just wanna think about, generally speaking, I mean, are you seeing kind of maybe underestimation, from investors in terms of that cash pay market?
It's a great question. I'll ask Aaron to jump in too. From a repeat testing, so Galleri, we recommend Galleri be used annually. That's what we've learned. You know, again, if you think about the natural history of cancer, everything we know about it is really based on imaging. And now we know that imaging is far from perfect. And when you look at this biological signal from cancer, cancers are moving through their stages much faster. So if we don't test annually, we're really gonna miss a lot of early-stage cancers that are progressing very rapidly. We've seen really good uptake on annual testing. If you look at Cologuard again as an example that's been out there, it took them many, many years to get to 30% repeat testing rates every three years. We're already well above 30% annual repeat testing.
So we think that's come on pretty well. It's too early in the digital health space with Function and others to really know. But you're right that it tends to be a very wellness-focused population. And we expect the compliance with annual testing to be very good. And then, are we underestimating the market? I think you've got some points of view about that.
Yeah. So, you know, as Josh said, you know, we've got partnerships already with Function Health, Everlywell, some of these digital health platforms. And, you know, we're seeing some uptake there. But that's definitely an area for future growth, that we'll learn and we'll see how that goes. They've got a subscriber base that is health seekers, and used to paying cash prices. So, you know, we see it as a way to grow in a pretty capital-efficient way, as we march toward the big picture, which is the broad reimbursement for the test in the U.S. and internationally.
The other way, David, to think about the self-pay market, excuse me, is, it is evolving. I mean, what you're seeing is more HSA accounts as people get more into high-deductible health plans, employer-sponsored health plans. They're giving the consumer more money.
Mm-hmm.
To do with that money what they want to do with it rather than prescribing them specific health benefits. And that's a different kind of self-pay market, right? Because it's not money out of their own pocket. It's usually pre-tax.
Yep.
Earnings, but they're deciding how to spend that money. And so if they should choose to spend it on Galleri, that to us, that feels like a self-pay.
Mm-hmm.
Kind of momentum.
Yep. That makes a lot of sense. You know, sticking to the topic, what percentage of Galleri tests have been reimbursed by third parties in 2025? And where do we come to some of the key milestones or expected timelines to get some of the private payers to improve ASP?
Let's see.
Yeah. First, so, you know, we currently have about 60% of our business is self-pay. The other portion is what we call enterprise from the enterprise channel, which is things like self-insured employers, life insurance partners, first responders, and so on. So, you know, you could say somewhere between, you know, 30%-40% of our business currently is reimbursed, being paid for by employers or life insurance partners. And, you know, it's important for us, you know, to get to the FDA approval. We believe, you know, that's where we'll start to have more conversations with national payers. And then, you know, with MCED legislation and FDA approval, we'll be able to access Medicare. And that's how we'll get to the broad access picture.
So right now, Medicare Advantage has the ability to pay for whatever they want. Fee-for-service Medicare needs statutory authority to pay for screening tests, which is why special laws have been introduced for colonoscopy and mammograms. And so there is a bill in Congress around providing Medicare the authority to provide coverage for MCED tests. And we're hopefully got a lot of momentum that'll get passed, you know, hopefully either at the end of this year or early next year. That's our hope. And then again, along with the FDA approval combined with Medicare coverage, I think we're really off to the races with reimbursed payers.
Gotcha. Well, you know, and the, the MCED market, I mean, I would argue that you know it has been building momentum over the year and, probably increasing a lot more competition here. So what are the strategic initiatives in, in kind of fortifying the product that, that helps you maintain your first mover advantage in, in this market that is probably gonna attract more and more?
You know, first of all, we welcome new market entrants. We've been out there alone for years building this market. So we now have Exact Sciences entering the market, Guardant Health entering the market, and others talking about entering the market like Caris and Freenome. So but having Exact out there, for example, with their share of voice is actually quite helpful for the field, and it's quite helpful to us. We have by far the best-performing product without question. We're the only company that has done the validation in actual screening populations, not just case-controlled studies. So we have a sizable advantage in being able to train our machine learning algorithms with massive data sets from both our clinical trials and real-world use. And our performance is pretty unparalleled.
I mean, you know, Exact Sciences tests, their false positive rate is five times higher than the false positive rate of Galleri. That's a key safety metric. Guardant Health is three times higher than that of Galleri, a key safety metric. Exact Sciences test has no ability to return a predicted signal origin. Something the FDA said is a critical element to any MCED test. So we feel like we've got by far the best-performing product in the market. Our lead with data is giving us a tremendous advantage to version the test. We're already on we have a PMA version now that's even advanced from the version that we had in our trials. And we're continuing our R&D efforts. We're finding ways of continually improving our performance and driving down our cost of goods, all, you know, with a very strong scientific underpinning.
Our R&D has proven its capabilities already.
Wanna go to the NHS- Galleri study anticipated for mid-2026. First, confirm that date is right. And how do you see the result influencing both, you know, in U.K. or even outside of U.K.? Because, I mean, I believe you can take the data set to any country in the world and basically show the benefits there. So, you know, what can we expect from that readout? What can we expect for using that data across the world?
It's a great point you're making. One of the reasons we chose to do this study in the United Kingdom was for two reasons. One, you know, they're a single-payer system and had the ability to work with us to enroll the study. In fact, this was a 140,000-person randomized controlled trial that enrolled in 10 months. Imagine doing that study in the United States. We'd still be enrolling it five years later. That was really incredible. The data from the UK is very rigorous and is often applicable to many countries outside the UK. We believe that once we get our hands on that data, the full data set, that'll be very relevant for conversations with the European markets, all the single-payer system markets, even markets, you know, in other geographies like Australia, Canada, and Asia.
So the data will read out in the middle of 2026. It will be three consecutive years of testing, with one year of follow-up afterwards. It will be a clinical utility study. The primary endpoint of that will be an absolute reduction in late-stage cancer detection, and it will have full performance data as well. So hugely meaningful data set that, combined with an FDA approval, with clinical utility data, should be sufficient, to achieve the broad market access and the broad application that we're looking for.
I mean, I'm sorry. I'm involving Aaron enough here, so we'll ask one for him. You know, with the FDA approval and, you know, the NHS study reading out, you know, we could see, you know, pretty significant demand for Galleri in 2026 and 2027. You know, what investments are you making in the Research Triangle facility to help optimize cost? Is there a significant CapEx still to go there? And then, you know, how are you thinking about gross margin progression as you start to get the volume? I did notice when I was at the research facility, you were not on NovaSeq X yet.
Mm-hmm. Yeah. So Dave got to go to our lab and see the whole automated thing. It's pretty impressive. It's still awesome every time I see it. As Josh mentioned, Dave saw, like, we've built this facility that can run a million samples today. And I think what you kind of saw last quarter in Q3 was our adjusted gross margins at 55%. Compared to the prior quarter, our adjusted gross margins were 45%. That 10% increase is due to volume. It's the fixed cost leverage that's built into the platform that we have. So we've made the investments to get to scale. So when the NHS or broad access in the U.S. or anywhere else internationally happens, that volume just comes onto the platform. We don't have to invest anything further. You also saw that building wasn't completely built out yet either.
So we can scale that building to be another 6 million-10 million samples, total. So we see a lot of opportunity to grow and expand there. And I think we really showed last quarter that we can get to our goal gross margins of 50%-60%, even at lower ASPs than we have today.
Let me. We also need to talk about the reimbursement after you receive FDA approval. So can you talk about the timelines for Medicare coverage following an FDA approval?
Sure. Well, provided that the MCED legislation gets passed, before the FDA approval, all the legislation basically states that Medicare shall have the authority to provide coverage for an FDA-approved MCED test, only an FDA-approved MCED test, and as far as I know, we're the only ones right now seeking FDA approval. There may be other companies. I think, others are on record saying they're gonna wait for Medicare approve, you know, coverage authority before they invest in those trials, which I've heard before as well. So the timeline looks like, we're gonna submit our final package to the FDA first quarter of 2026. We think it'll be a 12-month review. So let's say you get reimbursed in 2027. Medicare coverage will be about a year probably of getting a national coverage analysis and a national coverage decision. And then the legislation actually kicks in in 2028.
Mm-hmm.
It phases Medicare in over time, and it has a price in it that was set to Cologuard's price at the time as well. So that's how reimbursement will work. They'll set a coverage policy. They have pretty broad discretion about how they set their coverage policy, irrespective of how the legislation is written. So we'll see what that coverage policy looks like. But that's kind of the timing we're looking at right now if all the legislation gets passed before FDA approval.
The dynamic, I mean, I would think that your test would look more like SHIELD than it would Cologuard, but.
Yeah.
Anyway.
You can imagine a world where Medicare is paying $508 for a multi-cancer early detection test.
Mm-hmm.
Imagine them paying $1,400 for a single cancer blood-based test. That's highly unlikely. So I think the pricing, if this actually gets passed and put into law, I think the whole pricing for the field of blood-based diagnostics is gonna change.
Yeah. Which we're built, you know, with the platform that we have today to have the margins that we want to, at those lower price points. At a $500.
Yeah.
Yeah. Okay. You know, you have a strategic collaboration with Samsung in the Asian market, introducing Galleri in Canada. What are the revenue contributions for international markets in 2026?
Yeah. Great question. So we launched in Israel toward the end of last year. We launched in Canada most recently. It's a small contribution to 2025. 2026, we'd expect those things to ramp up a bit, but to still be in the early days of international expansion, especially compared to being in the U.S. for, at, you know, next year, we'll have been in the U.S. for five years at that point. And then, you know, the Samsung opportunity, you know, will, which we're very, you know, excited about. It's really an opportunity for us to accelerate expansion into Asia much faster than we thought we would, with a partner that we, you know, we could not have picked a better partner. You know, that'll start as a test send-out model where they're gonna approach the South Korean market, cash pay, build up their capabilities.
but they really want this to be a large part of Samsung's business, so with success there, you know, they'll, we'll expand onto likely Japan and Singapore. We're in the process of finalizing or turning that binding term sheet that we have into a commercial agreement, and hopefully we'll have more to say once that's done, but it's a partnership we're very excited by.
Gotcha. I'll stick with the run rate in terms of cash burn. I think you're down to, what is that, $290 million for this year is the expectation. How should we think about the cadence over the next few years?
Yeah. So, you know, we've just completed a successful PIPE, brought in a bunch of high-quality investors that we're very happy with. We raised $325 million there. If Samsung closes and they pass the regulatory hurdles, we'll have another $110 million from them. This will push our cash, you know, into 2030, you know, which gets us well beyond the dates that Josh talked about getting broad reimbursement in the U.S., beyond NHS making a decision whether they're gonna roll the test out or not. So, you know, we'll provide guidance for what burn in 2026, and it will be, you know, once we get through the annual planning process and into next year. But, you know, we believe we're set up appropriately from our burn right now to really capture the opportunity that we have in front of us.
I think it gives Aaron more flexibility to see opportunities. If there's a big opportunity in digital health or in an international market, we now have the flexibility to really take advantage of that, should we see the opportunity clearly.
Thank you so much.
Thank you.