Good day, ladies and gentlemen, and welcome to today's GRAIL analyst call. At this time, all participants are in a listen-only mode. After the speaker's presentation, there will be a question-and-answer session. Please be advised that this conference call is being recorded. GRAIL Investor Relations, please begin.
Thanks, Operator, and thanks everyone for joining us today to discuss detailed PATHFINDER 2 study results presented over the weekend at ESMO. On this morning's call are Bob Ragusa, our Chief Executive Officer, Aaron Freidin, our Chief Financial Officer, Joshua Ofman, President, and Sir Harpal Kumar, President, International Business and BioPharma. We'll be making forward-looking statements on this call based on current expectations. It's our intent that all statements, other than statements of historical fact, including statements regarding our anticipated financial results and commercial activity, will be covered by the safe harbor provisions for forward-looking statements contained in Section 27A of the Securities Act of 1933 as amended, and Section 21 of the Securities Exchange Act of 1934 as amended. Forward-looking statements are subject to risks and uncertainties. Actual events or results may differ materially from those projected or discussed.
All forward-looking statements are based upon currently available information, and GRAIL assumes no obligation to update these statements. To better understand the risks and uncertainties that could cause actual results to differ, we refer you to the documents that GRAIL files with the SEC, including the risk factors section on GRAIL's most recent annual report on Form 10-K and quarterly reports on Form 10-Q. Bob, please go ahead.
Thank you. Good afternoon, everyone, and thank you for joining us. We were very pleased to share results from the first 25,000 participants enrolled in our PATHFINDER 2 study at the ESMO Congress this past weekend. This analysis is a registrational dataset, which will be included in our PMA submission to the FDA in the first half of next year. Our first interventional study named PATHFINDER was completed in 2021, and results were later published in The Lancet. We subsequently undertook the 35,000 participant PATHFINDER 2 study, a much larger, similarly designed prospective multi-center interventional study. PATHFINDER 2 is intended to assess the performance and safety of Galleri in a large population representative of Galleri's intended use group. PATHFINDER 2 and NHS-Galleri trial make up our registrational clinical program for Galleri.
At 35,000 and 140,000 participants respectively, these are two of the largest screening studies completed to date and the two largest studies in the multicancer early detection field. From the beginning, GRAIL has focused on producing high-quality clinical evidence. Our evidence base includes clinical validation in large case-controlled and interventional studies, as well as real-world evidence studies, REACH, and other observational datasets. The ongoing clinical program supporting Galleri totals nearly 400,000 participants. Importantly, Galleri is the only MCED available which has reported results from people being screened in the intended use population.
Thanks, Bob. I'll now briefly walk through recent announcements that have strengthened our balance sheet via two transactions totaling $435 million. First, we announced last week a strategic partnership with Samsung, which includes an equity investment of $110 million by Samsung into GRAIL, subject to closing conditions. GRAIL and Samsung C&T will work as exclusive partners to commercialize the Galleri test in South Korea, with a possible extension into other Asian geographies, including Japan and Singapore. Samsung C&T will undertake key activities to drive adoption of Galleri, while GRAIL will initially perform the Galleri tests at our central labs in Research Triangle Park, North Carolina. In addition, we and Samsung Electronics plan to explore additional potential strategic and operational collaborations.
Samsung has considerable expertise operating businesses in South Korea and more broadly in Asia, and we look forward to working with Samsung to bring our multicancer early detection technology to these new key markets. Second, we announced this morning a private placement of $325 million of equity. This financing includes participation by new and existing institutional investors, including Deep Track Capital, Farallon Capital Management, Hims & Hers, Braidwell Limited Partners, three life science investment firms, and a tech and life science-focused family office investment firm. We are very pleased to partner with these quality investors to help realize our mission to detect cancer early when it can be cured. Net proceeds from these transactions are intended to support commercial activities and reimbursement efforts, as well as for working capital and other general corporate purposes.
As a result of today's PIPE financing, we believe we are very well capitalized, with an expected runway now into 2030. That runway does not include the agreement by Samsung to invest $110 million into GRAIL, which is subject to closing conditions. Now, let's move to Josh and the PATHFINDER 2 dataset.
Thank you, and hello, everybody. Great to join you today to walk through the data that was presented this weekend at the European Society for Medical Oncology Congress in Berlin. As Bob mentioned, PATHFINDER 2 is a registrational study for Galleri, which enrolled more than 35,000 adults over the age of 50 at multiple centers across the U.S. Blood draws were taken once at enrollment, and a Galleri test result was returned. If a cancer signal was identified, participants received a workup based on a signal-of-origin guided recommendation, what we call a CSO guided recommendation, but it remained at the physician's discretion. PATHFINDER 2 is intended to evaluate the performance and safety of Galleri in a large, diverse patient population. Enrollment demographics and characteristics were intentionally diverse, as any successful population-scale MCED innovation needs to demonstrate performance in a large intended use representative screening population.
The dataset we just shared at ESMO this weekend was from the first 25,000 participants of the 35,000 enrolled that completed one year of follow-up. 25,000 was a pre-specified cutoff for results to be included in our Galleri PMA submission to the FDA. Now, let's turn to the findings. The most important metrics for an MCED test, we believe, are the cancer detection rate. How many cancers does Galleri detect in a population being screened when added to standard-of-care screening? The positive predictive value, or PPV, which means if you have a positive test result, what is the likelihood you actually have cancer? The higher the PPV, the more trustworthy and actionable the test result is. In PATHFINDER 2, Galleri's observed PPV was 62%. Just as a reminder, the observed PPV in the first PATHFINDER study was 43%.
This represents a meaningful step up now in a large, more representative population. Specificity for the test was 99.6%, which means the inverse of specificity, or the false positive rate, was 0.4%. That's less than 0.5% . We also report the negative predictive value, or NPV, and episode sensitivity, which is a different metric than test sensitivity, which would be normally reported in a case-controlled study. The NPV, or the negative predictive value, was 99.1%. In interventional studies like PATHFINDER 2, we measure episode sensitivity based on how many cancers are detected within a predefined period of time. In the case of PATHFINDER 2, we set this time period to 12 months. We're very proud that our episode sensitivity for all cancers, including cancers with low DNA shedding, where we expected lower sensitivity, was overall 40.4%.
This represents a significant improvement over the current state, where screening sensitivity for detecting asymptomatic cancer is effectively zero for every cancer where there is no screening option available today. If you look at certain subsets of cancers, the episode sensitivity is much higher. Here we look at Galleri's episode sensitivity in some of these different predefined subgroups. Starting at the bottom, in cancer types without recommended screening options, Galleri's episode sensitivity was 55%. That is 55% of cancers where today there is no available screening. In all cancers, excluding breast and prostate, Galleri's episode sensitivity was 56%. Moving up to the most aggressive cancers, in a group of six of those aggressive cancers that have the highest five-year mortality, Galleri's episode sensitivity was 72%. Finally, in the 12 aggressive cancers that are responsible for 2/3 of cancer deaths in this country, Galleri's episode sensitivity was 74%.
Now remember, Galleri's episode sensitivity performance is coupled with a very high specificity, or a very low false positive rate, of 0.4%. We feel very good about the balance that our technology has struck between episode sensitivity and specificity. Critically, these episode sensitivity numbers should not be compared with test sensitivity numbers from case-controlled studies, where diagnosed cancer cases are used and controls are selected. As described in my recent blog post on the grail.com website, those case-controlled studies can be subject to many biases if not very carefully conducted. Another way to evaluate a cancer detection test is to look at the cancer detection rate, or how many cancers are found in the population when deploying the test. In PATHFINDER 1, we found that when you added Galleri to standard-of-care screening, it more than doubled the number of cancers detected.
In PATHFINDER 2, adding Galleri to USPSTF Grade A/B recommended screening tests, we detected 7x more cancer. How did we measure this? Of the 329 participants diagnosed with cancer in PATHFINDER 2, 133 of them were MCED detected. 20 were detected by breast, cervical, colorectal, and lung cancer screening. If we compare the 153 detected by MCED plus the A/B screenings to those detected just by A/B screenings, that is 153 versus 20, or a 7x difference. If we add in USPSTF Grade C recommended prostate cancer screening, then we see a 3x increase in the number of cancers detected, or a threefold benefit. This is a significant benefit, and this is only possible because Galleri can detect so many cancers that do not have screening options today.
You can see here that the cancer types detected by Galleri in PATHFINDER 2 represent a broad range, including cancer types with existing recommended screening. We know that Galleri can find interval cancers where recommended screening tests exist today. Even when participants are compliant with other screening, Galleri may still find a cancer. Most importantly, 73% of Galleri-detected cancers in PATHFINDER 2 do not have recommended screening options. This test is truly a breakthrough and is enabling patients and providers to find cancers that would rarely be diagnosed until after symptoms arise. Now, you see here the stage distribution of the Galleri-detected cancers. Remarkably, more than half of the cancers identified were in stages one and two. These include head and neck, liver, pancreas, kidney, lymphoid, anal, colorectal, lung, breast, plasma cell, bladder, and testicular cancers.
As you may know, with the advent of new therapeutic and treatment approaches, stage three cancers have become more treatable with better outcomes. Almost 70% of Galleri-detected cancers were in stages one through three. We believe an accurate, built-in cancer signal-of-origin capability is critical to guide rapid and appropriate diagnostic workups after a positive test result. Feedback we receive from the physicians and regulators is also very consistent here. In PATHFINDER 2, the Galleri test's first CSO prediction correctly identified the cancer signal of origin 92% of the time, leading to very efficient diagnostic workups. This is in line with what we saw in the first PATHFINDER study. We also look at the time to diagnostic resolution. Following a positive Galleri test report, it's important that we and physicians make maximal efforts to achieve full resolution as quickly as possible to avoid any prolonged diagnostic journey.
The median time to diagnostic resolution among true positives was just over one month, 36 days. Among all MCED-positive participants, full resolution occurred at a median of 46 days. Patients that received a positive MCED test and where cancer was not found had these cases resolved at a median of 75 days. We are very pleased with these results, and we note these are much shorter timelines than was observed in the first PATHFINDER study. Now, let's turn to the safety findings. Galleri was safely implemented in the intended use population in PATHFINDER 2. Of the 25,000 participants analyzed, 0.6% had an invasive procedure done as part of their diagnostic workup. We define invasive procedures broadly to include endoscopy, as well as biopsies and surgery. Invasive procedures were 2x more common in participants with cancer than in those not found to have cancer.
There were no serious study-related adverse events reported as of the date of this analysis. Let me summarize these results from the PATHFINDER 2 study. Adding Galleri to standard-of-care screening for breast, cervical, colorectal, and lung cancers yielded a more than seven-fold increase in the cancer detection rate. More than half of the Galleri-detected cancers were found in early stages one and two, when cancers are more treatable and potentially even curable. Nearly three quarters of the cancers detected by Galleri do not have recommended screening tests today. Galleri's ability to accurately identify where in the body the cancer signal is located helped guide an efficient diagnostic workup. Importantly, there were no serious study-related adverse events. These results from PATHFINDER 2 mark Galleri's ability to provide asymptomatic adults and their loved ones with a way to detect cancer early when cancers are more treatable and potentially curable.
This represents a really exciting next step in the evolving evidence base to bend the cancer mortality curve through multicancer early detection. Now back to you, Bob.
Thanks, Josh. Operator, let's open the lines for questions.
Thank you. At this time, if you would like to ask a question, please click on the Raise Hand button, which can be found on the black bar at the bottom of your screen. You may remove yourself from the queue at any time by lowering your hand. When it is your turn, you will hear your name called and receive a prompt to unmute. As a reminder, we are allowing analysts one question and one related follow-up today. We will wait one moment to allow the queue to form. Our first question will come from Subbu Nambi with Guggenheim. Please unmute and ask your question.
Hey, guys. Thank you for taking my question and walking us through the data presentation. 18 of the 133 MCED detected cancers were recurrent. I thought that was a bit high. Could you tell us if it was in line with your expectations? If not, why? I know how many years out from their cancer were these patients. Do we have information on that? They had to be three years out to be included in the trial, right?
That's probably one of the things.
Yeah, you want me to take that role, Josh?
Josh, you could, yeah.
Thank you. Good question. We don't have the detailed information on each of those recurrent cancers, but the criterion for this study was they had to be at least three years post-treatment. They're not actively treated cancers. These were cancers that were in remission or had been cured. The number of recurrent cancers in the population did meet our expectations. That will be variable in any given study, but we do expect and have found throughout all of our clinical studies and our real-world experience that we are finding recurrent disease in people who've been long cured or post-treatment for their active cancer. This did sit with our expectations and also with what we saw in the original PATHFINDER study.
Thank you for that, Josh. Josh, safety is an important metric, at least in the NHS-Galleri trial and then even in this trial. 0.6% of your patients had an invasive procedure. Are there any pre-specified endpoints for this safety metric with the FDA, or did 0.6% make the cut?
No, there were no pre-specified metrics. We know that for a given CSO-directed workup, endoscopy is a very appropriate first test. If you get a CSO that says esophagus, stomach, colon, or larynx, they often go right to endoscopy, which is categorized as an invasive procedure in the study. The fact that only 0.6% of all subjects had an invasive procedure is very positive from a benefit-risk profile perspective. Of the 0.6%, there were more than twice as many of the invasive procedures that occurred in people ultimately diagnosed with cancer, as opposed to the false positives. That is a very low number of invasive procedures, particularly given that endoscopy is often the right first test.
Perfect. Thank you so much, guys.
Your next question will come from Kyle Mikson with Canaccord.
Hey, guys. Thanks for doing the call. I wanted to just focus on the addition of PET imaging because I don't think you've done that in previous studies. First, I was wondering if that helped PPV or any other of the detection metrics. Also, does the addition of the imaging kind of reduce the importance of your CSO prediction accuracy?
That's a great question. Let me go back as to the reason why we even introduced that part of the study. In commercial practice today, Galleri CSOs are actually reported. One CSO is reported, and the workup is very effective at resulting in diagnostic resolution. However, there are cases where the initial workup does not reveal the cancer. In those cases, commercially, we offer a free repeat Galleri test to bring that to rapid resolution. We do not recommend whole-body imaging. It was for that reason that we introduced the PET-CT into that step of the study. If in the study, people got a positive Galleri result in a CSO and the initial workup was unrevealing, they got both a PET-CT and a blood draw. The purpose of that was to show that the blood draw is the best way to achieve diagnostic resolution.
In fact, that analysis is still ongoing and will be reported out shortly. The initial preliminary results are that the PET-CT revealed very few cancers and did not meaningfully impact our PPV.
Good. Thanks, Josh. Appreciate that. On the topic of cost, just using the PATHFINDER 2 data here, we estimate that you would need to price Galleri at around $200 to meet the QALY cost threshold for NHS. Clearly, much, much lower pricing right there. Do you think that the economic benefit analysis that's going to be done by the NHS right now is going to be a bottleneck and cause some friction when they roll that out, roll Galleri out, even if all the data is positive?
Let me comment on your math first, and then I'll turn it over to Harpal. Some things are not adding up with your math. We've published robust cost-effectiveness analyses already, looking at cost per QALYs, and there is, you know, it's based on modeled stage shift and interceptions. We've shown that it's highly cost-effective, even at the list price. I'm not sure you're doing the economics quite right. You may want, we can talk to you about that offline. I'll turn it over to Harpal to comment on the NHS.
Yeah, thanks, Josh. I mean, look, we are waiting, as you know, for the results from the NHS-Galleri trial. That will give us updated information on what proportion of participants do end up with a cancer diagnosis. We will be able to look at that by different subgroups and so on. That will determine how the NHS chooses to move forward in terms of deployment. That then will also, if you like, trigger the conversation that we will need to have with the NHS about how much the test is priced in order for it to be cost-effective. Those are discussions to come. We haven't had them to date, but we will absolutely be following up and initiating those once we have results from the NHS-Galleri trial.
Your next question will come from Yuko Oku with Morgan Stanley. Yuko, your line is open. Feel free to unmute.
Hello. Thank you for taking my question. PATHFINDER 2 demonstrated a shorter median time to diagnostic resolution than the original study, particularly for the false positives. What factors would you attribute to this improvement?
That's a great question. I think there are two main factors. The first factor was going back to the original PATHFINDER study. That was the first real interventional study done of an MCED. I think physicians were less aggressive in working up the CSO than they are today. The original PATHFINDER study was conducted during COVID, where lots of endoscopy suites, clinics, mammography suites were not open regularly, and patients were not visiting the clinical sites nearly as frequently. I think between the COVID effect and just the experience that the clinical community now has in working up CSO -directed signals, those two things combined with the much shorter median duration to diagnostic resolution.
Great, that was really helpful.
If I could just add a comment, Josh.
Please, Harpal. Maybe jump in, Harpal. Go ahead.
Yeah, I mean, I'm actually speaking to you from Berlin, where I've been at the ESMO Conference. It's really notable that many of the clinicians we talked to, to Josh's first point, with such a high PPV, there's now such a high degree of confidence that this is more likely than not to be a real cancer. There is a real effort now on the part of clinicians to say, if I get a positive test, I really want to look to make sure that this patient does indeed have cancer. That process of investigation is more intensive and more comprehensive and thorough now than perhaps would have been the case with the first PATHFINDER study.
Harpal, do you want to just mention what we saw in SYMPLIFY that kind of compounds this issue?
Sure. Happy to do that. The SYMPLIFY study, for those of you who don't recall, was a study that we conducted in the U.K. in symptomatic patients to understand if this test could play a role helping clinicians guide investigation when patients are presenting with symptoms that might be nonspecific. Originally, what we saw was that there was a PPV of around 75%. What we've done, and indeed the data has just been presented today or published today, is that we have followed up all of the patients who were originally designated as false positives. We found two really, really important pieces of information. First of all, over 1/3 of those patients initially designated as a false positive, 35%, have subsequently been diagnosed with cancer within two years of that blood draw. More than 1/3 of the false positives have turned into true positives.
Secondly, of those, all but one patient, the CSO prediction was correct. That really offers the opportunity for a very effective and efficient diagnosis. It takes the PPV in that symptomatic population up to 84%.
Your next question will come from Doug Schenkel with Wolfe Research. Doug, your line is open. Feel free to unmute.
Okay. Thank you. Thanks for doing this call. Good morning. Good afternoon, everybody. A few topics. On clinical utility, as we've seen in the past, FDA approval does not always translate to CMS reimbursement. Ultimately, the key dynamic will be likely stage shift, as you acknowledged in your prepared remarks, as a proxy for survival benefit. Is there anything in PATHFINDER 2 that makes you more comfortable about a positive outcome on this metric when we head to NHS-Galleri, or ultimately, do we just need to wait for the readout? Building off of that, given FDA requirements outlined at the FDA panel on MCED in 2023, what boxes have you checked with the FDA and what remains to be done? Essentially trying to get at the clinical utility question, which would be key to reimbursement, and then separately the regulatory question in the U.S. with the FDA.
No, thank you, Doug. Really, really good questions. Let's take the clinical utility one first. You know, there's nothing directly that can be inferred from the PATHFINDER 2 study to the stage shift or the reduction in late-stage detection that is the primary endpoint of the NHS-Galleri trial. There are many aspects of PATHFINDER 2 that give us more confidence in the overall performance of Galleri. Those specifically are the dramatically increased cancer detection rate when added to standard-of-care screening, and secondly, the much higher PPV that we've been observing, as well as the episode sensitivity, which is quite high. For those of you who may not have been following this, when case-controlled studies report high sensitivity, they very rarely translate into that level of performance in actual interventional studies.
We've seen that even in the MCED field with the one interventional study that was done in 65-year-old women, where the performance of the assay from the case-controlled study simply did not come close to replicating. We're very pleased with that episode sensitivity, those numbers coming out of PATHFINDER 2 relative to what we had seen in prior studies. I think those three things together, Doug, give us a lot of confidence in the performance, but they don't directly speak to stage shift or the reduction in late-stage cancer because that will much more be related to the case mix of cancers in that individual study and the stage distribution in that individual study. I'll turn it over to Harpal in a minute to comment more on the NHS CU .
On the FDA, based on the 2023 panel, it was quite clear what the FDA said in respect to the meaningful necessity of having a CSO-directed workup. Given what we've seen now with Galleri and PATHFINDER 2 with very high CSO accuracy and very rapid diagnostic resolution, we feel really confident about that finding. The other thing the FDA emphasized was the false positives being one of the biggest harms related to screening. With our 0.4% false positive rate, we feel very good about that. The other thing they mentioned, of course, as it relates to safety, was the overdiagnosis. We've published now multiple times that low-shedding tumors that are indolent are not typically detected well by Galleri. Galleri is very unlikely to contribute to the problem of overdiagnosis of indolent cancer. On those three dimensions, we feel very good.
I'll ask Harpal to comment more on the clinical utility question.
Yeah, I mean, I think you've largely covered it, Josh, but you're right, Doug, that the primary endpoint in NHS-Galleri will be looking at that reduction of late-stage cancers. In order to find a reduction, you have to have a randomized controlled trial. That's, of course, what NHS-Galleri is. I think just to add one point to what Josh said, one of the things that encourages me greatly from PATHFINDER 2 is that more than half of the cancers were found at stages one and two, and that's in a cohort of cancers where three-quarters are currently unscreened. I think, to the extent that you can take any guidance from a study that doesn't have a comparator arm, those points really do encourage me as well.
Okay, thank you for that. I guess I will admit a completely unrelated follow-up. I think it's more for the finance team, so unrelated to the studies. I am curious, were investors in the private placement provided access to any data as part of their investment evaluation that is not currently available in the public domain?
Hey, Doug, good question. No, I mean, all the information that they've seen is now public.
Okay. All right. Thank you, guys.
Your next question will come from Kyle Mikson with Canaccord.
Yeah, thanks for the follow-up. On the topic of follow-up, in the study design, do you guys think that the one-year follow-up is enough for this type of a test given, you know, obviously some cancers spread or appear more slowly than others? Have you talked to the FDA about follow-up and just kind of longer-term data and if that's necessary?
Great question. On the one year, you know, we have to find a period of time to represent ground truth because there is no effective ground truth. I mean, one thing we've learned through Galleri implementation is how imaging, you know, is not really the right ground truth. Time is the best source that we have. We think one year is enough. We did a large study with the American Cancer Society that's been published that looked at kind of the dwell times of cancer across stages. It came out to be very consistent with about a one-year time period on average across tumor types that they are advancing from stage to stage. We do think one year is about the right timeframe, both for defining ground truth, but also for the frequency of testing that we will be required to detect early cancers.
That's the most important finding as it relates to the finding of ground truth. The FDA has reviewed, you know, all of our studies, all of our protocols under IDE and approved them, and they've all had that one-year timeframe. I think, you know, most, at least in my experience, agree that that's a very reasonable way to define the episode.
Perfect.
Sorry, I was just going to add a comment. We do think that that 12-month period is the right period to be looking at this sort of episode sensitivity metric. Just to be really clear, PATHFINDER 2 is going to be following up all the patients for three years. We will see if later cancers emerge, you know, following that initial 12-month period. As you know, NHS-Galleri is three rounds of screening and will be following up those patients over longer periods as well. We will see the extent to which cancers later emerge. Indeed, if you were listening to my comments on SYMPLIFY earlier, you will see that we do indeed see that. There is that possibility, but it doesn't detract from the point that in order to report on this episode sensitivity metric, 12 months is the right period.
We have not shared any of that long-term data with the FDA yet because we haven't completed its collection. That will all be shared with the FDA in time.
All right, great. I believe that 14% of the cancers detected here in the study were recurrences, so like 19 cancers. If you take those out, you get a high 30% episode sensitivity and a high 50% PPV. The question is, do you think these cancers would be found using an MRD test in the real world, or are they appropriate for a screening test like this?
Remember, these are not people with active cancer. Whether they should have MRD tests is really when they're actively being managed for cancer. These are people who have completed all their treatment, are out of their cancer active status. They are not being treated for cancer. They're just normal people who are cancer survivors who are at elevated risk for cancer who are being screened. Let's not confuse that with getting MRD testing. That's not what was happening in this study. I don't know why you would want to remove those people from this study because that is a very important population to be doing cancer screening in, people who have completed their cancer therapy or have been cured or are in remission for cancer. Very different setting than MRD testing or anything like that. That's not what was going on in this study.
These were people being screened for cancer.
Yeah, very helpful. Thanks, guys.
Thank you. There are no further questions at this time. I will now turn the call back to GRAIL for closing remarks.
Just like to thank everybody for attending the conference call today.
Ladies and gentlemen, this concludes the call. You may now disconnect.