Guardant Health, Inc. (GH) Earnings Call Transcript & Summary

I'm Tejas Savant from the life science tools and diagnostics team at Morgan Stanley. It's my pleasure today to host Guardant Health. And from the company, we have Helmy, Amirali and Mike. So thanks, gents, for doing this. Let's kick it off. So maybe, Helmy, just to set the stage, could you share your vision for the company when you started it, where are you in that journey today? One of the things that sort of jumped out to me last Friday at the Analyst Day was how far you've come as a company since we first met sort of all those years ago? What's gone well? And where has progress been slower than you had hoped for?

Yes. It's actually very interesting. When we started the company, we had almost exactly the same vision that we've really turned into reality today, 11 years later this idea of really following the continuum of care towards a simple blood test for early detection of cancer, working from therapy selection, MRD and then screening. And from the 40,000-foot view, you can look at our pitch decks from 2012, 2013, and there's really nothing that has deviated from that initial vision. Of course, all the details are very different in terms of execution and some technology aspects and some different twists and turns we took. But I think it's very rare that you have a company that has sort of actually executed and accomplished everything that it set out to do. And obviously, there's a little bit more work to do to get there, but it's really an exciting sort of juncture we are at now. And that's why we really wanted to sort of take a number of hours at our Investor Day and really sort of refresh the vision we have for the next 5 to 10 years. And I would say that in most places, we've gone faster than we expected. I think around screening, we brought that up earlier than we expected. In some areas, we've gone slower, like reimbursement for 360. I don't think we thought it would be this year that we're finally getting coverage from some of the large payers. And so those things that are challenges end up being most though for the business. Once you get through, you then become the entrenched incumbent that really has a sort of advantage of the new entrants in this space. Yes, we feel very pleased in terms of all the progress we made. But even I think more excited about really what we could do with this platform for not just cancer, but potentially even other disease areas in the future.

Got it. So definitely sharing that excitement. In fact, I was so excited that I forgot to read out the disclosure statement. So please see the Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. And if you have any questions, please do reach out to your sales rep. So with that, let's start with therapy selection, Helmy. You're clearly the leading player there, despite others having tried to launch products into that market. What is your secret sauce in that market that is fueling such a sticky first mover advantage? I mean this is not a new market in sort of the grand scheme of things, you've been on there for almost 5, 6 years now.

Yes, that's a great question. So I wouldn't say it's sort of 1 element, which is what the difficulty is in executing in oncology diagnostics. It really does take 5 or 6 sort of components to be able to execute well and really be a leading company in this space. Certainly, it starts with the technology in terms of the performance, the clinical validation. We have many CDxs. We have hundreds of publications, the sales force really having the multi-hundred sales force that we have in oncology diagnostics. I think one of the largest teams and certainly the highest rated team in all of oncology diagnostics. And then operational excellence, there are fast turnaround time, the user experience that feels very frictionless and very simple and so it's really all of those elements that come together to equate to something that has a stellar and really sort of top of market product market fit. And that's, I think, what we've established over the years. I think we had an advantage when we started because we really were the first serious mover in the liquid biopsy space. And so we had the luxury of some time to iterate in some of these components. But now the bar is extraordinarily higher today in 2023 than it was in 2014 when we launched our first product.

Got it. So in a sense, it's back to the past situation this year with a focus returning to G360 after MRD, it's screening getting a lot of investor attention for the last 12 to 18 months or so. Really good momentum there. One question we occasionally sort of get is, there's recently this Lancet paper, I don't know if you saw it, about an increasing likelihood of being detected with late-stage cancers because of this post-COVID catch-up effect at play. Is that a factor at all in some of the recent volume trends you're seeing? And if so, could there be a little moderation perhaps on the cards despite that long runway for growth over the medium term?

We don't think so. If we look at sort of what has been the growth driver in 2023 for us, I think it's pretty easy to pin down. We're seeing certainly a big bump up in our breast volume almost overnight. We saw a 30% to 40% increase because of this ESR1 approval. Now there is another therapeutic option for a very large subset of breast cancer patients that physicians need to test for, so that immediately drove the need to test a lot more patients, a lot more breast cancer patients with Guardant360. I think some of the private payer coverage that we received with United and Aetna and Anthem certainly made a big difference. And then I think there's the -- we're at a sort of really nice part of the S curve in terms of a large percentage of that middle majority now realizing that comprehensive genomic profiling is really a must-have for every patient that they serve. We recently had the JCO article that came out of the University of Pennsylvania that showed essentially a 4x improvement in overall survival, if patients are tested upfront, with comprehensive genomic profiling. I think in that study, over 90% of the tests that were used were Guardant360 test. And so it's a really nice publication that demonstrates with [ OS ] survival curves really, the undeniable benefit that knowing how to treat, what to treat upfront makes a world of difference. I mean if you had a therapeutic that has 4x overall survival benefit, it would be a blockbuster drug. So yes, I think we have just lots of elements going our way in terms of increased usage and increased penetration of therapy selection in this market.

Got it. And then at the Investor Day, you noted that you're hearing physicians looking for a one-stop shop and you're starting to see good attach rates for TissueNext and Response. How often do physicians today order more than one assay from Guardant? And where do you see those sort of attach rates going over time? And sort of on a related note, are almost all physicians ordering those newer test current G360 users?

Yes. I mean our penetration in 360 is by far the highest in terms of our portfolio. I would say that our other products are still very much in the early innings, but they're growing extremely rapidly. And I think that's really one of the benefits we're seeing of just being so strong in that space is that we can drive really uptake of these newer products in a very efficient manner with high leverage. Yes. I think with tissue, that product, is one where we're seeing this paradigm of what we call "blood-first, tissue-next" where I think it's really resonating with the field where this idea that I make sure I don't miss and so I order both essentially on day 1. And then the idea is you get your liquid result in 5 to 7 days. And if that's negative, you can still have the option for the tissue result, which has been in the queue, and you're not wasting any time on. And so to ensure that sort of magic bullet of finding a genomic alteration in a cancer patient is not missed since neither modality is perfect. And I think that's resonating well. I think the idea of response is resonating well now that we have Medicare coverage for immunotherapy for all similar types. But I would say that we're still very much in the early innings in terms of the percentage of all oncologists that are ordering from the entire suite. I think something like 90% plus have ordered a 360, which I think bodes well for what we can get with the other products.

Got it. And before moving to MRD, I want to quickly hit on Japan. An important opportunity for you. You've got approval and reimbursement in place. How comfortable are physicians in Japan in using sort of liquid-based methods versus tissue? And is there anything you need to do in terms of your go-to-market strategy there to really ignite penetration?

We're very excited about the Japanese market. It's a market that is around 2/3 of the number of cancer patients in the United States, a single care system. The rates that we have there for Guardant360 is in the $2,600 to $2,700 range, which is similar to our U.S. ASP. And so it's really a matter of how fast can we go and how fast can we drive volumes. We are -- we've spent a lot of time planning for the sort of commercial launch of that test there. We are in the major EMR systems in Japan. We are partnered with the group that has a very large sort of distribution in all the genomic medicine hospitals there. And so we think we'll get to a very fast start. I would say that it's still sort of early innings there. Right now, the approval we have is in the sort of last line setting. But there's a lot of excitement, and I think a lot of push to try to get this type of testing to the front end. I think this JCO article about the 4x survival. We have other articles that were published locally that show the benefit of 360 upfront, really helping to improve outcomes. And so we think that as this testing moves to the frontline setting, it's going to be a market that is, I think, not too distant second from the U.S. opportunity. The nice thing is we've been in that market on the research side for 5 or 6 years. Obviously, SoftBank was a big investor, and so we have offices there with a lab there. So we've done a lot of the groundwork over the last 4 or 5 years, really building the right relationships with the top cancer centers. And so we're a known entity and a trusted, I think partner. So we're very bullish on our future there.

Got it. And then switching to MRD, could you give us a sense for how often you run into sort of tumor informed users and try to convert them over to tumor naive. How does that conversation typically go? What do you lead with? And I guess on a related note, some of these tumor informed players are now looking to launch broader panels, so 50 markers, perhaps even up to 1,000 in some cases. Does that sort of change the competitive dynamics at all in your mind?

No, not at all. I mean I would say that it's still very much early innings. It's really more about portraying the benefits of the assay than necessarily competitive selling at this point. And the fact of the matter is there's just no one out there who has a tissue-free solution that's very simple. So it's a fairly, I would say, straightforward sales that is really based on the quality of the data that you can present with the assay since it's a much simpler assay to adopt and to test the patient with and so on. So I think that's one of the benefits we have of a tissue-free solution. I think we wanted to sort of talk about the power of the technology we have in more detail at our Investor Day, and the fact that our assay is able to look at thousands of specific tissue markers and the epigenetic side, I think, is why it's such a powerful approach, why we're seeing the great sensitivity on the early detection side and why we're seeing, I think, very good performance with MRD. And yes, these assays in the tumor informed side are trying to play that same game in terms of catching up in terms of going from 16 to 50 to eventually 1,000. I think the other point, though, is that despite how many mutations these tumor-informed approaches go to, they're actually not essentially delving into the biological significance and underpinnings of the tumor, the markers we look at and the breadth we're looking at the tumor are very informative from a clinical point of view. We're going to be able to do things like define where the recurrence is. Is there a liver metastasis, tell you where the tumor site is, what the tissue of origin is, is it a recurrent tumor of a different type. And so there's really apples to oranges in terms of comparing these different assays. One is generating in terms of our Guardant Reveal and Guardant Reveal and smart liquid biopsy will generate, we think, something like 10,000 to 100,000 more informative information for a blood test than even the sort of larger tumor-informed assay.

Got it. I want to hit quickly on the COBRA early stoppage. You talked about using an older version of the assay and a study design that perhaps wasn't what you do today if you had to do it all over again. What sort of drives your confidence that it was not the assay, but the study design? And do you think there's a need to clarify that data among the physician community just so that's to prevent any confusion or muddying of the waters as it relates to the -- your push into MRD?

Yes. I think to reiterate, I think we had very high specificity for the assay that was used there. I think it was a challenging sort of interim analysis point with a very few number of patients. And so it's one that really has sort of no near-term impact on your business and would have been slightly positive from a utility point of view, but in 2027. So we look at it as neutral to maybe even slightly positive from an OpEx and sort of lines of operation point of view and the fact that we can now focus on just the new assays that we have. So I think it was very much overgrown in so many different dimensions. Very -- I think, happy and confident with the assay that was used there and certainly even more so where -- in terms of where we're going with the technology. And obviously, our conversations with physicians, both from our medical affairs teams and sales force and so on. We're going to essentially -- and we are, as we speak, informing them about the details of the study and the fact that it has very little, I would say, read through in terms of our assays today and reiterate the confidence that we have.

Got it. So speaking of apples and oranges, I'm going to ask you to do exactly that with the smart liquid biopsy enabled reveal data that you showed last Thursday. 80% longitudinal sensitivity, 99% specificity at the sample level. Could you help us think through what the appropriate sort of metrics were on your older assay that we should be comparing those numbers to? And how should we think about that surveillance sensitivity of 91% that you have on the older assay?

Yes, great question. So we have data sort of internally from some of these studies where we're able to see sort of how the old assay would perform versus this new one. And this new one essentially exceeds the performance in almost every metric, sensitivity, specificity and so on. So we know it's a strong improvement. When you look at the specifics from study to study, a lot of it is dependent on the number of patients and really the types of patients that are tested. Are they low-risk stage 2, high-risk stage 2, do they have stage 4 in the sample set and so on. So you really have to sort of look sort of both the types of patients that were there and then look at the confidence intervals. And from what we can see, this assay performed as well or better than even that study that you're citing. I think as you start looking into some other aspects of the data that wasn't sort of released last week that hopefully will be released early next year, such as landmark performance and so on. I think it will be very clear in terms of really how well this assay performed in this indication.

Got it. I want to switch to screening here with AmirAli. AmirAli, maybe could you elaborate a little bit more on the differences in the algo for Shield V2 versus V1? Were there differences in the training set, if any? And at what point would you consider adding other biomarkers, you sort of dabbled with protein for a little bit in ECLIPSE and perhaps immuno markers or something like that as well.

Yes, sure. So as a reminder, with Shield V1, we've got our ECLIPSE readout at 83% and 90% specificity, and we continue to believe it's above the bar for FDA approval, CMS reimbursement and guideline inclusion by ACS and USPSTF. But just as we always believe that wouldn't be the best that blood can do. So our team worked on Shield V2, a few months after almost a year after giving the V1 readout. Now we have some validation, clinical validation for V2 in a U.S. screening cohort prospectively collected that we won in a new set of 45 CRCs, down 84% sensitivity, which is very similar to what we found in ECLIPSE trial. And then V2, detected 40% to 50% of the CRC that got missed by V1 and ended up with a 91% blended sensitivity with 91% specificity. So delta was definitely on the algorithm side, on the assay lab chemistry, always the same, same NGS data, and then we get V1 performance versus V2. Really, the big delta is just more time, which means more data went into the learning system, more just higher number of samples and more representative samples coming from especially those malignant polyps stage 1s that typically they get actually caught during colonoscopy. So they don't get representing blood samples post-colonoscopy. They have more of those data sets now in the database of the algorithm actually made itself more sensitive. Analytical performance of the test has improved by 2x, meaning when the tumor level is half, we can detect them. And now we have this evidence that clinical sensitivity is better.

Got it. Your favorite question, AmirAli, on data drift. It came up at the Analyst Day, but I think it's worth sort of laying out, right? So you have that sort of 7-point data drift in ECLIPSE -- in this case, you had a 7-point improvement relative to the prior version of the assay. So just walk us through what gives you the confidence that you won't see something similar here?

Actually, there is a point in that data degradation that like sometimes real-world samples are not exactly the same set of samples that you generate your case control data. In case of ECLIPSE, some of the learnings that we got from ECLIPSE trial, there are these like very tiny malignants that you think they are polyp, you cut them, and then during centralized pathology and local pathology review, you figure out there what looks like some tiny CRC there, and they wouldn't have been like diagnosed at CRC cases actually. So that generated actually some of the drift that we've seen the data. I think the most important thing about this V2 data that gives us confidence is comparative performance between V1 and V2 are the same cohort. So first, V1 of 84%. It's very similar to ECLIPSE, meaning that the kind of CRCs we are dealing with are very similar. And then in the same cohort, there is a delta for V2 that we are detecting more. So there is no doubt in our mind that V2 is more sensitive than V1.

Got it. Advanced adenoma performance, I mean, stayed flat versus the ECLIPSE version of the assay, do you think until you start sort of going back to protein marker that's really kind of like what's going to drive that inflection in advanced adenoma. And is that something you're actively working on?

So there are different kind of biomarkers that we are looking in the research setting. I think there are opportunities still for cell-free DNA over time to get better. When you're looking at kind of different subclasses of advanced adenoma higher-grade 1 even in situ CRC stage 01, there are early indications that even more sensitive algorithm can cut more into the tails of it. So we see, over time, as again, more data gets generated, then algorithm gets better, how much cell-free DNA only device could potentially detect those cases? On the research side, we've been looking at different biomarkers. We have that experience with protein. Still, we are working on it, but there are a couple of other classes of biomarkers that they are interesting, but still they are more on the research side for us. I'm pretty sure V2 would not be the last version of the blood, just continuously this kind of technologies, as we generate data and understanding of biology performance, will get better.

I want to go to that survey result you showed at the Analyst Day earlier on 7 out of 10 individuals would not choose a stool-based test for CRC testing. Doesn't this result suggest that there could be sort of meaningful switching dynamics from stool to blood. Yet you're largely assuming penetration primarily in the unscreened patient population at launch. Isn't it sort of more difficult in a way to convert the unscreened population versus get people who are adhering to a screening protocol to switch from stool to blood.

The reality is there are pros and cons. So the Shield LDT commercial experience gave us a lot of actually good indication that a bunch of unscreened patient populations are seeing their doctors and in fact, we are drawing blood from them. So that's very exciting to see occurrence rate of 90% to completing Shield in unscreened patient population. When you look at the whole opportunity, still 50 million unscreened patient population versus maybe all in all, for stool-based testing, it's about 15 million annual testing still the biggest part of opportunities in unscreened patient population that would be huge for us to penetrate. But definitely, there is some kind of patient preference that plays a role here. Like when we look at the survey results that you mentioned earlier, when the people are given blood testing as an option, 7 out of 10 of them are gonna choose blood versus stool test. But we'll see. Definitely, we are focused on unlocking that unscreened patient population, but switching opportunity would be also considered.

Got it. One for Mike here, just in terms of Guardant Health in '24, in light of the OUS and MRD ramps underway, is there any reason why you wouldn't be able to grow your therapy selection and MRD clinical volume and perhaps even sort of modestly higher than the high 30s to low 40s kind of like growth rate that you're baking into the '23 guide?

Yes. I mean, I think at the Investor Day, we laid out a lot of opportunities particularly on therapy selection on how we can grow the volume. Definitely outside of the U.S.A. is a large opportunity for us. I think just even outside of Guardant360, we saw over 100% year-over-year growth on TissueNext so far this year. So I think that's going very strongly. We just mentioned on response. It's early days, but I think we start to see increased attachment rates with response as well. So I think there's really good opportunity on therapy selection for those volumes to grow. On the MRD side, we have talked about how we're sort of managing that volume growth based on reimbursement, and we're sort of running -- because we wanted to manage our cash burn because a lot of these tests at the moment are unreimbursed. I think if we can get CRC surveillance reimbursement from Medicare, that will allow us to unlock our MRD volume. And if that comes in 2024, yes, that could really help drive those clinical volumes and keep them at a similar level. But overall, again, our therapy selection, I think we're very bullish on the ability to continue to grow that volume.

Got it. What margins -- other than the ASP improvement in the smart liquid biopsy rollout, what are some of the other levers that you can pull in the near term here on the base business? Has the switch to the X already happened? And on OpEx, Mike, you did mention at the Analyst Day, you're confident in your ability to reign in R&D expense. Maybe just sort of like walk us through that because you're still going to continue investing in research and data generation, both in MRD and in screening.

Yes. I mean, on gross margins, no, the switch to the X hasn't happened yet. But I think that over time and over the 5 years, that can definitely be a driver for improving the gross margin somewhat. And really, our focus on the gross margins at the moment is on automation and driving efficiencies. And we were at pains to point out at Investor Day, all of the infrastructure that we've already built. So I think there's a lot of leverage we can gain now on the gross margin side. With research and development, yes, I think we said ECLIPSE has been a major expense for us over the last few years. That's winding down now. SHIELD LUNG has been an expense for us over the last 12, 18 months or so. and we're getting close to the complete enrollment there. So yes, I would expect, particularly on the screening side that R&D expense can come down next year. And we said we want to manage screening to this $200 million burn a year. So where we can reduce in R&D on screening next year is gonna allow us then to focus that spend -- similar level of spend in total, but we can start to focus that spend on the launch, and the sales and marketing is where we're going to need it when we get the FDA approval.

Got it. One final one for Helmy and AmirAli. Over the past year or so, there have been at least 2 occasions, I can recall, where there was a pure disconnect between investor perception of what had happened versus your own internal expectations and reaction to the catalyst. What part of the Guardant story do you think is sort of just grossly underappreciated by investors today?

I think a lot of parts that are underappreciated. I would say that -- maybe it's a multipart answer I'll give you. So I would say just kind of how much growth we still have left for therapy selection. I think that's an amazing business that now has very favorable unit economics with a lot of room for growth over the next decade, frankly. Not just because of increased penetration, but where the market is going in terms of earlier cancer and more indications and globally and so on. I would say -- the second area is really around kind of the sort of highest bubble that exists in liquid biopsy. I think because of our success, this has fueled billions of dollars of investments of other companies, other people chasing this. And so there is this sort of expectations that kind of get way ahead of their skis in terms of what's possible because you have a lot of unsubstantiated claims out there. And I can tell you that what we've been able to achieve has been sort of very extraordinary from a technological performance point of view, from a clinical validation and so on. And it's one of these things we continually see whether there's this sort of doubt, inflated expectations and then a trough of disillusionment and then kind of just getting to straight market adoption and the sort of needed things. So now we're there with therapy selection where we went through those undulations and now we're in the sort of great business point of view. We're going to do that again with MRD and certainly do that with screening.

Got it. Well, this was great. So thanks so much for joining me today.

Thank you.

Thank you.