Illumina, Inc. (ILMN) Earnings Call Transcript & Summary

Hello, and welcome to the 32nd Annual Piper Sandler Healthcare Conference. I'm Steven Mah, senior research analyst covering tools and diagnostics. I'm very pleased to have Illumina with us again this year and want to introduce Alex Aravanis, Chief Technology Officer at Illumina. Thank you for participating today.

My pleasure, Steve. I'm just going to start with my safe harbor statement. I've been asked to remind you that my comments today could include forward-looking statements. You should refer to our SEC filings for a discussion of the risks and uncertainties that could cause results to differ materially from our current expectations. It is our intent that all forward-looking statements regarding our financial results and commercial activity made during today's discussion will be protected under the Private Securities Litigation Reform Act of 1995. Okay. Steve?

Okay. Yes. Great, Alex. So yes, let's just jump right into it then. We'll start off first with some questions about the base business, pivot over to some COVID-19, then on some oncology topics and then some other general topics on maybe DNA storage and some of the trends that you're seeing going forward in synthetic biology. But yes, with that, let's just jump right into it. So let's talk about sequencing costs. So you guys recently announced the $600 genome, you're in process of launching NextSeq 1000, which will be at a lower price than the 2000. How should we think about sizing the markets that these new lower price points will open up?

Yes. Yes. So what we appreciate is that there's some significant elasticity in the market and some gaps in offerings that could enable more customers to do the type of sequencing. So these are, in particular, in the mid-throughput, people who'd like to be able to do a whole genome even if it's small numbers but difficult because of historic pricing in the mid-throughput. So in particular, we released the v1.5 agent -- reagents. This makes it possible for those mid-throughput NextSeq customers to now do a $600 genome, giving them that kind of pricing, not too far off than our high throughput customers. There's also customers who are doing single cell workflows and being able to do 2, 3 more times the cells, opens up projects that previously, they may not have been willing to do, but now that they can power them statistically. We've seen some nice anecdotes of that movement already. One example, I think that's already been shared is with Princess Maxima Center, which is the largest national pediatric oncology center in Europe. That customer was faced with a choice for their cancer sequencing in these children to guide therapies either to outsource it or to bring it internal. With this lower pricing, it's made them possible to do that internally, right? And that's a great thing because we know that when you bring the sequencing closer to a patient, it gets used more often and there's actually evidence that the outcomes are better. So it really gets to the patient and an opportunity and a placement for us that may not have happened otherwise. Regarding the NextSeq 1000, we appreciated that there were some customers where maybe the NextSeq 2000 was a bit out of reach for them. And so the lower price NextSeq 1000 could get them into that mid-throughput range, enable them to do some larger projects they couldn't do. So we think it's an important -- to create that kind of bridging opportunity. And of course, those customers with success there and have the option of eventually upgrading to a 2000.

Okay. Got it. And maybe just a follow-up question on kind of lowering the cost of sequencing. So one of the long read sequencing companies as -- they've said publicly, they're pushing down costs as well. Commented they would open up new markets, such in clinical areas, which historically are better suited for short reads. Can you talk about how you guys are positioning against long read? I know we've talked about kind of the lower price points. But maybe talk about what else you guys are doing? And if you guys are also considering developing your own long read technologies or other platforms.

Yes, it's a great question. Maybe just first to level set a little bit and then get into what we're doing to address broader customer needs. Today, on our platform, like we talked about, you can do on a NovaSeq, you can sequence 50 human genomes in about 2 days for about $600 a genome. So, right, my appreciation of where the long read technology is today, it's over $3,000, and it's 1 genome per 1 instrument for 2 days. So pretty big difference today, which is why most customers are continuing, for their bread and butter applications, to choose Illumina. If we think about the future, right, and let's give them the benefit of the doubt that they get down to $1,000, we've talked about our road map to the $100 genome. So in a future state, we're talking about potentially a 10x difference in price. In addition to the scale, the quality and cost. But in addition to that, the way I think about it as the CT of the company is customers aren't really interested in long read or short read. They want to do their science. They want to be able to detect the things they're interested in. Or in their patient, they want to be able to see the alterations that they care about. They don't care about long read or short read. So what we've been doing there is expanding the capabilities of our platform beyond what people thought probably was historically possible. So today, you can now -- we now have a tool called Expansion Hunter, which allowed you to do repeat expansions using our platform. It's a very important clinical application, especially in pediatrics. You can now do very accurate pharmacogenomics genes like CYP2D6. And so over time, we're going to continue to add on these capabilities. And the future we see is customers won't have to rely as much on niche technologies, expensive niche technologies, and they'll be able to achieve the applications that they want. A lot of this is through our breakthroughs in data science and machine learning. We made big investments there. We also acquired a team, a company called Edico. That's really paying off. And we're really kind of breaking these barriers of what people thought historically you could do with shorter read technology. And we're accomplishing those today. And so I'm not aware of a major application where we won't be able to address that. And then again, given the large cost differential, the scale, we think we're well positioned for -- to continue to be the platform of choice.

Okay. Well, great. And yes, maybe to take advantage of having the CTO with us today. So let's pivot to something that's more timely. And let's just talk about COVID-19. So you have a partnership with Genomics England, NHS, to study individual's genes and how they response to SARS-CoV-2. Could you give us a quick update on how that partnership is going, like the study design, when we might get some data? And then also maybe a sense of the clinical utility these results might give you and how Illumina might potentially monetize it?

Yes. We're funding our technology, it's playing a significant role in supporting the efforts against the pandemic in a few different ways. One is a series of research studies, some of which are quite large, like the ones with Genomics England and NHS, where they're going to sequence about 35,000 COVID patients. So they're going to do whole genome sequencing. They're also going to do viral sequencing. And the aim here is to really understand why there's such a wide spectrum of disease severity. Some people asymptomatic, some people develop severe disease, some even to succumb to it, right? And while things like age can explain a bit of that, most of it or much of it is probably related to other biology. So really important in understanding how to manage this long-term and future pandemics. So that project is going well. I can't speak to the exact timing of it. But it's full green light there. So we expect that to happen in the near term. The other -- one of some of the other applications are with the vaccines, right? So our technology was used to sequence the first COVID genome, which really defined it as a novel coronavirus on January 11, that began the vaccine race. So happy to play that part with our collaborators in China. And then -- but even post vaccine, right, it's going to be very important to understand, are the strains changing over time? Is there a selection pressure? Is there a new strain emerging that might have resistance to the vaccine or additional virility? So we think that's an important opportunity for us. Also, and then long-term surveillance, right? There's initiatives to set up periodic sequencing, early in outbreaks of new, for example, respiratory infections. We're excited about those. Imagine if we had known 3 months earlier that this new novel potentially devastating virus is in circulation, what we could have done.

Right. Right. Okay. Well, perfect. Yes. All right. So maybe let's just pivot on to the -- your oncology business. So you obviously, you guys are in the middle of acquiring GRAIL. And you also have a number of oncology-based content partnerships as well. So I know that's not necessarily competitive with the cancer screening and early detection technology that GRAIL has and a lot of these are more on the therapy selection. But how should we think about your oncology partnerships and your oncology focus going forward? Is it going to be more internally focused? Or are you going to continue to partner out more like a companion diagnostics partnerships with BMS and Loxo just as an example.

Yes. I mean we think our partnership strategy in therapy selection, in particular, is the right one. So we have great customers who built laboratory developed tests, done a lot of innovation for therapy selection, both in tissue and from blood. They're serving a significant portion of the market. But we appreciate that there's a need for broader distribution, right, for both tissue and blood sequencing for cancer, right? And so enabling more labs to be able to do these tests both for pharma trials and for patients. And then ex U.S., in many places, there isn't an LDT mechanism. So you need IBDs, you need distributed tests. So we continue to think that's a great opportunity for us. We've done many pharma partnerships around developing CDxs and continue and see the opportunity to do more of those to broaden the access of our technology beyond what some of our initial customers have done. And then we believe that cancer screening and monitoring opens up single biggest opportunity for us in the next decade, right? We think going out, that could be up to a $75 billion TAM. And our approach to open that market up because we think our participation can really accelerate it is through the acquisition of GRAIL.

Okay. Got it. Got it. Okay. So then -- yes, maybe just stepping back, as you think about R&D spend, how should we think about that? I know historically, it's sort of been in the 18% of sales range. Is that sort of the right way to think about it going forward? Or will you invest in these new opportunities?

Yes. I mean -- so every year when we do our budgeting, we think about, given the opportunities that are ahead of us, the technology innovations we have and can develop. What is the right investment to achieve those, to develop the best products, our long-term competitiveness and create shareholder value. In general terms, I would say you should expect comparable levels of investments and again, it will vary depending on the opportunity, but I would -- in gross terms, a comparable level of investment. The bulk of that will continue to go into our underlying sequencing technology to ultimately realize the $100 genome and to realize the additional benefits in scale and quality and some of the things we talked about earlier, additional applications, filling up the full menu. But there is a healthy investment also in downstream analysis, things like bioinformatics data science. We recently won the PrecisionFDA competition, for example, and in our clinical business, right? So we have TruSeq (sic) [ TruSight ] Oncology 500. That is comprehensive cancer profiling, we're doing the CDxs around that. We're going to invest to make sure that, that's successful. And then there's also our NIPT business on things like VeriSeq V2.

Yes. Okay. Great. Okay. So let me maybe ask a few questions on GRAIL that we've been getting from investors. So first question we've been getting is on the PATHFINDER study, so it's an interventional study, follows patients for 12 months is our understanding. So 1 question we've been getting is, are physicians actively going to direct treatment based on Galleri results? Or is it just limited to follow-up imaging and other follow-up stuff? And the other question we've been getting is, is there a specified endpoint that you guys are looking for in PATHFINDER? And then finally, when will there be a sort of like an interim or first look analysis of the data, given that it's only 12 months?

Yes. Happy to answer those questions, maybe I'll answer the last one first. That's easy, which is GRAIL has communicated that results from PATHFINDER at least initial results will be shared first half of next year. So that's our expectation. In terms of the -- how the test is being used, it's precipitating the workup that leads to a cancer diagnosis, right? So at sites today, right, and it's the only test today that's actually being piloted for multi cancer or has FDA approval to do this type of study. The test is being used at Cleveland Clinic, the Mayo Clinic, Sutter, Intermountain, Dana-Farber, Oregon Health Science, so really leading institutions. And they enroll patients, these are people without a history of cancer, essentially as part of their yearly physical. They're offering the test. Most, the vast majority of them get negative because of a very low false positive rate of the test. And those that get positive, then usually get a tissue of origin in the vast majority of cases, which then precipitates a workup. And so the treating physician there decides what to do next. In most cases, presumably, what they're going to do is follow standard of care. So if the cancer indicates that there's a head and neck cancer, they would likely order an endoscopy. If that confirms the diagnosis, then they would move to an oncologist to stage them and develop a treatment plan. So the test is not going to direct their treatment. That will be determined in -- based on standard of care, but it gets them to the point where they're afforded an opportunity for treatment that they wouldn't have been otherwise or not until the cancer was potentially at a later, more difficult stage to treat. You also asked about the endpoints. The key outputs of the study are going to be 2 types of things. One is there will be data on test performance, things like positive predictive value, false positive rate, tissue of origin accuracy. We expect those results will be comparable to their previous studies. And then there's another set of things which are quite important if you think about it as a pilot for a commercial launch. Things like how quickly do you get to a diagnosis, how many procedures does it require, what is the qualitative experience of the patient and physician. So GRAIL really wants to understand those things, not because there's necessarily a big risk, but to confirm that these things can be done in a very comfortable way and optimize -- take those learnings and incorporate them into the commercial launch.

Got it. That's very helpful, yes. Okay. So also, yes -- so maybe just pivoting on to your next study that you've announced. I know there's a pilot study, again, with National Health Service. I think it's 165,000 people. I think it's like a mix of people that are suspected of cancer and then a mix of patients with no cancer signs. And maybe discuss why that trial study was designed like that, whereas PATHFINDER was people with no history of cancer.

Sure. Sure. Happy to go into those 2 different use cases and what's the breakdown in mix there. But a few things just to say of sense, we were excited to see this partnership between the NHS and GRAIL. We think it's validating of multi cancer, right? The NHS is a sophisticated, very innovative, large health care organization. They really understand health care economics. They're very evidence-based. So they appreciate the need for multi cancer and that they see GRAIL as a leader to partner with, we think, is validating of our belief on the value of GRAIL and their leadership in the field. The partnership they're doing is really interesting. It has a pilot phase, which is going to have these 2 different cohorts. We'll talk a little bit more about that. But then it anticipates larger access. So it's not just a pilot or a study in which the NHS is going to contribute significantly to, but it anticipates a larger rollout. And that's exciting, too. The numbers that are discussed are up to 1 million patients. And that could happen as early as 2024. That's obviously a very significant opportunity to decrease cancer mortality in the U.K. and obviously, a meaningful business opportunity also. So I think it's a proof point to our hypothesis about GRAIL about these large opportunities that exist. And again, that's just one -- an opportunity in 1 company. With respect to the particular cohorts, it's about 165,000 people who are average risk. They have no history of cancer. So it's really a yearly screening test for them and of not too dissimilar from PATHFINDER, but much larger. And then there's a really interesting group, about 25,000. And these are people who have a suspicion of cancer. And so what the community, both in the U.S., U.K. and worldwide is appreciated is, unfortunately, people often go a long time with signs and symptoms of cancer, but their diagnosis can take a long time. And during this period, they'll get misdiagnosis, they'll confuse the cancer with something else. They'll get lots of additional tests and procedures. And that delays the diagnosis, right? And so potentially could have been at an earlier stage, had you kind of clinched it right away. It also means that there's a lot of unnecessary cost potentially. So those particular value propositions, which is in these individuals where there's a suspicion of cancer but it's not overwhelming yet, could you accelerate their diagnosis 6 months, save cost, save burden on the system and get them to their treatment faster and to analyze those types of metrics in that important population that they're interested in.

Okay. So the subset of 25 patients that have -- so these are people that are being like CT screened or higher risk or have some sort of familial risk for cancer?

No. It can be a mix of -- so it's a mix of things, right? It could be people who have mild weight loss, fatigue. It could also be people who have some family history, right? It's really people where to the physician, cancer is on the differential diagnosis, right. And physicians have this dilemma, right, which is, do you want to send people on to a CT scan? Do you want to get them a PET scan? Do you want to do additional testing? They're probably at higher -- we know that people who have these symptoms are higher risk for cancer by a significant fold, but many of them still are not going to end up having cancer. So what is the trade-off in resources spend, subjecting people to the testing. And there's a lot of subjectivity in that. So it could really help in that dilemma, which is do the GRAIL test. If it finds cancer, we know the accuracy, then you go right to getting that patient treatment versus either deciding to watch and wait, which could potentially be bad if it turns out it is cancer or subjecting them to test, which are unnecessary.

Okay. Got it, yes. So yes, that makes a lot of sense because, yes, I think the price points that have been kind of thrown out there. And maybe if you can remind us of the price points that have been -- that Illumina has thrown out there. At that price point, it's enough to be used as sort of like a triage test to either, yes, to your point, go to watchful waiting or to intensify the surveillance methods to CT scan and PET. Is that the right way to think about it?

Yes. So I think you should think of it as an additional test, a diagnostic aid, right, that could, for certain patients, quickly get them to a diagnosis that they wouldn't have otherwise or that would have taken much longer. And so it uses very similar underlying technology to the screening test, but it is a different test and has some different features to it for this population. And it's also classified as a diagnostic, for example, in the U.S. versus screening test. So that underlies some of the differences in price. It also has a different type of evidence and a different reimbursement path.

Okay. Got it. Okay. And so could you remind us again what the price points you think the test will go out at?

Yes. Yes. So what GRAIL has talked about, I believe, for the screening test is $600 and $900. And then I believe for their diagnostic aid in cancer test, it's on the order of -- I probably don't know the exact number, a couple of thousand. We can get back to you on that. I believe it's in their filing.

Yes. Okay. Yes. All right. And then maybe in the last couple of minutes that we have here, and maybe pivoting away from this, but going into more kind of the long-term growth drivers and long-term trends. There's been a lot of talk now about doing DNA storage. So you recently announced a collaboration with Microsoft, Twist, the Western Digital, to form a DNA data storage alliance. Could you expand on the role that Illumina is playing here? And when do you think this sort of data storage is going to be commercialized?

Yes. Yes. Yes. I mean this is an early -- so it's early days in DNA-based information storage, but it's exciting because it's starting to become real. If you think about any information technology, there's obviously a write aspect of the technology and a read aspect. We think NGS, especially given our incredible scale and throughput, the Illumina platform is a very natural read technology for DNA-based storage. So we're excited to build the foundations of this industry, start to build up some of the standards that could be used, along with Microsoft and Twist Biosciences and Western Digital. There could be early commercial applications in the not-too-distant future. So exciting, and we've been at this kind of beginning of these things in the past, be it single cell sequencing, things like that. And so this is one of several opportunities that could become a meaningful business for us long-term and playing -- helping build the foundation of that, we think is really important.

Okay. That's great. And maybe just the last -- we're a little bit over time now. But just real quick, just an overarching statement. So I know synthetic biology is sort of a very early burgeoning field. But can you comment on the long-term prospects for -- of Illumina in synthetic biology and where you think we are in maybe using a baseball analogy in terms of like innings? So are we in the first inning? Or third? Or kind of what's your sense to kind of level set people?

Yes. I mean I don't think we're in the first inning, but probably still early innings. There's real businesses being built on this that have real customers. So companies like Ginkgo, which we've invested in. So this is, for example, compared to DNA storage, this is a lot farther along. So this is a real business for us. We expect it to grow significantly over time. Many of these approaches to synthetic biology involve deep sequencing for machine learning, lots of combinatorial approaches. And so it's a great opportunity for us to support the development of these companies.

Okay. All right. Great, fantastic. So we are out of time, unfortunately. Alex, I really thank you for participating today and for Illumina, also participating in our conference again.

A pleasure, Steve.

Really appreciate it. Thank you so much.