Personalis, Inc. (PSNL) Earnings Call Transcript

I'm Tejas Savant, and I cover the life sciences and diagnostics sector here at Morgan Stanley. Delighted to have Personalis join us today, and representing the company are John West, CEO; and Aaron Tachibana, CFO. Welcome, gents. Before we get started, I have a quick disclaimer I need to read through. So please note that the webcast is for Morgan Stanley's clients and appropriate Morgan Stanley employees only. The webcast is not for members of the press, so if you're with the press, please disconnect and reach out separately. For important disclosures, please see morganstanley.com/researchdisclosures. And if you have any questions, please reach out to your Morgan Stanley sales rep.

So with that said, John, welcome once again. Maybe just to set the stage, could you share your perspective on the evolution of Personalis over the last 5 years or so? I mean maybe even before -- I mean, starting with your ACE exome technology through your ImmunoID NeXT platform?

Yes, I'd be happy to. Thanks very much for having us here today. I'd say that, I mean, my prior experience was really more in the sequencing equipment on reagent side. So I ran a company named Solexa that was ultimately acquired by Illumina and has become the basis for all of their sequencing technology today and ran the sequencing business there at the -- back at the beginning. We moved to creating a new company when we actually felt the platforms were doing well. But we said, the real question is, how are we going to use this? What's the medical opportunity here? And we have ended up picking cancer, because cancer is such a complicated disease that we felt it would really benefit from the deep data that was possible with next gen sequencing. But unlike some of the other companies in the space, they're doing great business, but there's a lot of focus on next-gen sequencing, and cancer genome must be a diagnostic company. And actually, what we saw as being the real unmet medical need is on the therapeutics side, because the diagnostics of today can only choose between the drugs that are available today. And sorry, in cancer, the drugs just aren't good enough. And what we saw, in particular, in the 2014, 2015 time frame, was the emergence of a new class of cancer drugs, where those drugs don't themselves attack the cancer. They enable the immune system to attack the cancer. And coming from a genomics background, I actually took one look at that and said, "Wow, this is a complete change in how we think about the genomics of cancer." Not only do you need to know all about all the genes, not just the cancer driver genes, but all the other genes can lead to mutations that can spur an immune response, but you also need to see the immune system. You have to care about the T-cell repertoire and the B-cell repertoire and the gammadelta cells and so forth. And so we set out to expand from a broad platform, which had -- we had always looked at all 20,000 genes and the things that we did, but we said, we need to add the immune system to that as well. And ultimately, it took us 4 years, but we were able to release in mid last year, our NeXT platform. And we believe it's the only platform so far that captures all 20,000 genes of the human genome in the cancer cells, but simultaneously from the same assay, can capture a comprehensive view of the immune system. So T cells, B cells, gammadelta cells and also the innate immune system to be able to see NK cells and others from the innate immune system, dendritic cells and so forth. And so this is the first time you have that kind of visibility. So we brought that out last year. And then the goal this year has been to take that into adding the cell-free DNA capability. And a number of companies that only have cell-free DNA, will treat cell-free DNA as if it's an alternative to tissue biopsy and almost a Coke versus Pepsi kind of presentation. And I'm sorry, we're scientists. It's nonsense. They -- from the tissue, you can see things that you don't see in that plasma. For example, we got beautiful data on the RNA from a tissue sample. We're not going to get that on a cell-free DNA sample. We can see the immune cells that are inside the tumor. Again, we're not going to see that in the cell-free DNA. What we do get from the cell-free DNA is the ability to monitor and see things changing over time. And so we see these 2 as complementing each other. And so the evolution here has been to start at 20,000 genes, add the immune system. And then, now we've added the capability introduced in early August, again, at the 20,000 gene level, to be able to look at how the cancer is changing over time using cell-free DNA. So Personalis is one of the few companies that actually has expert capability, both in handling tissue samples and also liquid biopsy samples. Most companies focus on just one or the other. But they really are powerful together. And so that's kind of how we've gotten to here. And then I think going forward, we'll be leveraging that even further with applications like minimal residual disease and monitoring cancer in detail with personalized assays.

Got it. That's a great overview, John. So maybe we'll start on the POPSEQ side, just given the news this morning that the VA extended your contract through next September, and they've also issued a task order worth $31 million or so. So how does -- first of all, how does the size of the task order compared to your expectations? And then how many samples do you expect from this task order? And how long of a runway does that provide you on the MVP side of the business?

Yes. So first of all, we're thrilled with the order. We're thrilled with the extension. We've been working with the VA since 2012. We've been their only provider for whole genome sequencing. And it's our understanding that in this case, again, yet another year, we're the only provider that they've selected for that. What they've done is the -- with these contracts, when we receive the original contract, it's a 1-year contract, and they have 3 years where they can add on a year at a time to the contract. And so this was adding -- they've elected to exercise once again 1 more year extension. So this is the fourth year on the contract that we have. This allows them to issue task orders. They can issue actually as many task orders as they want. Some years, it's been one. But some years, it's been multiple task orders. And so they have the ability under this new extension of the contract for almost another year now to be able to issue additional task order. So the first task order is $31 million. We don't know whether there will be additional ones or not. There certainly could be. The VA has a lot of samples. I mean we've -- even though they've had to slow down collecting samples during this COVID period of time, the most recent numbers we have are that they actually have 850,000 U.S. veterans approximately that are enrolled. And with this new order, we're under 150,000. There's 700,000 samples already collected by the VA, which they have not yet contracted for sequencing. So we see a long runway in that area. Obviously, as the cost of sequencing comes down over time, that will become more and more affordable. We also know that the VA recently was originally a 1 Million Veteran Program. They started collecting samples in 2011, and they were doing about 100,000 a year, and they're getting up over 800,000. They realized they were going to get to the 1 million and so they had a question, do we stop when we get to 1 million? And I think they've realized, this is way too powerful. So they made a decision to actually double the scale of the project. So now the target is to enroll 2 million veterans. And if you look at the VA as a health care system, there are about 9 million veterans that get their care from that system. So to start with, they were aiming at a little over 10%. Now they're aiming a little over 20%. You can kind of connect the dots and extrapolate a little bit. It's not hard to see that they were -- at some point, this should be part of health care broadly in the VA system, and you would sequence everybody who's involved there. So that may be the vision of where this heads to, but that's -- this has a long ways to go. We think it will become more clinical. It wouldn't be surprising to see pharmaceutical companies get to be involved. And actually, that's the most recent announcement we had in that space, we actually just hired somebody with that kind of experience to help advance the field there. So we're thrilled with the order. I think it gives us good runway through most of next year. There are thousands of -- we already have all the samples we need for the rest of this year. We had said we're aiming at being -- that we think, potentially the first for-profit company to actually sequence 100,000 genomes in the United States. And we announced the other day, we're up at 87,000 and still cranking. So I think we're optimistic. I think we're going to get to 100,000 this year.

Got it. Got it. That's great. And then, John, one of the things you've mentioned before is that, this year didn't involve a rebid, but next year does. How -- walk me through how that process goes? And is it sort of like, essentially, given that you will certainly be in the mix to be considered, given the work that you've done with the program? Or could someone else actually come in? Is it -- like, how do they even evaluate? Like, is it just a price-based thing? Is it sort of the quality of work you've done before and if you've had past experience doing it at this scale? And how many sort of potential bidders typically are part of such a process?

Yes. So we're open. This will be a competitive process next year. We've been through that several times. We've won every time. We will still be very careful to make sure we make the best possible proposal that we can and we're as competitive as possible. I think with that said, if you are running a project like this, you would not want to switch providers halfway through, because regardless of efforts to be -- to do what seems like the same thing, there will be subtle differences in the data between one lab and another lab. And so since we're often using this kind of data to look for subtle statistical differences between people who say, have a particular disease or don't or have different variations on the disease, the last thing you want is some other variable that's screwing the data. And so although the VA, of course, has the choice to pick any vendor that they want and other people, it will be open to people who can bid. In prior years, the criteria have included that the samples all have to stay in the United States, that the company that does this needs to be a U.S.-owned company, that they want to do it through a company, not through a university arrangement. They have a lot of understandable concerns around the data security issues because this is essentially VA, these are veterans' medical data. And even though it's in a research program, it still says a lot about the medical circumstances of the veterans, and so data security is a very serious issue. And I think also, you have to be credible as somebody who can actually operate at this scale. I mean, in the last 12 months, we've sequenced over 50,000 human genomes. I think the last couple of quarters, we've reported 14,000 a quarter. So maybe probably comes to 56,000 genomes per year rate and accelerating. So if the VA is going to select somebody, they will probably want to take the scale even further up. And so, if somebody says, well, I have a lab and we bought an Illumina sequencer. We think we could sequence tumor tissue and genomes as well. It's like, where's the 50,000 you already did? And so I think it would be partly what Personalis has learned over the 8 years that we've been involved in this, is that it's -- there's a lot to be learned by building that experience. And it's all the automation systems, but also the data quality systems and the -- all the data processing and delivery and backup and many different elements of this. And so we think that, that experience will be something that will be, that sort of anybody else that's going to compete for this would need to bring a similar kind of experience, and we think there actually isn't anybody else. There are no other companies that we're aware of that have done this scale of whole genome sequencing as a service on a commercial basis. And so, although, we respect, well -- there certainly could be and we will do the best we possibly can, we do think we're in a great position. And frankly, we've worked hard on it for the last 8 years to get to that position. So we do value the relationship a great deal. We're celebrating the new order we have and sort of step on the gas, here we go.

Got it. So on the last call, John, you'd commented about sort of this large batch of shipments from the MVP before they close their lab in April. Can you comment on trends -- in terms of sample flow from the MVP recently? And have they resumed sort of a normal cadence now?

Yes, they have. The VA lab is open again. The VA lab does at least 2 things. One is they receive blood samples that come in from VA hospitals all around the country where they're collecting those samples and they extract DNA from it. And they store it in an enormous liquid nitrogen-based cryogenic freezer. I mean, it's really enormous -- robots inside and everything. And the second thing they do is then as they select the ones that they want to have sequenced, then they send those to us. So as it turns out, during this period of time, during the pandemic, the collection sites are temporarily paused. They're not collecting new samples that way. You can register online. But most of the in-person collection that would happen at VA hospitals is currently on hold, which means that the only thing that, that lab has to do is send samples to Personalis. And so actually, we have a great relationship. They're very hardworking people, and we've been able to work with them so we've seen the flow of samples coming up. We actually anticipate with this -- just received this new task order yesterday. But we actually think it's likely that we'll start seeing samples from that soon. We actually -- we're still finishing samples from the prior task order. That will take us into next year sometime. But we anticipate that we'll start receiving samples under the new task order way before they're sort of needed. So we'll have -- it gives us a lot of forward visibility on revenue. As we map this out, I think it gives us a lot of clarity, probably at least until this time next year. And by that point, we'd be talking about a new contract, which we're optimistic will be at potentially quite a bit larger scale even.

Got it. And John, on that point, in terms of the new contract, do you think the sizes of the task orders could sort of increase dramatically? Because one of the things I wonder about is that you have all these samples in-house, but they've only seek -- they're only sending very -- a small fraction of those in small batches relative to the size of the overall project. Do you think that sort of cadence can pick up meaningfully? Or are there just certain, like, contractual sort of structures in place that prevent the VA from doing that?

No. I think -- I mean, we're -- certainly, we're open to doing whatever scale they're interested in, and we actually have more capacity than is currently being used to be able to be responsive on that kind of thing. And the VA certainly has the samples. So I think the issue is just the budget in the VA. Generally, this is a field that's been -- a project that's been supported by leaders of the VA all the way through the program. A couple of years ago when the VA's budget was being submitted to Congress, the White House issued a press release about this budget. It was going to be over $200 billion. And in a 1-page press release, which is mostly about hospitals and the medical care, they also mentioned the Million Veterans Program. It gives you an idea out of $200 billion budget, it still ranks being on the 1-page press release about what the VA is doing this year. They view it as being such an important asset to what the VA is doing. So -- and I do think it's really been visionary on their part. In fact, when they started talking about this, they were beginning to talk about sequencing 1 million people back in 2006. So I mean, in 2006, we were barely sequencing the first genomes on the first LexA systems, and the VA was already thinking about 1 million. So I have to -- hats off to them for being so forward-looking and realizing the power of the clinical data they have and how that could be combined with the genomic data.

Got it. And then I want to spend a couple of minutes on your sort of POPSEQ expansion plans, right, because -- can you talk to us about how the work in the MVP Program helps you relative to the competition as you bid for some of these projects? I mean, obviously, you get instant street cred because of the MVP. But beyond that, just on a -- from the point of view that these programs, you're likely going to have to either partner or build out your own lab in a new country, what exactly is it that Personalis, because of your MVP experience, can do much better than your competition?

Yes. So being able to sequence a few genomes is one thing. Being able to sequence tens of thousands is something else. And key elements that actually aren't so much to do with the sequencer, it's to do with the automation of the sample prep, it’s the data systems. We've had 8 years to develop the data systems that we have. And we've implemented them. Because our lab is a CLIA lab and CAP-accredited, even though this particular work we do for the VA is not a clinical test, we've actually been implementing things with the whole rubric of a CLIA laboratory. So everything is under SOPs, all the qualification of reagents. So there are many different elements of that. And so those are also elements that are helpful. When we talk with people at -- with other large programs like this, many of them actually are moving more quickly to establish a clinical side. They want in their countries, they're interested in having population sequencing, but they want it to be an integrated part of their medical system. And that's kind of what's happening in the U.K. now. Genome England has transitioned into the NHS system or is in the process of doing that. And so I think that's where a lot of these other countries and other groups will be headed. So the fact that we -- they may ask, we know you've done a lot of research sequencing, but how do you know that you could really do this from a clinical standpoint? And we can say, well, prior to COVID, we'd say you'd be welcome to come see us in the lab and see how we go about that. And what people can see is because we are also a clinical diagnostic laboratory, the way that we're operating things sets us up well to be able to deliver clinically as well if that's what's required. And the automation we've put in place, particularly the automation around the data, not just the sort of pipetting robots, but also the scale of the data -- for example, in the second quarter of this year, Personalis sequenced approximately 1,400 trillion bases of DNA sequence. So that's not something you're just going to attach and send to your -- with an AOL e-mail, right? That's managing data at that scale and keeping -- making sure that you have all the QA systems so that there's never a sample mixup and so that you can -- if those were clinical samples, you'd want to meet that scale. That's part of what's transferable. And we are in the process -- to the extent that people may ask, "How do we know you can do it in another country?" We're actually in the process of doing that in China. So we've announced that we're developing a capability in Shanghai, China. We're in the process of building that out now. That will be the leverage. And so for many people, the view is, okay, I believe you could do it in California, maybe you could do it somewhere else, but if you've also been able to build out a lab and transport a lot of that to China, there's a lot of other countries I could believe you can do that with as well. And so we're leveraging the experience of going to a second country and having the second country isn't Canada, which -- we'd be happy if it were Canada. But I think if people see us go to China, if you also then wanted to go to, what do I know? Brazil or somewhere else is clearly incredible that you could do that kind of thing.

Got it. And then, John, in terms of just the whole sort of U.S.-China geopolitical rhetoric, is that something you worry about? I mean, obviously, you already have presence there. And how have those conversations gone? Because every once in a while, we view -- especially stuff like genomic data, the governments of multiple countries view that as like sort of sensitive information that shouldn't leave the country's borders. And then China and the U.S. with that rhetoric is that's an even more sensitive consideration. How do you think about that?

Yes. It's actually the reason that we're building the lab in China is because we have large pharmaceutical partners who said, "Love the platform, but we can't -- we're running drug clinical trials in China, which we want to get approved so we can actually also market these drugs in China. And we can't get the samples out of the country, and we can't take the data. So can you please set up a capability to implement your next platform so we can have the same process for processing samples outside China and inside China?" And so that was the original motivation behind that. I would say that China has a long history of being involved in advanced genomics research. There were people in Beijing Genomics Institute before it was a company, when it was just a research institute in Beijing. They were involved in the original sequencing of the human genome, starting back in the 1990s, were involved in subsequent large genomic projects. They have others now. So I think it's an area where there is quite a bit of expertise and appreciation for the advantage of that. There is rhetoric between the current administration and the Chinese. We view this as a much longer-term issue than that. And I think although it gives us some headaches and it’s hard to fly to China these days because of the coronavirus, and there's issues about getting legal paperwork signed off because consulates are closed and things like that. But I think on the time scale that we're looking at here, working with pharma, there is cancer in China. And whoever the administration is here and in China, there will still be cancer in China. There will still be a need for cancer drugs in China. There's an overwhelming medical need here. And I think once the rhetoric is sort of aside, people are going to move on with doing what they can for their citizens to help move the ball forward in terms of cancer. And everybody knows that genomics is really key to that. So yes, there are some headaches, but I view that as a -- it's more of a speed bump than an obstacle. This is an important long-term need, and we plan to build that out.

Got it. Fair enough. And then on your hire of Kevin as head of pop-gen, I mean, he comes with a lot of sort of experience with Genomics England and then Genomic Medicine Ireland as well. What's -- so beneath him, talk to me about sort of the commercial build-out that you need to put in place as you go about sort of executing or bidding perhaps for multiple pop-gen projects across the world? And then secondly, just given the COVID-19 pandemic, has that pipeline kind of like slowed down a little bit because governments are focused on their COVID sort of response at the moment versus population sequencing?

I'd say that the people who were interested in population sequencing have often been interested for a long time. Sure. There will be some distractions because of COVID, but there are a fair number of groups that have budgets, who want to move forward, are interested in doing things. And so exactly what the time and scale is may be modulated a little bit. But I think one of the things we see in these projects going forward is that there may be multiple stakeholders involved in these projects, so whereas the original population sequencing projects were pure research projects funded by individual governments, so kind of almost a single stakeholder that would have the samples and get the data. Now we see this becoming much more of a multi-stakeholder situation where there may be research groups that -- and legitimately, they have a fantastic opportunity in research. Those may be government-funded, but there's also the clinical side. As I mentioned, many of the groups that are interested in this are not planning on waiting there. They plan on having clinical be involved, right, to start with. So now you have that group of stakeholders involved, which may include both the medical folks but also the -- whoever the payers are in the country that's involved, and then also pharmaceutical companies. And this is one of the things that Kevin brings. His most recent experience actually was with Genome Medicine Ireland, where although they modeled a lot of that after what was learned by the Genome England prospect, one of the things they've realized is, really should get pharma involved early. So they had a number of pharma partnerships there, still do. One of the ones that was publicly announced and was quite large-scale was with the pharmaceutical company, AbbVie. And I think that pharma is interested in this kind of data, because it's actually a good thing for many of the countries involved because they're interested not only in improving their medical systems, they see this as part of economic development for their country. And so they see genomics as part of the future of their country, but they also see biotech as an important part of their -- and the pharma industry as being an important part of their economy in most western, more modern countries. And so the -- those are opportunities then where you really want to bring together the researchers, the clinicians, and the pharma folks and say, how do we bring this together and have an effective program that can serve all those needs together? And the fact that he's actually done that, actually aren't very many people in the world who have not only thought about the idea, but actually done it. And actually worked with pharmaceutical companies in a large scale, they were aiming at 400,000 genomes. I think it's sort of still in process, but its -- he's been involved in actively doing it. And that's kind of our view of where this heads in the future. I think our view is, that it also, in time, moves on from just the germline genome to, as you see, Genome England now doing more and more in cancer. We anticipate that cancer will be an important part of this, particularly on the clinical side. And so this also, you start to see the convergence between what people have maybe seen as being 2, somewhat separate parts of the Personalis business. In fact, from our point of view, it's all 1 business. And it ends up -- these pieces are going to increasingly overlap over time. We think the cancer part will become an increasing part of population sequencing. We think the pharmaceutical companies that we've initially worked with, primarily on the cancer side of our business, are likely to be increasingly engaged in what we're doing from a population sequencing standpoint. We think our experience with whole genome sequencing, which has become more from the pop-gen side, we've already begun to offer that in cancer now. And we think that will be used more and more in cancer. So I see a lot of convergence between these different parts of our business.

Got it. So on the biopharma front, John, a couple of questions. I mean, one, obviously, you've seen some strong orders on ImmunoID NeXT. How does that pipeline look over the last couple of months? Have you seen the strength continue there? And then secondly, just in terms of prospective versus retrospective work and how that's shaped up as we reopen post pandemic, if you could speak to some of the trends there, that will be great.