Adaptive Biotechnologies Corporation (ADPT) Earnings Call Transcript & Summary

Good afternoon, everyone. I'm Salveen Richter, biotechnology analyst at Goldman Sachs. And thank you for joining us. With us, we're pleased to have the Adaptive team today, and we have Chad Robins, CEO.

To start, Chad, thanks for joining us. And maybe just to level set everyone, could you discuss the long-term vision for Adaptive? What do you think the portfolio will look like 2 years, 5 years, 10 years from now? And how much of a role will be therapeutics vertical play as a part of this business?

Sure. First, Salveen, thanks for having me back at the Goldman Sachs Healthcare Conference. I appreciate it. So I want to start the answer to this question with the premise that our immune system, it fundamentally does 2 things: it detects disease and it treats disease. And so Adaptive, at its core, is we're learning how to connect the immune system in our bodies to the diseases that it detects and treats. And then we can use this information as our source code to become what we call a clinical product development engine. And so this is kind of the vision for Adaptive. And we've been putting all of the pieces together to effectuate this vision. So if we -- let's discuss kind of the evolution of our product portfolio over time in both diagnostics and in therapeutics. So if we start with diagnostics, we believe that T-Detect franchise will change how diseases are diagnosed because then we can leverage the massive scale of the connections we're making between T cells and antigens. And so really, over the last 1.5 years, we've started to prove that the science is working, and we're building out the infrastructure to be able to kind of put this vision into place by being able to detect many different diseases at the same time with 1 sample. But if you look at the evolution, in the short term, we're doing this kind of one disease at a time. We've proven that we could do this with COVID. Now we're doing this in the Lyme setting. But as we move to the medium term, it's this concept of differential diagnosis or a patient coming in with the same symptoms where we can definitively diagnose one disease over another that is really going to contribute to the top line in the medium term. And in the longer term, it's really this concept of becoming part of primary care as an immune screen that we think is going to fundamentally change health care. But I think one of the things that's kind of the most important and I would also say is probably the most misunderstood part of our platform and part of Adaptive as a whole is it's these exact same underlying mapping data that will serve as a core of a target discovery engine for therapeutics. So it's that T-Detect, it's a diagnostic map. On the one hand, it would yield kind of T cells as targeting molecules, and on the other, antigens as potential targets. So as you know, we're already exploiting T cells as targets in our Genentech deal in cell therapy in cancer, where we can see -- but we can also see in real time the antigens that are driving a whopping T cell response across many different diseases. And it's those antigens, which are rich source of targets for therapeutics in several different modalities, such as vaccines, by informing more effective and efficient vaccine design, and in antibodies, which can essentially block the driving antigen. So in the near term, we're looking to leverage this data and capability and partnerships to be able to contribute to the top line. But if you kind of like take it one step further, which is kind of where you're asking, how are we looking to -- where is this going to become part of our therapeutic portfolio, as you know, we recently developed our own -- it's a really powerful one, it's a proprietary antibody discovery engine. So while it initially can be used in partnerships over time, we're continuing to invest in and develop the capabilities and infrastructure so that we're essentially able to capture more of the value chain in therapeutics. And while, this is just 1 example, if we have an antigen target, now we also have the ability to discover antibodies to block the target. So I guess to your question of the value of therapeutics, it's certainly an integral part of our business and one that we expect to drive significant value in the medium to long term. And then just to kind of round out the portfolio, as we shift from T cells to B cell diagnostics, in the short to medium term, we're growing clonoSEQ into the gold standard for minimal residual disease in certain type of blood cancers, and it's a universal test for multiple blood cancers that's using the same technology. And it's that organizational muscle that we've been developing for clonoSEQ and regulatory clinical development, market access, marketing and also our pharma partnerships, which we are leveraging to the T-Detect franchise. And finally, as you know, our underlying research business, while it's a nice source of near-term revenue, it's also this incredibly rich source of data that's fueling this clinical development engine.

Then -- so let's start with clonoSEQ then. Could you remind us of the growth levers at play for clonoSEQ throughout 2021?

Sure. So let me start with the critical success factor for the next several years, and then we can kind of hone in on 2021. So in order to be successful in making clonoSEQ the de facto gold standard in MRD monitoring in certain blood cancers, we have to: one, penetrate the community account setting; number two, we have to drive to the use of blood-based testing; and number three, we have to expand into this large category of non-Hodgkin's lymphoma. So the CLL clearance that we obtained last year in the middle of the pandemic, that will actually be one of the main drivers to get us into the community oncology setting. But more specifically, our efforts in, which you asked in the second half of 2021, we'll focus on leveraging the expanded and now trained up field force, our peer-to-peer education programs that we've built out to drive this demand inflection point. And if we look at it by disease indication in multi myeloma, where the momentum surrounding MRD continues to strengthen, we -- what we have, what I'll call many of these dabbling or initial users of health care providers, and we're taking step to accelerate that so they really incorporate it into pathways into their practices so that you get repeat usages for each patient and for a certain subset of types of patients. And then in CLL, really, this is a new area. So traditionally, CLL clinicians weren't performing MRD testing. So it's really on kind of motivating and explaining the value proposition with a new set of therapies, which are showing greater efficacy than ever before, why MRD is really an important tool for kind of patient monitoring and patient care. And so the ability to get into those accounts is critical to be able to hit our 2021 numbers.

Could you just talk about how COVID is currently impacting the uptake of clonoSEQ and just your confidence with regard to kind of achieving that 2021 number, is that...

Yes. So yes, if you look at kind of the beginning of the first quarter, it was slow coming out of the pandemic. Then March started to pick up as the highest month in Q1. April was strong, and the trends are continuing into May and June. And if you go by indication, we're seeing a faster uptake in ALL, where kind of 25% of ALL cells are in blood. We're seeing strong growth in multi myeloma. And we're seeing a slower-than-expected uptick on CLL, which, to your question, it's really related to the COVID impact because, keep in mind, most CLL patients are on oral therapy, and it's really not that as aggressive of a disease. And these patients are immunocompromised, and they've been kind of told essentially to stay at home and delay treatment until you get to certain vaccination rates. So we do expect, with the vaccination rates rising, that the patients will come back in, and we'll have an opportunity to access those patients and kind of start MRD testing. We do continue to expect the overall volume to double by year-end, with most of the growth, as we've talked about, coming in the second half because we do expect there to be some lingering impact from COVID that -- from the first step, we're still not in all of our accounts yet. And we know it's an ambitious target, but if all of the work and investments we put into place and our initiative plan, we absolutely do believe we can get there. So if you kind of break it down from external to internal, what we need to do to get there, externally, we need COVID restrictions to lift so that we can get patients back in, and we can get reps back in front of clinicians. We also from not only Adaptive, but in general, this concept of MRD in cancer treatment, whether it be solid tumor or liquid tumor, to be increasingly accepted as part of routine care. And so I think we are benefiting from a lot of the initiatives in MRD not just from Adaptive. And then internally, we got to get our newly trained field force out there selling in-person in the doctor's office. So we have almost all of our 40 reps that we've been talking about for a while now, they've been trained. They're ready to roll, and they just started getting access into accounts. And then other thing, we need to reach more patients with our direct-to-consumer outreach. We've done some really nice piloting over the last 12 months, and we're starting to see some good results. What we do know is that when a patient goes in and asks her doctor that says we need to know our MRD status, then they're generally successful in having the clinician order an MRD test.

And when you look at the competitive landscape, the current landscape and the evolving kind of future landscape, how does clonoSEQ fit or compare versus the other approaches?

Yes. So let's just first talk from a technology standpoint and then we can kind of hone in on different aspects. So all of the other companies working in, I'll call it the NGS, or the next-gen space, they're looking at mutations in the tumor that are going to shed in some capacity. The difference is our technology looks kind of specifically at -- we're able to count cancer cells because we're looking at the T or B cell. So just to be clear in MRD more broadly, clonoSEQ is not designed for, and it cannot be used for solid tumors, if you look at kind of the competitive landscape in the space. It's used for a certain set of blood cancers because these cancers are T or B cell-mediated where it starts with that one cell, and it can be done from both the bone marrow or in the blood. And -- but also keep in mind that clonoSEQ is the only FDA-cleared test to be able to monitor MRD in blood in cancer patients. And oftentimes, what happens is the clinician will have to use a variety of different techniques, different testing for different types of indication or disease state, whereas clonoSEQ has a major advantage. It fulfills a major need in hematology because it offers the clinician to use 1 MRD test across all patients with a set of lymphoid malignancies that the test applies to. And it's extremely patient-specific and it's sensitive to 1 in 1 million cells.

And what about -- like if you look at other technologies coming down the road, whether it's GRAIL or others, like how do you think they could play in that -- a role in this space?

Yes. So if you look at our technology versus a GRAIL or a Signatera, it's a different technology. It identifies a genetic sequence for each patient's unique T or B cell, and it uses those to count the number of cancer cells at regular time points both during and after treatment. And so for example, we're tracking, and we're paying close attention to the competitive landscape, including the Natera's Signatera assay that's used in multi myeloma. I mean, it's early, and there are small studies basically on retrospective data, and there's some promising prognostic data. There's a long way to go to be able to commercialize. In multi myeloma in general, I would say, there's a long way to go for everybody, including Adaptive in terms of -- and we're working on studies in the blood setting, whether it's the peripheral blood or whether it's circulating tumor data. I do think it's interesting to see these early results, and I think it validates, in general, that MRD's heating up. And multi myeloma is a huge market. I do think we expect that there's going to be more than one player in the market. But it's a fundamental kind of selling proposition of, especially in the community, that you're going to have 1 test across kind of multiple diseases and the clinician is going to be used to be able to read that test, which will be a differentiating factor. But I do want to point out, Adaptive has an assay. We're able to run on circulating tumor DNA as well. But the -- if you look at kind of blood-based testing in multiple myeloma, what we're doing, it's much more practical, we believe, and quantitative. And also as a reminder, our blood testing for multiple myeloma is already available in a CLIA-based setting. And in addition to kind of our ongoing study with Dana-Farber, we're also partnering with other collaborators to analyze kind of retrospective data sets for corroborating data between the marrow and the blood. And so we continue to monitor this, but we feel like we're incredibly well positioned. And we're also really the partner of choice for pharma as the only FDA-cleared test.

And maybe to that point, you partnered with Pfizer recently to use MRD as a clinical endpoint in their programs, could you speak to the growth opportunity that these types of partnerships could provide on the forward?

Yes. And I guess I should mention, it's not just Pfizer, that's just our most recent one, I mean, clonoSEQ is, I'll call it, the test of choice in clinical trials that incorporate MRD as a clinical endpoint by almost every key pharma company developing lymphoid malignancy drugs. We -- there's really a continuous pipeline we're working with, and we'll continue to disclose new partners as they come over time. And as I said, Pfizer was just the last one. And if you look at how the economics of these partnerships work, we obtain a mix of sequencing revenue, some upfront payments at times and regulatory milestones. We've been pretty clear that we've got a potential sitting out there on the balance sheet right now of over $300 million of these milestones, which are comprised of kind of regulatory milestones in the United States or the EMEA for the use of MRD as a secondary endpoint or as a primary endpoint in registrational trials. And I think the number that were run by almost kind of like dozen pharma, so these $300 million are spread between about a dozen different partnerships. And we continue to kind of penetrate within those accounts. So when they have a new drug, either within an indication or move to a new indication, they're very familiar with us and confident and have built up the trust with us and our partners as kind of the partner of choice. We started these deals about 40 years ago. And what we're starting to see is kind of these earlier secondary milestones start to hit. As you know, we hit $7 million in the first quarter. This was planned for the year, but it came earlier than anticipated. And then we've got a whole bolus of milestones that are kind of associated with this primary endpoint, which we're working with CDER on to be able to, essentially in myeloma -- particularly in myeloma to have CDER say, "Yes, myeloma can be a new primary endpoint in trials." It's not just us, obviously, the pharma companies are super interested in this happening as well. So we do believe that, that is -- those are real dollars and will be made available to us over time.

All right. So maybe pivoting to T-Detect. So we saw your T-Detect COVID become the first T cell-based diagnostic approved under an EUA. And clearly, a meaningful step for the platform. Could you speak to what this means in general for the platform and how you're able to leverage the infrastructure that you're starting to create around that test?

Yes. I mean, first of all, we've been talking about the T cells that we can learn how to kind of read how T cells are naturally detecting disease. And so this proved that we could do -- absolutely do that, with 99% specificity and almost 99% sensitivity. And so I think you're framing that in exactly the right manner in that, the EUA authorization, it was just a key milestone for the T-Detect franchise. So it's not -- it wasn't just a validation of the platform and demonstrating that T cells can detect the virus as well or better than standard of care. We have the opportunity then to educate the FDA as the first T cell-based test with an EUA approval. But that T-cell-based test, it's that same T cell-based test that's going to detect not COVID, but any disease. And so what it also allowed us to do is really hone in and improve our algorithm development. We were able to put kind of all the e-commerce in place, the virtual prescriber in place. We activated kind of LabCores for kind of blood draws and capacity. And we also started to build brand awareness through -- from a DTC, digital and social campaign. So we got a really great start to kind of accelerate what we talked about in your first question of our long-term vision of bringing T-Detect to market for kind of many different diseases.

And how is the launch progressing with T-Detect COVID? Do you have any metrics on patients treated at this point -- or not treated, but tested and kind of your breadth of exposure within the country?

Yes. We haven't released metrics yet, but it's gone pretty well. I will say we've actually gotten more -- because remember -- let me just take a step back, we kind of intentionally didn't go after kind of reimbursement in this setting to be compared -- or cross-walked into a serology-based testing. So this was done as a self-pay to consumers at a $150 price point, which ultimately doesn't support the margin profile that we're looking for, and we believe we'll be able to get in kind of these unmet diagnostics, for example, in the autoimmune space. That being said, even with the digital and social program, we're seeing, I would say, more tests come through than we thought based on the amount of kind of marketing dollars that we've put on it. But we aren't releasing -- Salveen, we're not releasing specific metrics on the test yet or I'm not sure we will.

Perhaps you could talk about the Lyme disease test, which is being developed and we're going to see an offering in the second half, I guess, by year-end this year. Could you discuss the signal validation that you've seen in Lyme to date and what that kind of final outlook you're expecting in the study? And what these patients -- I guess, when you talk to the clinical community, what they really want out of the test? So what's the clinical bar here?

Yes. So let me just first characterize this. Our plan is to make the T-Detect Lyme test available through a CLIA setting at the end of this year. And initially, because you won't be in the acute Lyme season with new cases, it will target what we call the PTLDS, which is the post-treatment Lyme disease syndrome patients. And at the same time, we'll start kind of preparing the market development for the advance of next year's Lyme season, which we can actively target patients that are in the acute setting. We do expect to share data from our control set with Johns Hopkins throughout the year, and we will show -- showing 2 different points. One is a 2x kind of sensitivity in the acute setting and also kind of a percentage of diagnosed PTLDS patients that still show Lyme disease positive per our classifier. And if you just look at the study, the ImmuneSENSE Lyme study, it's a clinical validation study that you know we started last year. But because of pandemic, we weren't able to recruit a certain patient population that had the rash. But the whole study is about 990 individuals. We've already enrolled 70% of it. We do expect to complete the enrollment this year, including the 60-patient rash cohort. And then at that time, too, we'll evaluate whether or not it makes sense to go for an FDA filing in this setting or not. Through our market access work, our payer outreach and all the studies that we've done, we've -- some of these things are obvious, but we know that autoimmune disorders are going to be reimbursed at a higher price point than infectious diseases, number one. We also know that this concept -- payers don't like panels, but they do like differential diagnosis. So if you can give a definitive diagnosis based on shared symptomatology, they are willing to pay for that. But back to your question, your question was on clinical bar. And so what we'll do is we're going to evaluate the specificity and sensitivity of T-Detect Lyme versus what's called STTT, or standard two-tier serology-based testing. We're also going to look at it versus kind of the MDTT, which is kind of a multidimensional-tiered testing as well. And then we'll -- the current standard of care serology, it's really poor, with a high false negative rate of kind of 60% to 70% in the acute Lyme setting and also kind of the long-term effects of the kind of underdiagnosed and/or the untreated Lyme disease, it's really terrible for patients and costly for the health care system. So we absolutely feel like there's a need to have a much more accurate, more sensitive and more specific kind of Lyme test out there, and so we're continuing to pursue it. I do want to make sure that this -- keep in mind because how you characterize the question that a full product launch with reimbursement in the diagnostic setting takes time. So you make a test available, you have to get clinicians using the test, patients wanting the test. So it's necessary to demonstrate that clinical utility and uses of the test to then be able to get reimbursed for the tests, especially at the rate that we're going for, which will support a very nice margin profile for the T-Detect Lyme test.

What is the commercial opportunity here at peak?

Well, if you look at kind of the -- right now, there's 2.4 million Lyme tests that are performed each year, with about 600,000 of those kind of the initial target market in the acute setting. So again, we haven't -- we're not kind of releasing and we're still doing market research on the payer, but over -- initially, we'll go -- over time, we hope to be able to access that entire patient population. But in the shorter term, we're going to go for a subset of that population.

And you've seen, I think, initial proof -- or we've seen initial proof of concept for T-Detect GI in Crohn's disease. How are you thinking about the clinical bar here for sensitivity and specificity in Crohn's and GI diseases in general?

Yes. So that's right. We had an early signal in Crohn's, which we validated. And at the same time, we were able to distinguish that from colitis samples. So in other words, kind of what our data shows, we're demonstrating that T cells can recognize Crohn's and are distinct from the T cells that recognize colitis or celiac disease and COVID and anything else. So specifically, the data that we showed -- that we haven't shown, but we will show so far is that in 500 ileal Crohn's, which is in the small intestine, 500 ileal Crohn's patients are classified -- it's already 70% sensitive at a 99% specificity. When I say already, remember, and this happened with COVID, the more samples we get, this is a self-improving diagnostic, so the TCR classifier is going to continue to get more sensitive as the sample size increases. So we already have another 5,000 extremely well-characterized samples in-house, both from Crohn's and colitis, that's going to help us flesh out the clinical development plan. And our expectation from these samples is to be able to not only evaluate and improve our signal, but we'll also be able to understand how our signal performs in patients with other types of Crohn's disease that occur in other regions of the digestive tract, such as the large intestine. And it also, hopefully, will not only have -- be able to distinguish between Crohn's and colitis, but will have sensitivity in the vertical of colitis as well. But if you kind of look at the bar of -- and which is really honing in on your question, I just wanted to kind of give you some context around what we have, if you look at the bar for sensitivity and specificity, while our current signature is very specific, that means we can rule in and identify Crohn's patient positive result. However, right now, a negative test will not necessarily rule out a patient for Crohn's yet, so we need to improve the sensitivity. But if you look at, right now, done either by a stool test or a scope, that bar, to be able to perform that with a blood test, while I'm not giving you a direct -- this is the sensitivity we have to hit yet, but if you're talking about a much less invasive test and if you have a high enough degree of sensitivity at a very high degree of specificity based on a blood test, again, we're honing in on what that is based on market research, advisory boards, et cetera, we believe there will absolutely be a product market fit with the clinical community to be able to use this test upstream, hopefully, of a scope.

And I guess, strategically, how do you choose which diseases to pursue next in your pipeline with T-Detect? Would it -- I mean, can one assume if you're able to derisk Crohn's disease that you just opened up the GI gamut with celiacs and as you said, colitis. But what's next for you?

Well, yes, that is a great assumption and exactly what we're doing to kind of build it around a nucleus of kind of differential diagnosis, example too, in Lyme, to be able to differentiate and distinguish between tick-borne illnesses that have a shared sort of symptomatology and then kind of putting those groups of differential diagnoses together. And remember, anything that we've mapped, we can diagnose with that kind of same sample. So we can always diagnose COVID. We can always diagnose Lyme once you have it. But to add to your question kind of on this prioritization question, we've got different stages throughout our R&D prioritization funnel, which start with diseases that have kind of a key TAM or market attractiveness. And we look at that by kind of high unmet medical need. We have to know really something about -- a lot about the science. So we need to have something where the antigenic space is. So each disease has a set of antigens. We have to know, for the most part, what those antigens are for that disease. But we also have to have an access to samples with metadata. So if you kind of put those things together, that kind of forms this framework for evaluation of -- and we've got many going on at the same time. But what we move forward, it really depends on those 3 or 4 factors combined. And then -- so initially, we -- there's kind of 3 disease areas, right, which is infectious disease, autoimmune disorders and cancer, oncology. And infectious disease, it is the quickest and easiest and the diagnostic space is -- you're pretty core for the most part. So that's why we started disease by disease and kind of that single disease diagnostic. We started with COVID and kind of moved into Lyme. Autoimmune is -- it's very difficult to diagnose these diseases. And what happens is your patient journey, there's this diagnostic odyssey where a patient goes around for often a period of a couple of years in specialist to specialist, costing the health care system upwards of $10,000, $20,000 to confirm -- or I'll say settle on or rule out to get to a diagnosis. And then -- and so we've got a very significant effort around kind of multiple autoimmune diagnostics that we can essentially change the patient paradigm. And then cancer is going to be most challenging because just what I talked about on the antigen space, there's many different antigens for different tumor type and neoantigens. It's very complex. There's hundreds of mutations. So it's going to take some time and it will take a lot of data to generate, but we absolutely do believe that we're going to be able to crack this. And it's a whole different concept of earlier detection than a Guardant or GRAIL or Thrive is looking to do. We have a pipeline in ovarian cancer, which continues to advance. It's just earlier, and we're taking those machine learning learnings, if you will, and applying those to the different spaces. So that's why kind of infectious disease and autoimmune is the first that will be kind of made available and then commercially launched with hopefully, cancer to follow in multiple tumor types over time.

Great. And then just touching base on the drug discovery side of the company. You're working now with Genentech to advance the kind of next shared product candidate. Just curious on how much decision-making Genentech has in terms of the target itself and the construct? And how do you think about, given it's solid tumors, which is difficult, the group of targets that you had identified initially right in the context of how difficult this is, but then secondly, with managing the tumor microenvironment. Like how much work is being done together? And how much work has already been done around the first shared product just generally?

What will be the first shared product -- what was the first shared product? Yes, yes. No, right. I mean -- so I think it's -- yes, it's worth mentioning yet. Cell therapy is extremely complex, both in efficacy, safety. There's always going to be a risk. But we have a very integrated partnership where -- with Genentech. But I do want to be clear, ultimately, the target selection is Genentech's choice. We have a joint research committee, we have a joint steering committee, and we work hand-in-hand to -- what we do actually is we've characterized many different T cell receptors against many different targets, and then we essentially elect to move forward in terms of what will be our new kind of first shared product based on what we're seeing in the data. And so that's -- it's a very collaborative partnership. But again, I do want to be clear, it is their choice based on -- with our input based on what targets to go after.

Okay. And then on the private product, I think you announced you've done some initial -- completed some initial proof-of-concept data here on 15 patients and then tend to process blood from about 60 patients by year-end. Could you just discuss the proof-of-concept data that could support a path to IND filing here?

Yes. I mean we completed, as you said, initial proof of concept using blood, the specific tumor mutation -- plus specific tumor mutations from 15 cancer patients, and we're on track to establish the prototype with at least 60 additional patients, but we're looking at potentially more. And we're learning a lot from each patient data set that we get. And our goal is then to establish a representative set of results, kind of which essentially allow us to quantify expectations and set the specifications to be able to define the end-to-end product -- private product prototype. And from there, we're going to need to define, test and refine over time this end-to-end process with Genentech all the way up through manufacturing prior to our IND readiness and all that's going on kind of right now. I will say, we're extremely encouraged by the data we've generated in our south San Francisco lab, which is fully in our control in Adaptive, and then we work closely with Genentech along the way and advancing this. This is really -- it should be, kind of noted, this is not just -- this is a cornerstone project for both Adaptive and for Genentech, where we're both putting a significant amount of resources into this above and beyond, which are readily apparent, for example, in the manufacturing process. So we are continuing to move full speed ahead with the shared product approach in conjunction or in parallel with the private product. So -- and these 2 programs, they're not completely independent of each other. There's been significant learnings that we're leveraging from our shared products that are informing the private product approach, and this includes taking Adaptive discover and fully characterize TCR into different studies and testing out our innovative kind of gene engineering that also be implemented kind of once we're in the clinic with our private product to go back to the shared product. So they're working really nicely in conjunction with each other.

Perfect. And then finally, what is next for Adaptive's antibody discovery platform?

Yes. So maybe I'll kind of take this full circle back to the opening question, which was kind of what's kind of next in therapeutics. We built over kind of the -- when COVID hit the last 1.5 years, this really unique proprietary special antibody discovery platform. And we are in discussions, and I know -- I'm fully aware this is a show me, not tell me, but we are in discussions to how to leverage that in partnership with the biotech and pharma community. And at the same time, we're continuing to invest to kind of move us up the value chain, so that once we find antigens that are driving a robust T cell response, then we can find antibodies against those antigens as -- the antigens as therapeutic targets and the antibodies as a therapeutic. And these are areas that -- this is kind of all coming together. And of course, it will take time to play out, but what we're seeing on the data side, we're extremely encouraged by, and it's something that we are looking to invest behind and to continue to build out our capabilities for our platform.

Great. Well, with that, Chad, thank you so much. Really appreciate the time today.

Thank you so much, Salveen. It's great to be here.