Personalis, Inc. (PSNL) Q2 2020 Earnings Report, Transcript and Summary

Good afternoon. My name is Jerome, and I'll be your conference operator today. At this time, I would like to welcome everyone to the Personalis Q2 2020 Earnings Conference Call. [Operator Instructions] Thank you. I would now like to turn the call over to Ms. Caroline Corner. Please go ahead.

Thank you, operator. Welcome to Personalis' second quarter 2020 earnings call. Joining me on today's call are John West, President and Chief Executive Officer; and Aaron Tachibana, Chief Financial Officer. This call will include forward-looking statements, including statements regarding the markets in which we operate, including potential market sizes; trends and expectations for products, services and technology; trends and demand for our products; Personalis' expected financial performance, expenses and position in the market; and the impact of the COVID-19 pandemic on our operations and our customers' operations. These statements are subject to risks and uncertainties that could cause actual results to differ materially from our current expectations. We encourage you to review our most recent filings with the SEC, particularly the risk factors described in our 10-K filing for fiscal year 2019 and in our 10-Q filing for our second quarter ended June 30, 2020. The forward-looking statements we provide during this call, including expectations for future performance, are based on our reasonable beliefs and expectations as of today. Personalis undertakes no obligation to update these statements, except as required by applicable law. Our press release with our second quarter 2020 results is available on our website, www.personalis.com under the Investors section and includes additional details about our financial results. Our website also has our latest SEC filings, which we encourage you to review. A recording of today's call will be available on our website by 5:00 p.m. Pacific Time today. Now I'd like to turn the call over to John for his comments on second quarter 2020 business highlights.

Thank you, Caroline. Personalis has been responding to COVID-19 in – began responding to COVID-19 in Q1 before government's shelter orders. That early response helped us have a record quarter in Q1. Our continued response and execution has now led us to another record in Q2. In April, we coordinated with the VA MVP program to take the largest shipment of samples from them that we have ever had just before their lab in Massachusetts was closed in response to the pandemic. The automation that we've put in place over the last three years then allowed us to efficiently maintain our whole genome sequencing work, even when we had a reduced workforce in our lab as a result of operational changes we implemented to comply with state and local orders related to COVID-19 and to protect the health and safety of our employees. As a result, in Q2, we were again able to sequence about 14,000 whole human genomes, up 70% from Q2 last year and matching the Personalis population sequencing revenue record we had set in Q1. On the oncology side of our business, we also had to adapt rapidly. Some prospective clinical trials that our biotech customers are involved with were either slowed or stopped due to the pandemic. And also revenue from our leading biobank customer fell 98% from Q1. We were able to overcome these challenges by beginning to ramp revenue from our next platform. This allowed us to achieve a sequential revenue increase for biopharma and all other customers. Most of our biopharma business is for retrospective projects, which have been resilient during this pandemic. Over the last four quarters, biopharma customer orders have significantly exceeded revenue, driven by adoption of our NeXT platform and predominantly by large pharmaceutical customers. As we saw in Q2, these orders are now beginning to convert to revenue. Another challenge we faced during the quarter was the timing of biopharma sample receipts. Many were delivered to us later in the quarter than expected so we were not able to process all of them for revenue. This does, however, put us in a good position heading into Q3. Our laboratory operations team executed well during Q2 and carefully arranged the sample processing schedules to focus on VA MVP samples early within the quarter when labor capacity was lower due to shelter-in-place guidelines. That allowed us to utilize more capacity later in the quarter for biopharma sample processing. As a result, our revenue from pharmaceutical customers increased by more than 20% sequentially. This more than offset the dip in revenue from biobank and biotech customers and helped us achieve quarter-on-quarter growth in our oncology-focused business. Driven by this growth of NeXT, our overall revenue increased for a 16th consecutive quarter to $19.5 million, up 23% from Q2 last year. All of this was accomplished, while more than 80% of our employees were working from home. Shifting focus to our progress this quarter, there are several factors which give us confidence in the long-term growth of our business. In the month of June, we were able to bring our liquid biopsy development team members back into the lab. You may have seen our recent press release launching NeXT Liquid Biopsy, which is now available for customer orders. Both our tissue and liquid biopsy-based products have been designed specifically to meet the needs of our pharmaceutical customers and provide data on all 20,000 human genes. With the addition of our liquid biopsy product, we expect to enable comprehensive monitoring of patients across 20,000 genes at multiple time points. Many new cancer drugs extend the lives of patients but do not eradicate the disease. So a growing segment of the cancer survivor population consists of patients who are still undergoing active disease management. Our first liquid biopsy product tracks the evolution of what can be hundreds or even thousands of cancer mutations in a single tumor. Importantly, we can also detect new mutations as they emerge under therapeutic pressure. Note that some cancer patients develop a second genetically independent cancer while being treated for the first. Our liquid biopsy product's ability to see new mutations may also provide early detection of some so-called second cancers. I'm extremely proud of our team. We pulled together during this challenging time to get this exciting product into the marketplace. We expect to receive first orders in the coming months. And while we expect that liquid biopsy revenues will be modest initially, we think this new product puts us in an even stronger competitive position as we go forward. Our liquid biopsy product is designed to be used together with our tissue biopsy product, not instead of it. We expect initial sales to be to customers using our tissue-based testing, who are also seeking liquid biopsy capabilities to monitor the same patients over time. We believe that our capabilities to offer both tissue and liquid biopsy-based products and to leverage the synergy between them, positions us favorably relative to companies who only offer one or the other. In the future, we will be broadening our liquid biopsy product line to include personalized tests designed for a specific patient's tumors. The team is hard at work here and barring any COVID-19-related interruptions or other unexpected delays, we expect to launch this additional offering in 2021. We have continued to gain traction with biopharma customers, realizing strong order levels once again in Q2. To frame the pace of customer orders in a different way, in the first half of 2020, we received about the same dollar value of biopharma orders as we received for the full year of 2019. As we've explained before, it takes time for orders to convert to revenue and the actual revenue recognized from an order may be less than expected due to failure of individual customer samples to meet our sample quality requirements and other factors. But this growth in orders gives us confidence in our future revenue stream. We are winning business based on the value proposition of our NeXT platform. Our NeXT platform can identify biomarkers, including neoantigens across all types of cancer, providing comprehensive data across all 20,000 human genes, all from a small tissue sample. Our customer base has also grown substantially over the last year. While we can't disclose all of our customer names, I can tell you that we have now received orders from a majority of the top 10 oncology-focused pharma companies. As of the end of Q2, we had 32 different customers who have placed orders for NeXT, up from 26 at the end of Q1. In addition, you may recall that new customers typically evaluate or run a small pilot before growing into larger contracts. Up until the last few quarters, a pilot order could be around $50,000. But with our recent progress, we are now receiving initial orders that can be several hundred thousands of dollars. And in a few cases, approaching $1 million. We believe that this growing adoption of NeXT further highlights the power of our platform and that the comprehensiveness of our platform is important to our customers. In June, we announced a research collaboration with Sarepta Therapeutics, a leader in precision genetic medicine for rare disease. As part of the collaboration, Sarepta is working with us to characterize immune response precision genetic therapeutics, utilizing our advanced proprietary neoantigen analytics. We are excited that the value of our platform is now being recognized in areas beyond cancer. We also announced in June that we are establishing a lab and commercial operations in the People's Republic of China. Several global pharmaceutical companies have asked us about China, underscoring our belief that it's a good time to expand there. Additionally, we recently announced a partnership with Berry Genomics, a Chinese company focused on the development and commercialization of genetic test technologies in clinical applications with approximately 1,500 employees and seven different clinical laboratories. We believe that our partnership with Berry Genomics will be complementary since our customer focus is pharma and theirs is clinics and hospitals. We have continued to expand our commercial team in both the U.S. and Europe. In particular, we have added business development staff with extensive commercial experience in companion diagnostic development. And as a result, we are now engaging with an increasing number of biopharma companies for potential companion diagnostic development programs. We've also expanded our quality and regulatory team to support these efforts. And in June, we had our first FDA pre-submission meeting for a single-site PMA for NeXT, which went well. I'd now like to update you on the population sequencing part of our business, which is sometimes referred to in the field as population genomics. As you may have seen in our June press release, our population sequencing business hit a big milestone with the 75,000 full human genome sample sequence, and we are well on track to reach 100,000 by the end of this calendar year. This could make Personalis the first for-profit company ever to sequence 100,000 whole human genomes in the United States, and we are really looking forward to achieving this milestone. Our work with the VA MVP represents the largest population sequencing effort within the United States. The VA now targets enrollment of 2 million veterans and over 825,000 veterans have enrolled so far. Personalis has been contracted so far to sequence over 116,000 VA MVP samples, with approximately 41,000 remaining to be sequenced. Having already received population sequencing orders totaling over $145 million from the VA MVP, we are now expanding our commercial team to address and extend our reach into what is projected to be a multibillion-dollar population health market. We expect our unparalleled experience with the VA MVP program to position us well for new opportunities in population sequencing. Given our clinical experience and work with pharma, we also see an opportunity to help transition population research to population health and to involve pharma in the future. Our population sequencing and biopharma businesses share a unifying theme in the comprehensive and large-scale genomic characterization of human samples. Both our whole genome sequencing and our NeXT platform cover all 20,000 human genes. Because of their shared underlying technologies and operational implementation, we achieved considerable synergy between the two businesses. I would now like to expand on the synergy between the population sequencing and oncology parts of our business, in particular, synergies that go beyond the operational and cost synergies already achieved. Our extensive experience with whole genome sequencing, combined with our deep expertise in cancer, has allowed us to launch whole genome sequencing from cancer samples. We believe that this will be increasingly important in the future, particularly in cancers such as breast and prostate, which have relatively low mutational burdens. Using our cancer whole genome technology will identify up to 20x somatic variants to serve as the basis for personalized cancer assays. We believe this will let us achieve high sensitivity, even in cancers which have low mutation rates, which have been surgically resected or which had relatively low amounts of cell-free DNA into the blood. We believe this can be a leading technology in some very large market opportunities, and we'll have more to say about this as these product developments progress. Before I wrap up, I'd like to give you an update on another pipeline product. As many of you know, we have been working for over three years on a combined laboratory and informatics project to advance our neoantigen characterization capabilities. We believe that neoantigens are the crucial centerpiece of a new generation of companion diagnostic biomarkers. We expect our biopharma customers will apply this capability to mainstream cancer drugs such as checkpoint inhibitors and also to experimental personalized cancer therapies, which explicitly target neoantigens. Our new advanced capability may also be used for development of drugs that are designed for use outside of cancer, such as gene therapies. This Personalis R&D project has required multiple proprietary technologies, including genetic engineering of proprietary human cell lines, mass spectrometry to identify and quantify peptides binding to HLA and the training of novel machine learning algorithms. Our data shows that this project has achieved a leapfrog advance in this field. We expect to launch this capability as part of our NeXT platform in the fourth quarter of this year. In summary, I'm very proud that our combined business has shown strong resilience throughout the pandemic. Customer adoption for NeXT has been excellent, and our pipeline of compelling new products is rich. We believe these factors, among others, put us in a strong position for long-term growth. With that, I will now hand it over to Aaron for our financial results.

Thank you, John, and good afternoon, everyone. Revenues for the second quarter of 2020 were $19.5 million, up 2% from $19.2 million for the prior quarter and up 23% from $15.8 million for the same period of the prior year. The $19.5 million was a new record high for quarterly revenues. The quarter-on-quarter revenue growth was driven by an increase in volume for genomic testing services provided to biopharma customers. Biopharma and all other customers accounted for revenues of $4.7 million for the second quarter, an increase of 8% from last quarter. In the second quarter, revenues from our NeXT platform began to ramp and exceeded $2 million. Also to note, the revenue increase from NeXT more than offset the sequential decline of over $1 million from both our biobank customer that had sample collection and shipment delays due to the pandemic and also biotech customers that were impacted by the slowdown of clinical trials. For the second quarter, the VA MVP revenue of $14.8 million was flat from last quarter and was 73% higher compared with $8.5 million for the same period of the prior year. The VA MVP unfulfilled orders at the end of the second quarter were $39.3 million and based upon current estimates, we expect the unfulfilled orders to convert to revenue over approximately the next two to three quarters. The balance of the unfulfilled orders will decline from the fourth quarter of the prior year through the second quarter of the current year since the annual new order is typically received late in the third quarter. Gross margin was 24% for the second quarter compared with 21.1% for the prior quarter. The VA MVP gross margins continued to be solid and once again were higher than the corporate gross margin reported for the second quarter. The VA MVP being higher volume and a single-service offering has been automated and does not require a significant amount of labor and has a very efficient sample test process. In the second quarter, the negative impact to gross margin from COVID-19 was approximately 80 basis points from higher labor costs related to overtime pay. Also, we had a 100 basis point impact from higher facility costs related to an increase in lease rates. As a reminder, we may see gross margin variability in the future as there are a few moving parts, such as sample receipt linearity from customers, the mix of customer projects and capacity utilization of labor and equipment. During our last conference call, we mentioned our initiative to build out a lab operation in China this year, which will add start-up expenses beginning in 2020, and most of these expenses will be classified as SG&A this year. We expect revenue from China to begin ramping throughout 2021. And during this ramp-up period, we expect to have gross margin headwinds from the under-absorbed labor and overhead. Operating expenses were $14.2 million in the second quarter compared with $10 million for the same period of the prior year. R&D expense was $6.5 million for the second quarter compared with $4.5 million for the same period last year, and SG&A expense was $7.7 million in the second quarter compared with $5.5 million for the same period last year. Net loss for the second quarter was $9.3 million compared with a net loss of $5.9 million for the same period of the prior year. The net loss per share for the second quarter was $0.29 and the weighted average basic and diluted share count was 31.7 million compared with a net loss per share of $0.89 and a weighted average basic and diluted share count of 6.6 million for the same period of the prior year. Now on to the balance sheet. We exited the second quarter with a strong balance sheet with cash and short-term investments of $105.2 million. Second quarter cash flow from operations was a usage of approximately $15 million, primarily due to the net loss and working capital needs. During the quarter, we maintained our buffer inventory level of approximately $1.5 million to help mitigate potential supply chain disruption. Now for discussion about our guidance. During our May conference call, we did not provide any 2020 revenue guidance due to the uncertainty from the pandemic. And although we had a solid second quarter, too much uncertainty remains about whether or not work conditions for us, our customers and suppliers will remain the same as today or change in the near future. Potential changes could have an adverse effect on our financial results. And therefore, we are not providing revenue guidance for fiscal 2020 at this time. We plan to provide an update to this information during our next earnings call. Now I will turn the call back over to the operator, Jerome, to begin the Q&A session. Operator?

Thank you. [Operator Instructions] Your first question comes from the line of Doug Schenkel from Cowen. Your line is now open.

Hey guys, this is Subbu on for Doug Schenkel. You posted a great update intra-quarter on the VA update. Given that you're on track to complete the project, how long – so most of the database, given that you're on track to finish 100 case sequences, most often, we notice that database that has integrated electronic health record data has more value. Do you envision that to happen with the VA MVP project?

Yes. This is John West. I'm happy to answer that. Actually, a part of – key part of the advantage that the Million Veterans Program has is that veterans have all been on the same electronic medical record system for years. And so they already have that in electronic format. And unlike other population sequencing efforts, they don't have to try to harmonize results from many different electronic medical record systems. All of the patients that are handled by the VA MVP are patients who would be – have been on the same electronic medical record system during their time as patients with the VA. So it's actually really a major advantage that they have. I would also say that we're coming up on the 100,000 patients. That's nowhere near finishing the project. The project is the sequence – to enroll 2 million patients. There are actually 825,000 samples that have already been collected. So actually, we're just beginning to scratch the surface of the project. We've been contracted to sequence 116,000, and we anticipate that there can be an additional order still this year that will extend the program even further.

So if I heard it right, you're already doing that. You already have data with the electronic health record data of all these individuals? Or do you plan to do that in the future?

Let's see, our customer already has that information in the VA. They're the ones who have that health record data. They're the ones sharing it. It will be their database. They're the ones paying for it. But they do already have all of that health record data.

Got it. And you recently partnered with Berry Genomics, and you also said that, that will serve as your local subsidiary opportunity. How do you plan to strategize that, given that NGS testing has just begun gaining traction there, and there are already a few high-profile local players? So what would be the key advantage? Would you call it your core technology? Or is there some other strategy that you're working on?

Yes. So thank you for the question. So the key reason that we are expanding to China is that there are international pharmaceutical companies that we already work with in the U.S. and Europe who are conducting international clinical trials. And so they have patients who are being enrolled into those clinical trials from China now in addition to other countries. And we've already been sequencing samples from patients that are from countries outside of China, but our pharmaceutical customers came to us and said, we're not allowed to export the samples of patients from China. We can't send them to you in California. They have to be sequenced inside China because of the regulations of China. And so they've asked us to set up this additional capability using our proprietary technology, our NeXT platform, to be able to run that inside China.

Got it. Thank you, guys.

Yes. Thank you.

Your next question comes from the line of Kevin DeGeeter from Oppenheimer. Your line is now open.

Hey, guys. Congrats on a really nice quarter. Thanks for taking my question. Can you just talk a little bit about the recently launched liquid biopsy program? Specifically, in addition to breadth of coverage, how else do you envision differentiating the product in the market, at least initially? And then in terms of the, what I'll call, incremental or second-generation product that would provide additional functionality for really kind of personalized precision oncology programs, just kind of walk us through the development steps that may be necessary to have that functionality?

Sure. This is John. So, let's see, our liquid biopsy, first of all is paired together with tissue biopsy. And both liquid and tissue have advantages. And because we have both, that's actually a relatively unusual offering. Many companies that offer liquid biopsy products don't actually have a tissue biopsy product. But the tissue is the only way to see RNA in addition to the DNA, and it's the only way to see the immune cells that have actually infiltrated the tumor. And so having the combination of the tissue and the liquid is an important step. I'd say that in addition to that, the – we have designed our product because it covers the entire exome, it doesn't just see mutations that people might expect because it's intended to be used in pharmaceutical research. Many pharmaceutical companies are interested in understanding when a patient responds to a drug for a period of time, but then eventually stops responding, what changed about the tumor? What were the mechanisms by which the tumor began to escape from the effects of the drug? And so being able to see the new mutations that occur, wherever they are in the genome, where they are in the exome, is really important to them. Most other liquid biopsy products are about 40 times smaller in terms of the number of genes. They look at – they traditionally just look at the cancer driver genes, the targeted therapy genes. And that's fine for – as a diagnostic. But if you're trying to understand the biology of the tumor, which is often the goal in pharmaceutical research, then it's important to be able to see all the biology, not just the, to some extent, the biology that we've known about for decades. So I think that, hopefully, that answers your question on the exome side. In terms of the personalized product, Personalis has designed a variety of tests over the years, both for ourselves and for customers. We have actually become quite expert at that. And so we anticipate that the ability to make personalized assays will be something that will leverage those same skills. So we're pretty optimistic about that. If you think that a key skill in being able to manage that kind of a program is actually the informatics of being able to keep track of it because you have a different assay for every patient and different patients may have different time points and so forth. And so there's a sort of a logistical complexity to all of that, and that's also an area that Personalis has a lot of strength. We've developed a system we call our Symphony system, which keeps track of all the internal informatics of our laboratory and data processing operations, and that's been a fantastic asset for us at this point, and we think it will also help us move quickly to bring these personalized products to market.

Okay. Great. Then maybe as a follow-up question. Can you just comment on the potential impact on average contract size from the addition of liquid biopsy more explicitly? Should we think about the customer that wants both tissue and liquid biopsy functionalities, essentially doubling the potential economics of Personalis? Or do you anticipate more of a bundled economic model where there's increase in functionality, isn't necessarily going to translate to a linear increase in economics?

Yes. So our expectation is that this will lead to a significant increase in the economics and also in the biological value of the data that we can deliver. We have a few customers now where we actually already provide data at multiple time points from tissue. And in those cases, we've already had cases where customers have paid up to $16,000 per patient because they're getting data from multiple time points via tissue. We think that kind of thing can expand quite substantially as we have more and more customers where it's easier because then they have multiple time points. And so when we look at the clinical trial protocols of many of our largest pharmaceutical customers, what we see is that they're often actually already collecting those kinds of samples. It wouldn't be unusual in a clinical trial for there to be 20 or 30 different time points at which the patient would be being dosed. And often when the patient comes in to be dosed with the experimental drug, there are laboratory tests that are conducted at each of those time points. So we think this can fit quite well with the way that clinical trials are already being conducted, and it will provide dramatically more insightful information on the trajectory of each patient.

Great. Thanks for taking my questions.

Great. Thank you.

Your next question comes from the line of Derik De Bruin from Bank of America. Your line is now open.

This is Dionne [ph] for Derik. I'm just wondering on NeXT, what is the average order size that you're seeing? What is each order encompassing? Is it just DNA sequencing? Or are they ordering more?

Yes. So this is John. Our NeXT platform includes both DNA and RNA sequencing. And from that combination, we also capture the complete T cell and B cell repertoire. So it's really quite comprehensive. That's for each tissue sample, we have that. Now that we have multiple time points with liquid biopsy, that could be included as well. So that gives you an idea of the data for patients. Typically, the orders we receive will cover all of the samples that the pharmaceutical company has collected during the course of a clinical trial. So that can be hundreds of samples, in some cases, 1,000 samples or more. And many of the orders we receive are retrospective. So the pharmaceutical company already has all of the samples. They understand what happened in the trial, at least in terms of the clinical outcomes and now they want to understand at the molecular level, what was different about all the different tumors and how they progress. So that's often the scale of what we contract for. With some customers, there are larger blanket orders, which are set up basically to put a certain amount of money sort of against the account and then individual projects essentially are just charged against that account. And as the account gets used up, then they'll put more money against it. So yes, that's a little different contracting framework. But in general, the contracts are for groups of patients who were the patients of a particular clinical trial or groups of clinical trials that are our related trials of a particular drug.

Great. Thanks.

Your next question comes from the line of Tejas Savant from Morgan Stanley. Your line is now open.

Hey guys, this is Edmund [ph] on for Tejas. Thanks for taking the question.

Hi.

Hi. Just in terms of your NeXT Liquid Biopsy, I know you guys talked about the economics a little bit earlier, but you guys mentioned that solid tumor serial testing is about $16,000 per sample. So in terms of the liquid biopsy product, would the ASP be somewhere within that range or even higher maybe? And in terms of your customers that are still on the old platform, can you remind us what percentage of that is remaining today?

Yes. This is John. So the – when we think about the NeXT Liquid Biopsy, we anticipate that different customers will have different numbers of time points they want to look at. We anticipate that early on, some customers will just do a straightforward comparison of what they see with tissue and what they see with a liquid biopsy. Of course, we don't expect those to be exactly the same. There are – the biology is different for the two and so there'll be some genetic variance and overlap, but as you'd expect, there would be quite a few that do not overlap. But people may want to characterize that and get a feeling for it and begin. As they start to explore the numbers of time points, our sense is that people probably will start with modest numbers of time points. And then as they see the utility of that that the number of time points per patient may go up and up and up. There's a lot to be seen from what we understand the dynamics of tumors can actually be reasonably rapid in terms of changes. You can have – when people have sequenced the metastases of a patient and then also sequence the primary tumor of the same patients, you can end up with twice as many genetic variants in the metastases as they were in the primary. So that's a lot of new variants to be able to be seen. And so it's a lot of dynamics to understand the biology of. And so those are all the things that we think people will start to get used to being able to see. It does add up to quite a bit of sequencing. I'll comment just as an interesting anecdote that is an internal calculation. We were looking at information that Illumina published on their – at an investor conference earlier in the year. They indicated that last year on Illumina platforms worldwide, the total sequencing was about 150 petabytes. And if we look at the amount of sequencing that Personalis conducted in the second quarter, we were probably in the range of 3% to 4% of that kind of a number. So the scale of sequencing that we do tends to be very large. This comes with the fact that we're doing a lot of whole genome sequencing and that our assays tend to all be 20,000 genes. So we have become a pretty large-scale sequencing center, and we think it's part of the power that we bring to the field is to have become truly expert of being able to manage that and the data for it.

This is Aaron. And so you had asked a question about the liquid biopsy. And so just to make sure that you're calibrated. So our current pricing today for tissue with NeXT is ranging between $3,200 and $3,500 per patient. For liquid biopsy, the pricing is probably going to be somewhere between 15% and 20% more than that. And as John was explaining, we're going to have the opportunity for multiple time points. So the average dollar per patient could be 5, 6, 7 times of what we're getting for tissue when you look at it from a per patient standpoint. And then the other question you had about what percent of our customers are on the old platform versus NeXT. So I had mentioned in the prepared remarks that NeXT revenue exceeded $2 million in Q2. So the NeXT revenue was a little more than half of our total biopharma revenue of $4.7 million. So we're starting to ramp NeXT now. We'll always have customers on the older platform because they'll need to remain and keep having data consistent for clinical trial purposes on the old system, but – or the old platform, but we're starting to see more and more convergence to NeXT at this point in time.

Got it. And in terms of the MVP side, I kind of just want some help understanding the increased enrollment to 2 million participants. They haven't provided a specific time frame for that, I don't think. And I'm just trying to understand how that impacts everything on your end? Does that mean more samples received per batch? More frequent sample receipts? Or maybe a bigger task order expected in the next quarter?

Yes. So this is Aaron. Go ahead, John.

Yes. So the MVP, their increase in target enrollment up to 2 million is likely to take something on the order of another 10 years. They started enrolling patients in this. In the year 2011, they've been able to collect about 100,000, enroll about 100,000, collect samples from about 100,000 per year since that time. And so by early this year, they are up to about 825,000. They have samples; they have hospitals and clinics all over the United States that they've been collecting at. During COVID, they have had to shut that down. So they've stopped collecting samples live at the hospitals, although people can continue to sign up online, but collections will be way down during this period of time. But as we get past this, they've indicated that they expect to go back to collecting samples. And it's really the fact they originally were targeting 1 million samples, but they began to get closer and closer to the 1 million, and they said, "What are we going to do when we get to a million? We're really going to stop? This is so valuable." And so they concluded that they would expand the target and they said, "Oh, we're going to keep going, and we're going to set a new target of getting to 1 million." And if you think about it, the number of veterans who receive their medical care from the VA health system, depending on exactly how you count it, it might be something in the 7 million to 9 million range. So the original 1 million target probably represented maybe 14% of the patient population. The 2 million target, you're probably approaching 25% to 30% of the patient population. You can imagine that as the cost of sequencing keeps coming down, they continue to gain experience with this, you can imagine, in the future, it's likely to be that they'll want to sequence all veterans in the VA system. I think that will just become the standard. So we may have some more years for that. Right now, they have so many more samples than they've contracted us for. Even if they didn't connect – collect those sample for several years, they could continue to contract with us because they've selected over 800,000 samples, but they've only contracted with us so far for 116,000, really related by the budget. So we do continue to see ongoing bipartisan support for veterans. Both Republicans and Democrats tend to be outspokenly supportive of veterans and the VA. And so we anticipate that the budget for this has a lot of potential to continue to grow over the years. And this is a program that has many years to run. As we've said, we also have this thing right now; this is our one customer in this category. Given our experience here, we're beginning to explore how we would expand into other population sequencing projects that could be in countries all over the world. And so we think this actually could be a pretty nicely growing business for us as we look out over the coming years.

Got it. Thank you. That’s very helpful. Thanks, guys.

Okay. Thank you.

Your next question as from the line of Mark Massaro from BTIG. Your line is now open.

Hi, Mark.

Hey, guys. Thanks for taking the questions. My first one is just kind of elaborating on that last question about POPSEQ. Over the last several years, we've had, clearly, Genomics England, all of us. As you just alluded to, I think at one point, there are approximately 60 countries in the world with population sequencing initiatives. So certainly, a lot of genomes or exomes to go around. You talked, John, about potentially expanding, looking for new opportunities in POPSEQ. Can you give us a sense for whether or not your initial focus will be more on global governments? Or could there also be some with health systems, Geisinger and Healthy Nevada Project have looked within health systems? So could you just give us a sense for how large of an opportunity you think this is? And where you think some of the low to intermediate hanging fruit is?

Sure. Thank you for the question. So yes, we certainly see a lot of projects. Our sense is that the – certainly, there are national projects. What we find is there are a few countries that already have well-established, large-scale sequencing infrastructure, typically in academic settings, whether it's the Broad Institute in the U.S. or the Sanger Centre in England and that kind of thing. But actually, it's not that many countries. And so many of the countries that aspire to have this kind of a program, they want to have it in their own country because they're trying to build up. They understand genomics is an important part of a future high-tech economy. But on the other hand, they need to have people trained. They need to have a way that this can be built. And so often, we find that these countries have funding for this kind of thing, but don't necessarily know or have anybody with the experience to know how to go about doing it at this scale. And so we've been beginning to engage with countries outside the United States. And some of these programs are actually remarkably well funded even just from the government point of view to – but what they're looking for is somebody who actually has a track record where they can count on you being able to actually do the job and not essentially have to learn on the job. And since we've already done so many genomes, we're one of the few companies that would be in a position to do that. I'd say that the – you bring up about health systems. I think that's certainly a great point. And what we're also interested in, and we get inquiries about this as well is the role of pharma. Because often what you're trying to find, in addition to the benefits for the patients who are sequenced, often this is actually research to try to understand how you advance pharmaceutically. And so the – having the involvement of pharmaceutical companies upfront in helping to figure out how you're going to prioritize this, what the goals are and yes, helping to fund this, we think can actually make these really quite meaningful in terms of their medical contribution and also help solve some of the financial equation to make this work over time. So I can imagine there being consortia that involve governments and also health systems and also pharmaceutical companies all working together with Personalis as perhaps the company that's actually implementing this and setting up the capability where it needs to be geographically and all those other stakeholders also participating in terms of their expertise, their use of the data and their funding of the program.

Great. That's helpful. And congrats for rolling out the liquid biopsy product. Can you give us a sense – obviously, we're seeing a lot of liquid biopsy products, not a lot, but some that are having a pretty significant impact pretty early on. So can you give us a sense for how impactful you think this product will be? And then as you roll it out, it seems like, especially the majority of your top 10 pharma companies probably will have some interest. To what extent have you talked to these pharma companies? Many of them have probably already used a liquid biopsy solution. So can you just walk us through some of the conversations you've had with some pharma partners about the incremental utility of looking at 20,000 genes over some of the existing products?

Yes, we have. Actually, we have been under CDA, talking with a number of our pharma partners for some time about this product. I'd say there's a lot of enthusiasm for it because it's a different kind of product. I mean they also understand that will take a little while to figure out how to learn it and what the best uses are. But I'd say the – one of the nicest phrasings ever that I heard the other day from one of these pharma was, we need to not just look under the lamppost, where we've been looking for 30 years that we already know what's in EGFR, KRAS and so forth. What we need is to see the new biology that we haven't seen before. We know that there are things we don't understand because if we understood them, we already would have solved a lot of these problems. And so the ability to really, almost for the first time, be able to look all the way across all the genes and see what's changing and to do it, like at a price that's now not in the stratosphere, makes it something – sort of opens up a new opportunity for them. So I think that the fact that they've worked with other small panels that are involved in the liquid biopsy space gives them an understanding that, yes, this is practical as possible. But I think they also feel like they've had blinders on. And it's like, can we now take the blinders off and actually open it up and instead of looking through a straw, can we actually see the whole view and see the whole landscape. And so that's what we see is a lot of excitement about the potential of what are we going to see when we see the whole landscape. And so I think it will be an exciting time in the next year or two to see the discoveries that happen using this new kind of capability.

Okay. And then on the biopharma side with liquid biopsy, I recognize the market potential is significant, especially when you look at global clinical trials. But there are obvious use cases of liquid biopsies for monitoring in the clinic. So it's not super clear to me what your intentions are on generating clinical utility data and potentially taking this through a clinical pathway?

Yes. Great question. So I mean you heard us mentioning that we're beginning to take our tissue-based NeXT through the FDA. That's part of building a corporate capability and familiarity sort of at the 20,000 gene level with the FDA. They actually have been, I think, almost quite enthusiastic about what we're doing, but you can imagine that if we're going ahead into clinical opportunities, those will also be things where we're going to want the appropriate regulatory credentials as well. And so that's part of what – kind of how do we look forward at this. Certainly, the potential for monitoring, whether it's after surgical resection or if it's after – if you're monitoring a patient that's currently being treated or even looking for recurrence of cancer. Those are all things. We think that there will be two types of products. One will be ones which have a fixed footprint, and that's what our exome-scale product where we've introduced now is, we have a – it's a huge footprint. We can look at all of the genes. But there is an expense to be able to do that. We think that when we look at some of these clinical opportunities, that may be where the more personalized product can fit in, the ability to design an assay that just captures the variants for that person. Obviously, we have a great platform upfront to be able to identify the variants. And for many cancer types, the exome-scale that we can handle with NeXT will be great for finding a reasonable number of variants for that purpose. But if you look at cancers like breast cancer and prostate cancer, as I mentioned, there just aren't a lot of variants. The mutational burden is low. And so if you make a personalized assay with just a small number of variants, you could be limited by the number of molecules that are available to look at. By our being able to eventually kind of marry our population sequencing, technical expertise and whole genome sequencing with what we're doing in cancer, we would anticipate being able to start looking at finding the variants using cancer whole genome. And the footprint is about 20 times larger. And so in cases like that, you can take these low mutational burden cancers or ones that are low shedders or ones where that – the level of the tumor may be really, really tiny because you tried to just take – remove it all by surgery and you might actually have been successful with that. So the levels are really super low. So if you want to be maximally sensitive for that, you may want to look at 1,000 variants, even several thousand variants. And for that, the power of whole genome to find that many variants, which you can then bake into a personalized assay, I think is likely to be on parallel. So Personalis will be in a position to do both. We have both the exome and the whole genome, and we'll have both the tissue side of this and the liquid biopsy side of this. So we think we're going to be in a position to do things both on the pharma side and ultimately, on the clinical side.

That's great. And I guess, maybe a question on the gross margins. Obviously, there's likely a lot of room for upside over the next several years. Given that you do pretty darn deep sequencing of the genome and the exome, maybe, John – maybe strategically, how should we think about gross margin expansion over time? With the observation that you probably do some of the deepest sequencing of anyone in the marketplace. Can you get to 60%, 70% gross margin over the next three to five years?

I'll let Aaron speak to some of the specifics. I'll just stick to maybe two elements, and one is scale mix, an enormous difference that they – an operation like we have, there's a certain fixed cost associated with that. And in the early years of any company in this space, you always see this. But as the volume grows, you just amortize the fixed costs much more efficiently. So I think that's one step. The second step is, for the last while, maybe the last several decades; actually, the cost of sequencing has continued to come down. We are optimistic that there's a lot of technological headroom for that to continue to be the case. And so in Personalis, a part of our intentional strategy has been we're usually designing assays not for the technology of today. At the time we're designing them, we're often designing them for the technology that we anticipate will be coming in the coming years, which will bring the cost down. So often, it's true. When we bring out these products to start with, you say, "Wow, you're using close to, I don't know, 1 trillion bases of sequence on one sample that could be kind of expensive." It is. But in a few years, I don't think 1 trillion base is going to cost as much as it does today. And we sequence in the second quarter, I want to say, something like 1.5 petabases just in three months. And so that scale is going to keep going up. We think that the cost of that is going to come down dramatically over time. And the – I think that just sort of heads us in the right direction from a gross margin point of view. So maybe that's my broad picture of it. Aaron could certainly speak more articulately to the – some of the specifics.

Sure. Mark, yes, so in terms of gross margin, so if you look back into 2019, we had gross margins in the mid-30% range, and we actually hit 37% a year ago. So simply put, today, we're in the mid-20s, primarily because of underutilization of labor and overhead. And that's where John's description of scale matters. So today, when you look at our VA business, it's very automated. It doesn't require a lot of labor and overhead. So most of the costs are variable versus fixed. On the biopharma side of our business, most of the costs are going to be fixed. And so we're probably running, call it, 50% of our total capacity on the biopharma side of the business. So volume matters. And another way to look at it would be for every maybe incremental $1 million of biopharma revenue. It would add close to 120 basis points and up to 150 basis points, assuming similar mix as we saw here in Q2. And so you can see that by adding $4 million, $5 million, it gets us back into the 30% range pretty fast. In terms of longer term, as John mentioned, scale matters. So we see ourselves getting to a 50% gross margin as we get to scale, meaning $200 million of revenue or more. Once we get to $300 million or so, there's no reason why the gross margins can't exceed 60%. In terms of deep sequencing that you mentioned, today, our VA margins, whole genome sequencing, it's very comprehensive sequencing, and it's very profitable. The gross margins are higher than the corporate average. And so I think we can get there, and it's just going to be a matter of time with some volume.

Excellent. Thanks for those responses and congrats on a good quarter.

Thank you, Mark.

Thank you.

Your next question comes from the line of Sandy Draper from Truist. Your line is now open.

Thanks very much for taking my question and a lot of my questions have been asked and answered. So maybe just a couple of quick ones. First, on the biobank business, you said it was down 98%. Obviously, didn't stop the sequential revenue increase, but just maybe a little bit more context on the magnitude there? And then has that business started coming back on? Or what are the mitigating factors? And when would you expect that to be getting back up to speed?

Yes. So this is John. There's a customer that is – a lot of the work that they do in collecting samples for the biobank, they are collecting live sample, well, collecting them in hospitals from patients who are having surgical resection of tumors. And that kind of supplementary activity was just shut down in hospitals all around the world in the last six months here. And so a lot of their ability to collect those samples, which essentially were prospective samples, came – was very substantially cut back. We do see that beginning to change. Sample collection is happening in countries all over the world. And other parts of the world actually are recovering from the pandemic a little sooner than the United States. And so some of the sample collections that are happening in some of those other countries are actually beginning to continue. I won't speak for their business, but our sense is that there is progress in those other countries. And that in time, as the U.S. gets past this, that's a business that will come back as well. And we think it actually can be a great business. Its part of the – when you think about the data strategy for a business like this, obviously, there are choices where we can sequence and pay for everything and build our own database. And we're doing some of that. But part of our strategy is also helping other people who want to build their databases because they have specialized expertise of various kinds. And so this customer, in particular, they have some areas in which they are really world leaders. And so as they're able to get the samples back on, again, we'll be helping them to build their database. So that's a business I would expect to – could easily recover 100%. And they'll probably have some catch-up mode that they're going to want to do unless they get to a point where they can get back to processing samples.

Great. That's really helpful. And then maybe one just final follow-up. In response to – or a follow-up to a question asked earlier about the size, and you talked about in a clinical trial, it could be maybe 100, up to 1,000 type patients. When I think about the big cancer clinical trials, I'm thinking 3,000, 5,000 global patients. Is there only a subsegment of those patients that the NeXT is applicable for or the samples are applicable for? Are they only going to do a portion? I'm just trying to sort of match up the numbers you said when – first, and when I think about a big cancer clinical trial, I think about 2,000, 4,000, 5,000 global patients. I'm just trying to see where the difference is there in the numbers? Thanks.

Sure. Yes. I'd say that there can be trials of a lot of different sizes. We work with some people who have – they're doing things in Phase I, and it might be a dozen patients and some of them get up into the tens and hundreds and then some of them – the 1,000 number comes to mind. I'd have to think about whether we've handled ones that are larger than that. There's not a capacity problem for us to be able to handle that. So I mean we've sequenced 75,000 human genomes for the VA. So if somebody has a clinical trial and they need us to sequence 5,000 tumor samples, that wouldn't be a problem. I think it's a question of how the individual pharma is handling it and so forth. In general, what we do see are all of the samples that they've collected in a given trial.

Got it. Okay. Very helpful. Thanks for taking the questions.

Okay. Thank you.

Your next question comes from the line of Swayampakula Ramakanth from H.C. Wainwright. Your line is open.

Great. Thank you. Good evening, John and Aaron. This is RK from H.C. Wainwright. Most of my questions have been asked, but just a couple of them. So you were talking in terms of future products, you're talking about how you're coming up with the platform which can combine both lab data and some informatics data within an AI piece and plan to launch that in the fourth quarter of this year. Outside of such a launch, what are the milestones out there that we can expect in the next six to 12 months?

Yes. So this is John. And I think the – perhaps the piece you're referring to is our new neoantigen ranking capability, and that's very – yes, so there the data is actually not mostly sequencing data. It's mostly mass spectrometry data, and we've actually had to genetically engineer our own cell lines, our own human cell lines to put just a single HLA into each cell line and then to see which peptides bind to them. So I'd say that's an example of building content. That content is being used for training neural networks. And so this is the artificial intelligence part of what we do. I try to not use the term artificial intelligence too much because it's probably overused, but it is – we actually do have pretty advanced algorithms in that case. And the really key thing is also having the data, the laboratory data to do the training because nobody else has the kind of human cell lines that we have. It took us years to figure out how to actually genetically engineer the cells and have them grow properly with the right features that we wanted and so forth. So I'd say it's a content strategy, but ultimately, that content is data, that it comes from having proprietary developments with human cell lines in the laboratory. So I think that kind of concept is one. We're not – we don't have any other programs that I'd like to announce at this point. But I guess, you can imagine, if we feel like we've been successful with one of those kinds of things, we would be open to exploring additional kinds of opportunities. We do use machine learning in quite a few different places in our informatics. Sort of once you have a team that gets good at that, they often can see many different ways to apply it. And so we don't always necessarily announce all those things. We just end up with better performance. But we haven't necessarily described all of those areas. If we – if you're interested in that, we could go back and look at some of the scientific presentations we've made at conferences, and I think we may have had some posters on some of those topics, the get out the machine learning kinds of things and all the different areas that we have applied machine learning to, to optimize the algorithms that we have.

Okay. Thank you. And then the last question for me is, how do you see your discovery relationship that you initiated with Sarepta to grow over the next couple of years? And also, do you plan to make such additional relationships, like the discovery kind of a relationship in the near future with other companies?

So I'm sorry, you're talking about Sarepta?

Yes, Sarepta.

Yes. Great. So I think the Sarepta one is a great example of opportunities for us that are outside of the cancer area. And I think people are just beginning to realize that this concept of neoantigens that we've been working with now for quite a few years is something where new antigens can be generated in medical settings that are not cancer. And particularly as you start having genetically engineered autologous products being used as part of a drug therapy, this gives a lot of room for all that genetic engineering ends up in like – in cancer, you'd call it a mutation and when you're making a drug, you call it genetic engineering, and it sounds like a good thing, but it's both things we're you're introducing different genetic sequences into the body which were not there naturally to start with. And so those will look like mutations, and they could spur immune responses, either ones you want or ones that you don't want. And so being in a position where you can help people think about that and have data to be able to do that, we think can have opportunities – additional opportunities that are outside cancer. Sarepta is quite a leading-edge customer with us. So frankly, congratulations to them for being at the head of something of a new traffic, but we think that there will be additional opportunities like that. It's probably too early to – for us to say anything specific commercially on that, but it is definitely an interesting area for us.

Thank you. Thank you, John for taking all my questions.

Yes, very happy too.

There are no further questions at this time. I would now like to turn the call over back to Mr. John West. Please continue.

Thank you, operator. And I'd like to thank our customers for their business and partnership. I also want to thank our employees for their hard work, particularly during this COVID period of time and putting us in an excellent position for long-term growth. This concludes our call for today. We would like to thank everyone for attending, and we look forward to talking with you again in another three months. Thank you.

This concludes today's conference call. You may now disconnect.