SOPHiA GENETICS SA (SOPH) Earnings Call Transcript & Summary

Good morning, everybody. My name is Luke Sergott. I cover life science tools and diagnostics here at Barclays. It's my pleasure to introduce Jurgi Camblong, SOPHiA GENETICS.

So -- let's kind of walk through here. You guys printed last night. Let's just walk through the puts and takes of the quarter and really how the year shook out versus your expectations then we can start digging into the guidance and then the business.

Sure. And thank you for the invite and good morning all. So indeed, yesterday, we reiterated our guidance, right, that we shared beginning of the year for 2022. And to give you some context, I think we're chatting about that. The business we have at SOPHiA makes that our visibility and forward-looking performance, it's quite good, right? Given we are being paid on consumption by players who are producing data themselves, and are paying us as they are loading the data into the platform. And so when it comes to forward-looking performance, we can really leverage on multiple pillars to have an idea of how business will be like in the future. So first being consumption of existing customers, which positively travels, right? And so we announced yesterday that for 2021, our net dollar retention of 142%. So which means that our existing customers grew 42% last year. Then we have a good visibility as well on what have been the contract we signed the previous year and are going to move into production. And those ones are going to produce in terms of forward-looking performance. And the third one is obviously our pipeline that's growing, thanks to new money and new content we are building and delivering into the platform, and that is going to enable us to sign new contracts this year and will produce revenue from end of the year to next Q1.

Okay. And then so as you're thinking about the 30% growth, it's more of just -- is it what's the key driver of that? Is that new customers? Or is that where most of that's going to come from your existing customers and new wallet share gain on the expand portion?

So historically, we never had net dollar retention below 125%, so which means that at minimum every year, we would grow our business 25% with existing customers, and we've been growing over 40%, right, this last year. So this gives you a sense of what's going to contribute to most of the growth next year, so which is having more consumption for existing customers. but as well increasing the ASP as our data analytics capabilities deliver new insights, more contact insights to these customers. We've seen that [ BP ] would grow. And so most of the growth really comes from existing customers versus new customers that we have to sign.

All right. And so it's like -- so -- and you're talking about consumption here, and this is what we're talking about earlier, and you're giving me the 101 on it. So as you continue to innovate and add to the platform, you guys grew your analyses up to 66,000 like 6% sequentially. Give us an idea of how the analyses themselves have changed over the last 3 years? So if you go back to where you came from, and where you are now, I'm sure you have a much more robust portfolio?

Indeed. And I think this is not a surprise, right?

Right.

Given that we are in a world where more and more data modalities become important in the chemical market. And when we speak about genomics solely as a single modality for the complexity of the signatures and the size of the panels get figure. So to give you a sense, when we were selling in 2015, our average sale price was about $50 per patient, and we were covering an average, let's say, 5 to 10 patients, right? While today, our ASP is rather around $140 per patient, and we're covering it that require our customer sequence panels of about 100 genes.

Okay. And then can you talk about how the actual panels themselves have changed as well from going from a targeted panel to -- so you guys are trying to switch now to the exome, and how you see that playing out for the whole genome?

I think the trend is very much defined by the market itself, right? And so about the needs in [indiscernible] disorders or in oncology, which are the 2 markets where we are reactive ourselves. And that are defining on the base of the new knowledge and on the new signatures, one needs to sequence, right? And so to take a step back, in the 2015 years for hereditary cancer testing, one would look at only at 2 genes. And almost of the labs are looking at 40 genes. When it comes to rare disorders, most of them would have indeed targeted panels and now they are going to full exome sequencing, so which are about 24,000 genes, right? And on top of that, they may be adding additional information such as mitochondrial DNA. So we are in a world where not only, I would say, you have more and more users of these technologies, right? In hospitals and reference labs. So where digital technologies such as genomics are being more and more adopted, which makes the market bigger. Not only we have more volume because now we know how to better use this data for multiple disorders beyond the hereditary cancer and our solid tumor oncomatology required genomic testing on routinely base. And last but not least, the signatures become more and more complex, right? Someone needs to look at more genes or one needs to look as well at very sophisticated signatures, such as HRD scoring, which are becoming a standard for prescription of PARP inhibitors to take one example.

Okay. And so as you're really just throwing off more and more data, and it seems like the whoever is producing the most data is winning right now in the market, but you're also just really following what the market demand is. So where are you seeing it going in 3 years?

Indeed, so our intent from beginning, Luke, was to be at the heart of a data game, right? And to end up seeing deep platform that would sit on top of the data that would be produced by others and by doing so, be streaming more data than anyone else. And so this creates kind of snowball effect, right? In the tech world, we talk about direct and indirect network effects. And so my opinion, it's not about who is going to produce the maximum number of data, but it's rather who's going to be streaming and computing the maximum number of data, right? And so in that sense, over the years, I think one of the very impressive proof points that we made is extending the size of our network. Today, we have about 800 customers, which are really Q1 university hospitals, academic centers, cancer centers as well as reference labs across 70 countries. They're all using our platform and connect it to each other, right? And so this, by itself, I think, demonstrate the value proposition that we have at SOPHiA and the power of this snowball effect and direct network we create as more data are being computed in the platform. So taking a step back, we started with genomic, right, to build this network. And then, as you know, we moved beyond genomics as well now with what people call phenotyping data, combining radiomics data with genomics data to be able to follow longitudinal patients as well. And this again hits back the system and make it even more smart to deliver outcomes to the clinical market, but enable us now as well to leverage on our network, on our technology and on the data we capture to serve the biopharma market.

Okay. And so the first pillar that you guys talk about is really about the land, right? So you have to land the new customers and you just went through 800 of the IDNs and leading hospitals. Give us a sense of the growth there among these institutions, but then also as it goes further downstream to regional players and small and smaller hospitals?

Yes. So when we think about the genomics market, we see that as about 3,000 centers around the world that are equipped today with genomics capabilities. So now eventually with new sequencers going into the market, the market will grow with as well a better understanding on how to leverage on genomics data for chemical perspective, the market will grow also, right? So this gives you a sense of the size -- the potential size of the network in the clinical market for genomics modality. Now beyond genomics, for radiomics where, as you know, we have a partnership with GE. This is a much bigger market, right? Where indeed, you have some hospitals that are equipped with CT scans or MRIs that are not yet equipped with sequencers. So this year, we really think about the market and about the market growing step by step. And today, our focus is very much on serving the ones who have the capabilities for producing the data -- and as they load the data, they are paying us on using the platform. So really like kind of [ small lake ], if you make on consumption. But of course, the more this field will be democratized, the more we can expect the market to grow as well.

And so when you're thinking about the partnership with GE and like you said, not almost every hospital has some type of radiography application, but not everyone has a sequencer. So if you're one of those smaller ones with just a CAT scan, could they ultimately license the genomic data that you guys have? Or is it more of just buying the access of like a digital genomic readout that marries it to the -- whatever the scan say?

That's a very good question. So basically, what you're touching is as a tech player once you build a network, how you can leverage as well on your network to basically serve maybe 2 clients now that through the platform can collaborate, right? So these are things that we already covered through a model, which, all integrated, where we can serve as well some centers that don't have sequencers, but have imaging capabilities to get access to the sequencing power of our customers that are using the platform day in, day out in oncology, in particular, for genomic testing and tumor profiling. But ultimately, what we are really trying to do through the execution of our 6 pillars, given, you mentioned one of our pillars, right? Which is to land the customer is to end up really being detect platform that will stream real-time, real world, the most precious data in oncology and rare disorders. And we believe that to end up being displayed, you need to have the biggest size of the network, hence the importance of the 800 customers we have across 70 countries, you need to be exposed to high volume of data. This year, we will hit 1 million genomics profile computed in our platform, right? So which will make our platform not only the most adoptive in terms of the size of the network, but as well in terms of consumption. And then it's very much about interactions of users with the platform and across users. And so this is typically what we are animating now as well through a model called the CarePath where beyond the pathologies that we serve with our genomics capabilities, beyond the radiologists that we start to serve thanks to our radiomics capabilities, we are connecting all together so that the oncologists can have a longitudinal perspective of how the patient has been responding to a treatment. And then the more data being gathered into the platform, the more the oncologist starts to see a cluster of patients to anticipate how the next patient that suffers from a similar cancer may as well answer to this type of treatment, right? And so this is really the intelligence we're building. And operationally, we're building that first landing indeed customers into the clinical market, so with comprehensive cancer centers, academic centers or reference labs that adopt our platform for a first application that are growing with us because our volumes are growing over here to whom then we can sell a new application. And then indirectly, as we are growing within the clinical market, we can leverage on these capabilities on this network and on this data to serve the biopharma as well. And then one fits the other, right? And this is the virtuous cycle that we've started to create since 2015.

Okay. And that's a great segue into the biopharma partnerships, right? So I think when I first met you guys I made a mistake in saying that it sounds like you're a CRO, and that was clearly not the case.

Okay.

So you educate me on that pretty quickly. So really talk about how the biopharma segment has been growing with you? And really, because as you think about innovation, you need a clinical utility on the one side and the diagnosis on the others. So you're working on the hospitals and the clear path in the radiology side on the diagnoses. And then so how does pharma help fuel the innovation cycle and companion diagnostic opportunities that you guys see? And then I want to follow up with a Deep IV or DEEP-Lung-IV?

DEEP-Lung-IV. Okay. Great. So as you said, one fits the other, right? And indeed, often companies start to serve the pharma before eventually they go to the real world for the diagnostic market. And we intentionally did the opposite because for us, again, it was about data. And the data that pharma needs are the data that are as well being produced real-time real work, not only their own data, right? So now that I think we've been able to build this network in the clinical market, we start to see being in a position which is pretty unique to serve the biopharma on the data we capture in the market to, for example, help them compare their data versus the one in the market to accelerate approval of drugs, right? Which is a so-called synthetic control arms. But beyond that as well, we can help the pharma, for example, interrogating hospitals that are identifying patients that would benefit from the new drugs that are being just approved, which is an awareness exercise, or on enabling them to accelerate and onboard patients that do fulfill specific criteria for participating into clinical trials, right? So these are the kind of activities that one can leverage on when you own this position in the clinical market, and you capture this data real-time, real world. Beyond that, I can give you another example, which is about HRV. As you know, PARP inhibitors will become very important in the world of oncology. Certain people are even telling that they might be as important as [indiscernible], right? So they could be transformative in the way we cure patients from cancer. And in that sense, I think our model has proven to be very effective. So from one hand, we launch a product, which is still RUI in the market to serve some centers like Dasa that's announced fifth largest lab in the world based in Brazil. SOPHiA in Taiwan, Ambry that we announced in terms of a letter of intent, which is signed, right? Which are the customers that are adopting this HRD scoring capabilities in the clinical market. But of course, this indirectly serves the pharma like AG, who need as well to have more and more labs and hospitals that are good and accurate at testing patients on the HR discourse so that PARP inhibitors can be prescribed to these patients on right? So I think that's a nice way to show how one works with the other, and then, eventually, how you can start working with the pharma for the next generation of CDx, leveraging on this technological capabilities, but now maybe for a new type of cancer that will require as well this type of [indiscernible].

That makes sense. And so give us a sense of how different the economics are on the biopharma side versus the diagnostic side? And the diagnostic side, it seems pretty clear, right? When you're talking ASPs continue to climb as panels getting more complex and they're just consuming more and more. So how does it -- is it similar on the biopharma side? Or is it -- you have some type of economics on the back end load?

So the economical model is different for us in the biopharma, but as you can understand, any data that is out there, it's important for the pharma, right? And they tend not to only favor one source because if you are sourcing the data with one partner, but not with the other, you may be missing some data that may make a difference. So as you know, drug development being so expensive, I think the pharma now really understands that they need to get access to diverse data sets to be able to go into the market more quickly. So our model, taking a step back in the clinical sector, it's a pay per use, right? And this is why our forecast can be quite predictable is because we know how customers are using the platform every hour, and we know what the ASP like. And in the clinical markets, indeed, we grew from 50 to 140 enough for HRD-like applications, our ASP is above $500, right? So which gives you a sense that even in that market, the more complex will be the markers, the more data analytics will be valued, the more we should be rewarded. On the biopharma market, it's a different model. So we are rather working with the pharma on the access of the insights or on specific projects, right? And so on some discussions that we may have regarding co-development efforts or eventually, future CDxs, right? That we haven't signed so far. Obviously, there is a milestone-based contracts, which are very different than the contracts we've been signing until now.

Okay. That's helpful. And then so let's talk about DEEP-Lung-IV, the retrospective study. And give us a sense of what -- what this is -- what the readout will portend or say, assuming everything works? And then I guess I had a question because it's on late stage. And as we're trying to get earlier and earlier down the diagnosis pathway, how easy is it for you guys to just move down that pathway with medicine?

Yes. So first, I would say, unfortunately, as you know, often it is a late stage, right? This was true with antibiotics and penicillin a long time ago because you start to serve patients that don't benefit from the other treatments. And indeed, if you demonstrate utility then, maybe you can go to earlier stage. So DEEP-Lung, really, I think, is a synthesis of what SOPHiA will be in the future, which is this player streaming real-world multimodal data day in, day out to create a comprehensive and evolving intelligent system for the oncologists that will be able to see how the next patient compares with other patients. And so in the study, we are -- we will be following 4,000 patients, and it's retrospective and prospective. We're combining genomics data, radiomics data. So we extract about 250 radiomics features from CT scans, plus about 200 clinical data points. We know that the algorithm that has been built before we would launch the study is promising when it comes to sensitivity and specificity. And so now we are being [indiscernible] November, we're putting the sites which have been 16. I joined our effort to be able to start already collecting this data and validate, if you like, real world what we have seen in our proof of concept as we were completing about 200 patient data points prelaunch of the study. Okay? And so the end game will be that eventually we will be able to predict who will respond or not respond to immunotherapy for patients that are suffering from non-small cell lung cancer, metastatic non-small cell lung cancer, stage IV, where, today, the treatment is expensive and where there is no predictive biomarker that can tell who is going to respond or no response, right? And so this is, if you like, I would say, a synthesis of what SOPHiA will be like as we are able to gather this multimodal data to serve the clinical work, but leverage on the data and technology as well to serve the biopharma industry, right? For example, to put in the market earlier more targeted drugs. And we started with lung cancer Stage IV because we know that there is a big demand. It's very expensive, but we have similar efforts on other type of cancers.

All right. Okay. And so as I'm thinking about your -- the company trajectory here, it's really -- you're still in the really early phases of just building out the network of the platform, right? And at the same time, you're working in parallel to bring on the content creators that are going to eventually come up with new therapies and you'll have new companions. And so that would be the final leg of growth taking you to the next stage?

Correct. So I think it's early, but, in the meantime, we are leading by far, right? So it's very, very difficult, as you know, to win the trust of clinical customers, right? And so we've owned this trust because of the accuracy of our algorithms, because of the quality of our platform. And as well, I would say our sensitivity to data privacy, and that's where we're leading versus others by far, supporting with the tech platform, this decentralized world. But of course, it's just the beginning, right? We're just seeing the tip of the iceberg.

And so it's a great segue into the competitive dynamics in the competitive market, right? I mean the other 2 companies have been around for similar amount of time in development. How do you guys differ and fit into the landscape? And give me a sense of your win rates. Do you even see the other competitors in a lot of these deals?

So I think the competitors you refer to are more central models, right? And they produce rather the data by themselves. So which is a different model and which is, I would say, complementary to our model. And when we serve the biopharma industry, the biopharma customers wants to have their source of data and our source of data, right? It's not one or the other. So it's rather complementary. In the clinical market, these are customers we can serve as well. The last you were thinking about because they may be very sophisticated in one disease application, but maybe they don't have HRD capabilities, maybe they don't have exon-sequencing capabilities. So to us more like partners with whom we can work rather than direct companies. And I think Ambry is a demonstration of that, right? Because Ambry is very similar to these companies. And now, hopefully will be a strong partner to SOPHiA to create this word, which is [indiscernible] world of centralized and decentralized world multimodal to serve the maximum number of patients in the clinical sector and, again, leverage on that with biopharma.

Right. Because I remember, I was before talking we were talking at Konica Minolta acquired Ambry and they were going to use their radiography data and a very similar model, but then I saw that you guys did the licensing. Give me -- so is it just for HRD, but then gives me a sense of this is another customer or a deal where you're going to try and expand and give more content?

Definitively. So the idea is first to demonstrate them the superiority of our technology for their own needs in some of these areas such as HRD, but there are others in the list. And in parallel as well, leverage on some of their technologies, including what they have been doing with Konica Minolta in the imaging field and our technology to eventually build new content as well, right? And I guess, under side serve clinical market, pharma as single [indiscernible], but as well benefit from us being a decentralized player to serve a more diverse set of customers. I think this speaks very much to the pharma coming back to your initial question because, as a pharma player, you want to be able to have an accurate outcome. That's why often you start in single site [ TNH ], but in the meantime, you want to have a broader reach. I think this is something that Ambry understood that by partnering with us, they can bring this value proposition to the pharma, even preapproval of a drug, and not only tell them, we will serve you preapproval and then as a single site [ TNH ], and with very high quality using SOPHiA technology. But then the benefit is at the same technology we do feel as well to serve other type of clients.

Makes sense. That's great. We're out of time. I could keep going, but I appreciate the time. It's been a pleasure.

Thank you. Thank you so much.