iCAD, Inc. (ICAD) Earnings Call Transcript & Summary

Good day, and welcome to the Second Annual Winter Wonderland Best Ideas Conference. The next presenting company is iCAD. [Operator Instructions] I'd now like to turn the floor over to today's host, Stacey Stevens, President and incoming CEO of iCAD. Ma'am, the floor is yours.

Great. Thank you so much, Matthew, for the opportunity to be here, and thanks for those listening on the call. Again, I'm Stacey Stevens, I'm the President of iCAD, and very seem to be, in the next couple of weeks, the CEO of iCAD. And I'm really looking forward to telling you more about our company today and really appreciate the opportunity to be here this morning. So just get the slides up here. So let me introduce you to iCAD. iCAD is a world leader in bringing to market what truly is transformative technology for the detection, risk assessment and treatment of breast and other cancer types. And very simply, our technology is saving lives and improving the quality of life for countless patients along the cancer care continuum. There really is no other company that does what we do today or that can match both the clinical and economic value proposition that we bring our customers or the life-saving technology that we're bringing to patients globally. So for those who may be brand new to the story, there are 2 segments to iCAD's business, each of which have some pretty exciting growth dynamics. The first is an AI software business, and that makes up about 2/3 of the revenue of the company historically, where we have developed a very robust portfolio of breast cancer risk assessment and detection software that gets used in conjunction with mammography to find cancer earlier and faster. We were the first to market with an AI solution called ProFound AI. And today, we are the dominant market leader for this type of mammography AI software. And this product represents the majority of the revenue of this segment of our business. We're also significantly improving workflow for radiologists, which, of course, is especially crucial during this time of critical staffing shortages in health care. And we're also leading a major paradigm shift in mammography from what today is an age-based screening paradigm to what in the future will be a risk-adjusted screening paradigm that's very unique and personalized for each one, and that we'll talk more about in the presentation. The other segment of our business is a unique radiation therapy modality called electronic brachytherapy. We actually own the IP on the smallest x-ray source in the world. And this incredible technology allows us to deliver really revolutionary treatments that offer both excellent clinical outcomes and really improved quality of life for patients with different types of cancer. Today, we're primarily focused on treating skin cancer and breast cancer. But we're running a multinational clinical trial to prove the efficacy of the technology and treating brain cancer, in particular, glioblastoma, which, of course, is one of the deadliest cancers where patients have a very poor prognosis, so we're very excited about some initial data published for the treatment of brain cancer and looking forward to this being a potential breakout opportunity for the company. So before we get into the details of our AI products and the value that we bring, let me just first put into context the fight against breast cancer. So in 2021, breast cancer surpassed lung cancer as the most common women's cancer worldwide. And we know that 2.2 million women will be diagnosed by 2025. 1 in 8 women will develop breast cancer. And some of these cancers are very aggressive cancers that can develop and rapidly grow between an annual mammogram. And despite what has been just a ton of advancements in screening programs globally and advancements in technology, there's still a staggering statistic, which is 20% to 40% of cancers are still missed, right? So there's just a huge need for tools that can help clinicians get these cancers detected and treated earlier. So in order to understand the unique value that iCAD brings, I think it's helpful just to take a quick look at the evolution of mammography. So going back 50 years or more, you had radiologists reading film on light boxes. And then in and around the 2006 time frame, the market began moving into a digital format, but still in the 2D realm. And we sold the product back in those days for 2D mammography, and we still have a large install base of 2D customers. And then in the last 6 or 7 years, the entire market started transitioning from 2D mammography to 3D mammography or what we call tomosynthesis. The good news is that 3D is better at finding cancer, but it also creates some significant clinical and workflow challenges for radiologists. And that's exactly where iCAD comes in with a very strong value proposition to address these challenges. So if you look on this next slide, in the top left of the slide, you see an example of a 2D mammogram. So in the 2D world, radiologists were looking at 4 static images of the breast. And that radiologist could look, and very quickly, in a matter of 15 seconds, determine whether there was something that looks suspicious for cancer. Now today, in the 3D world, on the photo on the bottom left here, that same mammogram in the 3D world now contains up to 300 1-millimeter slices of the breast. So imagine a radiologist sitting at a workstation scrolling through all 300 of these slices. And you just don't know if a subtle cancer is hiding in slice 36 or 75 or 102. And it's just a very tedious and time-consuming process, and it can be super, super easy to miss something. So you can see it's just a tremendous amount of data to get through. In fact, if you look in the chart there, you can see the same number of images that radiologists would read in the 2D world in a year is now produced in less than 3 days in 3D. So here's where the value of our AI algorithm comes in. So it's always on. It's a concurrent read. And what it does is that it individually analyzes each of those 300 slices of the breast. And it does that based on the training it has received on millions and millions of images, right, to know what looks like cancer. And it calls the radiologist's attention to those slices that looks suspicious for cancer. Some of our KOLs will liken it to having a second radiologist reading over their shoulder. So you can imagine the impact of this. And it truly is the emergence of AI that's been driving the next era of mammography. So we were actually first to market with the solution right at the beginning of 2019, and we were a couple of years ahead of anyone else in the market. Since that time, we've continued to build on the clinical and workflow superiority of the algorithm. And we're now getting ready to launch the fourth generation of this algorithm in 2022, which will be yet another leap forward in both clinical and workflow performance. So in order to get the FDA clearance on this product, we had to conduct a really rigorous clinical reader study. And the way it worked is that we had 24 radiologists reading 260 mammograms with and without the use of our AI software. And then we measured the difference in clinical performance and reading time. And the results were pretty impressive, right? So you can see here that when radiologists read with our AI software, they saw an 8% improvement in sensitivity or cancer detection accuracy. So while this may, on the surface, sound like a low number, when you put it in context and consider that there are 40 million women who are screened every year in the U.S. alone, that equates to several million women with cancers that would potentially go undetected without the use of our tools. So it's staggering, right? At the same time, we're able to lower false positives. This is a big issue in mammography, right? So we're able to lower false positives by 7%, and now eliminate the unnecessary stress and anxiety for women that's associated with a biopsy. And we're able to do all that by cutting in half the time it takes to read those mammograms. And this has become a particularly important piece of the value proposition of the product, particularly during this crazy COVID time where there's so much stress on clinical resources and health care facilities. Here's another way of looking at the impact of the tool. So you've got 2 graphs here. And on these graphs, you have sensitivity or cancer detection accuracy on the vertical axis and false positives on the horizontal axis. And the circles that you see on the graph represent the radiologists that were in our clinical reader study, and the size of the bubble is the time it took to read the mammograms. So what you see here on the left represents how they performed before the use of our AI software. So what you notice is that there is just a ton of variability, right, in radiologist's performance. Some are very good at detecting cancer but they produce a lot of false positives, right? And vice versa, some produce very few false positives, but they're also not so great at detecting cancer. So ideally, you want a radiologist who is up into the left of this graph here, right? Now if you look at the right side, this is what happened to these radiologists in terms of clinical performance when using our AI software. So very noticeable. We improved the performance of all the radiologists. They all moved up into the left of the graph, right? While at the same time, the bubbles get smaller and we're reducing the reading time to read that mammogram. So this is the real value of this product. Coming up here on this next slide is an actual case. An actual mammogram that illustrates the power of the tool. So this is a real mammogram of a woman with a very subtle cancer that was hiding in one of the slices of the mammogram, right? So even if you're a radiologist, it can be very hard to see this cancer, right? You look at it, and it just looks like a lot of white kind of snowy effect. And in fact, without the use of our AI software tool, 12 out of 24 radiologists missed this cancer in our reader study, right? So half of them missed the cancer, right? Now when they were using our AI tool called ProFound AI, 22 out of 24 radiologists found it. Actually, the tool found all of the cancer, but 2 still decided not to call it cancer. So this is just absolutely staggering when you consider that this woman, right, would have had a very high chance of having her cancer go undetected if our tool wasn't used and she would have left her health care facility and walked around for another year with cancer growing in her breast until her next annual mammogram, right? So you really can understand the power of this tool in terms of detecting cancer earlier, getting patients diagnosed earlier and getting them treated in the much earlier stages of cancer. So that's the clinical side. Now let me talk for a little bit about how we think about the addressable market opportunity, and how do we monetize this product as a company. So at the highest level, today, we're playing in a very large underpenetrated market with a lot of runway to go. So there are a couple of ways to think about our addressable market opportunity. There's the market today and there's a new evolving model that's pretty exciting and makes our company look more like a software -- a true software company. So today, what we sell is predominantly on a capital sales model with the sale of perpetual software licenses. And keep in mind, what you're seeing on the chart here is 1 product in 1 geography, right? It's our ProFound AI product in the U.S. market only. And the way to think about it is that there are about 11,000 3D systems today and another 2,500-or-so that are in the process of converting to 3D, right? So that's what we're going after, right? We have, up until this point, sold around 1,200 or 1,300 licenses in this market, so we still have 90% of the market yet to go, right? So the market is 13,000 units, about $400 million of opportunity in that 1 product alone in the U.S. market. And if you layer in another product that I'm going to talk about later on in the presentation, which is relatively new, which is our short-term risk assessment solution, along with our software that assesses breast density. You get a market in this capital world that we're operating in today that's around $800 million, right? So a lot of the times when our customers buy the capital, we won't see them again for another few years, right? Because they will have used their capital budget, and it will be a few years until they come back to us to buy the next algorithm upgrade, right? So while there are some differences in the TAM now versus in 5 years, it's essentially the same when you pull out kind of service and some upgrades that we would get during that time. So what's pretty exciting though is where the market is heading, right? So beginning at the end of 2021, we started to see increasing demand from our customers for an operational model, or what you'll hear us refer to today as our subscription model. And what is driving the desire of our customers to find -- what is driving this is the desire of our customers to find a way to deploy the technology faster without having to access capital budgeting process, which, of course, we all know has been very constrained and difficult during COVID. So they're looking for a way to get this technology earlier and to not have to take it out of the capital budget. So we've now fully operationalized the monthly subscription model as a new way to acquire our technology. And the value of the customers that move to this model will be much greater to iCAD over time, and I'll show you a 5-year comparison in a minute. And you'll see the comparison of how we will recognize the revenue on this side between our capital model and our subscription model, if you look at kind of the middle line here. So what would have been on average a $30,000 recognized upfront for a perpetual software license, and perhaps some recurring revenue from service, although our attachment rate has been fairly low, is now approximately $1,000 per month per mammography system, and we'll recognize that ratably over 30 months approximately. But what's great about these customers is that we expect them to be very sticky, right? So nobody in the history of selling this AI solution has ever tried to return it or say that it didn't perform as expected, right? So now we have a very nice track record of launching a major algorithm improvement at least once per year, and we'll get these customers to stay on the subscription model, maybe in perpetuity right? And we'll upsell them new algorithms, we'll upsell them new functionality like risk over time. And what we'll be doing is creating a growing installed base of customers that are generating just a ton of return -- of recurring revenue each year. So now if you go back to the same TAM slide that I showed you in the capital model, what used to be $800 million for a onetime purchase, and essentially, you don't get a lot of future revenue for years to come. Now we're in a situation with the subscription model where we're generating annual recurring revenue. And as our install base grows, that ARR compounds and you get a 5-year TAM that is double what it would have been under the traditional capital model. So we're excited in 2022 to start to make this transition. This is not going to happen overnight, right? Our customers -- for every customer that wants this model, there are still plenty of customers that want to buy a capital perpetual model. But we're expecting in 2022, on a blended average basis, to have around 15% or 20% of our business shift to this model and then it should accelerate from there. So we're going to be talking a lot more about this dynamic, this transition when we report our full year 2021 results later this month. And as we go forward in 2022 and report our quarters, we will break out the capital versus subscription so that there's full visibility to it every quarter. So the other important news and to move to a subscription model like this is that there are other changes going on in our market that we are swiftly adapting to as AI becomes a more important part of every health care facility strategy. It used to be that customers bought AI either from an imaging company or iCAD, right? And now the market is evolving and there are many channels that are selling AI. A customer may want to buy it from a patch company like a Change Healthcare or a Setra or even one of the emerging AI aggregators like Arterys, where we have a partnership. So the challenge for us is that we need to be everywhere the customer wants to consume the AI with the appropriate business model over time. And in the next year-or-so, we will move to more of a true SaaS hosted model once we overcome some technology hurdles and getting all these volumes of data to the cloud. And the other benefit of this is that we can access other budgets outside of the clinical department, the IT department as an example. The other change that we've adapted our commercial strategy to is the emergence of more enterprise deals that we've referenced in the last couple of earnings calls. Historically, iCAD sold perpetual licenses to a radiology department. Our average deal size was 1 to 2 licenses. Now, given the compelling value proposition of the solution and our customers' desire to provide equal care to all patients, we're seeing more hospital systems and networks interested in deploying the solution system-wide. So while that leads to a more complex sales process with more decision makers and it can elongate the sales cycle a bit, well worth it, right, for the much larger deal size and ability to get faster market penetration. And then finally, we are changing our approach in 2022 to just go after more of the market opportunity due to some system compatibility challenges and some time that it takes to gather data as the imaging companies improve their systems and our historic focus with particular partners, we've really been generating the majority of our revenue from a small segment of the business, less than 25% of the market. So in 2022, we've made some changes to our sales team to now aggressively go after the entire market with a superior value proposition in thousands of 3D gantries that have no AI at this point. So up until this point in the presentation, I have been talking about 1 product, ProFound AI. But now we are really entering what I call the next frontier of mammography that's going to completely change the landscape for breast cancer. And this is the area of short-term risk assessment. And so what we've done is we have developed the world's first short-term breast cancer risk assessment AI software that came as a result of a multiyear collaboration with researchers from the prestigious Karolinska Institutet in Sweden, working with a database of 70,000 women over 8 years, looking at the risk factors that lead to short-term breast cancer diagnosis. So why is this important? Well, if I'm a woman and I go in for a mammogram every year, I'm given a lifetime risk score. And that risk -- lifetime risk score is based primarily on filling out a questionnaire about my family history. The problem is these lifetime models are no better than flipping a coin, they're 50% or 60% accurate. And they're not very actionable, right? What would I do if I'm told that I have a 50% chance of being diagnosed with breast cancer in my lifetime, right? So -- and we tend to think that breast cancer is running in families, but the reality is 85% of breast cancers occur in women that have absolutely no family history. So physicians absolutely need a more accurate and standardized tool to assess a woman's short-term risk of developing breast cancer, right? And this is where our next high-impact AI tool comes into play. So we've developed the world's first clinical decision support tool that can assess, with greater than 80% accuracy, which women will be diagnosed with breast cancer in the next 1 year. This is super important, right? Because women who are in the highest categories, the radiologist is going to send these women on for supplemental screening like MRI or ultrasound, get these cancers diagnosed earlier, get these women treated earlier with obvious benefits to the patient, but also with an enormous cost savings to the health system. As we know that treating breast cancer in the earliest stages is much less costly than treating them in the later stages. So we're taking a very clear leadership position here in transforming breast cancer from what today is an age-based screening paradigm, kind of a one size fits all, right? To now a risk-adjusted screening paradigm that's truly unique and personalized for each woman. So that's our AI business, and we're really just getting started on what is truly an enormous opportunity from a revenue standpoint and of course, from the standpoint of the impact on patients. Future growth in this business will come from driving more revenue from our existing products, including risk, selective geographic expansion and then, over time, leveraging our core competency and algorithm development to expand to other clinical application areas. So we're starting to run short on time here. So let me just give you a very quick overview of the other segment of our business, which also has some exciting growth dynamics and that came to us via acquisition at the end of 2010. So this technology is called electronic brachytherapy. You can think of it as a high-dose x-ray-based technology. It's very different from conventional radiation in that it's not a live isotope and doesn't require special shielding or security. It can be used in virtually any treatment setting. The device is FDA-cleared to treat all cancers. Today, we're focused on treating breast cancer intraoperatively with an 8-minute treatment that replaces 6 weeks of radiation. What's driving the growth in this segment is primarily our U.S. dermatology business, where we can treat several million patients a year in the U.S. alone with a 5-minute procedure twice a week for 4 weeks that replaces a surgical procedure. And then as far as the future, we're currently conducting a multinational clinical trial to look at using this treatment on brain cancer patients, in particular, glioblastoma. And we have some very encouraging early data out of a study done in Russia that proved that we could extend both overall survival and progression-free survival of these patients. So we're anxiously awaiting the outcome of our multinational study and expect to have data in about a year on that. So I'm going to kind of switch over to the wrap-up real quickly here since we're running out of time. If I can get the slide to advance. So just to kind of sum up here. First of all, I am absolutely thrilled to be taking over the helm of this company in just a couple of weeks at such an important inflection point in the company's history. We have a world-leading differentiated product portfolio, and we're selling into very large underpenetrated market, so there is a huge runway of opportunity in front of us, right? If we think about 2022, our focus is on driving revenue growth, fully penetrating these large enterprise opportunities, transitioning the business to a subscription model over time, which will give us a large base of recurring revenue like a true software company. We've got attractive margins that will even improve going forward and many adjacent and untapped markets to expand the business globally over time. So all of these really gives us tremendous confidence in our ability to achieve our primary goal, which is delivering significant increased shareholder value. So thank you for the time today. I appreciate it very much, and we'd be more than happy to do a follow-up with anyone listening to the presentation today that wants to get into some more detail.

Thank you. That does conclude the iCAD presentation. You may now disconnect. The next session will begin shortly. Please consult the conference agenda for the next presenting company.