Epigenomics AG (ECX) Earnings Call Transcript & Summary

Good afternoon, ladies and gentlemen, and welcome to the Epigenomics AG Study Results Conference Call. [Operator Instructions] Let me now turn over the floor to your host, Mr. Frederic Hilke. Go ahead, please.

Thank you very much, Julia. Hello, and welcome to the conference call. Today, we will discuss the recent publication in the Journal of the National Cancer Institute titled Comparing the Cost Effectiveness of Innovative Colorectal Cancer Screening Tests. My name is Frederic Hilke. I'm the Investor Relations Manager for Epigenomics. With me here on the phone is the CEO of Epigenomics AG, Greg Hamilton; and Dr. Jorge Garces, Epigenomics President and Chief Scientific Officer. Before I hand over the presentation to Greg, let me point out some of the usual legal information. The following presentation will contain expressly or implicitly certain forward-looking statements concerning Epigenomics AG and its business. Such statements involve certain known and unknown risks, uncertainties and other factors, which could cause the actual results, financial condition, performance or achievements of Epigenomics AG to be materially different from any future results, performance or achievements expressed or implied by such forward-looking statements. Please find some of these factors described in the safe harbor statement. And now I would like to hand you over to CEO, Greg Hamilton, who will continue with the presentation.

Thanks, Frederic. Good morning, everyone, and thank you for joining our call today. We are excited today to discuss the recent publication in the Journal of the National Cancer Institute. Dr. Jorge Garces, our President and Chief Scientific Officer, will guide us through the publication, the key highlights and the implications for Epi proColon. Afterwards, we will both be available for questions. And now I'd like to hand the presentation over to Jorge.

Thank you, Greg, and everyone hear me okay?

Yes.

All right. Good morning, this is Jorge. I'm happy to be presenting today on this exciting new development. Really, this work, I want to start, was published in the Journal of National Cancer Institute, which is a very highly prestigious journal. It ranks in the top 1% among the 31,000 scientific journals that are in publication, and it ranks #6 among all cancer research journals out there. The study looked at modeling the long-term benefits, harms and the costs associated with a number of CRC screening strategies. FIT and colonoscopy have been proposed as screening strategies for colorectal cancer really since the late 1960s. And so this is the first study to evaluate new or innovative screening methods such as Epi proColon in terms of their cost effectiveness. The analysis found colonoscopy to be clinically the most effective, FIT to be the least expensive and Septin9 to be the test of choice amongst all other methodologies. So when Epi proColon is administered annually, it was found to be clinically more effective than both Cologuard performed every 3 years and FIT performed annually. Among the tens of millions of individuals that are not willing to undergo FIT or colonoscopy, which is about 1/3 of the population, the test of choice really was found to be the blood test, Epi proColon. So what is the importance of this publication? Well, first off, it was developed by CISNET. And CISNET is a cancer intervention and surveillance modeling network. It's really a consortium of NCI-sponsored investigators that work on simulating the mortality and incidence of cancer and then understanding how various screening methods can play a role in preventing and treating those cancers. These models have served to really guide public health research and to set policies in terms of reimbursement or clinical guidelines. For example, CISNET has been critical in informing USPSTF as part of setting their guidelines for colorectal cancer screening. It did so in the 2016 guidelines. More recently, it served as the key input to lowering the recommendation to start screening at the age of 45 by the American Cancer Society. And as importantly, CISNET has played a key role in really serving as a data point for determining CMS coverage for many of these tests. So one of the key things in the paper was to look at the clinical benefits of these tests. And as you can see on the left side of this figure, when you have a no-screening population or population that's unscreened, you expect about 110 events or colorectal cancer cases for every 1,000 people that are screened over the lifetime of those patients. And the number of cancers is about 42 or 43 cancers over that same time period. When you apply any of the screening methods, you can see that you lowered both the incidence and the mortality of cancer. Now the orange line on the bottom shows where Septin9 done every year can reduce the mortality associated with colorectal cancer. You can see that it does a much better job than every other method to the left. So it outperforms FIT done annually. And it outperforms Cologuard done every 3 years. Same thing for the incidence of colorectal cancer, you can see a very, very sharp reduction in the number of colorectal cancer cases, which, again, beats other methods. It would only colonoscopy can slightly improve over Septin9. And you can see that you would have to do Cologuard on an annual basis to, again, to beat the performance of Septin9, but that would be extremely cost-ineffective to do that since Cologuard is about a little over $500 a test. And then the PillCam done every 5 years would match the performance of Septin9 done yearly, but the PillCam is over $1,300 per test. And so again, that would make it cost-ineffective. So among all the tests, really Epi proColon stood out as clinically providing great benefit. If we look at the benefits and harms of the test, and I want to address this really upfront, one of the criticisms to our test has been that we have a high positive rate and a high referral to colonoscopy, about 20% positivity rate. Now some people have mistaken this that you can achieve this type of benefit by just sending 20% of patients randomly to get a colonoscopy every year, and that's just not true. If you were to do that, you would pick up randomly about 20% of the cancers. But with Epi proColon, that 20% actually contains 70% of the cancers. So we are picking up in that 20%, 70% of cancers, and therefore, sending a very enriched population to get a colonoscopy. So that's the one difference that I wanted to point out. Secondly, if you look at the number of colonoscopies that Septin9 done yearly yields, according to the model, it yields about 3,827 colonoscopies among 1,000 patients screened, again over the 25 years of screening. If you look at colonoscopy itself, then every 10 years in that same cohort, it actually yields 4,715 -- I'm having trouble reading, 4,735 colonoscopies. So much, much greater number of colonoscopies are seen using what's considered the gold standard. And so the idea that Septin9 leads to too many colonoscopies when you measure that against the gold standard, it actually yields 20% less colonoscopies than what's considered the gold standard today. So 2 things that I thought are very important coming out of the data. If you look at in terms of incremental costs, so one thing that often confuse people about these analysis, you can see that if you include FIT and colonoscopy in this column, where it says ICER dollars per quality-adjusted-life-years gained, you can see that FIT when it compared to no screening, actually is cost saving. That's because if you were to screen the population with FIT, you would save on the treatment costs that are incurred by patients that are not screened. And then you find those patients have -- will detect their cancer at a much later stage, which, again, is much, much more expensive to treat. Now when you do these cost analysis, for example, with colonoscopy, you then don't compare colonoscopy against the no-screening population. What you do is you compare it against the best next alternative. So this $48,155 is really the incremental cost that you get by implementing colonoscopy versus FIT to show that, in fact, it is a cost-effective method over FIT. And so you -- why would you want to do that? Well, because you would increase the quality of life years gained from 199 to 209 over FIT. So this delta in additional quality of life years that are gained costs about $48,000, which is acceptable given a cutoff of $100,000 for every life year gained. So same thing here. So when we look at -- when we ignore again, colonoscopy and FIT, which again have been methods that are broadly accepted for colorectal cancer screening, when we ignore those and say, "Okay. For the 1/3 of the population that is resistant to these 2 screening methods, what are the most cost-effective?" And you can see that for CTC, it would be cost-effective over no screening because you have about 177 life years gained over 0 for no screening at a cost of $1,092. Septin9 then has to be compared to the best alternative, which is CTC. Now if I apply Septin9 every year, I actually increased from 177 to 194 at a cost of $63,000. So again, a cost-effective method over the next alternative. And then really, those 2 methods dominate all the other ones. So I'll explain in more detail what that means. And Cologuard, for example, when done as prescribed every 3 years, noticed that it only yields 175 life years gained, which makes it a weaker method than CTC. And also, it's more costly at nearly $8.9 million versus CTC, which is only about $7.5 million. So it actually has less benefit and is more costly. So we want to find technologies that have greater benefits and are either less costly or only cost $100,000 more or less for quality life years gained. So that's really the purpose of the analysis. So what this is showing really is that Epi proColon is a very cost-effective method among those patients that refuse FIT and colonoscopy. So this is really what's called an efficiency frontier. And we've looked at data like this, where we looked at benefits in terms of quality of life years gained. And on the x-axis, we usually look at harms, right, benefit to harm analysis where this would be the number of colonoscopies. And clearly, you want to minimize the number of colonoscopies while maximizing the quality life years gained. And in previous papers that we have discussed, including the one published in Cancer Medicine by the Harvard Medical School showed that Epi proColon had a very attractive benefits-to-harm ratio. Here, we're looking at the benefits over cost. And so you can see that the cost of no screening here being set at the baseline, and FIT provides greater benefit, obviously, but it also provides that benefit at a lower cost. So FIT is actually a very attractive method from a cost effectiveness despite the fact that it is less clinically effective than, let's say, Septin9 done annually, which is much higher in terms of quality life years gained. So when we look at this efficiency frontier, again, we discussed that colonoscopy can yield a great benefit but at an increased cost of about $48,000, which, again, is cost-effective. Now when we exclude these methods from the analysis and then look at all the other alternatives, CTC, again, although it provides a greater benefit than no screening is not as clinically effective as Septin9. And so we can do much better than CTC in this patient population by doing Epi proColon every year at an additional cost of $63,000, which again makes it cost-effective. Now anything that's to the right and below these points or below the frontier is really dominated by this line. So for example, if we look at this inverted red triangle, this would be Cologuard done every 3 years. And you can see that Cologuard every 3 years not only yields less benefits than Septin9, but it's also more costly. So this, again, would be a nonattractive, not very cost-effective method because you're getting less benefit at a much greater cost. Now the only way that we can get greater benefit among the other tests would be to do Cologuard every year. But you can see that, that would be extremely costly. So for a very small gain in benefit, you have to take on a huge burden in terms of cost. And so you would have to pay $200 -- nearly $215,000 for every quality-adjusted-life-years gained, which makes it a very cost-ineffective method. So really, Septin9 stands out as kind of the sweet spot for maximizing benefits and being able to do so in a cost-effective fashion. Now when we look at this analysis, you could say, "well, what were the assumptions behind the model, and can this impact the final outcomes?" And there are a number of assumptions that are made in terms of test performance and in terms of the ability to detect cancer, for example. So what was done is really a probability analysis to show, well, is the method that's shown as being most cost effective, what's the probability of that actually being true if we modify some of these inputs? So for example, for Septin9, if we weaken its adenoma detection from 23% down to 0% and we weaken its colorectal cancer detection from 70% down to 40%, what would this do to the model? And so that's what, what was done. In essence, the parameters that were input into the model handicap each of the testability to really detect precancer and cancer. So you can see that, at $0 to $50,000 willingness to pay, really CTC dominates as the strategy. However, if we wanted to detect more cancers and more adenomas, then we're willing to pay more. And so if we go somewhere between $60,000 to about $200,000, you see this green line, which is Epi proColon done every year. And you can see that Epi proColon, despite being handicapped, really dominates as the optimal strategy about 54%, 55% of the time. You can see that. This is the probability that it dominates. At $100,000, which is the threshold that was used for the paper, you can see that Cologuard done every 3 years actually has 0 probability of being cost-effective at $100,000 cutoff. That's this line here, which runs along 0. So the only strategies that are cost-effective are Cologuard done every year, CTC done every 5 years and Septin9 done yearly. And Septin9, which is the green line, clearly dominates over these 2 other methods. Now when we reach a much, much higher cost-ineffective dollar value, then you can see that if we did Cologuard every year, it would start dominating in terms of the strategy, but then again, by definition, that would be cost-ineffective. So over the widest range from $60,000 to $200,000, Septin9 is the clear winner. So in this paper, what made it really interesting is that they looked at 5 different scenarios. And in all but one of the scenarios, Septin9 came in as the winner. So the base case scenario, they model a cohort of 50-year-olds that are screened from 50 to 75, and you can see that annual Septin9 is the most effective strategy, again, among patients that are resistant to FIT and colonoscopy. If you include FIT and colonoscopy, then colonoscopy done every 3 -- 10 years is the most effective strategy. But again, we're talking about those that 1/3 of the population that is not getting screened. So not only does the blood tests have the ability to really improve adherence among these patients, but if implemented, it would lead to great clinical efficacy and very high cost effectiveness. So in the base case, again, 70 years and older, annual Septin9 is the winner. If we follow the ACS recommendations, the recent ACS recommendations where they recommend screening starting at the age of 45, again, Septin9 comes out on top. If we adjust adherence from perfect adherence down to about 95% -- 90% return rate for those that get the test the first time and 15% for those that do not get the test, and we assume only an 80% referral rate to colonoscopy, then even with an adjusted adherence or nonperfect adherence, Septin9 comes out on top. And then if we handicap Septin9 to say that 12% of adenomas are systematically missed, meaning we'll never find them, and 18% of cancers are also systematically missed, even when we do that, Septin9 still comes out on top when we handicap its performance. So these are all very telling. Now in the one model where Septin9 didn't come out on top, it actually came in #2. And it may -- it missed the threshold of $100,000 only by $19,000. That was the model that USPSTF used from CISNET, from this group in 2016, so 4 years ago. And what makes that model different is that it has a much, much lower CRC incidence that is assumed in the model. It actually use CR data from 1975 to 1979. And the idea of using that incidence data was, well, let's use the incidence of cancer back when there was absolutely no screening. And that way, we can see well, what would be the impact? Well, if you fast track 4 years later, that assumption actually has been much, much improved. And the incidence of cancer it -- has been calculated to actually be nearly 1.6x greater than it was between 1975 and 1979. So if you forward 40 years to today, that -- those are the incidence numbers that are used under all the other scenarios. So when the USPSTF meet to redo their guidelines, they will be using the revised model, which -- with the higher incidence, which will probably look very much like the base case shown here or the USPSTF has said that they're going to look at age 45. So very high likelihood that it will look, like, these first 2 models. So when you look at -- even in the lower incidence, I wanted to give you some more tidbits on that, you can see that Septin9 done every year still outperforms Cologuard and still outperforms FIT. So the incidence does not really impact the outcome that Septin9 done yearly outperforms both FIT annually and Cologuard every 3 years. What it does impact is since the incidence is lower than when you have a much -- a population where the cancer rate is much lower, then it's more difficult, obviously, for those tests to become cost-effective. Up here and leave it for any questions, and I hope this was useful.

[Operator Instructions] And the first question comes from Hogan Mullally from Encode Ideas.

Greg and Jorge, can you hear me okay?

Yes.

Yes, we can hear you, Hogan.

Well, congrats on the [ data ]. I have a 2-part question around CISNET and the MISCAN model. I'm sort of curious how the journal authors and the MISCAN model are involved with CISNET? And then as a second part question, is the MISCAN the gold standard model used by CISNET? Or do they use any other CRC models as well?

I'd be happy to answer that, and Greg can provide more insight if needed. But -- so it's a great question, Hogan. So really, CISNET, as I said, is a consortium of investigators. There are 3 predominant models. MISCAN is one of them. It stands for Microsimulation Screening Analysis. And these are really computer programs that have been developed really looking at patient demographics, understanding the presence, the risk of cancer and the prevalence of cancer in a population and then tracking the natural history of disease over time. And then once you establish those baselines, you then use the same model to understand how different strategies impact that background incidence and mortality rates associated with that cancer. So MISCAN is, I would say, one of the most broadly used. It certainly was 1 of the 3 models that fed into USPSTF. It was the sole model that fed into ACS guidelines and has been used by Medicare actually to understand cost-effectiveness of many different tests. So for example, just a few years ago, MISCAN, along with another model SimCRC, did an analysis for Medicare showing that only a 10% increase in colorectal cancer screening rates would yield somewhere between $18 billion to $30 billion in savings for Medicare over the next 5 -- 50 years. And so yes, these models are used, as I said before, by clinical societies and government entities to set policy. MISCAN is 1 of 3 as I said. The other one is SimCRC, which is run out of the University of Minnesota. MISCAN is run out of Memorial Sloan Kettering and then Erasmus University in the Netherlands. And then the other 1 is CRC-SPIN, which stands for Colorectal Cancer Simulated Population model, let me see if I can remember, for incidence and natural history of cancer, and that's by the RAND Corporation. So typically, these models have been validated against real data. They typically yield very similar outcomes. But what's interesting about the paper that we've been talking about is that they use MISCAN for the analysis. But actually, in terms of authors, it included authors from both MISCAN as well as Sim-CRC. So Amy Knudson is typically one of the lead authors. She was the lead author actually in USPSTF and ACS models for their guidelines. And she's out of Harvard and usually is very involved in the Sim-CRC models, but was an author on this paper, which included a lot of the MISCAN authors. So I think that there's a very strong correlation between the models. And in this publication, in terms of authorship, there were authors from at least 2 of the 3 models. So hopefully, that answers your question.

No, that's great. And I think you've kind of already sort of touched on my next question, which is the relation -- relationship between CISNET and CMS. Clearly, you've got a huge decision coming up on August 28 with a preliminary decision on your NCD. So I'm sort of curious about whether CMS would have been aware of these data coming? Is there a mechanism within the NCD to ensure that they have seen these data? I mean, obviously, these data are very compelling and speak to the cost-effectiveness of Epi proColon. And how do you ensure that CMS sees these data as part of the NCD?

Yes. No, very happy to answer that question. So these data were actually available back in 2019. So we've been waiting on a year for the data to be published and available publicly. And it came at a very timely, can have been more timely, as you said, prior to the NCD decision. So we made Medicare aware of this data as part of our NCD process. We submitted to them the work. It was presented as an abstract at DDW, which is a national meeting, Digestive Disease Week that's held every year among GIs and other experts. And so we made CMS aware of the data. However, obviously, it carries a tremendous amount of more weight if the data is published, particularly in such a prestigious journal as JNCI. So yes, CMS is aware of the data. They're also aware that the data has just been published. And they obviously have a copy of the final paper in their hands as part of their evaluation. And the data is obviously very favorable in those data that's focused on cost-effectiveness, and it shows that Epi proColon is a test of choice among all other tests, excluding FIT and colonoscopy. It also shows that the test is highly, highly effective in terms of its clinical performance, which a lot of people had dismissed. And finally, the broad public is becoming educated that sensitivity and specificity is just one factor in overall test performance, right? Interval is a very important factor. Adherence is a very important factor. And so we always make the analogy that if you're trying to build a very fast car, you could focus all your attention on horsepower, but there's other variables like the weight of that car. You want to make it really light that could make it a very fast car. So again, we're very happy to have confirmation really of the data presented by Harvard just 6 months ago, showing that Epi proColon is more effective than FIT annually and then Cologuard done every 3 years.

Great. One last question, then I'll turn it over to anyone else in the queue, but -- and it's a bit leading, I guess, as far as I look at this and I say, are there any boxes that you haven't checked yet with CMS? Now we've got a second microsimulation study verifying the cost-effectiveness of Epi proColon. We've got the PMA. We're in the NCCN guidelines. Are there other check marks that we're missing as far as what CMS is looking for as far as reimbursement on August 28?

Hogan, that's a great question. I'll let Greg answer that.

Yes, Hogan, listen, we believe we've checked all the boxes. We don't know what more CMS could ask for. I mean, basically, the data is very clear and very compelling that Epi proColon or Septin9 outperforms FIT and Cologuard from a reduction of incidence and mortality of disease. So if those methods are covered, then by default, Septin9 needs to be covered. So we believe we've checked every single box. And at the end of the day, this test is going to save lives. And that's ultimately what Medicare's goal is to basically provide that benefit to Medicare patients, and we feel very confident going into this proposed NCD decision in the next 2 weeks.

Yes. And if I can add, I agree with Greg, but Medicare has been very collaborative with us in this process. It's been a very long journey for us. I mean, we've been in discussion with them for 4 years around this. And clearly, the developments in the last year or so have really improved our case in case of guideline inclusion, and this type of analysis by the world experts, I think, only helps our case.

And the next question comes from Dennis Berzhanin from Pareto Securities.

This is Dennis Berzhanin from Pareto Securities. Congrats on the study results, and I really appreciate the detailed talk about the results. So thanks for that Jorge. So I wanted to follow up with the early question about the communication with CMS. Just wanted to first confirm that just having a little bit over 2 weeks before this deadline in August 28, the CMS will have enough time to incorporate. I appreciate the fact that they had the data before, but the fact that the study was already published and they would be able to incorporate that into a decision. And also would CMS look at everything the same way? I mean, in terms of the, for example, using the 100,000 quality-adjusted-life-years gained as a kind of threshold for the benefit? Or is there something -- are there different aspects that CMS might be looking at this? Appreciate.

Yes. Dennis, this is Greg. Thank you for the call. In our discussions with CMS, they have told us that they are on time to issue the NCD by the statutory time line, which is August 28. And we do believe that they would have enough time to incorporate this in. But like Jorge said, the data was presented, much of the data was presented as an abstract at DDW so the -- they were aware that this was out for publication. And now that it is in print, we are hopeful that they will include it in the draft decision. Now that said, at the end of the day, we feel that we are confident that even without this publication coming out, that Epi proColon was going to be covered because that is ultimately what the data says is that it should be covered. And we just feel this is kind of icing on the cake. And although very nice and extremely positive for us, we just feel that at the end of the day, this is really a no-brainer for CMS. And ultimately, within the next 3.5 months, we believe Epi proColon's going to be a Medicare-provided test for CRC screening.

Great. And just a follow-up on that. Just -- Jorge, I appreciate that you mentioned the argument of high false positives for the test from lower specificity, and that was mentioned, the authors did mention that in the study a little bit. And do you think that CMS could weigh this factor against you? Or are there any potential challenges that CMS could say no that you can think of?