The rain doesn't help today.
All right, we're going to go ahead and get started. I'm Catherine Schulte. I cover Life Sciences and Diagnostics here at Baird. We're very excited to have Castle Biosciences with us today. From the company, we have the CFO, Frank Stokes, and from IR, Camilla Zuckero. Frank, Camilla, thanks for joining us.
Thank you, Catherine. Good morning. A quick reminder to review our forward-looking statement disclosures in our public filings. Good morning, Cath. Castle Biosciences is a diagnostic services company serving several market areas with a portfolio of tests, all intended to guide patient care and inform patient outcomes. Our tests are across a portfolio of diagnostic support tests, which are MyPath Melanoma tests. This is a genomic test that can identify melanoma or rule out melanoma in an indeterminate biopsy sample for suspicious lesions, suspicious pigmented lesions. We have a new program we'll talk a little bit about in a collaboration with SciBase to identify atopic flares ahead of the flare occurring to help the patient hopefully minimize that flare and get back to baseline more quickly. The bulk of our tests is in risk stratification or prognostics. We have our DecisionDx-Melanoma and SCC tests.
These tests are for two skin cancers, both of which are potentially fatal if they metastasize, and our tests provide a risk score to physicians on the likelihood of those patients metastasizing or not. We have a small test in ophthalmology, a similar prognostic in uveal melanoma. In our gastroenterology franchise, the bulk of that is our TissueCypher test for Barrett’s esophagus. The TissueCypher test is a spatial protein test that identifies the risk and scores the risk of a patient with Barrett’s esophagus progressing to esophageal cancer. Barrett’s is the only precursor to EC, and EC has about an 85% five-year mortality, which is almost always avoidable with ablation. TissueCypher provides a risk score that identifies which patients ought to be ablated. We have a recently acquired test called Esopredict. We obtained that through the acquisition of a company called Previse last summer.
Esopredict has a similar clinical utility to TissueCypher. We also have a pipeline test for atopic dermatitis. This is a test we've indicated we'd expect to launch by the end of this year. This is a therapeutic selection test, intended to be a therapeutic selection test for atopic derm. So far through the year, several key results. We've had continued growth in our portfolio across our therapeutic areas, continued strong gross margin. On an adjusted basis, we have one of the strongest gross margins in our sector. We generated cash nicely through the year and ended the year at about $275 million of cash, which is, we think, a comfortable position to be in to allow us to continue to run the business and potentially pursue other opportunities like the Previse acquisition last summer.
The SciBase collaboration, we'll talk a bit more about, but that was an investment in a public Swedish company that has a device and technology for atopic derm flare. Also, in the quarter, I think both happened in the quarter, we received a notice that both of our requests for reconsideration for coverage of our squamous cell carcinoma test had been accepted by the Medicare contractors, one by Novitas and the other by the MolDX program at Palmetto. Our capital allocation priorities continue to be most focused on commercial growth. We've expanded our TissueCypher team earlier this year, and that is approaching what we think is close to what we need to target that physician community. We think there are 12,000- 14,000 targetable clinicians there. I think at the end of the quarter, we were at about 65 area managers for TissueCypher. We continue to be aggressive on R&D.
Our R&D strategy is not only focused on our pipeline, but also on our existing portfolio of tests. That R&D is focused on both converting physicians who potentially aren't using our tests to date, continuing to build the moat against competition, and also driving reimbursement decisions with payers. We also have, as I mentioned, strategic opportunities. SciBase is a great example, and Previse is another great example. With Previse, we were able to get a very meaningful pipeline of potential tests in GI that would be very complementary to our TissueCypher test, which has grown very nicely since we launched it in 2022. We will skip that. Happy to take questions from you or the audience.
All right, great. If the audience does have questions, you can send them to session3@rbbaird.com. Frank, just starting first on your cutaneous melanoma business, can you just kind of set the stage for where we are from a market penetration standpoint today, maybe how you see that evolving over the next several years?
Yeah, sure. We think that on a run-rate basis, we are testing 30%, 31% of the patients that are suitable for melanoma testing. We use a number of about 130,000 melanoma diagnoses a year. SEER data shows about 100,000. There are four pretty strong independent publications that demonstrate that melanoma is severely underreported in the community. That's generally because melanoma is almost always diagnosed in a community derm setting. Many of those practices either aren't accustomed to reporting SEER data or are unaware of reporting SEER data or other factors. Those four publications suggest that melanoma is underreported by anywhere from 30%- 75%. We take the low end of that. We know SEER is low. We know it's not fully representative of the reality. We just take sort of a more conservative approach. We use about 130,000 melanomas a year.
On a physician basis, we think 55%, 60% of the eligible clinicians have used one of our derm tests this past year. The usage there can be a little different. Melanoma, a community derm in, let's say, somewhere close by Scarsdale may see four or five melanomas a year. It's fairly episodic. Doctors are not seeing melanomas in their practice every week. Good penetration there. The analogs for penetration range from breast cancer testing. I think the data there suggests maybe 85% or 90% of breast cancer patients who are appropriate for testing are getting one of the tests, Oncotype Breast from Exact, MammaPrint from Agendia is one. There's a couple others. Our uveal melanoma test, we estimate we test about 85% of those patients a year. That may be the high end of penetration. Maybe a more conservative approach would be to look at something like thyroid.
I think the companies that are in that space suggest maybe two-thirds of patients are tested, eligible patients, appropriate patients are tested for thyroid. In prostate, I think those companies suggest maybe 45%, mid-40s to high 40s of those patients are tested. We certainly see nothing clinically or clinical utility factors that would suggest melanoma couldn't be at that same level of prostate or thyroid. Still plenty of room to grow there. We certainly continue to have new physicians who are becoming aware of the test or beginning to consider the test. It's been on the market for quite a while. Internally, we sort of think, boy, that's odd that that physician is not familiar with our test. There certainly are no C practices or physicians who don't welcome diagnostic providers in for marketing meetings, etc .
We continue to slowly tip over those physicians that aren't using the test as well.
Great. For this year, you've talked about high single-digit volume growth for that test. Is that the right way to think about it long term as well?
I think so. We certainly, early days, had much higher growth off a much smaller base. We continue to drive new patient penetration. It's a much more mature market now. Really, by definition, the easiest doctors to convert are the ones that convert first. By definition, the ones that haven't converted yet are more difficult to convert. You see the pace of physician conversion slowing down, although we continue to have significant new ordering physicians each year. That's good to see. I keep waiting to see that playing out. We continue to grow the physician base. I think that's the right way to think about growth this year and next probably as well.
You've now had a few quarters below that high single-digit pace. Can you maybe talk to what you expect to drive the acceleration in the back half?
Sure. All of our tests, we should see seasonality. That seasonality is driven, we believe, by physician encounter days. It's not just business days in a quarter. For example, Q1 has a large number of business days, but dermatologists are avid enthusiasts of their two big medical meetings, which happen in the first quarter, one in Maui every year. I think the other alternates Maui to the Caribbean. The physicians quite see value in those meetings. That takes them out of their practice for a good bit of a week or more. Things like that drive physician encounters with patients. Melanoma has been very predictable. Really, the biggest step up sequentially is Q1 to Q2. It's driven by the largest number of physician encounter days in Q2. You don't have the meetings that we see in Q1. You don't have summer holidays.
There's also a big behavioral change that happens in the second quarter. Everyone goes from people broadly go from wearing more clothing to less. You go from hoodies and skiwear to tank tops and shorts as the weather warms. You're more likely to see a lesion that concerns you or one of your family members is likely to see it. That has a big driver in our melanoma volume. What we have less history in is, with the reimbursement dynamic around squamous cell carcinoma, our derm sales force is now focused exclusively on our melanoma test. We don't have great metrics for how to measure what we expect that to do. The last time we had only one dermatology test for our sales force to sell, I think we had 32 reps. Today we've got circa 70. We don't have a history to help us gauge that.
We would expect some wind at the back as we see those conversations. I wouldn't expect that the conversations around squamous cell are going to stop. The two tests in derm, our melanoma test, our squamous cell test, although they're for very different patient populations, they really are substantially the same clinical question. In the case of melanoma, it's what's the likelihood if you refer to a sentinel lymph node biopsy surgery that there's going to be a positive cell in a lymph node? The second question is risk of recurrence. Our squamous cell test is very similar. It's the appropriateness of referring the patient to adjuvant radiation therapy as a first order of clinical decision. The second is risk of recurrence.
These two tests, and one of the reasons that we've been so excited about them and we think we've seen such great success across that portfolio, is it really is one clinical question. It gives you the ability to go into a physician and say, we've got this utility for both groups of your patients. You don't have to learn a new technology. There's not a new clinical use. It's the same use. You've got a procedure you may or may not refer, and you've got risk of recurrence considerations that drive how you treat that patient. It's the same clinical question. I don't expect that our area managers would walk in and if asked about squamous cell say, no, I can't talk about that. We're only going to talk about melanoma. We expect there to continue to be dialogue and education around both tests.
The motivation and the direction of our reps is really driven toward melanoma while we wait for the reimbursement situation to clarify.
Yeah. Maybe on the reimbursement situation to clarify, it sounds like both of your reconsideration requests have been accepted by MolDX and Novitas. What are the next steps there? Maybe talk us through, as you think about the areas of pushback that they gave on their initial decisions, what's the incremental data and/or messaging from you that you think could change that, whether it's ART or you just put out some NCC and high-risk data? What do you think could move the needle there?
Sure. We did receive notice of acceptance. That's largely an administrative, I think what we said before is we would expect that. We would expect to have them both be received for reconsideration. That's a largely administrative milestone. I wouldn't wait too much on that. In the case of MolDX, in the original negative policy, there's language at the end of it that references data that was available but was not considered in the policy. The program integrity manual says that the contractor doesn't have to consider data that comes after the end of the comment period. The data that was referenced came after the end of it. It's clear that although they could have, as we understand it, it was clear that they elected not to consider that.
I would suspect that's largely, I'm supposing, but I would suspect that's largely just as a precedent, just to make sure that the process is maintained as a disciplined process. We've had significant publications since that period of time, including the two largest publications around SCC, risk of recurrence, and ART. We certainly expect that those are persuasive and have sufficient evidence to justify coverage as reasonably necessary. One of the key publications we had suggests that, and I think it's 2022 data that the study is based on, but it suggests that if Medicare had used our SCC test to appropriately guide adjuvant radiation therapy for all Medicare patients diagnosed with high-risk SCC, after the cost of our test, Medicare would have saved $1 billion. We think that, you know, it should be things like people who don't need adjuvant radiation therapy aren't getting it. That should be compelling.
If you had a parent who's 85 years old and they got a squame up here, how do you feel about sending them for adjuvant radiation therapy above the shoulders? The side effects there are going to be almost certainly going to have side effects that are notable. You've shot your bullet. If they come back in two years with a basal cell or something like that, you can't use radiation. It's expensive. That should be the determiner. We think the powerful economic data is also a big piece of the support. On the Novitas side, their policy was less focused on data. Clearly, there was data that was not considered. In summary, we've got a substantial body of data that's been developed since the close of the comment period and the drafting of the Novitas LCD, we certainly think is persuasive. That's our opinion.
Our opinion is the third most important opinion in that consideration. That's where it stands.
Yeah. Maybe as we move on to the non-derm side of your portfolio, I just wanted to talk about bigger picture strategy. Several years ago, you were very much a dermatology-focused company. You've made a few acquisitions outside of that area. TissueCypher has performed very well in the almost four years that you've owned that asset. I think it's going to be over a third of revenue going forward. As we think about your strategy going forward, what does Castle look like five years from now?
I think that the overriding characteristic of that would be our approach to diagnostic testing, development, licensing, acquisition, support, marketing has always been around identifying real clinical questions that physicians don't have an adequate answer for and providing an answer. We don't have a proprietary technology. Our derm tests are run on QuantStudio PCR technology from Thermo. We don't have a particular technology that we're trying to apply to different areas. The risk of that, of course, is when you have a hammer, everything looks like a nail. It's a bit of a sloppy analogy. Our approach has always been identifying a nail and then going to figure out how do you beat that into the wood. Our melanoma test was a great example. We never intended to have a cutaneous melanoma test. We had our uveal melanoma test that we had actually licensed.
Many of the physicians who treat uveal melanoma or who are in that area had knowledge or interlocks with cutaneous melanoma. They and our R&D team said, you know, it would make sense that there might be similar markers. It's a melanoma just in a different spot. I wonder if that's right. If it is right, who cares? We heard from physicians that 80% of cutaneous melanomas are diagnosed as stage one or two, low risk, 80%. The physician does typically a wide local excision, shows the patient up and says, you did the right thing. You saw something. You came in here early. We got it. Stage one or two, great. Just come back next year for your well-care skin check. We're worried about new melanomas. This one we got, it's OK. The data suggests that 66% of melanoma metastases come from that 80%.
That's because just population-wise, yes, the 80% are low risk. That's where the numbers drive 66% of your cases. There's this discordance between what a population-based analysis means and what the patient-specific analysis is. The company at the time before I was there said, you know, if we could identify an answer that re-bucketed these patients, doctor, would that be valuable to you? Overwhelmingly, dermatologists said that would be valuable because most derms, I used to say the majority, but I'll say many, have in their mind a patient that they had who had a low-risk melanoma. They did exactly what I described. Eighteen months later, they find out the patient is coughing up blood and they've got lung mets. They've now got a very serious cancer. Melanoma metastasizes to the lung and the brain. It's hard to treat, hard to rescue patients when they get to that point.
That led the company to say, I bet we can develop a test that might answer that question clinically. Sure enough, we were able to. That's the profile of the company: identifying a clinical need and deciding if we can provide an answer to it. Obviously, our focus, our priorities are in the therapeutic areas we're in. We have tremendous reputation, positive reputation in derm and GI. Other things that we can do in those spaces are extremely interesting to us. With Previse, I think it's going to be a very exciting part of the acquisition. We obtained a technology around a cell collection device. There are other devices that are similar, but it's a capsule with a brush inside. The patient swallows a capsule, the capsule dissolves, the brush is exposed, you pull it out, and you get tissue the length of the esophagus.
We believe, and there's early data at JHU from that acquisition, that suggests you may be able to diagnose Barrett's with the tissue that you retrieve. Why do we care? Everybody does an upper endoscopy and pinch biopsy. What's the need? That's the question we asked ourselves. What's the clinical need? The clinical need is that data suggests that somewhere between 33% and maybe as much as 45% of Barrett's patients are missed by surveillance endoscopy with pinch biopsy. The reason is because of technique, mostly technique, but also inconsistent cell placement in the esophagus. The technique is frequently because a full Seattle protocol takes 50 minutes. That's a big burden on a practice. Many times, the pinch biopsy is only done in irritated tissue, and maybe there's some dysplastic tissue somewhere else that's missed. That's one element.
It turns out not many people want to go in for an upper endoscopic surveillance biopsy if you just have risk factors, but you may well, in fact, develop Barrett's. There's another population of patients that could be tested with the cell collection device, which would, by the way, expand the number of appropriately diagnosed Barrett's patients, which would expand the opportunity for TissueCypher to help patients who currently are undiagnosed. That's exciting to us. There's a great opportunity there. There are other disease states we think that the cell collection device could be used for as well. That was a big priority for us because it was in our core GI space. It would make us more valuable to those gastroenterologists. It doesn't necessarily have to be. Really, the criteria for us would be less around is it the same doctor?
In derm, for example, we're typically not marketing MyPath to the same physicians as Melanoma and SCC. Yes, it's in derm, but it's a different group of people. We're less focused on that. I think the priorities will stay in those therapeutic areas.
How should we think about for derm versus non-derm, maybe margin profiles for each of those business and path to profitability?
Sure. Spatial omics is a little lower gross margin. Our lab and R&D folks have done a great job in increasing the efficiency of that test. When we acquired the test, the analogy I use, if you've ever been to a county fair and you did the spoon race with the egg, you got the egg on the spoon and you got to get it to the other side before the guy. It almost felt like the process was that manual and there were samples kind of manually being analyzed and carried around because it was a scientific-based development company at the time. We've done a lot to automate that. It is a bit lower gross profit. The gene expression profile testing is much more efficient. As you know, even with those lower margins, our blended adjusted gross margin is still very attractive.
It still gives us plenty of economics down the P&L for sales and marketing, for R&D, for EBITDA. That has been our focus as well. It's an important part of, and that's been a deliberate strategy. Our labs are in Phoenix and Pittsburgh, lower cost areas. We've been able to leverage that by not having our labs in Manhattan or something like that.
You talked about the sales force expansion on the GI side that you did earlier this year. Would you say those reps are fully productive now? What are some of the KPIs you look at for when you need an expansion? What do you think the long-term right size of that sales force is?
Sure. That expansion, I think, was completed by early to mid-Q1, maybe. We've always said maybe six months for a rep to get up to full productivity. Your question is, what does that mean? That means physician encounters per day, typically. That's how we sort of measure effectiveness. In terms of saturation for an Area Manager, we try to have a balance of about half the time the rep is working to convert new physicians who aren't yet using our test and about half the time caring and supporting and serving physicians who are ordering. When a territory gets too large in terms of not geography, but in terms of volume, then what you find is the rep is spending way too much time taking care of existing docs, which is important. They need to do that.
That means they don't have capacity to go to the physicians who aren't using the test yet and educate them and provide them information on the benefits of testing for their patients. When we see a territory get to that point, or maybe there's two territories that are that way, that's a good opportunity to take two and make it three or take one and make it two. The pin action on that is really good. Anything you do to reduce geography, of course, is important because I remember early days when we were first starting with melanoma, and I looked at the map and we had reps with four or five states. The tendency is you just don't have, you're burning a lot of time just getting from one place to the other. That has a negative impact on productivity, as you would expect.
When you tighten up territories, it makes the reps efficient. It also drives effort because if you've got a big territory and this doctor says no or isn't available or is not persuaded, you just go to the next guy. There's somebody over here. There's somebody over here. When you start whittling those down, it drives people to work harder on maybe that initial resistance. I got to get through that initial resistance. I got to work a little bit harder there. When you start whittling those down, you get tremendous pin action while also taking care of the physicians that you've converted. Of course, the art there, and the commercial team does a great job of it, is that handoff, how you handle that handoff. What you don't want is a physician saying, hey, you're my third Castle guy this year. What's going on?
You want to make sure that the physician has that touchpoint with not only the rep, but our Medical Science Liaisons, our clinical support team in the labs. That's a very delicate, and by the way, that's why our TissueCypher expansions didn't go from 14 to 65. You could look at it and say, and you have people did. They say, hey, it looks like you're probably going to need more than 14. It looks like this is similar to derm. It looks like you got to be closer to that. In a business school case, you just said, OK, 14 goes to 65, but you risk a lot of those relationships and you risk a fumble. It's harder to hire correctly that many people. It's been a measured expansion and I think an effective one.
Yeah. Growth for that test has continued to be incredibly strong, north of 90% in the second quarter. Obviously, that number will come down as the test matures. I guess how should we think about the growth outlook over the next 12 to 18 months? Any reason that we should see an absolute unit volume growth slow in that period?
Yeah. So you're right. The growth has been very, very attractive. A couple of things have driven that. When we acquired that test, the data that we had, the best data we had, suggested there are 280,000 surveillance endoscopies a year in this country. That's our patient pool. Now we have much better data that suggests it's north of 400,000. The patient pool opportunity is larger than we thought. The physician response has been as good as we had somewhere between hoped and expected. It's as good as we'd expected. I think at some point, like melanoma, we should see some seasonality in TissueCypher. I don't know what it would be, but it just makes sense that eventually those patient encounter days begin to be a factor. We're earlier. We've got more experience with melanoma. We're much better able, I think, to predict that.
TissueCypher is still a little bit of early days of a kind of a rocket ship ride that we're still working through. I think we'll continue to see volume growth. I think it'll continue to be very strong. I don't know. At some point, I expect we'll see something that says maybe it's seasonality, maybe it's reaching some kind of penetration level. For now, we certainly are pleased with what we're seeing. We certainly don't see substantial pushback from physicians on using the test. It's similar to melanoma. A lot of gastroenterologists, if you just think about it, most patients who have a biopsy are non-dysplastic or low-grade dysplasia.
That's where most of the esophageal cancers come from, because if the patient's high grade, they're going to get ablated, and the progression is going to stop because it's not 100%, but it's 99 point something percent ablations stop progression of Barrett’s to esophageal cancer. Those higher risk patients are being ablated. It's that low-risk population where the cancers are coming out of. It's a similar need to melanoma. A lot of those physicians have a patient who came in non-grade dysplasia, and they said, yeah, I mean, I guess come back in 15, 18 months. We'll do another endoscopy and just check again, and then they find out that they've progressed. The population answer is low risk for many of those patients, but on a personal level, it's not necessarily congruent.
All right. With that, we are out of time.