All right, great. We're gonna go ahead and get started. I'm Catherine Schulte. I cover life sciences and diagnostics here at Baird. Very excited to have the Castle team here with us today. From the company, we have the CEO, Derek Maetzold, and CFO, Frank Stokes. In terms of the agenda, Derek's gonna run through some slides, and then we'll go into Q&A. If you have a question, you can send it to sessionthree@rwbaird.com, and I will pass it along to the team. Derek?
Thank you, Catherine, and, thank you, Baird, for the, conference. Nice to be back here in person. I'm gonna go through just, part of our corporate presentation deck, just to make sure that we're all on the rough same page before we start Q&A, if that's okay. Disclaimer slide, we encourage you to look at our corporate presentation on the website for a full review of this slide in detail. W hat are we? We are a molecular diagnostic company that predominantly invests in and commercializes innovative, either first-in-class or best-in-class tests, and we focus on these four areas of disease or specialty counts. From a performance standpoint, we released our second quarter earnings in early August, and a couple of top-line comments to note here.
One of them is that we, again, showed tremendous growth year-over-year in total test reports that were issued or delivered, as well as total dermatology test reports, which is the core part of our business. Adjusted gross adjusted revenues came in at just over $50 million, up from $34 and change in second quarter of 2022. So again, very strong report growth and adjusted revenue growth on those top lines, and we did this with using minimal cash use. We had operating cash flow of just negative $3.8 million in the second quarter of this year, compared to $9 million of cash use last year. So ended the quarter, we think, in a very, very strong health position.
Commensurate with that, we also raised guidance from $170 million to $180 million, which was our original guidance, to indicate that we'll be coming in north of $180 million for the full year 2023. We have enjoyed over the last several years, since we became public in 2019, both before and afterwards, good, strong, consistent performance, both in terms of test report volume as well as in terms of total revenues. As I said earlier, we're expecting to be on track here for coming in north of $180 million based upon our performance in the first and second quarter of this year. Let's skip through a couple of slides here.
Just to kind of orient the audience again here, our predominant revenue drivers is really in the dermatology business. We have tests that focus primarily across the whole portfolio, really as risk stratification or therapy response test. What's that mean? Well, in the case of cancer management, as well as I think almost every human disease I'm aware of, one of the decisions one has to make as a patient and a physician upon a diagnosis of a new disease, or in our case, primarily a new cancer, is what's the what treatment pathway does it do I take my patient down? T hat usually is based upon coming to the best and most accurate area of prognostication or risk stratification. What am I facing?
If I face a low likelihood of my disease, or in the case of cancer, that it actually spreads and metastasizes, then patients simply go on a very, very low, I wouldn't call it watch and wait, but really clinical follow-up, not being overly concerned that that cancer will metastasize 'cause it is low risk, but actually just looking for overall health. In contrast, patients who have a cancer diagnosis and a higher risk of progressing, whether that's on clinical pathologic factors or in our case, based upon our test results, then those patients are typically escalated into a different care pathway.
That might include interventions, it might include the use of imaging routinely to pick up disease spread early, and as you can see here, most of our business focuses on the sort of risk stratifications, early decisions in the immediate post-diagnostic period of a cancer or disease state, where you're trying to make choices about which pathway of care to put a patient on.
The other area of work that we have right now is we do have a pharmacogenomic test in which we call IDgenetix, which combines in a single report, both drug-gene interactions, drug-drug interactions, and lifestyle factors. T hat a clinician, again, being faced with a patient who's been diagnosed with one of several mental health diagnoses, can use our test to go ahead and help direct or guide patients to a therapy choice that they'll will most likely have the best response to, as well as avoid those drug therapies that they're likely not to have an adequate response to.
F inally, up above there, we have an inflammatory skin disease pipeline program that is ongoing. We are looking to develop a test or a series of tests that will hopefully help a patient who's going from topical therapy for either atopic dermatitis or psoriasis, to systemic therapy, and help that patient and their clinician evaluate, understand, "I've got choices of what I can do with my disease of atopic dermatitis. Does drug A gonna have a great response or a poor response compared to drug B or drug C?" So that's our goal there.
We did announce earlier this year that we intend to release data prior to year-end to go ahead and give people a flavor of where this program is sitting and what we think we're gonna come out with it at the end of the day, s o that's the background on Castle from sort of a disease state standpoint, where we play. I'll just take one or two slides to dive into a couple of areas that I think might clarify questions later on, too. I f I go to our DecisionDx-Melanoma test, this is a test which really has two clinical uses to dermatologists, surgeons, and their patients in the early diagnosis of melanoma.
The two uses are, one, do I consider performing what is called a sentinel lymph node biopsy surgical procedure to assist in s eeing if this localized melanoma might have spread regionally to my lymph nodes. Our test is used predominantly to find patients who could be eligible under clinical or pathologic factors, but actually have such a low risk or a low likelihood of having any melanoma in their lymph nodes, that you could avoid that unnecessary procedure. The second use of our test is really looking here at the individual risk of recurrence.
Beyond that first choice of do we consider doing a additional surgical procedure or not, is: What are you facing as a patient? Are you in that low-risk train, where chances are just having clinical follow-up is the right choice for you, or are you on a high-risk train, such that maybe we should get you into medical oncology, or at least routinely image you, looking to pick up metastatic disease early?
A ll the current therapies that are used in melanoma, particularly the PD-1 inhibitors, we know that the best patient response, the best hope to actually have a patient even perhaps getting cure from PD-1 inhibitor therapy in melanoma, is if they get on when the metastatic spread is the smallest, lowest tumor burden. And that happens when you look at it and pick it up based upon imaging, as opposed to patient symptoms. I mention this point here because it's worth understanding, I think, the next piece of data. We had two very nice, strong publications come out in the second quarter of this year, one that we anticipated and one that was a nice surprise.
I'll go over those over in a couple of slides here. T he one we did anticipate, there was a study published by Dhillon et al. from three academic institutions in the US: Cleveland Clinic, Northwestern Chicago, and Oregon Health & Science University. All three of these centers had adopted DecisionDx-Melanoma on an individual physician basis for clinical care in their patient population, and at these same institutions, some other doctors had not adopted our test either, s o you had this intra-institutional control group. Patients walked in the same door, were in the same clinics. The only difference that we could see was that in the population that were sentinel lymph node negative, some patients had their subsequent care guided by our melanoma test results, and some patients did not, in the untested control group.
The important takeaway here is that at the end of the study, which involved 634 patients from these three institutions-
What we found was that, in fact, patients who had their care escalated, so they went on routine imaging protocols for high-risk disease spread, in fact, had their metastatic disease spread picked up earlier, about 10 months earlier than those that waited for symptomatic metastasis to occur. The tumor burden was smaller, and at the end of the study, 76% of the patients who had metastasized and had their imaging guided by the Castle Biosciences DecisionDx-Melanoma tests were alive, compared to only 50% of those patients who did not have clinical care changed with our test results. That is a very nice outcome to show from a, "Does this test make a difference in outcomes?" And I think the answer is yes.
Now, that being said, this is a nice precursor to the next study, which is essentially an ongoing collaboration that we have with the National Cancer Institute's SEER Registries Program. What is this? SEER is the group which is mandated to basically track tumor registries on about 48-49% of the U.S. population today. And this tumor registry, it gets reported up in the states that are in part of the SEER program, for all patients who are diagnosed with cancer, and tracks not only baseline characteristics, but also outcomes like survival, both disease-specific survival, or in our case, melanoma-specific survival, and also overall death or overall survival. Now, this study was published earlier this year, I think in maybe June first or June thirtieth, second quarter. And what's important, I think, is the top-line number here.
We saw that just like in the previous study from Cleveland Clinic, Northwestern, Oregon Health & Science University, that patients who had their clinical care guided, I guess you would say, or underwent testing with DecisionDx-Melanoma, had a 29% improvement in terms of three-year metastatic-specific survival, 17% overall survival, death from all causes. Put those two sentences together, you have back-to-back studies with different methodology from patient groups, showing you that the clinical use of our test is now linked directly to actually improvement in outcomes, not just improvement in change the management, which we think will lead to outcomes. Now, why is this relevant and important to us? I think this slide here is a busy slide.
We look down to the bottom rows. What you see here is that, actually, Genomic Health undertook a similar collaboration several years ago with the NCI's SEER program, and they had their three-year breast cancer-specific survival data analyzed, had an absolute mortality difference at the three-year mark of 0.5%, compared to 1.3 for us. So we showed, in the case of melanoma, in our test, nearly a twofold improved benefit in absolute survival change, compared to what Oncotype breast does for women diagnosed with breast cancer.
That's a nice piece of data to have, given the fact that you see today, probably the breast cancer test out there, 90-95% patient penetration, year in, year out, so very widely used. You see, widespread coverage taken care of, both from a Medicare standpoint and commercial standpoint. I believe this kind of data is what will hopefully lead us to where we expect melanoma to get to at the end of the day. The other one piece of data I'll talk about briefly is our DecisionDx-SCC test.
T his is also a gene expression profile test that we developed and validated at Castle. Its use is really to, its initial use was similar to DecisionDx-Melanoma, which is to say, in patients diagnosed with high-risk squamous cell carcinoma of the skin, we can provide further risk stratification. W e can help take existing pathways and help patients be comfortable in their doctors, that they actually can stay or go on to a low-risk clinical follow-up pathway only, o r perhaps they are a patient who has such a high biological risk, in addition to being risky, clinically and pathologically speaking, that they should seek interventions like, for instance, adjuvant radiation therapy, which is used and reimbursed today for patients with high-risk squamous cell carcinoma, a nd it's also recommended in all relevant society guidelines.
Now, I mention this because we recently completed and presented data here showing that our test not only risk stratifies patients, but if you take patients who are eligible for radiation therapy in our clinical studies and compare them to ones who received radiation therapy, and then look and see, does our gene expression profile test also help maybe pull out people who might have a response to radiation therapy? The answer is yes.
What you see up here is the cumulative probability of metastasis for a Class 1, Class 2 A, or Class 2 B test result. What's impressive here is that out of these 920 patients, which by the way, is the single largest study of its type in radiation therapy and SCC, that if you happen to have a Class 2 B test result and you did not receive radiation, you had roughly a 60% chance of having a metastatic event at over the 5-year period post-diagnosis, whereas that was reduced to less than 10% if you actually had received radiation therapy.
That's a great understanding of saying Class 2B test result from Castle, you're already eligible for ART. If we don't encourage you to take it, you have a 60+% chance of metastasizing, versus less than 10% if you go on ART. On the other end of the equation, ART-eligible patients who happen to have a Class 1 test result, we see almost no discernible benefit from adjuvant radiation therapy, which is important because most squamous cells of the skin are on the head and neck region. You can typically only initiate full-dose radiation in a location of the body one time.
I f you use it on a patient who might be high risk because they've got a large squamous cell on the top of their scalp, you've taken away that intervention for subsequent basal cell carcinomas, for subsequent brain tumors, for example. It's just off the table. One, can't use it, and two, maybe more important health economically, is that we've calculated the average cost of a course of therapy for squamous cell to be roughly $61,000 per patient. So you're talking about a large number of Medicare population who is eligible for ART by virtue of being high risk, clinical and pathologically. We can find people in Class one, which is about half the population, who actually will see no benefit.
That changes the kind of conversation a patient and their doctor have, I think, in terms of, "Do I do it now, or do I wait and watch?" W ith that, I'll stop, Catherine, and hope that wasn't too long of a discussion to get us going.
No, that was a great overview.
Mm-hmm.
Maybe just starting on DecisionDx-Melanoma, kind of your lead product within dermatology, had over 20% growth there in the second quarter. W here are you in terms of penetration today, from a market penetration standpoint, and how do you view that unfolding over the next couple of years?
Frank?
Yeah. O n a physician basis, we calculate. We use about 12,000 eligible clinicians to use that test, and I think last year we had in the past twelve months, we've had about 7,200-7,400 have used one of our derm tests. So on a physician basis, maybe we're 40-50% penetrated. On a patient basis, we think we're about 25% penetrated on a run rate, sort of second quarter run rate. If you look at other modalities, as Derek mentioned, we think about 90% of breast cancer patients are tested at this point. Veracyte thinks that about 60% of thyroid patients are tested. It looks to us like about 25%, maybe 30% of prostate eligible prostate patients are tested.
W e've got at least, you know, we think we should be at least where to, to where thyroid is, at least that much penetration, so the clinical utility is similar, avoidance of surgeries. There's actually some sort of benefits for our melanoma tests that are even more compelling, so conservatively, we ought to be able to at least get to that point, which would be a, what? 2.5 times our current volume, if we can meet that.
A lot of opportunity, even just from penetrating existing accounts more deeply. I f you look at order rates from earlier physician cohorts, how do you drive that increased penetration within an existing account versus, you know, increased penetration from the rest of the ordering base?
W e certainly have a number of current customers or clinicians, where from what we can see, based upon the pathology reports that follow our orders, they're likely ordering it or have adopted a use for almost all of their patients in their practice that they diagnose with, you know, with melanoma. We have other clinicians where we can see, based upon the kinds of patients we're not having them order the test on, that they sort of have focused on saying, "I get the best value of the Castle test in this kind of a patient." M aybe it's the ones that are around the edges of thinking about some of the biopsy procedure, s o one of our publication strategies and messaging strategies to those clinicians is to say, "Hey, looks to me like this is the sweet spot that you've self-defined.
T hat's a good choice and a good selection. Here's the data, why that is. Can we talk now about patients over here or patients over here, and why you've kind of come to a decision that maybe you aren't seeing value here, and can I share you with... share, share that data with you? O ne of them is to, I think, both provide us with, with data and publications that support the clinical value in the, in the non-sweet spot patients. I t's data like the Dhillon paper and like the NCI paper, which says, "Hey, all things being equal-"... across the board, if you use our tests, and you use those results to change how you would manage a patient long term, they live longer.
I f your goal of being a doctor is improve outcomes, I have a hard time seeing why you should play God about not testing patient A, but testing patient B because of the historical perspective. T hat kind of data will help drive us forward in terms of not only getting more appropriate patients tested within a current customer's practice, but also going to those 30, 40, 50% who we have not gotten to properly and saying, "This is some data here that's very hard for you to ignore." Two back-to-back publications showing that patients who receive our test, and as part of their clinical care, actually live longer. Now, what's the objection that we have left here?
Because that seems to me to be pretty compelling data about why you wouldn't want this information for yourself, for example, if you have a diagnosis of melanoma.
Yeah. I f we think about NCCN guidelines, you know, important for adoption, important for reimbursement. To your point with the Dhillon study and then the NCI SEER study, you now have... It's not just analytical validity, it's outcomes data, it's improving overall survival. Were both of those studies published in time so that they can be considered in this round of NCCN and any feedback that you've heard from NCCN around those studies?
Yes, they were published in time, and we made sure they saw it with our submission. The NCCN panel typically meets in mid-July for melanoma, and they normally update kind of their outcome from that meeting in the kind of December, January time period. M y hope would be that if we look at some of the critiques they have around the data that they have reviewed, which they haven't reviewed a lot of the data, but especially thinking about, we'd like to see this in a real-world, large population of prospectively tested patients. That's the NCI SEER publication right there. So I would hope that would have an impact in terms of improving the positioning of our test within the melanoma guidelines at this next December publication date.
Then maybe on squamous cell, can you just provide a latest update on reimbursement there? Do you think the study that you showed on the adjuvant radiation therapy can have a sway with Medicare, NCCN? Just curious, the path forward there from the data lens.
Oh, there's a lot of questions to unpack there. I guess, first of all, as maybe the audience is aware here, we received ADLT status on June thirtieth of this year. ADLT status stands for Advanced Diagnostic Laboratory Test, which means it underwent a separate review at Central Medicare or Central CMS by the Division of New Technology, and they rendered it as providing novel information not available through any other test or combination of tests, and it's a currently covered Medicare service. So as we indicated at our August fifth or sixth or fourth earnings call, we continue to be a covered service on the DecisionDx-SCC for the Medicare population.
In terms of these proposed LCDs going through their comment period and post-comment period, the Novitas comment period closed up last Saturday, the ninth. Palmetto's was earlier in August, I believe, if not late July or mid-July. If we look just back at historical trends, normally we see the MoIDX program takes roughly the full 11 or 12 months. There's a 12-month expiration date, and they usually kind of come in that time period. If we look at Novitas in the last year, that's when they also came up with their last proposed LCD. That, of course, was a pullback from being final.
I don't know if that's the right target to think about there, but probably by the end of the second quarter next year, we should see clarity in all those, both of those things. Now, to go to the more exciting question here, we believe there are a number of judgments done by both of those Medicare contractors that we hope our submissions will help them see a bit where maybe they were interpreting data in a slightly different manner and maybe missing the current treatment pathway for squamous cell carcinoma, that would change their approach in thinking about our test.
I do agree that this adjuvant radiation therapy data, we had done some rough calculations after purchasing data from IQVIA, and in the 12 months ending, I think it was May or June of 2022, so the mid of 2021 to the middle of 2022, we estimate there were roughly 29,000 Medicare patients who received adjuvant radiation therapy. If you just overlay the average cost of radiation therapy, that's about 1.8 billion dollars just in that population for ART. If you had physicians just use our test to rule out ART in Class 1 patients, we think that's roughly a $900 million savings, not to mention reducing complications, saving a therapy.
I'd hope all those things put together would help somebody who critically reviewed our response and then the data out there to go and impact that positively.
Yeah. And Frank, you had reaffirmed your 2025 revenue target, even excluding any squamous cell Medicare reimbursement. I guess, you know, what drove your, your confidence to do that, and how should we think about potential upside if, in fact, by that time you, you do have squamous cell reimbursement?
Yes. Excuse me. So we did say that we were not assuming squamous reimbursement for the from the period on when we updated guidance. So there is some upside there as we continue to be a covered service for squamous cell. I think the biggest driver has been the outperformance of volumes for all of our tests. When we set our original guidance, which was February of this year, since that period of time, each one of our tests has exceeded what we had modeled and internally predicted in terms of volume. So all of our portfolio of tests are performing extremely well, and we don't see that changing through the end of the year.
Yeah, and you've been involved in some tuck-in M&A over the last several years, getting into GI and mental health. I guess, how have those two assets performed relative to expectations, and what's your appetite for M&A going forward?
I don't think that M&A is certainly not the top of our capital allocation priorities. I think that we have plenty of opportunity with the portfolio we have. Each of the deals we did, we had compelling reasons to do them, and they have—those reasons have played out. Our thesis has been correct, and we're excited about the opportunity for TissueCypher. That's been, as I said, that outperformed what we had hoped it could do this year. To the extent that we needed to take a minute and optimize some volume capacity in our lab, which is a great problem to have, by the way. Just reinforcing and validating how compelling that test is.
On IDgenetix, the market is just tremendous, and we have a differentiated test. So what you see is, with each of our acquisitions, we, we entered larger market segments, larger patient opportunities. Melanoma at 130,000 patients, squamous cell, we think about 200,000 eligible patients, MyPath, about 300,000. We think there are about 450,000 surveillance endoscopies for Barrett's esophagus, and then pick your number on, on, mental health conditions. Unfortunately, it's a, it's a tremendously large number. So, each of those opportunities brought us bigger market opportunities, bigger, bigger addressable markets, and, I think we're just early days and seeing, seeing those opportunities and seeing what, what can be, what those tests can do.
I think that, you know, we look this time next year and the year after that, and they'll be viewed as less of tuck-in and much more strategic and compelling.
Yeah. We've got three questions that we're asking every one of our companies that's here, so I'll ask you both to-
Have they all answered?
They've all answered.
Okay, I guess we got to be quick.
I'll get both of your inputs on these three. So as we think about the next 12-18 months, what do you view as the two biggest opportunities for your company?
For Castle? I think that we'll continue to. All of our tests, so let's go ahead and take our growth test. Melanoma, we think, is around 20-25% penetrated. The other ones are all single digits or low single digits. I think the opportunity to continue to perform and outperform on volume is clear across the board. One, two, is that I think we've been very efficient at cash use, at cash use, and one of the expectations we have for ourselves, and so for the marketplace, is that we exit 2025 in an operating cash flow positivity financial framework. I think that's on track with DecisionDx-SCC, without DecisionDx-SCC. So that to me says that's very comfortable.
Now, the sort of behind door number two, I think, would be: When's the launch timing for our inflammatory skin disease test? When do we see that coming to marketplace? Is that gonna be late 2025 or earlier in 2025? M aybe some of the data we disclose and see later this year will help guide that decision-making. That could be a very exciting new test for us to launch. Reimbursement will, of course, lag, but it could hopefully solidify in everybody's minds what we should look like in 2029, 2030 from a company standpoint, from a financial perspective. T hen finally, I think what we're early days right now is we have seen a number of states adopt what I would call biomarker state laws.
For our little part of the world of diagnostics, what I see happening in the next year, year and a half, would be, do those biomarker laws, which are focused on commercial plans that are those commercial plans now stepping up to the plate and complying with what those statutes require? Which is to say that if you have well-validated tests that are meaningful, either Medicare coverage and guidelines, etc., that just because you happen to be diagnosed with, say, melanoma at age 64.5, we're not gonna pay for your test, but at age 65 and one day, we are gonna pay for your test, that disparity of care based solely on age goes away.
I think for the whole industry and for Castle, too, if those biomarker laws are implemented in a positive fashion and a compliant fashion, that creates great ASP upside for Castle as well as the rest of the players in the space.
Yeah. T hen all of-
Frank, anything to add or?
No, that's good.
On flip-
I'm gonna get the bad one right in the back.
Yeah, here you go. On flip side, two biggest challenges you see over the next 12-18 months.
T he biggest challenge we will see is the continued obstinacy of the commercial payer landscape. Commercial payers don't make objective decisions on whether to cover services or not. They don't make the best decisions for their covered lives. They make the best decisions for their MLR, and that's not gonna change. I think... So that's one. So what's another challenge? I think at some point we will see legitimate meaningful competition for many of our tests. Right now, we only have, you know, really organized competition for IDgenetix and but I think the areas we are in are so compelling, and we've demonstrated that the markets are so attractive.
T he great news of our volume exceeding our expectations is it's also exceeding the expectations of people that might wanna compete with us. So I do think that over the next year or two years we will begin to see the proliferation or more impact from some competition, and we've been prepared for that for years, and that's part of our strategy, and our data and evidence development is making it harder and harder to compete with us.
All right, and the last one would be: What is a question that analysts and/or investors don't ask you very often, but you wish that they would?
We get asked a lot of questions. I do think that we certainly wish that we could spend more time discussing with analysts and investors the power of some of the data we've generated, and less time on some of the noise around, you know, things like LCD processes and things like that. Because wherever those LCD processes go and come out, the way to make them come out the right way is with compelling data. And so ever since Derek founded this company, that's been one of the key pillars of what we do every day, is building compelling support for our test. Why? For physician adoption, for payer persuasion, and to make it even harder to compete with us.
W e don't get to spend enough time, I think, talking about the power of that data.
I would say certainly in the last year or so, there's been an inordinate amount of focus on reimbursement unknowingness, and that's okay. I appreciate that from a sort of short-term standpoint. I think what that's resulted in having us less time to do is to really say: Why did we start Castle in the first place? It was to really, at the end of the day, leave an impact on human health that was positive. If we did that, we would expect doctors to order our test, people to pay for them, and we have a nice business at the end of the day, not the other way around.
W e've lost an opportunity in the last year of some of these reimbursement challenges or perceived challenges of saying, "Hey, what are we here for?" This is a very, very nice, stable business. I think the acquisitions of both Cernostics and AltheaDx were done to make sure we had stronger pillars, so that if we did face reimbursement challenges, like everybody's gonna face, that becomes less material of an issue and more about it's a noise around the edges, as opposed to a death knell for the business. T hat's the part we've gotten away from as a company in the last year or so, responding to very, very important questions, by the way.
I t doesn't necessarily tell what the business is about. Tells you what the business may not be about.
Yeah. All right, great. Well, with that, we're out of time.
Thank you.
Derek, Frank, thanks so much, and thanks, everyone, for joining.
Thank you.