Good afternoon. My name is Derek Maetzold. I'm here with Castle Biosciences, and to my right here is Frank Stokes, the Chief Financial Officer at the company. Thank you to Piper Sandler first for the invitation to present at this year's conference, and thank you to those of you in the audience and listening for attending. I would recommend that you look through our disclaimers and forward-looking statements, both in this live presentation and our website, accordingly. There we go. We were gonna walk through a couple of different items today. One of them is just a quick snapshot on third-quarter performance that we announced a few weeks ago, and then spend a little bit of time organizing around the test solutions that Castle provides today and, hopefully in the future, assuming positive results from research, and then finalize up with any questions you may have.
So what are we? We are a molecular diagnostics company that focuses on proprietary algorithm-driven tests, advanced diagnostic tests that hopefully have solved or will solve the needs of patients seeing clinicians for their diseases. And by doing so, we are having an impact on transforming how patients are managed and for improved outcomes. From a third-quarter perspective, I'll just point to a couple of points here. We had two core revenue drivers in the third quarter of 2025: our DecisionDx-Melanoma test, which is our initial test in the dermatology marketplace, and our TissueCypher test, which is our initial test offering in the gastroenterology marketplace. Both of those increased or combined had a revenue growth over last year of 36%, which actually matched the volume growth of 36% last year as well. So nice, nice consistent volume and nice consistent revenue growth as a result of that.
Our adjusted gross margins, that was what we focused on as an important metric, came in at 77%, compared to 82% last year. So about what we expected given the loss of our squamous cell carcinoma coverage in the second quarter of this year. The net cash provided by operations was just under $23 million, and that compares to just over $23 million a year ago. So we feel pretty good about that health of the operations. We ended the quarter with cash and cash equivalents and marketable investment securities of $287 million, which gives us a nice healthy balance sheet to operate the business to meet shareholder needs and business needs. And we had a net loss of $0.5 million compared to an adjusted EBITDA of $9.2 million.
We also announced, commensurate with earnings, that we were rolling out on a limited access basis a new test in dermatology, one that we are calling AdvanceAD-Tx, which is a test designed to guide systemic therapy selection in patients with moderate to severe atopic dermatitis who are 12 years of age or older. And I'll get into a bit more detail on that test in a second here. So there's different ways of organizing, kind of how companies or how laboratories work along a patient's journey. This is just one way of organizing it from screening tests when you're really screening either, well, by definition, screening should be asymptomatic patients, but maybe patients who have some mild symptoms, you aren't quite sure if they have a disease yet, all the way through to risk stratification, guiding therapy selection potentially, and then what we would call rescue and prevention.
We organize ourselves really around dermatology and gastroenterology. We have a legacy test for DecisionDx which is in ophthalmology, which is used for a rare cancer called uveal melanoma. That's fully penetrated. We think we test around 85%-90% of a 2,000-patient population year in, year out. So not an investment thesis for Castle. From a dermatology standpoint, we have five tests currently available for clinical use: MyPath Melanoma, which is used primarily by dermatopathologists to assist in the differential diagnosis of a biopsy that's been taken, and the pathologist isn't highly confident they can rule out melanoma or rule in melanoma. So it's used in that sort of concerning area of a biopsy. We have two tests that perform risk stratification value or prognostic value on top of current staging systems: DecisionDx-Melanoma and DecisionDx-SCC. We now have two tests that help guide therapy selection.
Our DecisionDx-SCC test, which is used in patients with high-risk cutaneous squamous cell carcinoma, was shown in 2024 in two back-to-back propensity-matched studies to be able to predict a patient's likelihood of response to adjuvant radiation therapy in those patients. What we found in both studies was that roughly 50% of the patients who are eligible for adjuvant radiation therapy, we would predict to have no clinical benefit. That's a substantial number of Medicare patients we can remove off of the adjuvant radiation therapy, therein because of no predicted benefit from that intervention. We also have in a development stage a collaboration with a Swedish company called SciBase, in which we are looking to explore or exploit their EIS technology, electrical impedance spectroscopy technology, to see if we can predict flares in patients with atopic dermatitis.
The intention here would be to take a patient who might have mild to moderate to severe atopic dermatitis. Many of them flare on a monthly or bi-monthly basis. If we can go ahead and predict the patient's flare three or four days in advance, they should be able to apply their rescue medication and hopefully avoid that flare completely, or at least have it go from a moderate to severe flare incident to hopefully a mild and non-concerning flare. That's the focus of some ongoing clinical trial work, and we look forward to updating the investors in the future. On the gastroenterology perspective, I'll jump to the risk stratification segment. We have two tests offered in that marketplace. One of them is TissueCypher, which is our leading test for Barrett's esophagus.
We also have, through the acquisition of Previse, a secondary test called Esopredict, which is based on methylation technology, and we are exploring the opportunity to provide both differential diagnostic as well as screening support by looking at a potential tether with a capsule attached to it and a sponge in the side, where you'd have a patient with symptoms in the gastroenterology marketplace who, for whatever reason, doesn't want to undergo an endoscopy, isn't sure if they're comfortable undergoing an endoscopy procedure to rule in or rule out Barrett's disease. They could, in the gastroenterologist's office, hopefully swallow this capsule with a tether attached to it, let that capsule dissolve, pull it up, and assist in the differential diagnosis of Barrett's esophagus disease, so that's our portfolio, today in the near- term that we can talk about publicly.
So we turn this sideways a little bit here and think about opportunity for the company. I guess I would have you look at the fourth column over there about potential US-only TAM and population of patients. So we believe there are roughly 130,000 patients diagnosed each year with stage one, two, or three, which means localized disease or disease-limited spread to the sentinel lymph nodes, and that represents, based upon our current ASP, roughly a $540 million opportunity in the US. We believe we closed out the third quarter of 2025 with a run rate of around 30%-31% market penetration. So quite a ways to go. But already one in three patients are benefiting from having our test applied to their clinical case. DecisionDx-SCC, again, as I mentioned earlier, focuses on patients with high-risk cutaneous squamous cell carcinoma. We estimate there's roughly 200,000 patients.
If you take our ASP when we had Medicare coverage, you're looking at roughly $820 million TAM in the US alone. If you skip down two lines to AdvanceAD-Tx, we're quite excited about this launch, even though we're executing a very controlled launch. We estimate that there are roughly 13 million patients who are age 12 and above in the US marketplace who have what would be termed moderate to severe atopic dermatitis. That's the target population. It'd be awfully nice if we had those patients benefiting from our test. I think that's a very aspirational goal, but it shows you the kind of marketplace we're walking into right now, at least in dermatology.
In gastroenterology, with our TissueCypher-based test and the backup test of Esopredict, we think there's roughly a $1 billion U.S. TAM, based upon 420,000 patients diagnosed each year with Barrett's esophagus disease. And as I mentioned earlier, our rare cancer test for uveal melanoma. So let's talk about our atopic dermatitis test. We spent the last few years running a large prospective observational study. And in the arm that we're gonna talk about here, it was looking at patients who were diagnosed with atopic dermatitis, moderate to severe, either on or considering going on systemic therapy. And by systemic therapy, we have a few patients in our data set who were actually on or starting methotrexate or cyclosporine off-label. We were predominantly interested, though, in patients who were actually receiving a Th2-targeted biologic therapy or a JAK inhibitor therapy as part of their systemic therapy.
We ended up discovering and developing and then validating a 487 gene expression profile test that links these genes to 12 known inflammatory or skin barrier pathways. And with that, we were able to validate two results. We find people who either have what we would call a Th2 molecular profile phenotype, which we believe is a patient whose symptoms are largely driven by a Th2 pathway, which is the most common or all patients should have a Th2 pathway component. And then we have a second result, which is called a JAK inhibitor responder profile test or phenotypic pathway. These are patients who we believe have not only a Th2 component driving their disease but also have other pathways driving their disease such that they seem to be a super responder to JAK inhibitors as opposed to a Th2-only therapy.
I'm gonna go through that data in just a minute here. When we developed our study, we were thinking through what would be the clinical application of how this is used in practice. And one of the concerns we had was that, while it might be cleaner and neater and more likely to succeed if we required patients to have a biopsy of their atopic dermatitis lesion, that's not clinical practice today. Today, most atopic dermatitis patients are diagnosed clinically only by the eyes of the dermatologist, and patient symptoms or patient signs being communicated. Introducing a requirement for a biopsy to be performed seemed to be a workflow change that would have to have a very, very high level of burden in order to be adopted widely in the dermatology field.
So we therefore spent some time a few years ago thinking about how do we get a reasonable amount of living skin from the atopic dermatitis lesion that we believe will be able to preserve RNA to enable a gene expression profile type assay to be done. And we developed a non-invasive scraping technique that turns out is very robust across our 49 centers in the US who participated in our large prospective clinical study, with very, very little failure rates occurring on a per physician or per site basis. So no biopsy is required for this non-invasive sample to be collected. And as I said earlier, we have a JAK inhibitor responder profile and a Th2 molecular profile. Went through this before. So let's talk about the benefits of what we saw out of our identity study, which was this large prospective observational study.
The first thing to orient you to is to look at sort of when do we know or believe physicians and patients are making decisions around, "Am I getting a good response from this therapy or not?" And it turns out that's around three months or so of starting a new therapy when people are making a call of saying, "Am I satisfied with this new therapy I'm starting? Am I dissatisfied?" So we chose 12 weeks of follow-up as our clinical cutoff endpoint. We also thought, from a drug approval perspective, the atopic dermatitis drugs are generally approved based upon the ability of a drug versus placebo to achieve something called EASI-75.
EASI-75, EASI is a score for rating eczema that is used by therapeutic companies to go ahead and set a target at 75 would equate to 75% clearance of that patient's atopic dermatitis. If their atopic dermatitis had a score of X, you would have to see a benefit of X of 75%x or 20%x to go ahead and get to your efficacy level. We thought setting a target at EASI-75 doesn't really add much value. You can get that just by reading a product label or trial and error.
Instead, we, I guess, circled around our current clinicians who use our skin cancer test and said, "What would be an altruistic goal to get to?" and what came back was, "If you can have a test that will give us an identity of a patient who might achieve EASI-90, meaning 90% clearance, then we have a useful test clinically," 'cause that's a really satisfied patient who has seen their atopic dermatitis go from moderate to severe to quite mild, or in some cases here, complete clearance, so we set that as our target, and what we were able to go and demonstrate is that if a patient has a JAK inhibitor responder profile and they're placed on a JAK inhibitor compared to if they were placed on a Th2 biologic therapy, what's that differential look like?
And so on the first box there, what you can see here is that nearly one in two or 45% of patients who are started on a JAK inhibitor who have a JAK inhibitor responder profile achieve EASI-90 and also achieve IGA of 0, which is complete clearance, all occurring on the starting dose of the JAK inhibitor, not the escalated dose. In comparison, if those same patients who had a JAK inhibitor responder profile were started on a Th2 biologic therapy, you have 8.3% likelihood of achieving EASI-90 or one in 16, roughly one in 14 patients. Those are two far different patients. A one in two chance of being a super responder on a JAK inhibitor or a one in 16 chance of being a super responder on Th2 biologic. The rest of the scores across the way follow along the same pattern here.
The vIGA is the validated Investigator Global Assessment score system that is used in atopic dermatitis. 0 means complete clearance. Again, 36% versus 0% achievement of 0 clearance on Th2 targeted therapies. Did the patients remain flare-free? Again, about half the patients in the JAK inhibitor treated group with the JAK inhibitor responder profile were flare-free during the course of the study. At the same kind of comparison state. And importantly, from a sort of patient benefit standpoint, we found that the patients achieved an EASI-90 score roughly 3.8x faster if they were on a JAK inhibitor versus on a Th2 targeted therapy. So great results when you can present these to a dermatologist and say, "Are these actionable?" The answer is yes, it is.
So we did quite a bit of market research over the course of the summertime to support how we should discuss this with our clinicians, how we can make sure we don't overpromise our data sets. And the slide up here I'm showing you is essentially indicating that the likelihood to use the test in practice, this is after about a five-minute overview of the test profile, by the way, and what we see here is roughly 78% of the clinicians who we interviewed in this market research study indicated that they would definitely or probably use our test after just a five-minute discussion around what the product profile looks like. That is a fantastically accepting response from customers.
Now, it turns out that while there are some dermatologists who focus predominantly on atopic dermatitis, KOLs, academic individuals who can focus on just inflammatory skin disease, for example, the vast majority of patients with this disease condition are treated by medical dermatologists. It turns out that medical dermatologists, by and large, in the community practice settings, also treat skin cancers. So we anticipate that many of this, of the clinicians who will be beginning to adopt this test for clinical use are the same 7,000, 8,000 clinicians who have adopted our skin cancer test for clinical use. So no change in call pattern. It's a medical dermatology call point, and they're seeing, we think, of course, roughly 13 million patients a year with moderate to severe atopic dermatitis in patients 12 years and older compared to 130,000 patients with melanoma as examples.
So a very, very nice walk-in opportunity. We're done here? Okay. There you go. So with that, time is out. Thank you for your attention.