2023-07-11 16:45:41 ET
Summary
- Castle Biosciences makes tests for patients diagnosed with certain diseases to help evaluate optimal treatment options.
- The company's lead products are its DecisionDx tests for melanoma and SCC patients.
- Castle's share price plunged when it looked as though the CMS planned to stop reimbursing for the these products.
- That decision has now been reversed, leading to a share price spike, but a lack of profitability at Castle is holding back a longer term bull run.
- New tests launched in the gastrointestinal and mental health spaces may reverse the trend of growing losses, but it will take time to find out if they are sufficiently reimbursed to drive a profit.
Investment Overview
Friendswood, Texas based Castle Biosciences ( CSTL ) joined the Nasdaq in July 2019, its IPO raising ~$64m via the issuance of 4m shares priced at $16 per share.
The company is a provider of diagnostic tests, although not necessarily to diagnose patients with a particular disease, but to "inform disease management and improve outcomes", according to Castle's business overview in its Q123 10Q submission (quarterly report). The 10Q continues:
For the diseases that our portfolio of tests cover, we believe the traditional approach to developing a treatment plan for cancers and other diseases using clinical and pathology factors alone is inadequate and can be improved by incorporating the personalized information our diagnostic and prognostic tests provide.
We currently offer five proprietary multi-analyte assays with algorithmic analysis (“MAAA”) tests for use in the dermatologic, ocular and gastroenterology fields. We also offer a proprietary pharmacogenomics (“PGx”) test to guide optimal drug treatment for patients suffering from depression, anxiety and other mental health conditions following our acquisition of AltheaDx, Inc. (“AltheaDx”) in April 2022.
Currently, our revenue is primarily generated by our DecisionDx-Melanoma risk stratification test for cutaneous melanoma, our DecisionDx-SCC risk stratification test for cutaneous squamous cell carcinoma (“SCC”), our TissueCypher risk stratification test for Barrett’s esophagus (“BE”) and our DecisionDx-UM risk stratification test for uveal melanoma (“UM”).
At the time of writing, Castle's stock trades at a value of ~$20 per share, meaning anybody who bought shares at the IPO price has realised a 25% gain on their investment to date, although shares have traded as high as $95, albeit briefly, in February 2021.
Castle stock had fallen to ~$33 per share by the end of 2022, however, partly due to its slightly underwhelming financial performance.
As we can see above, the company has generated some good revenue growth - from $22.8m, to $137m per annum within five years, but the growth has come at a cost, with operating income of $7.3m in 2019 becoming an operating loss of $(91.1m) in 2022.
Meanwhile, in March 2022, Castle opted to acquire AltheaDX, "a commercial-stage molecular diagnostics company specializing in the field of pharmacogenomics ("PGX") testing services that are focused on mental health", in a cash and stock deal worth an initial $65m, that could rise by a further $75m if certain performance milestones are achieved. The deal did not seem to go down well with investors, with Castle's stock falling from ~$45 per share prior to the deal's announcement, to a low of $16 by July 2022 - a fall of nearly 65%.
Trading was flat for the next nine months or so, and Q123 earnings were broadly positive, with revenues increasing 57% year-on-year to $42m, although the net loss increased from $(24.6m) in Q122, to $(29.2m), implying earnings per share of $(1.10). By the end of May this year, shares had recovered somewhat, trading at ~$26.
Reimbursement Woes - MAC Novitas Threatens to Withdraw Support for DecisionDx
Castle's stock hit a new low of ~$13 per share at the end of June, however, as it appeared that the Medicare Administrative Contractor (“MAC”) Novitas Solutions - which makes Local Coverage Decisions ("LCDs") about whether to provide Medicare coverage for laboratory services performed in Pennsylvania laboratories, would rescind its coverage of Castle's DecisionDx-Melanoma and DecisionDx-SCC tests.
As we can see above the vast majority of tests carried out by Castle - whose laboratory is based in Pennsylvania and therefore falls under Novitas' jurisdiction - are either DecisionDX melanoma, or DecisionDx-SCC, meaning that Novitas' ruling would have had catastrophic implications for Castle's business.
The company earns nearly all of its revenues via reimbursements, and according to its 2022 10K:
Medicare accounted for 53% and 57% of our revenue from test reports for the years ended December 31, 2022 and 2021, respectively, with an additional third-party payor accounting for 12% of our revenue from test reports for the year ended December 31, 2022;
If reimbursement were withdrawn, it is unlikely that physicians would continue to prescribe the tests, or that patients would pay the entire costs of the tests themselves. Regarding reimbursement rates, I found the following statements in Castle's Q321 10Q submission.
Firstly, regarding DecisionDx-Melanoma:
On May 17, 2019, Centers for Medicare & Medicaid Services (“CMS”) determined that DecisionDx-Melanoma meets the criteria for “new Advanced Diagnostic Laboratory Test (“ADLT”)” status.
Beginning in 2022, the rate for DecisionDx-Melanoma is set annually based upon the median private payor rate for the first half of the second preceding calendar year. For example, the rate for 2023 was set using median private payor rate data from January 1, 2021 to June 30, 2021. Our rate for 2022 was $7,193 per test and continues to be $7,193 per test for 2023.
Secondly, regarding DecisionDx-SCC:
In the first quarter of 2022, we requested that Novitas Solutions (“Novitas”) conduct a medical review of our DecisionDx-SCC test. That review was completed towards the end of that quarter. In the second quarter of 2022, following the completion of a requested medical review and pricing of our DecisionDx-SCC test by Novitas, we obtained a PLA code and began receiving reimbursement from Novitas for DecisionDx-SCC at a rate of $3,873 per test.
Seemingly, then, CMS reimbursement for Castle's two most important tests is currently set at ~$7.2k and $3.9k - were it to be withdrawn, it seems likely that Castle would lose more than half of its business overnight. It was not just Castle feeling the heat either - according to Genomeweb:
The original final LCD , posted at the beginning of June, rescinded coverage for Castle Biosciences' DecisionDx-Melanoma and DecisionDx-SCC tests; Pacific Edge Diagnostics' Cxbladder Detect, Enhanced Detect, Monitor, Enhanced Triage, and Resolve assays; Interpace Biosciences' PancraGen; Clinical Genomics' Colvera; Abbott's UroVysion fluorescence in situ hybridization test; and the University of Pittsburgh Medical Center's ThyroSeq Cancer Risk Classifier and PancreaSeq Genomic Classifier.
Novitas Changes Its Mind - Castle Stock Surges
Luckily for Castle, Pacific Edge, Interpace et al, at the end of last week Novitas reversed its decision - originally shared in its draft policy documentation - to rescind coverage of these tests.
Novitas had made it clear that tests would not be considered medically reasonable or necessary if their clinical validity had not been established and if they were not sourced by three databases for all oncology biomarker tests: ClinGen, OncoKB and National Comprehensive Cancer Network (“NCCN”).
In fact, Castle cautions about Novitas' activities in its Q321 10Q, and the company is well aware that it's reimbursement depends on it being able to show that its tests can have a clinical benefit.
According to a recent investor presentation, DecisionDx Melanoma "provides answers for two critical clinical questions", and its "test results predict a patient’s individual risk of recurrence and individual risk of sentinel lymph node positivity using two proprietary algorithms."
As we can see above, the company says it has "direct evidence" from a 634 patient study that use of its test leads to superior patient outcomes. Castle has worked with the National Cancer Institute also, analysing real-world data from patients tested at three years versus those non tested, demonstrating that those tested had better survival rates.
As soon as the Novitas decision reversal decision was announced, Castle stock began to climb, jumping from ~$13 per share, to nearly $20 per share, a >50% gain that leaves Castle more or less back where it began - trading above its IPO price, but also at a near 80% discount to its all-time high, and down >55% since reporting its acquisition of AltheaDX.
Crisis Averted - For Now - But Where Is Castle's Business, & Share Price, Headed Next?
There are two ways to look at the short-lived Novitas' crisis. One is to conclude that whatever the market feels uneasy about in relation to Castle and its business, it may not be reimbursement related, or we may have seen a bigger spike after the rescinding of reimbursement decision was reversed, back towards former highs.
The opposite conclusion could also be drawn that reimbursement coverage is having a kind of "Sword of Damocles" effect on Castle's share price, with the market believing that one day Novitas will announce its intention to rescind again, and this time stick to its decision, causing Castle's business to collapse.
A third point that should be made is that Castle's business has not exactly been outperforming even while reimbursement has been in place, which may be the biggest worry of all for Castle and its shareholders, although the company is intently focused on growing its business.
57% annual growth in revenues in Q123, and 73% in total test volume are good numbers, and guidance for FY23 of $170-$180m implies a forward price to sales ratio of ~3x, which is relatively low, although I would not necessarily expect the share price to grow in response to that number given no guidance has been provided on net income/loss, and that losses have been increasing, not shrinking, in every quarter.
It may be that melanoma and SCC markets may not be large enough to inspire the market to buy into Castle's growth story - its CEO Derek Maetzold admitted as much on the Q123 earnings call , whilst discussing a dramatic change in the company's market opportunity, for the better:
We entered 2020 with an estimated in-market US total addressable market, or TAM, of approximately $547 million.
Through thoughtful planning and executional excellence on our near- and long-term growth strategy, we have seen significant organic growth in the top line, diversified our portfolio with strategic investment and increased our estimated in-market US-only TAM to $8 billion.
Here is how that increase is broken down in Castle's investor presentation.
As we can see above, it is not so much the dermatology market that is increasing, although there is room for growth in this market, rather the addition of gastroenterology and mental health testing.
For TissueCypher in gastro, Castle says in its Q123 10Q that:
Effective January 1, 2023, the published Clinical Laboratory Fee Schedule CLFS rate for TissueCypher is $4,950, which will remain effective through December 31, 2024. This rate is based on the median private payor rates received between April 1, 2022 and August 31, 2022.
The growth in this division does look impressive, with test volumes increasing from 56, to 1,383 in one year, and already >50% the size of the SCC test volumes. Ditto IDgenetix, the mental health therapy response test, which is processed in Castle's San Diego laboratory, meaning it is not under Novitas' jurisdiction, but another MAC, Noridian. Test volume growth has been impressive, although reimbursement for the test is lower, at $1,500 per test.
In my view, Castle may well have the product suite to drive some strong revenue growth in the coming years, and the market may not have recognised it yet, preoccupied with the Novitas' reimbursement decision. My concern, however, is profitability. Will the new tests lose the company money, as the DecisionDx tests have done in almost every quarter since launch?
Castle reported SG&A costs of $47m in Q123, R&D costs of $14.4m, and cost of sales $10.2m, resulting in an operating loss of $31.5m.
Concluding Thoughts - Castle Stock May Be Heavily Discounted - But One Fundamental Question Remains
In September 2021 I covered a med tech company called iRhythm ( IRTC ) for Seeking Alpha. The company was facing a very similar issue with reimbursement for its ambulatory cardiac monitoring device Zio, with its MAC announcing out of the blue that it was going to slash reimbursement rates.
When the MAC reversed its decision iRhythm stock surged from a low of ~$45, to a high of >$160 in a matter of months, but the company has struggled to maintain the upward momentum. Why? The answer is that the business is not profitable.
I see the same issue with Castle - the company can only turn the recent spike on Novitas' rescinding its decision to rescind reimbursement, if you like, if it can show that the business it is in is actually profitable. Is it simply too expensive to complete the laboratory testing with reimbursement rates at their current level?
With the likelihood seemingly being that reimbursement is more likely to disappear than increase in value, perhaps the market is right in its current valuation of Castle's business, and was wrong two years ago when the stock price was rampant.
Granted, all new businesses run at a loss while they build up their market share and bring new products to market that require intensive marketing campaigns, although Castle has been at this for some time - DecisionDx was first approved in 2013.
Castle reported current assets of $271m as of Q123, so it can handle a few more quarter's worth of losses, but if shareholders are expecting the stock to take flight now that the reimbursement crisis is over, they may be disappointed, as the key underlying problem at Castle may not be reimbursement, but profitability.
As such, I am not optimistic about Castle's ability to recapture former share price highs of >$50, for example, in the short term. I plan to keep watching the gastrointestinal and mental health test volume numbers, and the SG&A figure for a few more quarters, and if I see revenues outgrow costs and losses narrow at a rapid enough rate, I'd hope to buy at a price <$35. If that doesn't happen, my conclusion would be that Castle's business model is failing.
For further details see:
Castle Biosciences: Reimbursement Reversal Win Masks Profitability Problem