In my recent article, The Case For Celgene (NASDAQ:CELG), almost all the emphasis was on its operations, which are throwing off large amounts of free cash flow, with only moderate amounts of long-term debt relative to the fast-growing profit stream.
Here I try to bring order and updated information to the far-flung and confusing pipeline that Celgene has created. The two most important parts of this article include first, its very organization, which lays out a way of thinking about CELG's pipeline, and the CELMoDs, which I see as having a large present value that however cannot really be estimated quite yet, given their early stage in development.
As a way of thinking about the CELG pipeline, first, I want to go back to a description of the old Alza Corporation, a drug delivery pioneer that J&J (NYSE:JNJ) acquired in 2001 for $10.5 B in stock. In the late 1980s, Alza was a very high P/E public company. A brokerage report described it as an onion, with one layer of value and then, underlying that layer, another. The analogy, which proved accurate in this case, applies to CELG.
More mathematically, please consider this old paradox, which may also apply to CELG, principally for its CELMoDs but also to its partnering strategy. It's a way of thinking mathematically about a possible though rare outcome to provide a fair present value. In this little story, imagine a game of chance, such as a simple coin-flipping game, played by Jack and Jill. Jill flips a fair coin with a 50-50 chance of heads or tails coming up each time. On the first throw, if Jack guesses the outcome correctly, she pays him $2. She then tosses the coin again; if Jack guesses correctly, she pays him $4. This continues until Jack guesses wrongly one time.
The question is, what is a fair amount of money for Jack to pay Jill in advance to play this game? (In thinking about the answer, Nassim Taleb fans should please ignore Fat Tony; this is a pure math question.)
The answer is that the fair value of this game is infinite. This is, in theory, a winner for Jack no matter how much he pays Jill. The reason is that this is an infinite series. The first coin toss has a fair value of 50% of $2, equaling $1. The second toss has a fair value of 25% of $4, also equaling $1. This never ends. The fair value is $1 + 1 + 1 + 1... to infinity.
That's because of the incessant doublings of course. Keeping the payoffs constant at $2 would give a fair value of $1 + 0.50 + 0.25 + 0.125... ending with a value of $2, not infinity.
The many venture capital-type investment CELG has made and the CELMoD platform technologies may together comprise a nearly infinite series with slow decay. The theoretical fair value is likely not infinite, but it might be the sum of $1 + 0.98 + 0.96... or something like that with a slow decay that yields a high, very high or maybe extremely high fair value that Mr. Market is not appreciating.
Only time, and lots of it, will tell the truth in the case of CELG.
I'll begin with an analysis of the "onion" of the "layers" of CELG's pipeline.
Introduction to thinking about CELG's pipeline
There is no comparable company to CELG in the way it has organized its R&D efforts. At first, it appears disorganized and off-putting. But I now see three layers or types of its R&D. The most important swing factor, CELMoDs, are the least well-known.
I think of the company's layers three ways; there is some overlap between these classifications, but it's the best I've been able to come up with. Farthest from the core or center of the CELG "onion" are the partnered enterprises, discussed next. In the middle layer are the in-licensed drugs, such as ozanimod, discussed after that. Then there are the internally-developed products, of which the commercially most promising are the CELMoDs; these are discussed subsequently.
Possibly the best-known aspect of CELG's product development involves its extensive partnering. I reviewed this in some detail in a 2-part series last July. The first article was titled Celgene's R&D, Part 1: Background And The Juno Story. The second, some of which I update extensively in today's article, was titled Celgene's R&D, Part 2 - Revlimid, CELMoDs, And Avila.
The three partnerships that CELG has promoted most extensively are, first, Juno (NASDAQ:JUNO), in which company CELG purchased $1 B worth of stock in mid-2015 at $93/share. JUNO was trading around $50 at the time and is currently below $20. (I never liked the deal because of the cost and said so contemporaneously.) The second and third most emphasized by CELG over the past two years are probably the deals with Agios (NASDAQ:AGIO) and Acceleron (NASDAQ:XLRN).
Each relationship is different, but in general, the companies co-develop pipeline products with CELG and the junior biotechs having some shared marketing rights, and CELG takes an equity stake in the company.
Even collectively, all the many relationships of this nature that CELG maintains may have more of a general technology benefit than identifiable financial value relative to CELG's large market cap. Just look at the small market caps of those three stocks:
- AGIO - $1.8 B at a price of $42
- JUNO - $2.1 B at a price of $18.9
- XLRN - $0.9 B at a price of $24.
While all these collaborations are interesting, and some that involve privately-held companies may provide key technology to CELG, as a whole, their current value as a part of the whole of the value of CELG has to be small and thus need not be discussed further. See pp. 108-9 of the 2015 10-K, filed in February 2016, for further details on the extensive arrangements CELG has with these and many other companies. An independent review of many of the alliances and other deals that CELG has entered into the past few years, including the Receptos deal discussed below, will provide readers with a sense of just how active CELG has been.
What's all this equity in small, money-losing junior biotechs worth? Again, only time will tell. Often, there's no way even to theorize the per-company value, because there is no publicly traded stock. As far as the above three stocks go, a cursory look at their charts shows great volatility. And the prices can volatilize upward and just keep going up, as well as down or just meandering horizontally.
I monitor AGIO, JUNO and XLRN daily, and assume there is a single-digit number of billions of dollars' worth of value from these partnerships collectively, between equity stakes and future profits. But it's overkill to think too deeply as a CELG investor, given the role of chance and the need for lots of time - and investment by CELG and the companies - to create actual value in the form of profits.
Moving closer to the core of the pipeline...
Purchased R&D - why it matters; why the risk-adjusted prospects may be favorable but why the market has figured all that out already
If the market already knows what you and I know, then there is no special reason to learn more than the basics about certain aspects of a complicated company's business. That's how I think about the several high profile in-licensed drugs, which certainly "should" have a risk-adjusted, time value of money-adjusted value at least equal to CELG's investment in them, which exceeds $10 B. But there's no guarantee of that.
The most important of these is ozanimod, which stems from the acquisition of Receptos in 2015. Ozanimod, as the last four letters of the compound's name suggest, is an immune modulator. CELG purchased Receptos for $7.2 B net of cash acquired, but other expenses that CELG assumed pushed the actual cost to $7.7 B (per the 10-K, p.74 and the press release accompanying the deal). Add in interest due on the debt, loss of interest income on the cash expended on the deal, and the cost of clinical trials, and the cash invested already likely exceeds $8 B.
Ozanimod is in two pivotal Phase 3 trials for multiple sclerosis, a crowded field. Its mechanism of action is similar to that of the blockbuster Gilenya, a Novartis (NYSE:NVS) drug. It is hoped and expected that a cardiovascular observation period required when Gilenya is initiated will not be needed for ozanimod. Whether this will be sufficient to allow ozanimod to be a commercial success in the face of generic Gilenya (fingolimod) is unclear, though.
Gilenya has established a multi-year history of safe long term use with significant reduction in deterioration of the clinical and brain imaging status of MS patients, and the Phase 3 studies for ozanimod are not comparative against Gilenya. We do know that NVS, despite Gilenya going generic before this decade ends, did not purchase Receptos. Thus the commercial prospects for this drug for MS are unclear in my mind, and I believe that it is also unclear as to whether CELG will market the drug for MS or out-license it to another company. Biogen (NASDAQ:BIIB), the world's largest MS company, bid on Receptos, but CELG outbid it.
Then BIIB took what amounted to an option on a competing development-stage drug with a similar mechanism of action, anselimod, but ended up not going forward with development, leaving the CELG drug in better position.
So, the MS indication for ozanimod has a wide range of possibilities for CELG.
CELG also has this drug in Phase 3 for ulcerative colitis and Phase 2 for Crohn's disease. These are the two major inflammatory bowel diseases, where there is significant unmet need for an effective oral therapy. Top line data from both trials may be ready for public release in Q3-4 of 2018. If so, profits from these indications would not be realized until 2020 or later, as few drugs are profitable the quarter they are introduced. There's lots of promise for ozanimod in these indications, but they are highly competitive.
Ozanimod is highly material to CELG; but as an investor, I have trouble believing that I can out-think Mr. Market on its prospects.
CELG has done another well-studied deal in the same therapeutic category with Nogra, a private Irish company. Here is how CELG discussed the deal for mongersen after it was struck, from April 2014:
Celgene Corporation entered into a global license agreement with Nogra Pharma Limited... to develop and commercialize GED-0301 [mongersen], an oral antisense DNA oligonucleotide targeting Smad7 mRNA for the treatment of moderate-to-severe Crohn's disease and other indications...
Under the terms of the license agreement, Nogra Pharma Limited will receive an upfront payment of $710 million, regulatory, development and net sales milestone payments and tiered royalties. Aggregate payments for regulatory and development milestones could potentially be $815 million for multiple indications. Starting from global annual net sales levels of $500 million, aggregate tiered sales milestones could total a maximum of $1,050 million if annual net sales reach $4,000 million.
This is also an expensive deal for CELG, but I do like it better than the ozanimod deal and am hopeful there's a strong ROIC in it for CELG. Assuming that this drug is judged safe and effective by the FDA, CELG may have spent around $2 B on mongersen, broken down this way:
- $710 MM upfront
- $815 MM milestones through approval
- significant monies on interest costs, foregone interest on cash spent on the deal, and clinical trial costs.
Then there are the future milestone payments on sales to subtract from revenues assuming success in development.
And then there are "tiered royalties" on sales. Looking through the 10-K and elsewhere, I find no details on how large they are. 10%? So this is a deal that investors should in my view think about comprehensively in building financial models. It's a great deal for Nogra, which has locked in big returns from one Phase 2 success. Nogra was so far ahead of the field in developing mongersen that after the Phase 2 results were announced, JNJ signed a deal with Ionis (NASDAQ:IONS) for IONS to develop a competitor using the same mechanism of action for Crohn's disease.
There are more such products, especially durvalumab, which can be classified in more than one way, not exactly purchased as with ozanimod; this and related topics are discussed after discussion of the CELMoDs and another important internally-generated product, CC-486, which is discussed next.
So let's look at why I'm bullish on what CELG is now revealing about its internally-generated pipeline. I'll begin with a brief discussion of the follow-on product to Vidaza, then move to the larger class of CELMoDs, about which CELG is disclosing much more nowadays than it used to do.
Oral Vidaza could be superior to Vidaza
On Dec. 4, CELG made a presentation and then held a Q&A at the American Society of Hematology, or ASH, meeting. Most of what was presented and then discussed is highly technical and not closely tied to the stock price. (Slide 72 shows the ever-changing list of collaborators with CELG's R&D.) One of the most commercially relevant details that came from the remarks of the CELG personnel related to a next-generation version of Vidaza, which is given by injection. Vidaza, chemically azacitidine, is approved by the FDA to treat a variant of anemia and a rare type of leukemia. Sales were large but have been shrinking rapidly as the drug has gone generic in the US.
CELG has developed the pipeline compound CC-486, which it has been calling oral azacitidine. CC-486 is in development both for anemia and a variety of malignancies, per the CELG pipeline web page.
No certainties, but CELG insisted that perhaps the set of presentations about which they were most excited at ASH suggested that CC-486 was not just an oral equivalent of Vidaza, but was superior at least to some degree for at least one indication. Given that Vidaza had substantial peak sales and that CC-486 is being tested for more indications than Vidaza achieved, I think this could be a billion-dollar seller, perhaps reaching the $2 B level or higher at peak.
Some caveats to that. In one of the CELG presentations, it keeps its estimate for peak CC-486 sales lower than I do, and I don't think it's prudent to go far beyond what CELG is willing to say publicly. Also, of course it's always risky when writing about development-stage assets. The author's words cannot be unwritten once published, and the future may prove very different than hypothesized in January 2017.
Now, the main event is one I have followed for almost two years, and which CELG hopes leads it to greater glory in the 2020s and beyond.
Why CELMoDs have incalculable but large present value
In researching this article, when I did a Google search for CELMoDs, the first three links Google showed me were my own Seeking Alpha articles on them, two from 2015 and one from last year.
This shows that not much attention has been paid to what could be an explosive group of profit drivers. This is one of the reasons that I'm hopeful that alpha can be generated by CELMoDs.
CELG took a chance on the IMiDs, which are one specific type of CELMoDs, in the 1990s, when its senior people took a meeting to look at a completely different target and were directed to look into the anti-cancer (and other) effects of thalidomide, a drug that was banned for causing great damage to the fetus. Yet thalidomide, later commercialized as Thalomid, proved effective for leprosy and myeloma.
The chemically related IMiD lenalidomide proved superior for myeloma, and was commercialized after Thalomid for myeloma and two other indications. With the research that led to the development of Revlimid and then the more potent anti-myeloma agent pomalidomide (Pomalyst), CELG became a leader in exploring the protein system that its drugs was affecting.
This has led to an understanding of how Revlimid and Pomalyst work, which in turn has led to rational drug design and testing of the broad class of drugs called CELMoDs. These work on the same or similar proteins as do Revlimid and Pomalyst (which are called IMiDs or immunomodulatory drugs), but can affect more diseases.
CELG laid out the evolving case for CELMoDs on an R&D day in September with what it called a "Deep Dive" into protein homeostasis. It has persuaded me that the CELMoDs in the clinic, plus the at least two in preclinical development, plus presumably more to come, may (no guarantees) more than replace Revlimid and provide the best possible growth drivers for the company.
Here's the rationale.
How the IMiDs led to CELMoDs
CELG has found that by binding to a protein called cereblon (the "C" in CELMoD), lenalidomide (Revlimid) leads to a chain reaction within the cell that ends in the breakdown of one or two toxic proteins. This action causes several favorable anti-cancer actions and is most pronounced in myeloma.
CELG has found that this system, which involves a protein such as cereblon, a tagger called ubiquitin, and small intracellular structures called proteosomes, may be usefully enhanced by other agents, such as the three CELMoDs that CELG has in the clinic. In the Deep Dive presentation, the company goes into great detail publicly - for the first time that I'm aware of - in laying out the case for how these drugs work and why they may be moving fairly quickly to a regulatory filing.
This is rational drug discovery at its best. It requires not one but two discoveries. First, the company must identify a disease characterized by an excess of a protein that can be degraded via a proteosome. Second, it must find a drug that leads to the degradation of that protein (or proteins) in a way that is safe for the patient and represents an improvement in treatment of the disease over existing drugs.
So: IMiDs can be considered as a subset of the broader class of CELMoDs. In the Deep Dive presentation, while it's slightly obscure (in typical geeky CELG fashion), slide 136 provides some help for how the IMiD or CELMoD joins with a protein such as cereblon and a substrate ("S" = a protein) to lead to breakdown of different proteins, which is, so CELG believes, often an important therapeutic goal.
Slide 11 shows the range of possible diseases that the broad class of CELMoDs could address. These include:
- blood cancers (types of sarcoma)
- carcinomas (most cancers)
- lupus, other autoimmune forms of arthritis
- neurodegenerative diseases: Alzheimer's, Parkinson's, Huntington's; ALS
- genetic diseases: cystic fibrosis, Gaucher's disease.
Likely CELG has done some work on all the above to specify the above targets.
That neuro is on the field explains one or more recent CELG R&D deals.
And, clearly, while cancer is a very large field, it's intensely competitive, with breakthroughs coming frequently, unpredictably, and from all parts of the globe. Whereas, if CELG is indeed in the lead amongst drug developers with novel, often orally-administered, targeted drugs that will diminish intracellular levels of proteins where that is a desirable therapeutic goal, CELG will have an important competitive advantage that it should extend to any and all diseases.
Thus - the plan for CELG shareholders is that the first three CELMoDs, briefly discussed next, are just a start. No guarantees, but I believe that's the plan at the company.
CELMoDs - current clinical trials
CELG has three CELMoDs in early and mid-stage trials that it has hinted might be good enough for an NDA to be filed with the FDA, such as for a provisional approval pending Phase 3 study. It calls these by its Celgene symbols, not compound names yet: CC-122, CC-90009, and CC-220.
In the Deep Dive and at ASH, CELG focused on this drug for a form of lymphoma, diffuse large B-cell lymphoma.
Based on the data provided, I'm optimistic that assuming it's well-tolerated, this could be an important drug for DLBCL. The company is hopeful that current ongoing studies of CC-122 could lead to FDA registration for the DLBCL indication (Deep Dive presentation, pdf page 102).
DLBCL is an important form of non-Hodgkin lymphoma, about which the National Cancer Institute provides summary information.
Because this compound has many other potential uses, it's worth going back to comments made on its Q3 conference call (p. 12):
Michael Pehl - Celgene Corp.
And we think that there is a very distinct need for better therapies in CLL, and we think also that CELMoDs are an excellent drug class for that purpose. When you're thinking of patients who are relapsing out of ibrutinib [Imbruvica], when you think about the really low CR [complete response] rate and when you're thinking about the very high number and proportion of MLD-positive patients with CLL, there is a distinct need for better therapies. We are very excited in that context about CC-122. With think it has a really differentiated immune profile, and CC-122 is going to be the drug to be moved forward. [As opposed to Revlimid, which was the topic of the question.]
Thus, this compound may have significant value in CLL.
Then, whether it is '122,' '220' below or a CELMoD that is getting ready to enter clinical trials, CELG is expressing optimism that for nonresponders to Revlimid combination therapy, one or more CELMoDs could do very well. Also from the Q&A, p. 16, from Michael Pehl on that topic:
But even when you look at the best available data - and I was trying to say this - 25% to 30% of patients, even with REVLIMID-dexamethasone- daratumumab, are relapsing after 6 to 12 months. There is extremely limited options if a patient is relapsing after daratumumab or after pomalidomide. And the overall survival of patients with high risk such as genetic features, despite all these new available therapies, is still only three or four years. That's very bad. That's absolutely not good enough. We can define these patient populations very clearly. We can define them clinically. We can define them molecularly, and this is exactly where we want to test our new CELMoDs in, and we feel extremely good about that.
This is under study for acute myeloid leukemia. One of the secondary endpoints involves efficacy, and the ending date for the Phase 1 study is estimated to be August 2019. Whether strong clinical data and good safety could allow marketing approval for relapsed AML is unclear.
NCI's Info about AML may be of interest to readers.
The main focus now for this compound is systemic lupus erythematosus, or more simply lupus or SLE. This is in Phase 2. CC-220 is also in Phase 1 for myeloma. The sarcoid indication has been dropped.
In the Deep Dive presentation, pages 121-2 summarize a successful Phase 2a study, which will now proceed to a Phase 2b proof of concept study using a low dose and a high dose (also see p. 119 for a schematic of the study's progress).
As an editorial aside, it's refreshing to see CELG take the encouraging Phase 2a data and not jump to Phase 3 with it. There are numerous benefits to performing a careful 2b study. If the results are super-strong, the FDA can even count it as a pivotal study.
I believe that CELG expects two molecules to enter clinical trials by 2018. That's not a guarantee, and in any case, entering human trials is a milestone, with neither safety nor efficacy assured. Nonetheless, I anticipate a stream of CELMoDs to enter human testing in the future, and for this class of drugs, with their unique mechanism of action, to be very interesting and perhaps medically very important on a case-by-case basis.
Other in-house pipeline products
While some of these may have come by acquisition, CELG lists other "CC" drugs in the pipeline. These include:
- CC-90001: fibrosis - Phase 1
- CC-90002: anti-CD47 antibody; myelodysplastic syndrome, AML and solid tumors - Ph 1
- CC-90011: LSD1 inhibitor; solid tumors - Ph 1.
I do not know enough about these candidates to opine, but it's good to see a large contingent of them. Since they are all in Phase 1, probably not much information has been disseminated to analysts about them.
Before concluding, a word on the durvalumab line items in the pipeline list and the issues of the free-spending 2015 period.
The durvalumab deal and the lingering problems from 2015
The AstraZeneca (NYSE:AZN) late-stage immuno-oncology drug durvalumab, a latecomer to the I/O party in the PD-1/PD-L1 space, was the subject of an important deal that CELG struck in Q2 2015 under the leadership of Robert Hugin, now the former CEO. Durvalumab is listed in the greatest number of therapeutic categories in the CELG pipeline web page, namely under 5 separate diseases:
It's thus of some importance to discuss briefly, first on its own merits and then to provide some final context to how I think about the extensive CELG R&D efforts, before making concluding remarks about CELG's stock in view of all this information.
The deal's terms are important, as I criticized them from the standpoint of a CELG shareholder soon after this press release described the arrangement:
Under the terms of the agreement, Celgene will collaborate with AstraZeneca to develop the anti-PD-L1 antibody MEDI4736 in hematology and make an upfront payment of $450 million. Celgene will lead clinical development across all new clinical trials within the collaboration and be responsible for all costs associated with these trials until December 31, 2016, after which it is responsible for 75% of these costs. Celgene will also be responsible for the global commercialization of approved MEDI4736 indications in hematology, and will receive royalty rates starting at 70 percent of worldwide sales from all uses in hematology. Royalty rates will decrease gradually to 50 percent over a period of four years after the first date of commercial sales.
I thought this was overly favorable to AZN, especially as the big income from a drug usually comes more than 4 years after the drug is introduced. CELG paid a large upfront fee, will bear most of the costs of the projects, and is not rewarded heavily for this investment in a drug that is still, in 2017, development-stage. The deal's main goals were laid out several months later in a follow-on press release. The title and main clinical trial goals are as follows:
The program will initially include four studies:
- MM-001 - A phase 1b multicenter, open-label study to determine the recommended dose and regimen of durvalumab as a monotherapy or in combination with pomalidomide with or without low-dose dexamethasone in patients with relapsed/refractory multiple myeloma
- CC-486-MDS-006 - A phase 2 international, multicenter, randomized, open-label, parallel-group study to evaluate the efficacy and safety of CC-486 alone or in combination with durvalumab in patients with MDS who fail to achieve an objective response to treatment with azacitidine for injection or decitabine
- MEDI4736 -NHL-001 - A phase 1/2 open-label, multicenter study to assess the safety and tolerability of durvalumab as monotherapy and in combination therapy in patients with lymphoma or chronic lymphocytic leukemia
- MEDI4736-MDS-001 - A randomized, multicenter, open-label phase 2 study evaluating the efficacy and safety of azacitidine subcutaneous in combination with durvalumab in previously untreated patients with higher-risk myelodysplastic syndromes or in elderly (≥ 65 years) acute myeloid leukemia subjects not eligible for hematopoietic stem cell transplantation
The above covers all the 5 listings in the CELG pipeline web page.
Now it's time to sum up.
CELG's pipeline - pros and cons, and summary
On balance, I have a positive view of the present value of the pipeline and R&D partnerships CELG. Because pipelines represent cost centers that, it is hoped, become profit centers that return far more than their cost of capital, they are highly uncertain, so what's most important to me about CELG are the current profit centers, discussed in detail recently by me in The Case For Celgene.
Some companies would rest more on their laurels given the commercial success CELG has achieved so far, but CELG is more aggressive than that, and has reinvested heavily in its pipeline.
Thus, it has created a complex network of relationships. Different people will think about its R&D program differently from the way I have organized it, but as stated, I like the approach I've evolved. It allows me to try to get a handle on the big picture and then focus on specifics in proper context.
I do look at the CELMoDs as assets that might represent as close to an infinite profit series as anything an operating company of CELG's nature might create. Unlike a true monopoly such as AT&T (NYSE:T) had in its pre-breakup days, pharma companies can only "rent" monopolies, due mostly to patent expirations but also to improving technologies. Within that, CELG may have developed assets in CELMoDs that could provide large and growing profits well into the 2030s. Again, there are no guarantees here for investors, but I'm thinking that the Street may be behind the curve on CELMoDs.
Ozanimod, mongersen and other acquired or in-licensed drugs may well add to those potential profits, though it's important to see how well they recover their cost of capital. Finally, the broad space of partnering is a swing factor that's difficult to assess, but along with the CELMoDs has a great deal of potential upside. As with the Jack-Jill coin-flipping example, it only takes one big, lucky win to really win big in pharmaceuticals.
When I combine the solid growth prospects from CELG's established products with the varied ways that its pipeline may help it "win the future," I find the stock, in the $114 range trading around 19X estimated 2017 GAAP EPS, to be more attractive than the other large cap stocks I follow. Thus, I am in an "accumulate" mode for CELG, understanding that the stock is volatile, risky, dividend-less, and in my opinion most suitable for growth-oriented, risk-oriented investors.
Thanks for reading and sharing any knowledge of this topic, and any other comments you may wish to contribute.
Disclosure: I am/we are long CELG,BIIB.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: Not investment advice. I am not an investment adviser.