Morphosys (MOR.DE) is one of the most unusual biotech companies, as it breaks three basic rules that apply to drug development companies:
Rule No. 1: Development-stage companies burn cash and therefore must constantly raise capital and dilute existing shareholders.
Rule No. 2: Development-stage companies are risky and volatile because they rely on a limited number of binary events.
Rule No. 3: Investing in cutting edge, growing segments of the pharmaceutical industry is associated with a high level of risk.
Morphosys is the only company I am familiar with that systematically breaks each and every one of these rules. It does not have any drugs on the market and is not expected to have any in the foreseeable future, yet it is profitable. It is involved in drug discovery which is associated with a high attrition rate, yet statistically, there is a very high chance that it will have commercial revenues at some point in the future. It is involved in one the fastest growing segments in the industry, but can be regarded as a conservative holding since it will never be dependent on a limited number of binary events. And finally, it has no need to raise cash in the coming decade in order to support its activities, as its costs are covered by other companies.
Morphosys has developed a unique technology for discovering and producing monoclonal antibodies. The technology, called HuCAL (Human Combinatorial Antibody Library) basically mimics the immune system, as it scans a vast repertoire of antibodies and identifies the ideal ones for a given target. This approach represents a shift from the traditional approach of developing antibodies, because it does not involve animal immunization.
The classic way of developing antibodies involves immunizing an animal (typically a mouse) with the target of choice and utilizing the animal’s capability to produce antibodies against the specific target. Morphosys’ solution bypasses the need of immunization, as the entire process of screening and selecting the antibodies takes place on the scientist’s bench. Although most of the antibodies on the market were developed through animal immunization, the “artificial” approach is well accepted and going forward, the industry will probably see plenty of drugs based on both approaches. At the end of the day, the only thing that matters is the ability to generate an antibody and get it approved, which can be achieved by both avenues, and in some cases, screening technologies have a clear advantage over the immunization. Morphosys’ technology is not the only screening technology out there, but it is certainly considered one of the best, if not the industry’s leading screening system, as underscored by this long list of collaborators.
Morphosys’ platform is the basis for three business segments, each representing a different risk profile and market potential: (i) Research and diagnostics, (ii) partnered pipeline and (iii) Proprietary Pipeline. It is the combination of these segments which makes Morphosys such a unique investment opportunity.
Research and diagnostics
The research and diagnostics division was built primarily by acquisitions and is currently serving research labs and diagnostic companies around the world. Of the three divisions, this one can be seen as the most “traditional”, as it has a relatively low growth rate and commercial potential. On the other hand, the research and diagnostics division is considered very safe, as it is independent of clinical results. In 2008, after several years of losses, the division finally became profitable, despite a sequential decrease in revenues, down 7% to $18.2M. The company expects the unit to grow 10% in 2009 and to maintain a modest level of profitability in the near future.
Morphosys is trying to reshape this division by focusing more on antibodies for diagnostics. The market of diagnostics, albeit much smaller than that of drugs, is expected to experience strong growth in the coming decades as part of the trend towards personalized medicine. Antibodies have an important role in the field and are well entrenched in the market, due to their high specificity and ease of use. Morphosys is addressing the market for antibodies for diagnostics, estimated at several billions of dollars annually, by licensing its platform to diagnostic companies. It is currently involved in over 20 projects, some of which are expected to mature into commercial products. Last year, Morphosys announced that Sweden based Phadia would use an antibody developed by Morphosys in one of its tests for auto-immune diseases. The financial potential of such projects is relatively small, but unlike therapeutics, diagnostic products represent a fast route to market and limited regulatory risk. Without a doubt, the big potential for Morphosys lies in its partnered therapeutic pipeline.
The partnered pipeline segment is Morphosys’ cash cow, responsible for the vast majority of revenue and profit. The capability to discover and screen antibodies is now very common in the industry and in research labs, but only a small number of companies mastered this technique and created an efficient platform that can feed a vast number of projects. Morphosys is considered one of the four horsemen of antibody discovery, along with Medarex (MEDX), Cambridge Antibody Technology, now part of AstraZeneca (AZN), and Abgenix, which is now part of Amgen (AMGN). Medarex and Abgenix have developed immunization-based platforms whereas CAT and Morphosys have developed screening platforms that do not require animal immunization.
The universal and highly adaptable nature of the HuCAL platform enables Morphosys to be actively involved in tens of projects every year. Typically, Morphosys licenses its technology to pharmaceutical companies and develops anywhere between several to tens of development programs, each program revolves around a single target. The typical deal structure entails an initial licensing fee, full cost reimbursement by the partner, milestone payments of $12-16 million per program and mid single digit royalties on sales. Each program by itself may be modest in size, but considering the fact that Morphosys currently has 55 partnered programs, three of which already in clinical testing, the overall value is obvious.
Morphosys has agreements with a large number of partners, including some of the largest pharma companies, but in 2007, the company decided to focus most of its research activity on one collaboration by inking a transformative deal with Novartis (NVS). The deal, one of largest ever deals in the pharmaceutical industry was a ten year collaboration which includes more than a hundred new programs. In return to full, but not exclusive access to Morphosys’ technology, Novartis committed to pay $600 million over the course of ten years on top of the standard milestone and royalties on future sales. Although Morphosys is not limited with respect to partnering with other companies, Morphosys does not intend to sign additional broad discovery deals. Therefore, going forward, the Novartis collaboration will account for the vast majority of Morphosys’ activity, and will unofficially turn it into Novartis’ antibody division.
In 2009, the company expects 20 new programs to commence, primarily with Novartis, as well as the advancement of 4 antibodies to clinical testing by its partners. Looking ahead to the coming decade, Morphosys believes it will be involved in a “triple digit number” of programs. Based on present collaborations, Morphosys could cumulatively be involved in as much as 180 programs, an exceptional number by any standard. Most of the programs will not be financed by Morphosys, which will carry the cost only for programs it pursues independently or co-develops with partners.
Suffice it to say, the vast majority of these programs will fail to reach the market. In drug development only a minority of drugs prove both effective and safe, with approval rates traditionally in the single digit range, depending on the indication. Luckily for Morphosys, antibodies are thought to have better success rates due to their excellent safety profile and the ability to identify patients who would derive benefit from the treatment. According to several retrospective analyses, antibodies have a ~3 fold higher approval rates in indications such as oncology and autoimmune diseases, making Morphosys’ prospects even better.
The company tried to illustrate this in one of its presentations last year (see figure), by applying its internal statistics and historical approval rates for antibodies in order to predict the number of programs that will get approved. According to its analysis, which should be regarded as an extreme form of forward looking statement, the company will eventually see more than 17 (!) drug approvals. Investors would gladly settle for half of that number. Assuming average peak sales of $500 million per antibody per year, ten years of peak sales and a royalty rate of 5%, the cumulative value of Morphosys partnered pipeline could reach $4.25B, spread over fifteen years. This figure excludes any revenues from Morphosys’ wholly owned pipeline.
Three partnered antibodies that were developed by Morphosys are currently in clinical trials, in the hands of Roche, Novartis and Centocor. To date, none of them generated proof of concept data, however, that might change during 2009.
Roche is developing gantenerumab, an antibody for the treatment Alzheimer’s disease. The antibody is similar, in concept, to Elan’s (ELN) and Wyeth’s (WYE) bapineuzumab (bapi) as both antibodies target Amyloid beta, a protein which is one of the hallmarks of the disease. Roche advanced gantenerumab to phase I in 2006 and since then completed the accrual of 30 patients. This trial is somewhat atypical for a phase I study because it is a randomized, double blind comparative trial, so there could be signs of efficacy in the data. The market potential for Alzheimer disease is estimated at over $10 billion, however, to date, no drug proved successful in changing the course of the disease. Until recently, antibodies against Amyloid beta were considered a very promising target, however, following disappointing data for bapi, investors’ excitement towards this approach waned. Roche is expected to publish data from the phase I trial during the course of 2009.
Novartis is developing BHQ880, an antibody against DKK-1, a protein that inhibits bone growth and has been shown to be involved in bone related conditions. By neutralizing DKK-1 with an antibody, it may be possible to stimulate bone formation. The potential market for BHQ880, providing it proves effective, is very large, spanning from osteoporosis to multiple types of cancer.
The concept of preventing breakdown of bones with an antibody has already been validated by Amgen’s denosumab (Dmab), currently evaluated in a battery of phase III studies. Last year, Amgen published very positive results from a study in post-menopausal women with osteoporosis, in which the antibody led to a meaningful improvement in fracture incidence and bone density. Additional trials showed that Dmab decreases bone loss in breast and prostate cancer patients who received hormonal therapy. A third potential use might be prevention or shrinkage of bone metastases in cancer patients, with data expected in the 2009-2010 timeframe. Dmab is expected to hit the market next year, and instantly become a blockbuster, due to the large addressable market (~5 million people in the US are receiving treatment for osteoporosis) and the substantial cost to society as a result of osteoporosis complications, such as fractures.
Novartis will probably pursue BHQ880 in the same indications Amgen’s Dmab is being evaluated, but the two antibodies should not necessarily be considered as competitors. Not only does each of the two antibodies binds a different target, they are involved in distinct biological signals. Dmab is thought to inhibit bone destruction whereas BHQ880 is expected to stimulate bone formation, so the two may even be synergistic. But first, Novartis will have to show BHQ880 is effective on its own and bring it to market. In order to do so, it picked a relatively small indication – multiple myeloma.
In February of 2009, Novartis started a phase I/II study in multiple myeloma, a blood cancer in which tumors colonize in the bone and degrade it. The vast majority of patients will develop bone lesions at some stage of their disease, resulting in bone loss, pain and increased likelihood of fractures. By stimulating bone formation, BHQ880 may decrease or even prevent bone loss that seems essential for the creation of bone lesions. This, in turn, may lead to not only better quality of life but also reduced tumor burden.
The concept of targeting DKK-1 is based on a growing body of evidence which shows that DKK-1 has an important role in multiple myeloma. For example, a study published in 2003 showed that multiple myeloma cells can create bone lesions by secreting proteins which lead to bone loss, and that one of the proteins they secrete is DKK-1. In addition, the investigators examined cancer cells from patients and found that cancer cells in bone lesions secrete high levels of DKK-1 whereas cancer cells from the blood of patients without bone lesions do not produce the protein.
The decision to start from a small indication like multiple myeloma as opposed to larger indications such as osteoporosis or even prostate cancer has its merits. Despite the significant progress with drugs such as Celgene’s (CELG) Revlimid and Takeda’s Velcade, no drug has been able to cure multiple myeloma, so new treatments are in high demand. In addition, multiple myeloma is not nearly as prevalent as osteoporosis, making it an ideal fast track indication, with a short time to market and a relatively low cost. A typical registration study in multiple myeloma requires less than a thousand patients, while in order to file for approval in osteoporosis, Amgen had to accrue 7800 patients.
Novartis is evaluating BHQ880 in a fairly large study (267 patients) with a placebo arm, which could make potential positive results more credible and serve as a proof of concept for the drug’s activity. This demonstrates again the advantage of having a large partner behind the wheel, as a company like Morphosys would never start such a large and costly trial at such an early stage. Novartis will probably initiate clinical trials with BHQ880 in additional indications in the near future.
Undisclosed antibody (Centocor)
Centocor, a wholly owned subsidiary of Johnson & Johnson (JNJ), signed a licensing deal with Morphosys in late 2000. In the summer of 2007, Centocor promoted one of its programs to phase I trial in solid tumors. Five months ago, it started another trial in an autoimmune indication, idiopathic pulmonary fibrosis (IPF) with the same antibody. Although Centocor did not disclose the identity of the antibody and its target, it seems that the mysterious antibody is CNTO-888, an antibody targeting a protein called MCP-1. Similarly to the two other partnered programs, CNTO-888’s trials are relatively large with 54 and 120 patients planned for accrual in the cancer phase I and IPF phase II, respectively. The phase II is a placebo controlled study.
MCP-1 plays a role in recruiting cells of the immune system by mobilizing them to specific sites, and is therefore believed to be involved in processes such as immune response and wound healing but also in autoimmune diseases such as multiple sclerosis and even metabolic diseases such as type II diabetes. Although MCP-1 has never been validated as a target, many studies suggest that molecules that block the actions of MCP-1 may be useful in treating a range of diseases. According to Centocor, MCP-1 is also involved in blood vessel formation, so targeting it may be useful in solid tumors, which must build new blood vessels in order to grow and invade distant organs. By binding MCP-1, Centocor hopes that CNTO-888 will starve tumors, similarly to the mechanism of action of Genentech’s Avastin.
Centocor is the only company actively developing an antibody against MCP-1 in the clinic, but it is not the first one to try, and prior experience does not leave a lot of room for optimism. Novartis was developing its own anti-MCP-1 antibody several years ago but decided to discontinue the program shortly after it got into the clinic. In a phase I trial in RA, Novartis’ antibody did not only fail to show benefit, but also led to a worsening of disease symptoms in some patients. This casts serious doubts over the prospects of anti-MCP-1 antibodies in autoimmune diseases, but one still cannot reject the entire concept based on one antibody. Other companies are trying to inhibit MCP-1’s activity by targeting its receptor (CCR2) with a small molecule rather than an antibody. ChemoCentryx is evaluating its compound in a phase I trial in vascular restenosis, a condition caused by blood vessel blockade following a stent procedure. Incyte (INCY) also has a CCR2 inhibitor, but at the moment, the development program is on hold, due to financial constraints. BMS is currently enrolling patients in a phase II study in diabetes for BMS-741672, another small molecule inhibitor of the receptor. The most advanced antibody in the field was Millennium’s (now part of Takeda) MLN-1202, an antibody against the CCR2 receptor, but the company decided to discontinue its development after disappointing phase II results in RA.
To my knowledge, Centocor is the only one who is evaluating inhibition of the MCP-1 pathway in oncology. This could help to differentiate CNTO-888 from other drugs, but only based upon concrete data from the phase I trial, which is still ongoing. The decision to do a phase II in an autoimmune disease can be interpreted as turning way from cancer indication, but perhaps this is part of the planned development program Centocor had originally laid out. The initiation of a phase II trial implies that Centocor already reached the maximum tolerated dose, so if investigators see signs of activity in the phase I, the data should be available this year.
Morphosys’ proprietary pipeline is the third and most recent initiative of the company. The basic idea is simple: Using the tens of millions that flow into Morphosys’ bank account every year to build a small, early stage clinical pipeline. Therefore, Morphosys does not expect to burn cash in the foreseeable future, even as it anticipates having a handful of antibodies in clinical testing. These wholly owned programs are the only chance the company has to generate additional meaningful revenues in the near term future through licensing deals.
Morphosys does not intend to independently commercialize its wholly owned products, but to out license them after proof of concept data. This strategy is similar to Isis’ (ISIS) strategy, which alongside Morphosys, is one of the few companies that are developing drugs without burning cash. There is, however, one critical difference between the two companies – the fields in which they operate. Isis’ antisense platform has the potential of revolutionizing the pharmaceutical industry by creating a completely new class of drugs, but to date, antisense drugs have not been fully validated. In contrast, monoclonal antibodies are a highly validated class of drugs, with over 20 approved agents to date, some of which have achieved blockbuster sales. Isis will have its first opportunity to prove that antisense drugs work with its high profile agent, mipomersen, a lipid lowering drug currently in several phase III studies.
Morphosys’ decision to build its own pipeline raises two principle issues. The first issue is the availability of good targets, as identifying the right target is the first and probably most important step of developing antibody-based drugs. Morphosys technology licensing deals usually revolve around specific targets, where the partner gets exclusivity for the target. In other words, Morphosys cannot license the same target twice or independently develop antibodies against an already licensed target. With more than 50 partnered programs ongoing and over a hundred future programs, the pool from which Morphosys can choose is rather limited. When I asked Morphosys’ CEO, Simon Moroney, how he views this issue, he admitted that a lot of targets are indeed taken, but claimed that on top of the substantial amount of available targets, there is a constant increase in the form of new targets every year. In addition, he added, the knowledge and experience gained through so many partnered programs compensates for the loss of potential targets.
The second issue is Morphosys’ ability to create and develop a clinical program from scratch, as this task requires a different set of skills on top the scientific know how. It is still too early to evaluate the company’s performance in this area, however, judging by the first clinical program, Morphosys’ management team know a thing or two about picking the right candidates.
Morphosys has several wholly owned development programs, only one of which, MOR103, is in the clinic. The company expects to promote a second antibody, MOR202, to the clinic in 2010. In addition, Morphosys expects to have several co-development programs in the clinic going forward, as part of the Novartis deal.
MOR103 is an antibody targeted at GM-CSF, a protein traditionally known for its ability to stimulate the production of certain blood cells in the bone marrow. GM-CSF is used as a drug to stimulate the generation of new blood cells in cancer patients who undergo chemotherapy and following bone marrow transplantation. Although it was discovered over 30 years ago, GM-CSF’s role in inflammatory diseases is still being elucidated.
The past years saw the accumulation of data that implicate GM-CSF in inflammatory and autoimmune diseases. High levels of GM-CSF were found in joints of rheumatoid arthritis patients, and in animal models, targeting GM-CSF with an antibody led to a decrease in symptoms in several disease models. But perhaps the most convincing evidence comes from anecdotal cases in a 1990 clinical trial. In the trial, ovarian cancer patients were treated with chemotherapy followed by GM-CSF. Some of the patients on the trial also had rheumatoid arthritis and investigators noticed that the administration of GM-CSF led to deterioration of the disease in these patients. This finding implies that GM-CSF has a causative effect in RA and that disease control might be achieved by neutralizing it.
MOR103 entered a phase I study in the first quarter of 2008 in healthy volunteers. The study was expanded to evaluate higher doses than originally planned after no safety issues were observed in the original cohorts. During the second quarter of 2009, the company expects to publish data that will include both safety results as well as biomarker data, but obviously, no efficacy data can be generated in healthy patients. MOR103 will probably enter phase Ib/IIa in RA patients later in 2009, with potential proof of concept towards the end of 2010.
As part of the development program, last year Morphosys acquired exclusive rights to a patent which covers the use of antibodies against GM-CSF for therapeutic use. The patent, which is valid only in the US, may block other companies from selling anti-GM-CSF antibodies in the American market, but it still remains to be seen how this patent holds up in court. If MOR103 becomes approved, this step may turn out to be brilliant, considering the fact that the US accounts for more than half of the worldwide RA market. Of note, there are additional companies that are developing antibodies against GM-CSF, including Nycomed with its partner Micromet (MITI). The two companies plan to start a phase I study with their antibody this year and do not seem too excited about Morphosys’ patent. Furthermore, Morphosys’ patent does not cover antibodies against GM-CSF receptor, such as Medimmune’s CAM-3001, which could compete with MOR103 if both get approved.
MOR202 is an antibody against CD38, a protein expressed by multiply myeloma cells. According to the company, it will start a phase I trial in 2010. By that time, two more anti-CD38 antibodies are expected to be in the clinic. The first is Genmab’s HUMAX-CD38 which entered the clinic early last year. In addition, Sanofi-Aventis is expected to advance its antibody to the clinic in 2009. Antibodies for multiple myeloma are one of the most active areas in the industry, following Rituxan’s (another antibody) success in various forms of blood cancers. Unfortunately, Rituxan is not effective in multiple myeloma, so developers are looking for the “Rituxan of multiple myeloma” by targeting additional targets such as CD38. Companies developing antibodies for multiple myeloma also include Genentech and Bristol-Myers Squibb, but to date, no antibody has demonstrated clinical proof of concept in the disease. The value of MOR202 as a preclinical agent is low, but if Genmab or Sanofi validate CD38 as a target, Morphosys might be able to license its antibody without any clinical data. At that point, the company may still choose to wait until MOR202 generates clinical data, probably in late 2011.
Earlier this month, Morphosys celebrated a decade as a publicly traded company, a decade which was mostly comprised of research and pre-clinical activities. The next decade will be characterized by intensive clinical activity for Morphosys’ proprietary pipeline as well as for its partnered pipeline. While the company expects dozens of programs to reach the clinic in the course of the next decade, the biggest value creating events for the partnered pipeline will arrive only in 7-10 years time, with the potential approval of some of the antibodies.
Although the company is not involved in any late stage clinical programs, Morphosys still looks like it has plenty of upside potential for the coming years. First and foremost, Morphosys is looking at several inflection points with respect to its proprietary pipeline, the first of which is expected next year. According to the company’s CEO, they are actively looking at acquiring products from other companies, which could instantly enhance the company’s pipeline. In addition, Morphosys’ value will definitely be derived from the size and quality of its partnered pipeline, which is expected to grow by several candidates every year. Even a single agent that demonstrates good activity in phase II can be translated into substantial stock price appreciation. Immunogen, for example, derives most of its valuation from a single drug with good phase II results: T-DM1. Similarly to Morphosys’ licensing agreements, Immunogen is eligible for royalties of ~5% on T-DM1 future sales, but due to the impressive phase II data and the clear blockbuster potential, Immunogen’s market cap is now $357M, slightly higher than that of Morphosys. Lastly, from a financial standpoint, Morphosys is a solid investment with positive cash flow and almost half of its market cap in cash, so it will remain independent of the capital markets in the coming years.
The main disadvantage stemming from the company’s business model is its dependence on Novartis. Morphosys decided not to start new broad collaborations, so in several years time, projects with Novartis will capture most of the company’s bandwidth, turning it into an unlikely acquisition target for anyone but the Swiss giant. Therefore, Morphosys is not a likely acquisition target, despite its valuable assets, and even if Novartis decides to buy it down the road, most chances that it would not encounter competition from other big pharmas, who would not want to have such a commitment to one of their competitors. Some may view this situation as an advantage because it will allow Morphosys to stay independent in the coming years and build value for its shareholders.
In summary, Morphosys can be viewed as a blend of pharma and biotech. On the one hand it has the innovation and upside potential of a small biotech, and on the other, it enjoys the diversification and risk mitigation of a large pharma. Morphosys still operates in the drug development field, where failures vastly outnumber successes, but unlike most of its peers Morphosys has statistics on its side. Does it mean that Morphosys is a risk free investment? Absolutely not, but for investors who would like to get exposure to the growing biotech field but with limited risk, Morphosys is as good as it gets.
Since its inception almost 6 months ago, the biotech portfolio, co managed by Ran Nussbaum and myself, generated a return of 21.6%, versus a decline of 6.1% and 10.3% for the NASDAQ and S&P, respectively. In addition, the portfolio outperformed the leading Life Sciences indices and ETFs, including the NASDAQ Biotechnology Index (^NBI), represented in the figure below by the iShares NASDAQ Biotechnology (IBB) ETF.
We decided to sell one of our two positions in Immunogen for a gain of 104% following the sharp climb the stock had experienced, as well as our position in Allergan (AGN) for a modest gain of 17%. On a less positive note, we are also selling our position in Synta (SNTA) at a loss of 70%, after its lead drug failed in a phase III in metastatic melanoma, one of the toughest indications in the industry.
On our buying list this time are Morphosys, which, unfortunately is not traded in the US, Cephalon (CEPH) and Array Biopharma (ARRY). In addition, we are also buying more of Morphosys’ neighbor in Munich, Micromet. Micromet has been suffering from negative sentiment following Medimmune’s decision to cease development collaboration for Micromet’s leading product, blinatumumab (MT-103). When a large company decides to abandon a drug, its decision is usually regarded as a red flag to investors. In this case, however, Medimmune’s decision seems to be based more on strategic issues rather than the quality of blinatumumab. After a long call with Christian Itin, Micromet’s CEO, we believe that the Micromet story remains intact and there is no change in the potential (as well as risk) of the company’s platform. More on Micromet in the next article.
Portfolio holdings as of March 29th, 2009