The Great Pharma Yard Sale

In advance of hearing their pitch, entrepreneurs are increasingly telling me that I’m going to think their start-up is “too early” and that my time is better served speaking to later stage opportunities.  Considering Osage does seed through late stage investing, I tend to get a little defensive about such comments. 

Then again, why does it seem like the whole life science community believes that VCs no longer invest in early stage start-ups?

There is some validity to that statement as few life science deals are closing and most are later stage.  Yet, amidst this doom and gloom a fair amount of Series A investments are closing.  So, something is clearly not adding up.  How can Series A and later stage investments be on the rise, but early stage investments be on the decline? 

The answer: Corporate spinouts. 

Merger Mania

In 2008 and 2009, many of the major pharmaceutical companies merged with each other, which resulted in the creation of massive single entities (e.g. Pfizer-Wyeth).  Subsequent to those mergers, a vast integration process began: consultants were hired; low hanging inefficiencies were resolved; programs were reprioritized; and hundreds of thousands of employees were let go.    

Program reprioritization is a way of life for drug developers at pharma companies as assets in non-strategic areas, and those with development challenges (e.g. unwanted tox profile), are cut while promising ones are kept.  Before the mergers started, pharma companies had the R&D budget to support numerous promising programs and were reluctant to license them out.  Therefore, VCs didn’t really pursue pharma assets since they knew pharma companies only wanted to sell them their hairy castoffs. 

Today is different.

As many analysts and pundits correctly pointed out, merging two large pharmaceutical companies would not necessarily drive efficiencies or create value in the short term.  The integrations have taken longer than expected, development teams lack direction and are sitting on their hands in many cases, and moral is low.  With patent cliffs approaching and topline growth largely stagnant, companies are searching for ways to create shareholder value and justify the mergers.

The easiest way for pharma companies to “create value” for shareholders in the short term is to cut costs in order to boost their bottom lines.  The challenge with creating value in this matter is that the mergers, integrations, and cost cutting measures of the previous three years already trimmed most of the fat from the companies and there simply isn’t anymore fat to cut.  At this point, if pharma companies want to lower their R&D cost, they will need to jettison some of their prized possessions’.

In Steps The VCs

Out-licensing has never really been a strategic interest for pharma companies, but it is fast becoming a core activity of their BD groups in the wake of the mergers and subsequent R&D cuts.  Because pharma companies are feeling pressure to get development programs off of their P&L, good assets are being put on the auction block at very attractive prices and deal terms for VCs.  Not only are the asset prices attractive for investors, but also the fact that many of the pharma companies will actually give VCs money (e.g. equity, research support) to take the assets. 

Call options for out-licensed assets are quite attractive to many VCs.  In exchange for the ability to purchase the asset back at a predetermined price, the pharma company will often agree to cover some of the program’s development costs and provide access to in-house research expertise.   While limiting their potential upside, many VCs welcome such call option agreements because it provides cheap capital, additional expertise, and a willing and interested buyer.  A great example of call option success story is when Cephalon exercised its call option to purchase Ception Therapeutics. 

How The Next 6 Months Will Play Out

In many cases, it can be a challenge for early stage start-ups to compete with pharma companies for VCs’ attention.  Right now, pharma companies are offering late stage assets at attractive prices, non-dilutive capital, research support, and an ability to buy back the asset at a profit for investors.  

While out-licensing is currently sexy, it is also too good to be true as corporate spinouts are not a sustainable model for VCs.  There are only so many assets that pharma is willing to part with and once those assets are scooped up there is a huge risk of negative selection.   It is unclear how long the out-licensing trend will last, but sooner or later, VCs will come back to what they do best – early stage investing in transformative technologies and those entrepreneurs intrepid enough to roll the drug development dice.

Academic Drug Discovery

Translational research has been a hot topic amongst the academic life science research community for the last couple of years.  As research centers across the country attempt to form their own translational groups, a common thread amongst those institutions has emerged - the desire to have in-house drug discovery and early clinical validation capabilities.  Academic research centers are known for their ability to identify novel biologic targets and providing in-house drug discovery services enables academics to take their work to the next stage of product development.  

While proponents of translational research are quick to laud the vision of universities in creating soup to nuts drug development capabilities, others are not so supportive.  

Many critics challenge the notion that academic research centers should veer from their mission of conducting basic research.  Politics aside, it is very much an open question as to whether or not academic centers can execute the drug development process at the same level as pharmaceutical companies or CROs.

The Old Days

Historically, universities created compound libraries through the generosity of pharmaceutical companies.  In return for gifting compound libraries, pharma companies would receive modest tax breaks and academic goodwill.  However, the reality is that most of those libraries were gifted for a reason - either having little strategic value for the pharma company, or they were already picked over (scaffold fatigue).  Universities also found that it was challenging to use the libraries (not your typical post-grad skill set) and often cost $10,000+ to run large-scale screens.  This is turn caused university screening cores to run smaller, more directed screens, which limited their ability to find attractive leads.  

New Models

According to a recent Nature Reviews Drug Discovery article, there are 78 small molecule–focused drug discovery centers at universities or nonprofit research organizations in the US.  The report discussed some interesting stats:

• 2/3 of the drug discovery centers have high throughput screening infrastructure
• 2/3 have hit-to-lead medicinal chemistry expertise
• 1/2 have in vivo efficacy capabilities

In looking at those numbers, roughly half of the drug discovery centers appear to have the ability to perform IND-enabling work.  Building a competency in preclinical drug development takes lots of time, money, and talent, and it might be a stretch to say that 35+ research centers truly have that capability.  Even simple (if there is such a thing) drug discovery programs require scores of medicinal chemists, formulation chemists, toxicologists, etc.  Pharma is able to support disease-focused teams with deep preclinical drug development expertise, a luxury that universities cannot afford. Therefore, universities must work extra hard to find super talented researchers who can wear multiple hats during the development process. A good example of one university that seems to be doing all the right things is Duke University.

In 2006, Duke founded the Duke Translational Research Institute (DTRI) to expedite the translational development of drugs that were being discovered within the institution.  DTRI became the umbrella organization that was to oversee the management of the university’s clinical research activities.  Aided by a $52 million NIH grant, the DTRI hired a strong management team with start-up, VC, and CRO experience.  The DTRI provides CRO-like services for Duke faculty, as well as for outside groups, and is now self-funded.  In 2008, Duke hired Allen Roses, the former head of pharmacokinetics for GSK, to helm the Duke Drug Discovery Institute (DDRI).  Folding DDRI into DTRI gave Duke the ability to provide drug discovery, lead optimization, clinical development services for all faculty members, as well as outside groups.  DTRI is a good example of the kind of university-wide coordination, recruiting of talent, and capital required to execute a successful drug discovery program.

Engaging Venture Capital Funds

Ultimately, the goal of university drug discovery centers is to advance technologies to the point where investors and pharma companies become interested in licensing those technologies.  Pharma companies are more likely to fund collaborations with individual PIs or research groups around a specific indication than VCs, which prefer to license technologies directly into a start-up.  For pharma companies, they have a vested interest in developing long term relationships with Key Opinion Leaders, whereas VCs focus more on investing in a specific technology that fits into that particular fund’s investment thesis and timelines.  

As pharma pipelines have dried up and VC dollars become more scarce, both groups are increasingly looking for similar attributes when they evaluate technologies that are derived from university drug discovery units:

• Viable biology 
• Novel target in a relevant disease area
• Target is druggable (at least theoretically) / responder studies

The Future

Generally speaking, academic institutions are good at identifying interesting biologic targets, while pharma companies are good at generating molecules against those targets (See Osage blog - June 13, 2011).  Creating drug discovery centers at universities challenges that paradigm and co-locates drug discovery and development functions within the university.  Going it alone for universities enables them to have greater oversight and ownership of the development process, but at what cost?

With the NIH budget being flat for the past five years, it is hard to believe that drug discovery units at universities will be able to grow through grants.  Instead, universities will increasingly need to serve outside customers to cover their overheard and to subsidize services they provide to faculty.  Targeting outside customers brings the institutions into competition with CROs which likely can beat them on service offerings and prices.  

Despite these headwinds, there are a number of roles that academic drug development centers can play:

1. University talent could be directed to solve problems that hinder drug development for all stakeholders, such as creating models that better predict human versus animal toxicology, or generating improved disease models.  
2. VCs are hesitant to invest in cardiovascular, diabetes, and obesity diseases due to increased FDA regulatory burdens.  These are diseases that effect a large part of the US population and academic drug discovery units could certainly play a role in driving innovation in those areas.

It is too soon to tell if the academic drug discovery effort will be successful.  Ultimately, I think a couple of models will emerge and the early success of Duke’s discovery program certainly provides a blueprint for others to follow.

Could vaccine adjuvants lead to new Lupus therapies?

Therapeutic vaccines are hot right now thanks to Dendreon’s recent approval for Provenge and the H1N1 scare of 2009 that resulted in the government pumping hundreds of millions of dollars into venture backed vaccine companies.  Many of the therapeutic vaccines now in development are subunit vaccines that require co-administration with an adjuvant to pump up the seroconversion.  As a result, Toll-like receptors (TLRs) have become one of the hottest areas of adjuvant development for venture and strategic investors alike. 

While I have seen a number of interesting TLR stimulant opportunities over the last year, I have only seen one TLR antagonist idea – and it happened to be a pretty darn cool one.

Lupus is a disease that occurs when a person’s immune system turns on its owner, resulting in significant inflammation and subsequent tissue damage to healthy cells.  One of the key mediators of Lupus-associated inflammation is interferon-alpha (INF-α), a cell signaling molecule that recruits immune cells to destroy pathogens, or in the case of Lupus, the person’s healthy cells.  The binding of pathogens to TLR, specifically TLR-7 and TLR-9, has been shown to increase the expression of INF-α, making researchers wonder if by antagonizing TLR activation could they decreased INF-α expression and subsequent destructive inflammation?

Dynavax Technologies Corp. and the Baylor Institute for Immunology Research published a study in Nature where they administered a TLR-7 / TLR-9 antagonist in two separate lupus prone mouse strains and then monitored INF-α expression levels.  What the researchers found was fairly astonishing, namely that by antagonizing TLR-7/9 they were able to decrease INF-α expression.  As the INF-α levels dropped, the mice became more sensitive to glucocorticoid therapy, the standard steroid therapy used to treat inflammation. 

This is positive news for Lupus patients because over time most stop responding to low dose steroid therapies, which in turn forces doctors to increase dose strength.  While increasing the steroid dose decreases inflammation, it is also associated with a litany of dangerous side effects.  By antagonizing TLR-7/9, there is a glucocorticoid dose sparing effect that could protect Lupus patients from many of the unwanted side effects associated with prolonged exposure to high dose steroids.

Toll-like receptors are ideal adjuvant targets because they promote the activation of cell-specific immunostimulation, but it would now seem the same principle holds true in reverse.  TLRs might also be a new class of drug targets for a number of inflammatory disorders including rheumatoid arthritis, Lupus, and Crohn’s disease, that could benefit from selective inhibition of molecules that mediate inflammation.  

RNAi, Don’t Be a Hater

It’s easy to join the “RNAi doesn’t work” bandwagon these days.  Roche and Merck have all but given up hope on the space, and partnering talks for existing RNAi startups have ground to a halt.  But, people should be reminded that pharma investment interests go in cycles, and well, RNAi might just be in a down cycle as opposed to being a write-off.

That being said, should RNAi really be in a down cycle?

Many of us haven’t given up hope on the promise of RNAi, a relatively new science that was only first discovered in 1998.  In a short amount of time, RNAi has established itself as a staple of research labs and become the go-to assay for modulating gene expression patterns.  While academic researchers typically do not have to worry about tissue specificity, stability, etc., the ubiquitous usage of RNAi illustrates the promise of the technology.  

In some respects it is not a surprise that pharma companies and startups have struggled to commercialize RNAi.  Many of the mechanisms and molecules associated with RNAi processing have only recently been discovered.  Such discoveries should empower researchers to invent new methods for RNAi delivery and tissue selectivity.

Al Doig of Insight Pharma Reports recently compared the current state of RNAi drug development to that of monoclonal antibodies in the early 1990s.  Al notes,

“It was a 1975 publication describing the technology for generating MAbs, for which the authors received a 1984 Nobel Prize, that ignited interest in the possibility of MAb therapeutics. In the 1970s and 1980s, MAb technology represented great science, and biotechnology and pharmaceutical companies were excited at the prospect of developing MAb drugs. However, the development of MAb-­based drugs lagged far behind the science. The first MAb drug, Johnson & Johnson’s Orthoclone OKT3, was approved in 1986, albeit with limited applicability. Interestingly, there were no other approved MAb drugs until 1994 and the “avalanche” of MAb drug approvals didn’t begin until 1997.”

To put that timeline in perspective, when compared to MAbs, the avalanche of RNAi drugs should come sometime around 2020.  There are numerous great startups, many of which are university startups that are poised to set off that avalanche: 

• Alnylam – Scripps, Rockefeller, MIT
• Tekmira – University of British Columbia
• RXi Pharmaceuticals – University of Massachusetts
• Calando Pharmaceuticals – Caltech
• Silence Therapeutics – spinout of Ribozyme Pharmaceuticals (University of Colorado)
• Quark Pharmaceuticals – licensed core technology from Silence Therapeutics
• Dicerna - City of Hope

I prefer to think of RNAi as in a lull right now.  It has gotten past the initial hype and enthusiasm, and now people are performing the long, arduous, and non-sexy task of making the technology work so that it ultimately can have clinical utility.  Positive clinical data from any of these companies should provide enough momentum for RNAi to become hot again.

Tau Protein & Amyloid Beta, a Love Story

The primary cause of Alzheimer’s disease has been debated for decades amongst researchers.  Some researchers staunchly believe that tau protein, a scaffolding protein that accumulates into tangles in neurons, is the primary driver of the disease.  While others believe that amyloid beta, a cleavage product that accumulates into dense plaques in the brain, is the main culprit.  Tao supporters have become increasingly convinced that amyloid beta is not related to disease progression due to the late stage clinical failures of several amyloid beta drugs over the past year. 

Recent evidence now suggests that Tau and amyloid beta could be linked. 

Bess Frost and Marc Diamond published a piece in Nature Rev Neuroscience (2009) that highlights their work in showing a direct link between amyloid beta and tau in causing neuronal toxicity in several animal and in vitro models.   The initial work by Frost and Diamond has been confirmed by outside laboratories, but the exact mechanics of the interplay between tau and amyloid beta, and how that relates to Alzheimer’s disease progression, remains poorly understood.

A recent review in Nature Reviews Neuroscience, Lars Ittner and Jurgen Gotz of the University of Sydney, posits a new mechanism of disease progression for Alzheimer’s disease. 

The new hypothesis, which Ittner and Gotz refer to as the “tau axis hypothesis”, proposes that increasing levels of tau makes neurons vulnerable to toxicity induced via amyloid beta.  To explain their hypothesis they break Alzheimer’s down into early and late phases.  In the early phase, amyloid beta plaques accumulate around the dendrite and begin to damage the dendrite.  Prior studies have linked amyloid beta to tau phosphorylation via amyloid beta binding to NMDA receptors, which is thought to be necessary for tau to aggregate into tangles.  Activation of the NMDA receptor then triggers the phosphorylation of tau via the postsynaptic protein 95 (PSD95) and tau begins to aggregate.  Tau is associated with neuronal toxicity, so as it accumulates it sensitizes the dendrite to further amyloid beta induced toxicity – essentially creating a feedback loop whereby tau and amyloid beta increasing push one another to become even more active.  This leads to greater aggregation of tau and amyloid beta and the eventual loss of synaptic function and subsequent neuronal death.  

The translational impact of linking amyloid beta and tau to Alzheimer’s disease progression is substantial, not to mention it also squares away a lot of recent conflicting data.  Researchers have recently shown that having both overactive tau and overactive amyloid beta leads to aggressive Alzheimer’s disease and death. What has also been shown is that mice with normal tau and overactive amyloid beta develop Alzheimer’s disease, although not as progressive and lethal as when both proteins are overactive.  Two additional models show that when tau is knocked out or down when amyloid beta is overactive, mice tend to have minimal memory deficits. 

Taken together, this data suggests that blocking the ability of amyloid beta to signal tau protein aggregation in the early phase of disease progression could lead to decreased loss of memory for patients.  However, it remains unclear what the drug target should be to prevent amyloid beta and tao signaling.

The tao and amyloid beta story is in its infancy but over time should lead to some new creative Alzheimer’s therapies.