The two experimental drugs that treat COVID-19

The FDA today announced it is working with private companies to bring two previously approved therapies to approval as treatments for #COVID19, the coronavirus first detected November 17 in Wuhan, China. Both are being tested for efficacy in humans, as part of the normal (but now highly expedited) process for seeking FDA approval.

One is remdesivir, the Gilead antiviral developed for use against Ebola and Marburg, both exotic filoviruses with case fatality rates of about 50% (vs. about 0.7% for COVID19). The National Institute of Allergy and Infectious Diseases (NIAID), the division of NIH that funds research on infectious diseases, is sponsoring its own randomized clinical trial using remdesivir against COVID-19 that has been given to 250 patients already.

The other is chloroquine, used since 1949 as an anti-malarial. It has long been known to work against SARS coronavirus, the closest prior pathogen to COVID-19. It works against a wide range of viral and bacterial pathogens — including Ebola — by inhibiting cellular entry. With remdesivir, it has already been tested in China against COVID-19.

Mikhail Shilman, my coworker (at KGI) and cofounder (at our startup) issued a statement about these two approvals:

The FDA decision this morning to fast-track approval chloroquine for treatment of COVID-19 coronavirus shows the potential for previously approved drugs to be deployed quickly to treat emerging infectious diseases that can rapidly infect an entire community or nation.

Chloroquine, an anti-malarial, has been safely used for 70 years around the world. The FDA has policies to encourage the repurposing existing drugs because it is quicker and lowers the risk of unexpected adverse effects. It is also a drug that is already being mass produced and is simple to manufacture, and thus can be quickly scaled up unlike complex large-molecule therapies.

Coronaviruses attack epithelial cells within the lungs. It can be life threatening because once it is inhaled, the infection inflames and swells the lungs in a way that makes it increasingly hard to breath. We thus need an orally administered drug that can be distributed in our system to reach and protect organs like lungs.

As we showed in our 2015 paper in Nature Scientific Reports, chloroquine is one of only two quinoline drugs that protects cells against certain types of viral and bacterial pathogens. It blocks the entry of these pathogens into the cell, including Ebola and SARS coronavirus, by binding to and inhibiting the function of the host Cathepsin protein, which mediates the uptake of these pathogenic agents into human cells. While we have not tested COVID-19 coronavirus, it is believed to work the same way as SARS and MERS coronaviruses.

Of the two drugs, chloroquine is the weaker of the two, better tolerated and better suited for treatment for the several weeks necessary for the body’s immune system to clear a virus such as COVID-19. The other drug, amodiaquine, is much stronger and better for immediate threats but has the potential for side effects if taken for longer periods; that is why I’ve been working to develop amodiaquine for inhaled anthrax, which kills the majority of patients in a matter of days.

Recent research has shown that chloroquine is effective against COVID-19 samples in the lab. Bayer said today they are donating three million chloroquine tablets for use as a COVID-19 treatment. Given the urgency of the pandemic and the proven safety of this drug, it is now appropriate to test chloroquine as a way to protect or treat high-risk populations.

Market Realities

Some of the coverage of the COVID-19 crisis has asked why the US doesn’t have a treatment for this coronavirus and why the government is taking so long to bring one to market. The two reasons are the regulatory process, and the perception of market demand.

Well, the reason is that it typically 7-10 years after discovery of a drug to get it approved. DiMasi, Hansen, & Grabowski (2003) estimate the mean duration of a Phase 1 safety study as 1.8 years, a Phase 2 efficacy study as 2.1 years, and a larger scale Phase 3 study at 2.5 years. So that is 6.4 years, not counting the 2-3 years before that to justify getting into humans, and the 12-21 month delay (depending on therapeutic area) waiting for FDA approval (DiMasi, Grabowski & Vernon, 2004).

While the FDA allows off-label use of approved drugs, and compassionate use access when there is no standard of care, our society (like any other) expects a rigorous, multi-year process to prevent sale of unsafe or ineffective products.

The other issue is the commercial reality. The DiMasi et al (2004) study estimated that on average, clinical trials for an antiinfective drug took $362 million and lasted 10.3 years.

As the CEO of a (tiny) pharma company, it’s hard to make a case for pursuing a scientifically feasible drug unless a) it’s a big market or b) someone else (government, foundation) bears the risk.  I know for sure that I won’t get venture capital to fund my trials unless investors think there is a good chance of achieving sizable revenues — particularly given the 11% clinical success rate for all human therapeutics (DiMasi et al 2016).

So when there’s an epidemic, under normal circumstances firms will have something approved long after it’s over. If no one (or almost no one) is sick any more, there’s no market for treatment. However, insurers and public health authorities are willing to buy a vaccine so the disease doesn’t come back again.

Finally, even if approvals are accelerated, it takes time to be able to manufacture a drug at scale with FDA-approved quality standards. Repurposing existing drugs are the only ones likely to be ready for widespread usage this year. Thus, we should be excited by this good news, and the increased sense of urgency by the government to encourage further investment in deploying these therapies.

John, Paul and Eroom

Catching up on “In the Pipeline” over spring break, I followed Derek Lowe’s pointer to yet another article in Nature Drug Discovery attempting to explain the declining productivity of drug R&D.

Rather than start from scratch, the team from Sanford C. Bernstein (an investment research company) summarize many of the previous diagnoses, noting the lack of incommensurability of the earlier proscriptions:

They include: the FDA’s ‘Critical Path Initiative’; a series of prescient papers by Horrobin, arguing that bottom-up science has been a disappointing distraction; an article by Ruffolo focused mainly on regulatory and organizational barriers; a history of the rise and fall of medical innovation in the twentieth century by Le Fanu; an analysis of the organizational challenges in biotechnology innovation by Pisano†; critiques by Young and by Hopkins et al., of the view that high-affinity binding of a single target by a lead compound is the best place from which to start the R&D process; an analysis by Pammolli et al., looking at changes in the mix of projects in ‘easy’ versus ‘difficult’ therapeutic areas; some broad-ranging work by Munos; as well as a handful of other publications.

There is also a problem of scope. If we compare the analyses from the FDA, Garnier, Horrobin, Ruffolo, Le Fanu, Pisano, Young and Pammolli et al., there is limited overlap. In many cases, the different sources blame none of the same countervailing forces. This suggests that a more integrated explanation is required.

† The book by Gary Pisano (2006) of Harvard Business School is a favorite here at KGI — I think I’ve seen it on more shelves than any other.

The main point of the article is to introduce a so-called “Eroom’s Law”, in which the number of new FDA-approved therapies (per $ million R&D) fell by half every nine years. It is intended to be the inverse of the Moore’s Law that has driven IT improvements or more than 40 years, in which the number of transistors per chip doubles every two years.

The Eroom’s Law was the major new gimmick of the Scannell et al article, and Lowe devoted an entire posting to it. I’m far from convinced: unlike Gordon Moore, there’s no real explanation of the causal mechanisms — even post hoc. Instead, it seems more like Eroom’s Syndrome — a collection of symptoms with multiple root causes.

The four (possibly five) causes that the authors identified:

• “the ‘better than the Beatles’ problem;
• the ‘cautious regulator’ problem;
• the ‘throw money at it’ tendency; and the
• ‘basic research–brute force’ bias.

There may also be some contribution from a fifth factor, termed ‘the low-hanging fruit’ problem.”

Some of these are dog-bites-man stories. The excess of caution by the FDA is like the weather — everyone complains about it but nobody does anything about it. Even though they mostly dismiss it, the “low-hanging fruit” argument is also an old one.

Interestingly, they argue that “throw money” is easily reversed: if, in effect, big pharma R&D shops had become fat, dumb and happy (my term not theirs), a prescription of lean and mean can be applied with little impact on output. If true, this would certainly help reverse the trend.

It’s the two remaining ideas that to me seemed the novel ones.

Better than the Beatles

This was I thought the most clever argument — both in its identification, and in its framing:

Imagine how hard it would be to achieve commercial success with new pop songs if any new song had to be better than the Beatles, if the entire Beatles catalogue was available for free, and if people did not get bored with old Beatles records. We suggest something similar applies to the discovery and development of new drugs. Yesterday's blockbuster is today's generic. An ever-improving back catalogue of approved medicines increases the complexity of the development process for new drugs, and raises the evidential hurdles for approval, adoption and reimbursement. It deters R&D in some areas, crowds R&D activity into hard-to-treat diseases and reduces the economic value of as-yet-undiscovered drugs. The problem is progressive and intractable.

As they point out, content-based IP loses its novelty and so people seek out new things. After everyone saw Gone with the Wind, there was room for Casablanca, On The Waterfront, The Sound of Music, and a couple of Godfather movies. In resource based industries, the consumption of the old content (e.g. coal from a mine) makes the remaining holdings more valuable.

They point to a specific area — anti-ulcerants — as an example where drug discovery is held back by the back catalog. Two families of (now generic) solutions are already available, and while the third approach “would probably be safe and effective,” it is unlikely that any healthcare reimbursement system would pay for a new class of patented medicines, except for those rare cases not treated by the first two.

Their conclusion is even more depressing: “This general problem applies in diabetes, hypertension, cholesterol management and many other indications.” They also note that some of the decline of R&D productivity may be because pharma companies shifted from crowded (but high approval rate) therapeutic areas to less crowded (but lower approval rate) areas.

Basic Research/Brute Force

In this argument, the authors contend that the whole attempt to solve health problems through basic research and a molecular understanding of disease has been a disappointment, if not an outright failure. In addition to the molecular basis — and brute force analysis attempting to find new molecules — is also the assumption that the best way to solve a disease is to make a single molecule that binds to a single target.

Based on his industry experience, Lowe shows great sympathy for this argument

This gets us back to a topic that's come up around here several times: whether the entire target-based molecular-biology-driven style of drug discovery (which has been the norm since roughly the early 1980s) has been a dead end. Personally, I tend to think of it in terms of hubris and nemesis. We convinced ourselves that were were smarter than we really were.

I lack both the science and the industry experience to assess the validity of this complaint, except to note that this critique is obviously not universally shared. It would be nice to think that this article would start an honest conversation, but my guess is that too much money (private, government, university) has been invested in molecular approaches for this idea to gain traction until the evidence is undeniable.

Conclusions

The issue of declining R&D efficacy is a huge one for the industry, and also for investors in pharma companies big and small. The authors list a number of reasons why they are optimistic in the next five years:

Flat to declining R&D costs, as well as a bolus of oncology drugs, more orphan drugs and 'biosimilars as BLAs', might put an end to Eroom's Law at an industry level. Whether this improves things enough to provide decent financial returns on the industry's R&D investment is a different question. Financial markets don't think so. Industry executives do.

Their final idea is for each big pharma company to create a Chief Dead Drug Officer (CDDO), who is incentivized to analyze the failure of R&D investments and report the board and the public about the reasons beyond Eroom’s Law. A provocative idea — but I’m not holding my breath.

References

Gary Pisano, Science Business: The Promise, the Reality, and the Future of Biotech, Boston: Harvard Business School Press, 2006.

Jack W. Scannell, Alex Blanckley, Helen Boldon & Brian Warrington, “Diagnosing the decline in pharmaceutical R&D efficiency,” Nature Reviews Drug Discovery 11 (March 2012): 191-200. Doi: 10.1038/nrd3681

New markets, new models for old pharma

One of the most knowledgeable healthcare economists, Scott Gottlieb (M.D.) of AEI has weighed in on big pharma’s dying business models. While no Pollyanna, Dr. Gottlieb finds the glass surprisingly half full — while (as he often does) identifying areas where the FDA is a barrier to innovation.

Writing at the WSJ (and the AEI website) Dr. Gottlieb summarizes the success and limitations of the old business model. The former was represented by Lipitor, which earned $13b/year at its peak. The latter “mostly involved screening millions of random compounds against a molecular target.”

However, he argues that big pharma is learning the proper lessons from the end of their traditional models.

Lesson number one is to use a biological (rather than scattershot chemical) approach:

They took advantage of new science that allowed drug discovery to become a much more targeted endeavor—focused around precise information about the molecular fingerprints of disease.

In this kind of "rational" discovery process, size matters less than precision. Drugs are designed around the individual biological receptor they're targeting rather than being screened out of a library of random compounds. Deploying bigger tools no longer counts as much as having concentrated expertise in the particular pathway that causes a disease.

The next lesson has to do with the organization of innovation:

[B]ig bureaucracies were hostile to the kind of risk-taking and scientific focus needed to do good research. At most companies, the majority of new drugs are discovered by a handful of scientists with repeated successes. There's something unanticipated in drug research that can't be industrialized.

In recent years, pharmaceutical companies started to carve up their sprawling research enterprises into smaller, more focused teams. The right size of a research team is now said to be 20-40 people. To get at new science early, drug makers rely on collaborations with academic research teams and licensing deals with smaller biotech outfits.

(We call the latter process “open innovation”).

The final point is one that been often remarked over the past few years. Drug companies shouldn’t try for incremental improvements for conditions that have been treated in the past 30 years, but instead try to address other conditions (like Alzheimer’s) which lack an effective theraphy.

As a success story for the new, improved big pharma, Dr. Gottlieb cites Pfizer bringing Crizotinib (a lung cancer therapy) to market in 6 years instead of 10-15. Still, he points to the role of the FDA as increasing the costs and delaying the availability of therapies, by increasing the length of trials and the cost per patient.

While various regulatory reforms have been proposed, “none has been meaningfully advanced.” He concludes:

Drug makers are renewing themselves by realizing that their research success wasn't tied to the scale of those endeavors, but their precision. Today, the continuation of these scientific advances depends on their regulators reaching a similar understanding.

Charging towards a bleak future

Many reporters, analysts and other observers over the past decade have remarked on how the traditional big pharma business model has been running out of steam. Proposed solutions have included buying biotech companies (as Roche did) and forming generic divisions (as has Sanofi). Still, from outside, the (in)actions of big pharma resemble the controlled flight into terrain of other IP companies.

Last week, two consultants from Booz & Company published their own analysis of the problems in the Booz house journal, strategy+business.

Alex Kandybin and Vessela Genova deduced the strategic choices of 10 major pharma companies (Abbott, AstraZeneca, Bayer, GlaxoSmithKline, Johnson & Johnson, Merck, Novartis, Pfizer, Roche, Sanofi) through their acquisitions and divestitures from 2004-2010. Nine of the 10 have bet on biologics, five on OTC, four on generics and one (Sanofi) on animal health.

They draw an analogy to the choices of the computer industry:

Most industries go through periods of both deterministic and stochastic development. For instance, the computer industry in the 1960s and ’70s had all the characteristics of a deterministic process. IBM, Burroughs, Cray, and others pursued similar strategies, selling giant data processing machines known as main- frames. The personal computer changed the dynamics of the industry, triggering a turbulent stochastic period. It became impossible to predict where the computer industry was going, and in the early 1980s the incumbent players’ strategies diverged significantly.

This is, alas, an inaccurate revisionist view of the industry: in the 1980s, it was quite clear that the PC was democratizing computers and that standardized microprocessors enabled market entry and reduced margins.

More importantly, the computer industry of the 1970s has significant a priori heterogeneity: it was not for nothing that people referred to IBM and the Seven Dwarfs (or IBM and the BUNCH). Cray was in a very narrow and dangerous niche diametrically opposed to commoditization trends and desperately dependent on Cold War spending.

One place where the authors clearly have it right is that big pharma is fleeing from the highest margins in the life sciences (and among the highest margins anywhere) towards average or sub-average margins. The operating margins of pharmaceuticals is 29%, vs. 12% for generics, 8% for services and 2% for drug wholesaling.

This is utterly consistent with the 15-year-old observations of Clay Christensen: lower cost solutions eventually supplant higher cost solutions, destroying margins. Or, as my former colleagues Jason Dedrick and Ken Kraemer showed in their 1998 book, IBM’s shift from hardware to services dramatically grew revenues but cut margins.

What to do? The authors offer fairly generic (i.e. undifferentiated) advice: firms should embrace change, consider multiple scenarios, and assess the firm’s unique capabilities.

In the end, the old model of one-size-fits-all drug is breaking down. What will replace it? One prediction is personalized, genomic-based medicine. But even if that’s true, many uncertainties remain, including how quickly that future will get here and which part of the value chain will be the most unique and thus valuable.

Saving lives through data mining

At the Open Science Summit (@OpenScienceSum) this weekend, a major theme was applying the lessons of open source software (and other means of social production) to human health and drug discovery. WIthin this, my interest was (as in OSS a decade ago) new firms were organizing to take advantage of these opportunities.

Several of the sessions focused on creating (or simulating) a data commons in which the collective findings of dozens (or thousands) of researchers could be aggregated to allow researchers to spot patterns that would be otherwise unobservable in a single trial.

One example of that was Marty Tenenbaum, founder of Cancer Commons (cancercommons.org), which as its name suggests is seeking to be the clearinghouse for cancer data. Some of this is published research, but the goal is gather and share its own data, as the mission statement articulates:

Cancer Commons is a new patient-centric paradigm for translational medicine, in which every patient receives personalized therapy based upon the best available science, and researchers continuously test and refine their models of cancer biology and therapeutics based on the resulting clinical responses.

Representatives of two for-profit companies talked about their efforts to aggregate the terrabytes (or petabytes) of medical data to allow researchers to mine existing data for new insights.

NextBio aggregates public genomic data and allows (paid) customers to combine that with proprietary data to identify possible relationships. The freemium model gets academic researchers started for free — presumably so that grad students get hooked as grad students and take that to their new employers.

DNAnexus is preparing for the shift from a world with one sequenced human genome to millions, and the vast amounts of storage (and computing power) that will be necessary in only a few years to handle this avalanche of data. Earlier this month it teamed with Google (one of its investors) to host the Sequence Read Archive on the cloud for all researchers to have access.

Meanwhile, the consumer genomics pioneer 23 and Me has both genotype and phenotype data for 125,000 people: while the scale is much smaller, the uniformity is greater. One audience member called on the company to use the data for medicinal research, given the notorious lack of comparability for most phenotype data.

Data mining for medical research reminds me of data mining for social science research. Spending money to gather data is normally an entry barrier in the social sciences, so if there’s a free source of data (patents, open source repositories) this attracts a flood of new researchers and studies.

So if researchers don’t have to gather their own data to hypothesize — or evaluate — potential treatments, this is going to drastically cut the cost of entry into pharmaceutical research. We’ve seen this story before — again in open source — there will be a few big winners that move quickly and decisively to exploit this new technology, while other incumbents will find their traditional business models fail as barriers to entry crumble.

Of course, there are other entry barriers for pharma besides research — trials, funding, distribution, brand. Pressure will come on these fronts too.

An executive of Celtic Therapeutics talked about approaches to cut the cost of clinical trials,by using crowdsourcing for the design of trials and telemedicine for more efficiently processing patients. He claimed the costs of trials could be cut from $15-20 million to $1-2 million. Normally I’d take such dramatic improvements with a pound of salt, but the company claims as an advisor Karim Lakhani — one of the world’s leading crowdsourcing experts (and a personal friend).

At what cost diversification?

In teaching about diversification strategies, one of the main theoretical arguments supporting diversification comes when firms provide financing for a portfolio of bets that can’t be separately financed through the stock market.

In his latest posting to “In the Pipeline,” Derek Lowe notes that such diversification has also been one of the historic strengths of Big Pharma. Quoting a blog comment by (longtime Lilly executive) Bernard Munos:

(Arthur) De Vany has shown that the movie industry has developed clever tools (e.g., adaptive contracts) to deal with (portfolio uncertainty). That may come to pharma too, and in fact he is working on creating such tools. In the meantime, one can build on the work of Frank Scherer at Harvard, and Dietmar Harhoff. (Andrew Lo at MIT is also working on this). Using simulations, they have shown that traditional portfolio management (as practiced in pharma) does achieve a degree of risk mitigation, but far too little to be effective. In other words, because of the extremely skewed probability distributions in our industry, the residual variance, after you've done portfolio management, is large enough to put you out of business if you hit a dry spell.

Both Munos and Lowe ask if Pfizer — the largest drug company in the world — doesn’t have a big enough portfolio to diversify against patent cliffs, who does?

In its forthcoming August 22 issue, Forbes is also running an article on the ideas of Munos to reduce R&D and outsource innovation. It also quotes criticism from former Pfizer R&D head John LaMattina, who thinks the death of the pipeline is greatly exaggerated.

Despite LaMattina’s criticism, I find persuasive the argument by Munos — published two years ago in Nature Reviews Drug Discovery — that the current model is running out of steam.

I don’t have enough experience with pharma R&D enough to offer my own fixes, but my sense is that outsourcing inefficient search processes to external partners isn’t going to work. Perhaps startups will have better intuition as to where to look versus a systematic search by big pharma, but if the discovery paradigm is busted, outsourcing it won’t solve the problem. Instead, that would suggest we need a new search or discovery paradigm.

Still this suggests a new debate to bring to my KGI grad students next spring in the innovation management class. In particular, I’d like assign the 2000 article in Research Policy by Scherer and Harhoff (or the Scherer et al article in the Journal of Evolutionary Economics) if their earlier coursework prepares them for the math. (The RP seems more relevant, but the 2002 De Vany article on movie studios cites the JEE).

Wednesday I sent an e-mail to Harhoff, who I visited many years ago and was once dissertation advisor to a friend of mine. He wasn’t aware of the recent visibility of his earlier work, so I guess I’ll have to use the forward cites from Google Scholar to help students make the connection.

Is the death of blockbusters greatly exaggerated?

John LaMattina, former head of R&D for Pfizer, has a provocative post that argues that “The Death of the Blockbuster Has Been Greatly Exaggerated.”

He lists some contrary evidence including a recent WSJ article on potential blockbusters in the pipeline. And of course the title is a deliberate allusion to the famous Mark Twain line.

It’s hard to argue with his credentials and experience, and he makes some important points about the problems about why it’s hard to make predictions about the future. Some drugs will do worse than expected, and some will do better: Lipitor was predicted to peak at $0.8b/year but peaked at $13b/year.

He also argues that there are some major diseases left to be cured, and these cures will be lucrative by any measure.

I can't say I’m complete convinced. Even if such blockbusters remain, on the recent trajectory the cost of getting those blockbusters is getting higher — at some point firms may no longer try.

Perhaps the more serious problem is that the returns from the losers may also be lower, since incremental therapies are having trouble commanding a higher price than established off-patent drugs. (As an allergy sufferer, I can say that the generic versions of Claritin and Flonase are pretty good compared to what was available 20 years ago.)

So the glass is not entirely empty even if it’s not quite half full. The boom times of the old paradigm are clearly in the past. Perhaps some new paradigm (computational biology? personalized medicine?) will enjoy its own boom, but we haven’t seen it.

If pharma R&D is broken, what will replace it?

The last few weeks, I’ve been watching some disconcerting news that adds to existing concerns about funding pharma research.

At Forbes, Matthew Herper presented two graphs. One shows the number of approved new molecular entities (NMEs), which has wild swings (and an impressive peak in 1996) but no overall trend despite the huge upswing in pharma and biotech R&D.

The second graph is the one that (as Derek Lowe noted) is the depressing one:






Two parts are depressing. First, of course, that we’re getting fewer not more drugs per dollar as time goes on. But almost as depressing, the obstacles are not just the limits of science, but also the regulatory regime which at times can distort the curve (good or bad) with fairly arbitrary decisions.

If that’s not depressing enough, then there was this post (at the NY Post) by former Wyeth chemist Josh Bloom, which opens:

The folks at Scientific American have launched "1,000 Scientists in 1,000 Days" -- a program to bring together scientists, teachers and students to improve America's "dismal" showing among wealthy countries (27th out of 29) in graduating college students with degrees in science or engineering. I'm sure they mean well -- but, at least as it applies to the field of chemistry, "1,000 Unemployed Scientists Living With Their Parents at Age 35 While Working at the Gap" would be a better name.

In addition to the soaring costs of bringing a NME to market and the FDA, Bloom points to offshoring:

To trim expenses, companies began to outsource research to India and China. It started as a trickle, but soon became a tsunami, leaving many thousands of highly intelligent and well-trained professionals with nothing to do -- a shameful waste of talent.
My colleagues and I at Wyeth watched helplessly as one company after another shed employees in huge numbers -- 300,000 since 2000. When Pfizer -- facing the looming expiration of its Lipitor patent and a poor research pipeline -- bought Wyeth for its portfolio of products in 2009, it cut about 25,000 jobs, with more to come.

Now Ph.D. chemists can’t even get a job teaching high school chemistry — or at least in New Jersey, where they’re a dime a dozen.

Bloom seems to think that US drug discovery is not cyclical, but in fact in a permanent decline. Let me offer two historical analogies.

For now, IT seems to be cyclical. I remember after the end of the dot-com era, software engineers were also laid off as IT companies died while the survivors offshored most of the routine work to Bangalore. Now, experienced engineers are once again in high demand in Silicon Valley “as they benefit from a tech job market that hasn't been this overheated since the dot-com bubble in the late 1990s” (as the San Jose Mercury put it a week ago.)

When I was growing up in San Diego, the demand for aerospace engineers was cyclical — layoffs and shortages all throughout the 1960s and 1970s. The Reagan buildup brought it back one last time, but eventually the SoCal aerospace industry died after the end of the Cold War. The cycle has ended and the industry is gone.

What’s interesting is that (iPhone Apps notwithstanding) the sale of software products is also in a steady decline. The growth is mostly in software-as-a-service (SaaS) models like Google and Facebook and the rest of “The Cloud.”

People have been talking for years about a similar shift for medicine, from mass-producing little pills to providing personalized medicine. In 2004, IBM’s consulting arm (the former PWC) released a report entitled “Pharma 2010,” which envisioned such a shift. (Such complex service-oriented businesses would play to both IBM and PWC’s strengths). It didn’t happen by 2010, but that doesn’t mean it won’t happen.

So what is the future?

1. Less spending on R&D
2. Spending on R&D but mostly offshore
3. A shift from mass R&D to custom (bioinformatics-based) research

I don’t know what the future holds, but the employees (and shareholders) of the big pharma companies hope that their brands, distribution, manufacturing and (yes) R&D will give them a significant role to play in that future.

Where are the heroes of pharma research?

Since accepting my new position with Keck Graduate Institute and switching my industry orientation from ICT to the life sciences, I’ve been struck by the difference in how the two sectors are treated in popular culture — particularly drug discovery, whether “big pharma” or the scrappy biotech startups.

One metric is news articles. When I pick up my copy of Business Week every Friday morning, there will be three or four articles about the major computing, communications or other electronics companies, but many weeks nothing about pharma, big or small. Reading the Wall Street Journal or the New York Times (back before they started charging) suggests a similar pattern.

It’s even more pronounced with books. The weekend brought news of the death of Tom West, a longtime engineering manager for Data General. Most people have forgotten his name, but every computer engineer of my generation read about his success creating the Eclipse minicomputer in The Soul of a New Machine — the book that both created Tracey Kidder’s reputation and also the first to apply “new journalism” to study a complex engineering design problem.

I own at least three bookshelves of memoirs and histories about PC, software, Internet and other ICT companies. (I say at least three because some are in boxes and some are mixed in with other books). These books were written in the past 30 years — since the IPOs of Apple and Microsoft made these firms a household name — and include at least 10 Apple books and 5 each on IBM and HP. (Some of these books are on companies that were never consumer brands, including DEC is Dead, Long Live DEC about the once-great Digital Equipment Corp.)

However, when I went to find similar books about drug discovery companies in the San José public library, almost nothing was to be found. In fact, the only book held by more than one branch was Poison Pills: The Untold Story of the Vioxx Drug Scandal, which as the title suggests is a one-sided “exposé” about Merck & Vioxx.

Along the same lines, watching CNN in an airport waiting lounge — the only time I ever watch CNN — today I saw a similar breathless exposé about allegedly FDA lax approval processes. In this case, the guests were a couple of geriatric contributing editors of Vanity Fair, who wrote an article (“Deadly Medicine”) about Avandia, arguing that prescription drug deaths mean that the FDA standards are too lax.

From the TV interview was no evidence that either man knew anything about small molecule drug discovery, PPAR ligands, risk assessment or the steps necessary to gain a Ph.D. in any of the life sciences. In contrast, before my friend Randy Stross wrote Planet Google, he’d lived in Silicon Valley for decades and written six other books, including a biography of Steve Jobs.

One would reasonably assume this sort of media coverage impacts legislative and judicial decision, class-action lawsuits and even the career intentions of high school and college students. The folks that create privacy-invading breakthroughs like Google and Facebook are lionized while people who try to save lives are (often) demonized.

One notable exception was the 2010 movie “Extraordinary Measures,” based on the memoir of serial entrepreneur John Crowley about his efforts to find a cure for Pompe disease. As one promo put it:

An uplifting, inspirational drama, chronicles John Crowley, the man, who defied conventional wisdom and great odds, and risked his family's future to pursue a cure for his children's life-threatening disease.

Of course, any success of the movie is due more to Brendan Fraser in the lead role with two cute ailing kids, and of course Harrison Ford playing a cross between Harrison Ford and Scottie (“I’m givin’ it all aye got, captain!”).

So are the companies that produce therapeutics and diagnostics as evil as one would think watching CNN or browsing books in the personal health section of Barnes & Noble? If not, what can be done about it?

Lost in translation

Any big pharma CEO or shareholder would love to see a stronger link from basic research to the bottom line. But it’s not something that can be achieved via administrative fiat.

Writing at the WSJ Health Blog, Shirley Wang reports

When former NIH head Elias Zerhouni ran the $30 billion federal research institute, he pushed for so-called translational research in which findings from basic lab research would be used to develop medicines and other applications that would help patients directly.

Now the head of R&D at French drug maker Sanofi, Zerhouni says that such “bench to bedside” research is more difficult than he thought.

When Dr. Zerhouni tried to make Sanofi more like a nimble small biotech, he found it didn’t work because (as the WSJ put it) “small biotechs are no more successful than large drug makers at coming up with new drugs.”

Blogger Derek Lowe remarks that Zerhouni “was, in all likelihood, living in sort of a bubble at NIH.” And, Lowe suggests, knowledge of how to fix this is more likely to be found in an industry veteran than a government refugee.

On the one hand, Lowe notes that current NIH head Francis Collins wants to create a new NIH translational research institute. On the other hand, Wang cites its earlier posting, in which the ex-head of Merck said NIH is not well suited for translational research and instead should stick to basic science.

BTW, Lowe and Wang don’t seem to understand “open innovation” is or why Zerhouni is recommending it.

Defining open innovation is much easier than fixing the difficulties of pharma commercialization. A definition can easily be found with Google and there now are a large body of research, an online community and blog on the topic.

Spending on millions on lottery tickets

The recent difficulties of big pharma creating new blockbuster drugs are well known in the industry. Every time they bring a compound to trials, they’re placing a multimillion dollar bet that it will both be effective and also will make it through (increasingly difficult) FDA approvals.

However, these same sort of gambles also apply far upstream, when a new molecule, new technique or even basic science is being pursued by university scientists.

In his blog (In the Pipeline), Derek Lowe today highlights an example of this from the 1980s — one that brought a Nobel Prize in physiology.

He quotes from the new book Adapt by Tim Harford, as summarized in a recent Slate article.

In 1980, Mario Capecchi applied for a grant from the U.S. National Institutes of Health. . .Capecchi described three separate projects. Two of them were solid stuff with a clear track record and a step-by-step account of the project deliverables. Success was almost assured.

The third project was wildly speculative. Capecchi was trying to show that it was possible to make a specific, targeted change to a gene in a mouse's DNA. It is hard to overstate how ambitious this was, especially back in 1980. . .The NIH decided that Capecchi's plans sounded like science fiction. They downgraded his application and strongly advised him to drop the speculative third project. However, they did agree to fund his application on the basis of the other two solid, results-oriented projects. . .

What did Capecchi do? He took the NIH's money, and, ignoring their admonitions, he poured almost all of it into his risky gene-targeting project. It was, he recalls, a big gamble. If he hadn't been able to show strong enough initial results in the three-to-five-year time scale demanded by the NIH, they would have cut off his funding. Without their seal of approval, he might have found it hard to get funding from elsewhere. His career would have been severely set back, his research assistants looking for other work. His laboratory might not have survived.

The problem is, he did exactly what NIH didn’t want him to do, and its process was designed to screen out. Again quoting Harford:

The NIH's expert-led, results-based, rational evaluation of projects is a sensible way to produce a steady stream of high-quality, can't-go-wrong scientific research. But it is exactly the wrong way to fund lottery-ticket projects that offer a small probability of a revolutionary breakthrough. It is a funding system designed to avoid risks—one that puts more emphasis on forestalling failure than achieving success.

Both Lowe and Harford praise the Howard Hughes Medical Institute — in other words a foundation that can take big risks with its own money.

Once upon a time, big rich R&D intensive companies existed to take these sort of risks. In one 20 year period, AT&T invented the mobile phone, the transistor, error correction codes, solar cells, transatlantic telephone cables, the laser and the communications satellite. While the reconstituted AT&T is now a duopoly, the old Bell Labs is long gone.

A long time ago, big pharma made the great breakthroughs. More recently, they let university professors create biotech startups and then bought them. But now such startups look like a bad bet for VCs, and a new financing model is badly needed.

We don’t know who will fund the next round of great breakthroughs. By their nature, we don’t know what they are or where they will come from. Perhaps if we’re lucky, the new tools for computational biology and chemistry will reduce the time and costs of making such discoveries — reducing the amount of money that scientists have to beg for and improve the odds of getting it.

A new way to fund drug discovery?

In the past decade, two failings of the modern drug discovery have become apparent. First, the major pharma companies are unable to develop new drugs as successful as the ones going off patent.

This has been a major theme of Derek Lowe in various posts in his blog “In the Pipeline”. For example, in a November 2010 posting, he lists the commonly know problems: “lower rates of success in discovery, higher costs, patent expirations, etc.”

The other is that the biotech startup model is beginning to fizzle out. Whether or not the current science will make commercially successful biologics, the idea of creating a new company to develop a few compounds — ending with a successful IPO — hasn’t worked for many years. (It doesn’t help that the IPO market has closed overall for bio and non-bio startups alike.)

Into the breach steps Duane Roth, onetime industry executive and entrepreneur who now heads San Diego-based Connect. Roth gave a webinar April 27 where he discussed the problem and some ideas of how to solve it.

Roth’s suggestions were based on a paper he did with former Warner Lambert executive Pedro Cuatrecasas that was sponsored by the Kauffman Foundation and summarized in Roth’s op-ed last year in Xconomy San Diego.

Roth was pessimistic about the prospects for both traditional pharma and the newer biotech model. He said that the IPO market “is never coming back for pre-revenue companies,” stretching out the return for startups and their investors. Today, only one major biotech startup remains independent — Amgen — while both Genentech and Genzyme have become subsidiaries of big pharma.

Instead of the traditional vertically integrated model, Roth argues that the industry needs to separate out a new role in the value chain: a product definition company. In this model, the federally funded research institute (i.e. a university) would license discoveries to the PDC, which would prescreen these for the product development companies (which might also have the distribution channels to bring these products to market).
In this division of labor, the product definition company spends $3-5 million per discovery to characterize it and attempt to develop a prototype. The drug development companies are the ones that do the clinical trial and hope to bring it to market.

The difference in this new model is that the PDC quickly studies the possible compounds and then sells them. There are no 10 year waits for IPOs — quick exits for the winners and returning the compounds to the universities for the losers.

From his vantage point at Connect, he’s watched the mobile industry (led by Qualcomm) make a similar dis-integration between chip designers, contract fabricators and handset makers.

As a longtime strategy professor, I can think of few examples where industries naturally changed themselves and many where industry inertia prevent long-needed reforms. (Think of the record industry.) Thus my question to Roth was: how do we get there from here?

Roth thinks big pharma will invest in and support such companies. He also hopes that angels could fund some companies, given the relatively small capital needs and quick returns.

I feel more pessimistic. This sort of model requires creating a market — buyers, sellers, price and quality measures — and it seems like today’s market is spotty at best. I can imagine that CROs could try to do this product definition, but be unable to get a fair price from potential buyers. Or they might get to greedy, hoping for a later sale at a higher price rather than handing off the compound to a firm that can bear the cost and risk of clinical trials.

Still, there’s no denying the cracks in the current model. During its heyday, US firms were the winners in both big pharma and most notably for biotech startups. The industry — and universities and the country — need to find a 21st century model that will take advantage of our home court advantages: university science, risk capital, and a large affluent domestic market.

Webinar: fixing biopharma business models

Connect — the UCSD-affiliated entrepreneurship nonprofit — is hosting a free webinar on April 27 about fixing the business models of both biotech startups and integrated pharma companies.

The advertised topics are

• What killed the fully integrated model?
• Has the start-up model run its course?
• What will the next model look like?
• Where will future investments come from?
• What role should federal and state government play?

The speaker is Duane Roth, founder of Alliance Pharmaceutical, former manager at J&J and Wyeth and now head of Connect.

These are crucial questions for the pharmaceutical industry and associated firms across the value chain. I’ve already signed up.