When AstraZeneca revealed plans last month to step away from internal neuroscience R&D, the European College of Neuropsychopharmacology released a statement expressing alarm that yet another big pharma firm was pulling out of the field. “There is a growing sense that neuroscience in Europe is now facing a severe crisis,” the group said.
April 2009 Sanofi overhauls its pipeline, ending development of more than a dozen drugs and vaccines, including an antidepressant. The following year, the company cuts hundreds of sales jobs, citing its reduced activities in central nervous system (CNS) research.
January 2010 GlaxoSmithKline ends neuroscience drug R&D in Harlow, England, and abandons some areas of neuroscience research, including depression and pain.
July 2010 A year after its acquisition of Schering-Plough, Merck & Co. closes a neuroscience research lab in the U.K. and consolidates activities at two sites in the U.S.
December 2011 Novartis announces the closure of its neuroscience R&D site in Basel, Switzerland, and the formation of a smaller group in Cambridge, Mass., to study the genetics of certain CNS disorders.
February 2012 AstraZeneca announces it will lay off some 2,200 scientists, mostly in neuroscience R&D. The firm is closing a research lab in Montreal and ending R&D in Södertälje, Sweden.
AstraZeneca had joined its competitors Novartis and GlaxoSmithKline in significantly cutting the number of scientists working on diseases that affect the central nervous system (CNS). Others, including Merck & Co., have consolidated their efforts, leading to the perception of a mass exodus of big pharma from neuroscience drug discovery.
Not everyone agrees that the field is in crisis. But after years of expensive discovery and development campaigns and high-profile drug failures, big pharma companies are unquestionably reconsidering how they do neuroscience R&D. For companies like Pfizer that are sticking with internal research, the goal is to change the approach by using the tools and methodology that have made oncology R&D so successful.
Other big companies say they will rely on the outside world—academic labs and biotech firms—to fill their pipelines until their understanding of the science is better. But some industry watchers worry that without a solid internal commitment from big pharma, the health of the neuroscience drug discovery ecosystem is at risk.
The business argument against neuroscience R&D is straightforward: The rate of attrition—the percentage of compounds that fail to make it past Phase II or III clinical trials—is simply too high for many companies to justify risking research dollars in the field. The situation is underscored by high-profile failures like Eli Lilly & Co.’s semagacestat, an Alzheimer’s disease treatment that was pulled after two Phase III trials showed that cognition actually worsened among patients taking the drug.
The statistics are grim, says Bernard Munos, founder of InnoThink Center for Research in Biomedical Innovation. Munos, who previously served as a corporate adviser to Lilly, points out that for Alzheimer’s disease alone, companies have in the past 12 years poured $18 billion into testing 211 drug candidates; 95 of those have already been abandoned.
According to a recent study by the Tufts Center for the Study of Drug Development, just one in 10 CNS drugs that began clinical trials between 1993 and 2004 received Food & Drug Administration approval; for other classes of drugs, the rate was one in six. And getting to market takes longer for a CNS drug than for drugs in other therapeutic areas. Between 1996 and 2010, the mean time for clinical development and regulatory approval of a CNS drug was 32 months, 35% longer than for other drug types, the Tufts Center found.
Meanwhile, health insurers will not necessarily reimburse the cost of new drugs. For example, drug company executives argue that even if a new drug to treat depression is more effective than selective serotonin reuptake inhibitors (SSRIs) such as Prozac and Paxil, it might not be covered by insurance because so many reasonably effective generic alternatives exist.
“I think it’d be very hard to come up with a new Prozac,” says Richard F. Pops, chief executive officer of Alkermes, a biotech firm developing drugs for pain and depression. “Because all the SSRIs are off patent, the medicine is essentially free. Payers are not going to pay for a new Prozac.”
Most drug developers agree that it can be hard to make a business case for neuroscience R&D. However, opinions vary as to why the science has failed. R&D leaders who have stopped or reduced internal neuroscience activities cite several distinct hurdles to inventing new neuroscience drugs.
On the research side, the knowledge of the underlying biology is insufficient to generate reliable drug targets, according to Moncef Slaoui, head of R&D at GlaxoSmithKline. “The central nervous system is still a black box,” Slaoui says. The “interplay between pathways is enormously more complex than in other biological systems.” As a result, virtually no research efforts exist in areas such as stroke, despite a serious need for treatments.
In addition, inherent limitations exist in discovering drugs that act on the central nervous system: “Because in CNS you’re generally constricted to small, hydrophobic chemicals that get into the brain, it’s a relatively small universe of compounds,” notes Steven Marc Paul, head of the Helen & Robert Appel Institute for Alzheimer’s Research at Weill Cornell Medical College. Paul, a former head of R&D at Lilly, shepherded a number of CNS drugs through the pipeline.
On the development side, the lack of fundamental biological knowledge makes testing drugs difficult. Anxiety, for example, is hard to define both biologically and clinically. “Clinical end points are so poorly defined and so irreproducible,” Slaoui says. An individual’s sense of well-being “is deeply personal” and difficult to measure objectively.
Others believe the problem with CNS drugs is not one of science but of methodology. “Historically, in CNS, the successful drugs were discovered by accident,” says Michael Ehlers, chief scientific officer for neuroscience research at Pfizer. Only after stumbling on efficacy in humans would scientists put the drugs back into animal models to try to unravel how they work.
Furthermore, many of those drugs “were tested without knowing if they even got into the brain,” notes Darryle D. Schoepp, head of neuroscience and ophthalmology at Merck.
That approach “is not a great way to discover new mechanisms, because it is not a model of disease,” says Ehlers, who was recruited by Pfizer from Duke University a year and a half ago to shake up its neuroscience activities. “The message across the field needs to be ‘that can’t continue.’ When I look out and evaluate smaller companies or programs, they’re kind of stuck in that mind-set. That’s not a failure of science; that’s a failure of scientists to respond to biology.”
R&D executives at companies such as Pfizer and Merck that remain committed to neuroscience drug discovery say the key is revamping the way research is conducted. For starters, scientists need to rethink how CNS diseases are defined.
“In neuroscience, I think some companies might have gotten caught in the trap of thinking in classic diagnostic categories,” Ehlers says. The field needs to adopt the approach that has transformed oncology R&D, he says. “We used to talk about leukemia and cancers in terms of tissue of origin and their histopathology,” he observes. Today, whether a tumor is located in the liver or the breast is less relevant than knowing its molecular genetic profile.
The shift to better stratifying CNS diseases, Ehlers says, has been stymied by an inability to uncover the genetic clues that provide starting points for drug discovery and later help find the right patients to enroll in clinical trials.
He believes the bottleneck is starting to open up. To take advantage of advances, he has overhauled how Pfizer conducts neuroscience research.
In the past, big pharma often gave its scientists a mandate to work in areas such as Alzheimer’s or schizophrenia, regardless of tractable drug targets. Now at Pfizer, Ehlers says, his team is “indication agnostic.” Any program that Pfizer undertakes must have a critical mass of biological knowledge—for example, human genetics, human phenotyping, and evidence of dysfunctional neurocircuits—to convince Ehlers it’s worth pursuing. “We start there,” he says. “That hasn’t always been the case.”
Moreover, Pfizer no longer relies on mouse models as predictors for responses in humans. “We’ve for the most part stopped all rodent behavior as a model for disease and are much more about what’s happening in the brain,” he says. Scientists measure human responses to prove experimentally that a drug works.
Pfizer’s goal, according to Ehlers, is to tackle fewer projects but have more confidence in their potential for success. The result should be a drug pipeline “rooted in something more than optimism.”
Other firms with active internal neuroscience programs have also adopted new research strategies. “There are ways of working through this smarter to try to increase the chances of success,” Merck’s Schoepp says. For example, scientists now deploy positron emission tomography (PET) ligands to determine whether a drug is reaching the brain and interacting with the receptor of interest. Not knowing whether a drug reached its target was a fundamental problem with earlier drug discovery.
Drug companies that have cut back on internal neuroscience R&D are taking a variety of tacks to stay in the game. For example, AstraZeneca says it is not abandoning the space but rather changing how it works there. “Continuing to spend and invest our money the way we had been wasn’t the best way of working,” says Menelas N. Pangalos, the company’s executive vice president of innovative medicines.
AstraZeneca has created a virtual neuroscience unit comprising 40 to 50 scientists who will partner with academic and industry researchers on drug discovery and development. “The idea is that through collaborations, contract research organizations, and using AstraZeneca resources when they need to, they will be able to basically build a research pipeline and take it all the way through Phase IIb,” Pangalos says.
The goal is to shift from just following the science coming out of academic labs to actively engaging with the academic and biotech organizations making fundamental discoveries, Pangalos says. “Rather than reproducing everything in our own labs, we can go work with those academics” and design a plan to move ideas forward, he adds. “We will share risk and share reward.”
The AstraZeneca scientists tasked with forging external collaborations will not be encumbered with the bureaucracy that strangled partnerships in the past, Pangalos maintains. “The group is going to be fully empowered with senior, experienced people to make decisions on the spot,” he says.
Many R&D executives argue for an overhaul of how neuroscience research is conducted, but not everyone is convinced that individual efforts are the fastest way to find reliable new targets. GSK’s Slaoui believes the gaps in understanding many CNS diseases are simply too large for any one company to make a serious impact. That’s why GSK exited research in depression and pain two years ago.
Slaoui believes real progress will be made when companies join hands to unravel the basic biology behind CNS disorders. For a disease such as Alzheimer’s, for example, “we can only afford to ask reasonably short-term questions,” he says.
Today’s clinical trials test a drug when the disease has already manifested itself in day-to-day life, when it has already inflicted major damage to the brain. To make a truly effective Alzheimer’s medicine, Slaoui says, the industry needs to assemble a cohort of patients who can be followed for decades to track biomarkers, the impact of medicines, and other metrics that can help companies design better drugs and better clinical trials.
No single company can afford to support a cohort of that size, and Slaoui believes precompetitive collaborations will be vital to progress in diseases such as Alzheimer’s.
InnoThink’s Munos agrees. “There is no doubt that when drug R&D is held up by big questions about our understanding of basic science, the sensible thing to do is for companies to join hands and collaborate to resolve these enigmas. It makes no sense for each company to do this on its own,” he says.
“If we look into our pipelines, probably 70 to 80% of the targets we work on are the same, so we actually duplicate target validation 10 times,” Slaoui points out. Although this duplication of effort might make sense in some areas, others are so high risk, “why don’t we put our efforts together—us and public institutions—and discover targets?” After all, he says, ample room for competition remains because every company will tackle new targets in a different way.
Precompetitive collaborations do exist in some areas. R&D executives cite the success of the Alzheimer’s Disease Neuroimaging Initiative, a long-term effort funded by government and industry to generate biomarkers and improve clinical trial design. Companies continue to conduct drug discovery for Alzheimer’s disease, in part because of the availability of new targets and in part because of the enormous commercial potential for a medicine that demonstrates even minor improvement in people with the disease.
Not everyone has embraced the idea of working together in ways that could truly accelerate drug development, Munos says. “Companies are still not sharing the results of failed trials, which would help some of them learn from the setbacks of others and avoid avenues that are unlikely to work,” he says. “It can become very dysfunctional.”
Another area of concern is the health of the biotech and academic institutions that generate the cutting-edge discoveries that seed the pharma pipeline. Companies such as AstraZeneca and Novartis say they will rely more heavily on academia and biotechs; industry observers wonder, however, whether drug firms will partner at an early enough stage, and with sufficient funds, to effectively translate basic science into drug candidates.
“I do think a lot of the innovation in CNS is going to come out of smaller companies,” says Weill Cornell’s Paul. He’s a founder of the recently launched Sage Therapeutics, which is dedicated to working on drugs for schizophrenia, depression, and traumatic brain injury. At the same time, Paul warns that big pharma’s hesitation to take risks on early-stage programs will also give pause to venture capital investors. “I am concerned about the health of the ecosystem and whether there will be adequate funding,” he adds.
Indeed, biotech and academic scientists see significant opportunity down the road, as well as short-term challenges. The shift away from internal research at big pharma “provides a real niche for groups like ours to fill,” says P. Jeffrey Conn, director of the Vanderbilt Center for Neuroscience Drug Discovery. “On the other hand, if pharma really shuts down CNS efforts in a broad way, there may be no market for what we can do.”
Anecdotally, Conn has seen signs of a divide: decreased interest in partnering at an early stage from some pharma companies and sudden enthusiasm for partnering by others.
Conn is “not convinced that the risk is as high as a lot of executives in pharma see it,” but the current climate is causing him to rethink Vanderbilt’s strategy. “When I started, I didn’t expect to do full drug discovery through candidate,” he says. “I thought our role would be to validate targets and work with pharma to do lead optimization.”
Now, largely out of necessity, he is considering taking compounds to the point of early tests in humans. “With the changes that have happened, there may not be an appetite in pharma for really new approaches if you don’t have direct clinical data sets,” Conn says.
Others are also concerned about the gap between the basic science and the stage in drug development where venture capitalists or a pharma company will step in with funding. Nonprofits, the National Institutes of Health, nongovernmental organizations, and private donors “are going to need to play a larger role in funding translational research,” says Mark Frasier, director of research programs at the Michael J. Fox Foundation, a nonprofit focused on finding new treatments for Parkinson’s disease.
“I think we’re seeing the goal line pushed back,” Frasier says. For several years, the Fox Foundation has funded biotechs to collect enough early-phase data to entice big pharma to invest in its programs. Now, companies seem to want more clinical studies before taking on the risk, Frasier says. “That just means there’s a larger role for organizations like ours to fund larger, more expensive translational research.”
Meanwhile, executives at the big firms that continue to conduct neuroscience R&D believe the opportunity for breakthroughs couldn’t be better. Companies that gave up may regret that decision, Pfizer’s Ehlers suggests. Their exits, he says, “are shortsighted as to where the actual, fundamental state of biology is in neuroscience.” As the underlying genetics of CNS diseases become clear, the knowledge “will become catalytic.” In 10 years, Pfizer’s choice to stay in neuroscience could seem prescient, he says. “Our competitors may wish they’d chosen differently.”