Inside External R&D At AstraZeneca

The words “big pharma” conjure up big research infrastructure. But one of the biggest drug companies, AstraZeneca, recently decided to move nearly every aspect of neuroscience R&D outside its own walls. The scientists charged with managing the new network of drug discovery partners say the approach has brought swift access to breaking science, the freedom to rethink the fundamentals of drug discovery, and, most important, relief from the risks posed by heavy internal investment.

The virtual neuroscience unit, known in-house as the neuroscience innovative medicines unit, or neuroscience iMed, was born in February 2012, when AstraZeneca unleashed massive job cuts across its R&D operations. It wielded the heaviest ax against its neuroscience unit, closing a research facility in Montreal and ending R&D in Södertälje, Sweden.

The moves wiped out internal neuroscience research in favor of a strategy of building a new drug pipeline through partnerships. The neuroscience research staff was whittled down from 600 people to roughly 45 scientists tasked with establishing and maintaining a network of collaborators.

AstraZeneca’s overhaul coincided with neuroscience cutbacks at several other big drug firms, notably Novartis and Glaxo­SmithKline, that were prompted in part by the failure of drugs in late-stage studies.

“Our traditional internal model was not working,” recalls Menelas Pangalos, AstraZeneca’s head of small-molecule research and early-development activities. Unraveling the science of central nervous system diseases is one of the riskier pursuits in drug research. At the same time, “the most exciting scientific advances in this area were happening outside our own walls,” he adds.

The virtual approach helps the company lower investment, “while allowing our neuroscience group the flexibility to work with the best neuroscientists around the world,” Pangalos says.

But unlike a virtual biotech firm, whose staff generally operates out of a briefcase, the neuroscience iMed team can tap into AstraZeneca’s vast internal infrastructure to offer collaborators everything from high-throughput screening to proprietary compounds. And whereas a biotech firm must focus its energy on getting one compound to the finish line, the neuroscience iMed team is maintaining a portfolio of 26 projects, notes Samantha Budd, who coleads translational science for the unit.

In the 18 months since the neuroscience overhaul, AstraZeneca has assembled a team that primarily works out of an open-plan office in the company’s Cambridge, Mass., facility. The group is a mix of scientists who have spent most of their careers at AstraZeneca and new hires, many of whom worked with Pangalos when he headed neuroscience at Pfizer.

The team’s first partnerships reflect its deep contacts in the neuroscience community. For example, its most recently announced deal, a three-year pact with Tufts University, is rooted in a 20-year friendship between Pangalos and Tufts neuroscientist Stephen J. Moss. Similarly, scientists at Vanderbilt University, a partner since January, quickly found common ground with AstraZeneca’s neuroscience team since several key players on each side had earlier worked together at Merck & Co.

That common ground is essential, Astra­Zeneca executives say. Although the lab work is done outside the firm, AstraZeneca is deeply involved in guiding projects. In the case of the Tufts collaboration, for instance, two AstraZeneca scientists will oversee the seven to 10 postdocs being hired.

“We want academic partners who feel comfortable almost being like a seamless extension of our team,” says John Dunlop, a vice president of the neuroscience iMed. In fact, the unit has its own information technology infrastructure that sits outside the AstraZeneca firewall to provide full control over how collaborations are built.

For career-long AstraZeneca scientists who find themselves on the small, trailblazing team, the difference between the old and new ways of approaching R&D is palpable. As Budd puts it, “It feels a lot like I work in a different company.”

Breaking into new science is easier under the leaner model, Budd says. She recalls that several years ago when intriguing discoveries were made around tau, a protein implicated in Alzheimer’s disease, she explored starting up a research team in AstraZeneca’s neurology group. Her excitement was tempered when the neurology team leader began to tick off a daunting list of things needed to get a project going, such as hiring tau scientists and creating a mouse model.

“Today, we don’t say, ‘How do we build the capability in tau?’ ” Budd says. “We say, ‘Who’s doing it?’ ”

And the approach to science is less rigid than before. “We were very fixed in disease areas,” Budd says. Now, scientists ask questions like “What is pain?” instead of trying to fit a target to a therapeutic area.

AstraZeneca scientists say they have already seen the impact of letting biology guide their drug discovery efforts. In January, the neuroscience iMed teamed with Vanderbilt to develop compounds acting on the M4 muscarinic acetylcholine receptor to treat symptoms caused by diseases such as Alzheimer’s and schizophrenia. But after six months, the partners have learned their compounds may be relevant in disease areas not covered by the original pact—and outside of Vanderbilt’s area of expertise.

AstraZeneca has been “very nimble” about linking the Vanderbilt scientists to groups with the appropriate animal models to get studies up and running, says P. Jeffrey Conn, director of the Vanderbilt Center for Neuroscience Drug Discovery. Attempts to go beyond the boundaries of the agreement might have hit a wall in a more traditional pharma-academia partnership, he notes.

Collaborators say the neuroscience iMed has been quick to directly connect partners, be it encouraging links between academic labs with complementary expertise or introducing a partner to a research service provider. Thanks to a link made by AstraZeneca, Conn says, his team is working with a group in Sweden that needs biomarker expertise.

Less than 18 months into the neuroscience iMed, the culture of the neuroscience unit is bleeding into other parts of AstraZeneca, Pangalos says. The streamlined unit has been making decisions about discovery programs more quickly than its traditional counterparts, he notes. And although the company has no plans to replicate this model in other therapeutic areas, the group’s approach to R&D is having an impact.

“They’re helping to transform some of our thinking about clinical trials and early development,” he says. “It’s changing our mind-set about how we get work done.”