As Pfizer scientist Irina Apostolou heads out to visit the Boston labs of her academic collaborator, renowned Harvard Medical School systems biologist Lewis Cantley, her colleague Robert Martinez will occasionally rib her about the long journey ahead.
After all, if Martinez wants to visit Harvard neurobiologist Xi He, with whom he is trying to develop an osteoporosis drug, he only has to leave his 18th-floor labs and ride an elevator down to the 12th floor. Apostolou has a comparatively hellish commute: to reach Cantley’s fourth-floor labs she must get off on the 11th floor, walk some dozen or so feet across an elevator bank, and switch to another elevator. Apparently, the wait for that second ride makes all the difference when there’s pressing science to discuss.
Martinez and Apostolou lead research teams within Pfizer’s Centers for Therapeutic Innovation (CTI), a unit of the big pharmaceutical company dedicated to developing drug candidates through collaborations with academic scientists. Pfizer is banking on the immediate access afforded by labs adjacent to university researchers to enable closer, more efficient, and more fruitful partnerships.
More broadly, CTI is Pfizer’s experiment in building a different R&D ecosystem, says Anthony Coyle, chief scientific officer of CTI. Drug companies have tried academic alliances in the past—and not always successfully—but Coyle is counting on CTI’s more intimate approach. He says it can bring a deeper knowledge of basic science into the discovery process, allowing the company to make informed choices about the molecules it moves into the clinic, be smart about clinical trial design, and do it all in a shorter period of time.
Pfizer has set ambitious goals for CTI, which is focused on developing biologic drugs. By 2015, the company expects to have data from the first human efficacy studies of biotherapeutics discovered in CTI labs. And it believes its approach will save money by allowing fewer failures along the way.
CTI launched in November 2010 with a broad collaboration with the University of California, San Francisco. Since then, Pfizer has established sites in New York City, San Diego, and Boston, each linked to up to eight academic institutions. At each location, the firm has opened labs near its partners to facilitate scientific interaction.
Currently, Pfizer has 80 of its own scientists working in 23 project teams across the four locations, Coyle says. The teams generally include three or four Pfizer scientists and three or four academic researchers led by a principal investigator (PI). As drug candidates move closer to the clinic, scientists from Pfizer’s larger R&D organization are brought onto the project.
The PI’s lab receives financial support for the postdoctoral students assigned to the team. But the university doesn’t collect any other money unless drugs developed through the collaboration enter the clinic, when milestone payments start to kick in.
As the experience of Martinez and Apostolou attests, Pfizer has made an effort to locate its labs in each city’s science hub. Although not all partners are in the same building, most are within a few blocks. Looking out the floor-to-ceiling lobby windows of Pfizer’s Boston labs, a visitor can spot Beth Israel Deaconess Medical Center, Boston Children’s Hospital, and Dana-Farber Cancer Institute. Competitors are within view, as well: The rooftops of Merck Research Laboratories are a stone’s throw from Pfizer’s top-floor digs.
Pfizer and academic scientists appear to work side by side on a weekly, if not daily, basis, trading reagents, sharing equipment, and exchanging data and knowledge. “What’s so cool is they have badge access here,” Martinez says of his academic collaborators. “That proximity really fosters a quick exchange and ease of interaction.”
The same is true at Pfizer’s other CTI sites. “The CTI guys have UCSF IDs, and the UCSF guys have Pfizer IDs,” Coyle says. The access reinforces the notion “that this is a joint project team and we’re doing it together.”
That’s not to say the academic and industry scientists immediately formed one big happy family. Now in its second year, CTI still faces a learning curve. “Clearly we’re starting a group from scratch,” Coyle says. “There is a runway.”
As might be expected when trying to meld two cultures—one corporate, the other academic—Pfizer has had to figure out the right ingredients for a mutually beneficial collaboration. The first step was devising a general memorandum of understanding with each institution that laid out intellectual property rights and other terms.
The next step was finding the right PIs within the partner institutions. Pfizer put out a call for proposals at each site and has so far reviewed 400 ideas from academic scientists.
As the company winnowed the ideas down to the 23 projects under way, “we’ve learned that CTI is a model intended for individuals who are like-minded, who share a common aspiration to see basic science being translated,” Coyle says. “That is not for everybody.”
Like-minded can mean a lot of different things. Some of CTI’s academic collaborators, such as Cantley, see the Pfizer partnership as an outlet for ideas they don’t have the time or capacity to develop quickly. Although Cantley has successfully started biotech companies, he says the collaboration with Pfizer makes sense for one particular drug target discovered in his lab. Cantley’s lab has experience in developing small molecules, and this target called for a biologic drug.
“There’s no reason for me to re-create the wheel,” he says. “This to me seemed logical.”
Other collaborators, including emeritus Harvard scientist Harold F. Dvorak, want to see their ideas turned into drugs but just haven’t been bitten by the entrepreneurial bug. “If I’d have been 30 years younger, I might have started my own company,” Dvorak says. “But even then, I’m probably not the type of person who would do that. Pfizer brings to the table all sorts of things that I could not do by myself and even if I had founded a biotech company would be difficult.”
Dvorak says his lab had been “creeping along” at trying to develop an antibody against a particular cancer target. Pfizer stepped in and offered multiple ways of making it. “They have facilities that really nobody in academia has,” he adds.
The first year of CTI was largely devoted to setting up and populating the Pfizer labs with scientists, sealing contracts with academic partners, and reviewing and selecting projects. The second year has been about getting projects up and running as swiftly as possible. “By year three, we have promised the organization that we would start to deliver four first-in-patient studies per year across multiple therapeutic areas,” Coyle says.
As Pfizer gets deeper into its CTI experiment, the components needed for a successful collaboration are becoming obvious. In addition to connecting with academic scientists who want to see their discoveries translated into medicines, the company seems to have sorted out the mechanics of working together efficiently. “It’s been really important for all of us to articulate very clearly and to make sure the objectives and incentives are aligned,” Coyle says.
The collaborative process begins on day one of each project, when Pfizer and academic scientists create a “statement of work”—a detailed document laying out responsibilities, goals, and timelines.
To develop his timeline, Cantley says, he and his Pfizer partners worked backward from the clinical trial they envisioned. That meant defining what disease would be pursued—in his case, the target is in oncology—whether biomarkers could be used to define a patient population, and what tissue samples might need to be gathered in the study. From there, the scientists devised a step-by-step strategy for developing a drug that fits the clinical profile.
For some of the postdoctoral students, working with Pfizer was a wake-up call. “It’s teaching people in my laboratory who are involved in this collaboration a very different way of doing science than is typical in academic research,” Cantley says.
Academic scientists focus on making discoveries—understanding a pathway or how a cell works, for example—that might lead to a drug target, he explains. But that work is not done on the timeline that industry demands. “We virtually never have a milestone where this month this has to get done, this month that has to get done, and at the end of six months we make a ‘go or no go’ decision,” Cantley notes. The academic researchers were initially “kind of surprised” at the pace, he says with a chuckle.
“We do cutting-edge science, and it’s very focused,” Apostolou confirms. “They had to adjust a little bit.” But she adds that the academics have adapted quickly to how things are done in pharma and stresses that the industry and academic scientists complement one another.
Pfizer scientists are on their own learning curve. Martinez, who worked for years in Pfizer’s traditional R&D labs before moving to CTI, points out several key differences between how CTI teams operate and how the company’s traditional research labs are run. Because each step of the drug discovery process must be transparent for Pfizer and its academic partner, the detailed collaboration plan is a departure from the norm, Martinez says.
The CTI setup also offers unexpected opportunities for expanding the knowledge of Pfizer’s own scientists. Whereas the company’s traditional labs organize groups by therapeutic area, CTI houses projects spanning multiple disease areas. “There’s a lot of cross-learning going on,” Martinez says about his osteoporosis drug development work. “I’m seeing techniques that people are using in oncology that we are now going to start using.”
This is not the first time Pfizer has experimented with getting closer to the source of basic discoveries. The company has formed dozens of academic alliances, but they lacked CTI’s sense of urgency and teamwork, Martinez says. “This is really at the very beginning a very interactive event.”
Pfizer’s desire to improve its interactions with academia is shared by its big pharma competitors, although each company is taking a different approach to tapping university science. “Just about all the companies’ internal research budgets are getting slashed,” says Michael Latwis, an analyst with the health care consultancy Decision Resources. “They need to be more productive with fewer resources, so they’re trying to latch onto external research wherever possible, especially at the discovery stage.”
In the past several years, companies have shifted from funding single projects by individual professors to broader collaborations around specific therapeutic areas, often spanning several labs at a particular institution. More recently, big pharma has pushed the envelope further, bringing together researchers from multiple academic centers and opening labs on or near a campus.
Merck & Co., for example, has invested in the California Institute for Biomedical Research, a San Diego-based nonprofit devoted to turning ideas from academic scientists into drug candidates.
AstraZeneca, meanwhile, is relying entirely on external sources, be it academic labs or biotech companies, to develop its early-stage neuroscience pipeline. In one of its first pacts, AstraZeneca’s “virtual” neuroscience R&D team established the A5 alliance, which involves four academic research centers working to understand apolipoprotein E4 genotype biology with the goal of discovering better drug targets for Alzheimer’s disease.
But most industry observers consider CTI to be the grandest of the academic experiments. Among the four sites, Pfizer has committed up to $335 million over its first five years.
CTI fills a niche that had previously been ignored, says Noubar Afeyan, managing partner of the Cambridge, Mass.-based venture capital firm Flagship Ventures. “If you look out at the entire landscape of starting points for commercialization coming out of biology, some fit nicely into and have the credentials to get venture capital, and some fit into pure NIH funding,” Afeyan says.
The National Institutes of Health funds basic science, whereas venture capitalists generally look for a technology or group of assets that can be the basis of an entire company. But in between are single targets or molecules that lack a good vehicle for being pulled out of academia. They are the sweet spot for Pfizer’s CTI. “I really see it as a net positive to the biotech ecosystem,” Afeyan says.
As Pfizer starts in the next 18 months to put drug candidates from CTI labs into Phase I human safety studies, the rest of the drug industry will be watching. The real test will come in CTI’s fifth year, when Coyle expects to start proof-of-concept studies for some molecules. Then, the company will see whether it truly has a better understanding of a drug’s target early on in a development program and whether it “can reduce the attrition rate from Phase I to the end of Phase II,” he says.
Although industry watchers believe CTI can help translate targets into drug candidates, they are cautious about the prospects for the experiment. Pfizer managers have been “plenty innovative and have certainly brought value to this early-stage world,” Afeyan says. But drug companies too often give up on industry-academic collaborations before they have a chance to take off, he says. “I would hope this model is given a chance over five to seven years to really prove what they would do.”
Pfizer claims it is committed to the CTI model with the necessary amount of financial and scientific resources. And the project leaders say they understand the importance of their role in testing a new model for R&D. “If this works, and we are working hard to make it work, it’ll be transformative for the industry,” Martinez says.