How can the emerging field of data science help to fruitfully translate research and discovery into successful drug development? Academics and pharmaceutical executives at Mount Sinai’s School of Medicine discussed ways to build a bridge across the valley of death, and the growing importance of data across all business areas.
Academic institutions are good at identifying and planting seeds with the potential to flower, eventually, into life-saving medicines and devices. The National Institutes of Health, as a primary source of the fertilizer needed to germinate these seeds, has helped enable big pharma to focus on what it does best: middle and late-stage drug development and commercialization.
The path from discovery to product development, however, often detours into a “valley of death,” where academic research and early-stage clinical programs languish in the shadow of risk. Some academic institutions are addressing this problem by “building the infrastructure necessary to better demonstrate” the potential value of an early stage program, said Uwe Schoenbeck, chief scientific officer for external R&D innovation at Pfizer. Schoenbeck said Pfizer’s Centers for Therapeutic Innovation (CTI) is finding ways to work together with academic investigators to achieve better results.
Asked about the most frustrating aspect of collaborating with academic organizations, John Sninsky, VP, discovery research at Celera, a molecular diagnostics firm, said that academics can sometimes “overestimate the value of what has been done, in the context of what still needs to be done” to bring a drug or device to market. Noting that “IP is central to our existence,” Paul Stoffels, worldwide chairman of J&J’s Janssen Pharmaceuticals division, said, “If we’re spending between one and four billion dollars [to develop a drug], IP must be protected.” IP rights don’t come cheap, of course. In the antibody space specifically, Stoffels said “the cost of goods is becoming unbearable” when royalties start stacking up. Nowadays, companies pay 15% or more in royalties on the sales of most new drugs, said Stoffels.
Muzammil Mansuri, SVP, Gilead Pharmaceuticals, noted that a lack of exit options for venture capitalists (VCs) represents a key dilemma to financing early stage development programs. “Conditional exits mean less money up front,” said Mansuri. Pharma, VCs and academia “need to take the risk together,” he said.
In his keynote speech at Mount Sinai’s SinaInnovations conference this morning, Jeffrey Hammerbacher, formerly head of Facebook’s data team and co-founder of Cloudera, an open-source software firm, said healthcare is witnessing a “manifest destiny for data,” which will expand not from sea to shining sea, but from “atom to shining bit.” Ever-growing data sets and analytics are coming, so companies can either “ride the wave or complain about it, and get washed up.”
Hammerbacher, who recently joined Mount Sinai’s School of Medicine as an assistant professor, also stressed the importance of failure, a sentiment echoed later on by Stoffels. To illustrate his point, Hammerbacher referred to the Kremer prize in the UK, which offered a cash prize to the first engineer capable of designing an airplane powered by a single human being. Many engineering firms spent large sums of money to design and develop planes that, when tested for the prize, crashed into pieces. The engineer who finally won the contest, 18 years after it was first announced, won by taking a different approach to the problem, Hammerbacher said. “His approach was to design a plane that could be rebuilt quickly after it crashed, so that he could try out new ideas faster.” After failing in the initial tests, the engineer was able to quickly reassembled his plane, make modifications, and try again until he won the contest. “Building a harness for innovation, in order to fail quickly and try again,” is critical, said Hammerbacher.
The need for transparency and trust in producing and sharing data, whether it’s for licensing deals or regulatory approval, also necessitates a willingness to fail, said Stoffels. Pharmaceutical companies should “create an environment where [scientists and researchers] can fail, and leadership stands up for them.” If people lose their jobs every time they fail, it creates a strong incentive to fudge the data, which is bad for everyone.
The three-day SinaInnovations conference was held at Mount Sinai’s School of Medicine in New York City.