Crowdsourcing Cures

Drug companies are notorious for keeping information on their new product pipelines close to the vest. Even if a compound fails in development—and most of them do fail—scientific and clinical data are usually kept under wraps. But with R&D costs soaring and productivity falling, pharmaceutical makers are becoming willing to show their hands if it means a failed drug candidate might be revived.

A pilot program to find new uses for unsuccessful drug candidates recently launched by the National Institutes of Health (C&EN, May 7, page 6) is the latest sign that big pharma acknowledges the value of collaboration. Three drug companies have initially signed on to participate in the program, called Discovering New Therapeutic Uses for Existing Molecules.

“The triple frustration of long timelines, steep costs, and high failure rates has bedeviled the translational pathway,” NIH Director Francis S. Collins said at a press conference earlier this month unveiling the project, the first major undertaking by the National Center for Advancing Translational Sciences, or NCATS. “We must do better by working together as a community.”

Drugmakers Pfizer, AstraZeneca, and Eli Lilly & Co. are contributing 24 compounds—and all related data—that went through clinical testing but were shelved after proving ineffective for their intended uses. NCATS, which NIH established in fiscal 2012 with a budget of $575 million to help speed the discovery of new drugs, will kick in $20 million in fiscal 2013 for grants to academic researchers to study the molecules for new indications. If a promising new use is found, the drug company can opt to pick up the rights to the compound and do the costlier, mid- to late-stage trials required to bring the drug to market.

Drug repurposing is not new. Indeed, NIH points to several important medicines that began under a different guise. Azidothymidine (AZT), the groundbreaking HIV therapy, was initially developed to treat cancer. The morning-sickness drug thalidomide was taken off the market in 1961 after it was shown to cause birth defects; it was later found to be an effective treatment for leprosy and multiple myeloma.

And clinicians long have noticed side effects in drugs that have led to their use in new indications. Pfizer’s Viagra, for example, was originally developed to ease angina; after noticing a peculiar side effect in a chest pain trial, researchers began testing it as an erectile dysfunction treatment.

NIH officials hope offering well-studied compounds to the broader scientific community will prompt new ideas that can be swiftly tested in the clinic. “We need to generate more of these success stories in a more successful manner,” Collins said. “We need to crowdsource the solution.”

Industry watchers support the more systematic effort to comb through known pharmaceutical compounds in search of new uses. Because the properties of compounds that have undergone some human studies are well articulated, the time required to bring them to market for new indications will be considerably shortened, says Bernard Munos, founder of InnoThink Center for Research in Biomedical Innovation and a former corporate adviser to Lilly.

Furthermore, companies have already invested tens of millions of dollars in these compounds. In a period of shrinking budgets, pharma companies would welcome salvaging failed drugs to recoup some of their investments.

“The economics of drug repurposing are so compelling that it makes one wonder why the industry, which is desperate for innovation, has not embraced this sooner,” Munos says.

Intellectual property, or IP, was the biggest stumbling block that emerged from a workshop on drug repurposing that NIH held last year with representatives from drug companies, government, and academia. After all, the new NIH program would require drug companies to share data that are normally kept under lock and key. At the same time, new IP might be generated as researchers explore new indications for the drugs.

Traditionally, companies looking to collaborate with outside researchers or academic institutions needed to hash out the IP details on a case-by-case basis. “Too often, it has taken longer to negotiate the terms of the agreement than to actually do the research,” Collins said.

To overcome that hurdle, NIH, Pfizer, AstraZeneca, and Lilly came up with standard confidential disclosure agreements and collaborative research agreements that govern the relationships between the drug firm, NIH, and the grantee.

The template confidential disclosure agreement is a three-page document intended to protect companies’ proprietary information when they reveal data about drug candidates during the grant application process.

If NIH approves an initial summary grant application, a standard collaborative research agreement lays out how the drug company and grantee will work together and the basic terms of ownership of new IP. The agreement also allows a third party to pick up the project if the original owner of the compound isn’t interested in pursuing a promising new indication.

Pfizer was “instrumental in helping us shape these agreements,” Collins told reporters. The big drug company offered insight from its existing drug-repurposing relationships, notably a pact with Washington University in St. Louis that marked Pfizer’s first major effort to open up its molecular vault.

In 2010, Pfizer and Washington University inked a five-year drug-repurposing agreement under which university researchers can peruse a Pfizer database of more than 500 compounds and propose to study them for new indications. Once a project is approved, the academic scientist’s access is widened to include clinical data around the molecule and more information about Pfizer’s research.

Pfizer has been inundated with proposals since the database went live two years ago, according to Rod MacKenzie, head of PharmaTherapeutics R&D at Pfizer. One project in particular has yielded “very intriguing results by linking one of our molecules with an indication that we probably wouldn’t have thought of,” he says. Pfizer is now deciding whether to study the compound further, he adds.

The experience convinced Pfizer that sharing molecules with the wider community for evaluation “really does help expand the ideas we have,” MacKenzie says. “We kind of got inspired by that to think even more broadly about how we might do this on a bigger scale.”

AstraZeneca’s decision to join the NCATS program was also informed by its drug-repurposing efforts. In December, the firm teamed with the U.K.’s Medical Research Council to open access to a cache of 22 compounds—some that failed in clinical trials and some that are still in development—to the broader academic community.

In what Donald E. Frail, the head of Astra­Zeneca’s New Opportunities Innovative Medicines unit, calls a “game-changing element,” the partners posted basic information about the compounds online for all to see. In the eight weeks that followed, more than 100 proposals flooded in, says Frail, who formerly worked at Pfizer and helped set up the relationship with Washington University. With the help of AstraZeneca scientists, a subset of those proposals will be developed into full proposals. “Some we’re very excited about,” Frail says. “They do bring unique ideas, some things we hadn’t considered in the past.”

NCATS’s drug-repurposing program differs from the existing efforts in some key aspects. Compared with Pfizer’s agreement with Washington University, for example, NCATS is starting with a much smaller pool of compounds. “We wanted to have a narrow range of compounds that were at the same stage of development, so we could have a set of data that could be comparable,” says Kathy Hudson, acting deputy director of NCATS.

NIH also wanted to simplify negotiations by limiting the pool to compounds that are not in active development and haven’t been approved for another indication. In addition, only robust compounds with “really crisp and clear data on safety and a lot of pharmacology backbone” made the cut, Pfizer’s MacKenzie adds.

One or two more companies might join the program before the call for applications goes out next month, Hudson says. “It is quite an effort on the part of companies to go through and evaluate the status of a particular compound that’s not being actively developed and put together the data package,” Hudson notes. Maintaining patents and raw material supply for compounds otherwise lying dormant is not cheap, company executives add.

As applications are filed, NIH will match companies with researchers. If multiple applications come in suggesting the same indication for the same compound, Hudson says, NIH will likely try to get those researchers to team up.

After a multistep, peer-reviewed grant process, NIH likely will award the first round of grants next spring, after which it expects investigators to hit clear milestones. The goal in the first year of a grant is to validate the compound in a preclinical model; in years two and three, the compound will move into preclinical studies. If goals aren’t met, the scientists will not get additional funding, Collins told reporters.

The process will yield results even if a hypothesis doesn’t pan out, company officials maintain. “If there is a failed clinical study, that will in itself be important,” AstraZeneca’s Frail says. “We’ll learn from that and move on to more productive hypotheses.”

Although the companies don’t have any preconceived notions about what their shelved compounds might be good for, NIH officials believe that access to pharma’s vault might prompt research in diseases that lack industry attention. Collins said the agency is hopeful that the NCATS program will tackle central nervous system diseases, an area from which many drug companies have stepped away.

In an era of shrinking corporate research budgets, InnoThink’s Munos sees potential for an even broader impact: “As drug companies retrench,” he says, “someone has to step in for patients.”