There’s a broad consensus that widely sharing patient-level data from clinical trials is desirable, with potential advantages ranging from faster identification of side effects, easier confirmation of a therapy’s efficacy, and a reduction in duplicative efforts. But many in industry and academia are still reluctant to divulge their data, for a variety of reasons. Now, a new project from Boston’s Brigham and Women’s Hospital and Harvard University is trying to build a global, neutral platform for such data sharing, but some are skeptical it’s the right solution.
The idea for Vivli, from vivlíthiki, the Greek word for library, was articulated nearly 20 years ago in a dissertation written in the wake of the Human Genome Project. Informatics researcher Ida Sim, now a professor at the University of California, San Francisco, School of Medicine, was inspired then by GenBank, a website now operated by the National Institutes of Health that allowed researchers to easily upload and find full genomic sequences in one place. “I thought, ‘Why don’t we have that for human clinical trials?’” Sim said.
Despite her lobbying efforts, the idea didn’t take root, however. Clinical trial registries did emerge, such as ClinicalTrials.gov, but they typically offered only broad information about each trial and its findings, not the actual data from each trial subject.
After later helping found the World Health Organization’s International Clinical Trials Registry Platform, Sim took part in an National Academy of Medicine report last year urging the sharing of all clinical trial data no later than 6 months after publication. The report, which was sponsored by nearly two dozen pharmaceutical companies and government agencies, created the “joint vision” that Sim felt was necessary for such a monumental task. Meanwhile, earlier this year, the International Committee of Medical Journal Editors proposed requiring its member journals to mandate sharing clinical trial data as a condition for publication, and Vice President Joe Biden’s “moonshot” is expected to push for more data sharing to expedite cancer research.
Vivli’s founders, who described their efforts last week in The New England Journal of Medicine, are hoping to capitalize on this new momentum. Although many major pharmaceutical companies, government agencies, and foundations already require anonymized patient-level clinical trial data from studies they fund to be made publicly available, the practice is still far from universal, and the information from these trials is spread across different databases. These data sets are not easily combined, and many of them are difficult to search without knowing exactly what to look for, notes immunologist Barbara Bierer of Brigham and Women’s Hospital, co-chair of Vivli’s working group with Sim.
“Our hope is that if we offer something that is sufficiently flexible and yet comprehensive, that we will forestall the proliferation of independent silos,” Bierer says. To make Vivli more appealing as a resource for researchers, the project will try to aggregate the trial information from all existing clinical trial data platforms and help direct researchers on where and how to request data from them. Its founders also hope it will provide an easy-to-use place for academic and biotech researchers who don’t have their own clinical trial data-sharing options. “It takes phenomenal resources to do this kind of work, and individual investigators don’t have a methodology or a platform to do that,” Bierer says.
Vivli’s work to date has included working on an informed consent document for those wanting to submit data to give to their trial participants, and developing the early frame for a website. The project has received funding from several organizations so far, including the Laura and John Arnold Foundation and the Wellcome Trust, as well as several biotech and pharmaceutical companies. Vivli plans to raise more money from organizations like these, and potentially sees itself being funded long-term by biotech firms.
Pharmaceutical giant GlaxoSmithKline, which operates its own large clinical trial data database, has funded Vivli’s working group and sees great promise in the project if it can effectively sync with other existing platforms, such as its own. “We believe Vivli has the potential to partner with existing systems and deliver a positive change to the data-sharing landscape,” GSK spokesperson Mary Anne Rhyne noted in a statement.
Not all are sold, however. Trish Groves, editor-in-chief of the online journal BMJ Open, attended a March meeting in London about the project and cautions said that many authors and funders there had “grave doubts” about Vivli. “It seems simply too monolithic, too rigid, too focused on one study type, too commercial (potentially), and too U.S.-focused,” she emailed to ScienceInsider, expressing concerns over Vivli’s attempt to become a one-stop shop for clinical researchers. Groves suggests that more targeted and community-focused platforms, such as the Worldwide Antimalarial Resistance Network, will better serve researchers, particularly those in low-income countries. “The real life solution is almost certainly going to be an ecosystem of platforms that learn, over time, to talk to each other,” she contends.
Kay Dickersin, director of Johns Hopkins University’s Center for Clinical Trials and Evidence Synthesis in Baltimore, Maryland, likes the idea behind Vivli, and hopes that it will help improve the quality of trial reporting. But she worries that its working group is too skewed to industry users and those who already operate data-sharing platforms. “We need some sort of coordinating vision, but the question that I would have is—is this a coordinating vision?” Dickersin said.