Drive for Data Disclosure Raises Clinical Research Challenges

Pharmaceutical companies are jumping on the clinical trial transparency bandwagon, while also seeking to protect confidential information. As the European Medicines Agency (EMA) moves to make public reports from clinical trials that support market applications, industry hopes to retain some control over who can access proprietary studies and when open access occurs. A related goal is to restore public trust in biomedical research, which has suffered from charges that pharma companies overstate benefits of drugs and hide negative findings.

In January, the European Federation of Pharmaceutical Industries & Associations (EFPIA) and the Pharmaceutical Research and Manufacturers of America (PhRMA) launched a program to vet requests from biomedical researchers for patient-level data, protocols, and clinical study reports for new medicines approved in the US and EU, beginning this year. Johnson & Johnson also announced a collaboration with Yale University’s Open Data Access Project (YODA) that authorizes the university to review and release J&J research data to qualified outside researchers, under a program established by medical device maker Medtronic to deal with pressure to disclose more safety information. GlaxoSmithKline, Roche and other pharma companies have similarly established independent boards to review requests and distribute appropriately confidential research information.

IOM seeks input

Many of these issues are being analyzed by an Institute of Medicine (IOM) panel preparing a report on “Strategies for Responsible Sharing of Clinical Trial Data.” The group released a “discussion framework” in January to solicit input from all parties on proposed data-disclosure principles and activities. These were explored further at a February workshop on the pros and cons of different data sharing models.

Some academics and advocacy groups support total open access as the best way to gain more visibility into failed trials and to facilitate reproducibility and further analysis of research papers. This approach runs the risk, though, of compromising patient privacy, raising difficult informed consent issues, and opening the door to “rogue science” that produces biased and inaccurate analyses.

Controlled access, as provided by YODA and industry initiatives, permits researchers to obtain clinical data, while protecting patient privacy and intellectual property. Yet it raises questions about the independence of third-party reviewers and about who bears the cost of setting standards and building data platforms and infrastructure for responding to researcher queries. Data sharing programs can cost millions of dollars—equivalent to 15% of a total clinical research budget, according to some researchers. And costs may be much greater for retrospective data access, which can involve tracking down participants to redo informed consent. These issues are even more complex when dealing with data from multiple sources, as occurs with consortia and multi-sponsor programs.

Conflicting legal regimes in different countries and regions, a broad range of informed consent issues, and equally complex rules for protecting intellectual property further complicate the data sharing debate, explained attorney Mark Barnes of Ropes & Gray at the IOM workshop. Most requests for access to study data come from other companies, not from independent researchers, Barnes observed. And certain EU member states place limits on patient data disclosure, while some African nations require special permission for secondary uses of clinical trial data collected in that country. FDA policy also may be a problem, as the agency currently requires informed consent to assure participants that outside parties will have access to only a summary of the clinical trial—not to de-identified data.

Disclosure of clinical trial results can undermine the ability of a company to patent a test product and to retain data exclusivity, which can kill a small biotech company, warned Chimerix President Kenneth Moch. Roger Rosenblatt, Chief Medical Officer of Merck, said that all parties should be concerned about the “unintended consequences” of disclosure requirements for generating misleading data and erroneous conclusions about the efficacy and safety of new drugs. “This is not a theoretical risk,” Rosenblatt stated, citing how negative reports can have dire consequences for vaccines.

FDA proposal

Meanwhile, FDA has proposed to make available to researchers de-identified and masked data submitted in market applications. As part of its own transparency program, FDA issued a request for comments last June (2013) on whether disclosure of certain clinical and pre-clinical information held by the agency could help identify new endpoints and study designs for clinical trials as strategies for promoting innovation and public health.

Such disclosure may require changes in current laws that protect trade secret and confidential commercial information of regulated products. Pharma companies and research organizations commented that de-identifying massive amounts of clinical data would be challenging and may not be worth the effort, and that the risk of re-identification is real. Some data sharing might be useful in assessing highly prevalent chronic conditions that have multiple treatments, but that it’s trickier with specialized treatments for diseases affecting limited patient populations.

The IOM will consider these and other issues at another workshop in May and in a final report due by year-end. Meanwhile, pharma companies want EMA, at a minimum, to delay data disclosure until after an experimental therapy comes to market and to ensure that only bona fide researchers can access study reports. Industry critics regard such challenges as a sign that industry has something to hide. Sponsors may gain some relief in the final disclosure policy, but greater transparency is in the cards.