The decade-old drumbeat for clinical trials transparency grew louder with release of a paper in BMJ Open about the dearth of voluntary data reporting – sometimes even from late-stage trials – on drugs that were subsequently approved.
Lead author Jennifer Miller, assistant professor in the division of medical ethics in the department of population health at New York University School of Medicine, and colleagues at Harvard University and Yale School of Medicine reported that only two-thirds of trials, per drug, that supported approvals in 2012 were disclosed – a threshold the researchers called "below legal and ethical standards." Nearly half of the drugs reviewed – 15 approved drugs or vaccines developed by the industry's biopharma giants – were associated with at least one undisclosed phase II or III trial.
Miller also is the founding president of Bioethics International, a New York-based not-for-profit organization launched in 2006 to monitor the ethics and governance surrounding the research, development, marketing and accessibility of drug therapies. In 2009, the group launched its Pharma Integrity Scorecard initiative. After six years of development, the findings for the subset of approved drugs in 2012 represent the first published scorecard.
"We started with clinical trial transparency because of two downstream effects," Miller told BioWorld Today. "The first is the massive amount of distrust that the public is having around pharmaceutical companies. Part of that is driven by the fact that there have been some massive ethics failures in disclosing clinical trial results in the past."
She cited, as an example, the failure by London-based Glaxosmithkline plc to fully disclose findings on Paxil, which led to FDA and congressional hearings, the filing of a lawsuit by New York's attorney general and a settlement that included GSK's commitment to report future trial results. (See BioWorld Today, Sept. 10, 2004, Sept. 14, 2004, and Sept. 16, 2004.)
Failure to report trial results also results in "a huge public health problem," Miller added.
"We're supposed to have an evidence-based medicine system, but you can't have evidence-based medicine without all the evidence," she maintained. Bioethics International is the first organization to report hard data on data transparency in drug development, according to Miller.
The primary goal of the new study was to analyze the number of trials, by approved therapy, that were registered and reported results on ClinicalTrials.gov, that published findings in the medical literature and that complied with the 2007 FDA Amendments Acts (FDAAA). That law, which arose partly from previous suppression of trial data, gave the agency more authority to require safety labeling changes and postmarketing safety studies and to post quarterly reports listing marketed drugs under evaluation for potential safety problems. In addition, the law mandated the registration and public reporting of results from certain trials – mainly controlled trials of a drug or biologic conducted in the U.S. under an investigational new drug application. (See BioWorld Today, Sept. 30, 2008.)
The FDAAA framework contains a much narrower reporting burden for trials than does voluntary reporting, and the government's mandate is reduced further by conflicting interpretations of "controlled" and "interventional" trials, according to the researchers. Thus, they segmented their analysis of FDAAA-mandated trials into two sample pools and excluded any trials with unknown phases or any cited as phase I/II in at least two primary data sources.
The 15 drugs included in the analysis of voluntary, or full, reporting of trial results were listed in order of noncompliance. At the top of the list was Stribild (elvitegravir/cobicistat/emtricitabine/tenofovir disoproxil fumarate, Gilead Sciences Inc.), followed by Aubagio (teriflunomide, Genzyme Inc./Sanofi SA), Elelyso (taliglucerase alfa, Pfizer Inc./Protalix Biotherapeutics Inc.), Zaltrap (ziv-aflibercept , Sanofi/Regeneron Inc.), Stivarga (regorafenib, Bayer AG), Eliquis (apixaban, Bristol-Myers Squibb Co.), Zioptan (tafluprost ophthalmic solution, Merck & Co. Inc./Santen Pharmaceuticals Co. Ltd.), Xeljanz (tofacitinib), Pfizer Inc.), Bosulif (bosutinib, Pfizer), Perjeta (pertuzumab, Genentech Corp./Roche AG), Signifor (pasireotide, Novartis AG), Erivedge (vismodegib, Genentech/Roche), Inlyta (axitinib, Pfizer), Sirturo (bedaquiline, Janssen Therapeutics) and Menhibrix (meningococcal groups C and Y and haemophilus B tetanus toxoid conjugate vaccine, GSK).
Stribild met 100 percent compliance with FDAAA-mandated clinical trial reporting, however, as did four additional drugs. The rest fell short, with Elelyso, Stivarga, Perjeta, Signifor and Erivedge showing 0 percent compliance with the FDAAA rules, according to Miller and colleagues.
'CAN'T INNOVATE WITHOUT LEARNING THE LESSONS THAT WENT BEFORE'
Overall, the researchers identified 318 relevant trials associated with the approved drugs, involving 99,599 research subjects. A median of 57 percent (interquartile range [IQR] 32 percent to 83 percent) of trials were registered per drug, 20 percent (IQR 12 percent to 28 percent) reported results in ClinicalTrials.gov, 56 percent (IQR 41 percent to 83 percent) were published and 65 percent (IQR 41 percent to 83 percent) were either published or reported results.
Per drug, a paltry 17 percent (IQR 8 percent to 20 percent) of trials supporting FDA approvals were subject to FDAAA-mandated public disclosure. Of those, a median of 67 percent (IQR 0 percent to 100 percent) were FDAAA-compliant. Sixty-eight percent of research subjects (67,629 of 99,599) participated in FDAAA-subject trials; nearly half of those (33,405 of 67,629) were enrolled in noncompliant trials, according to the researchers.
Going forward, Bioethics International plans to rank every drug approved by the FDA, a year at a time, using the template established in this study.
Initially, "we didn't know if we could measure ethics," Miller admitted. "And we didn't know if it could make a difference." But a handful of contributors, starting with the Susan G. Komen Foundation and eventually adding university and other foundation sponsors, believed in the idea.
"We spent a lot of time mapping the areas of concern that people had around pharmaceutical companies," Miller said, and data transparency rose to the top.
The transparency data could be used by the FDA, she suggested, as a mechanism to monitor trial results, especially given that the Department of Health and Human Services and the NIH last year called for mandatory reporting of virtually all human trial results. (See BioWorld Today, Nov. 21, 2014.)
"We hope this scorecard provides a mechanism of support to the FDA going forward," Miller said, noting that Mark McClellan, former FDA commissioner and administrator of the Centers for Medicare & Medicaid Services, served on the nonprofit's advisory team.
And Miller made clear that the study results were not meant to serve as a stick to browbeat the industry. The data were shared prior to publication with each company on the list, and 100 percent responded to the researchers, suggesting that all companies have some form of commitment to transparency.
"You can't innovate without learning the lessons that went before you," she said. "If you don't put the clinical trial information into the public space, it's not generalizable knowledge, and that's a huge human rights concern."
'WE HAVE THE DATA TO HAVE AN INFORMED CONVERSATION'
Despite the enormous variation in reporting of trial results, best practices are emerging. True to its word, GSK disclosed all trial results and fully complied with legal disclosure requirements for its vaccine, Menhibrix, approved in 2012. Janssen – a unit of Johnson & Johnson, of New Brunswick, N.J., came close with Sirturo.
GSK's commitment to transparency has been "a journey" that began with the establishment of its trial registry in 2004, according to Murray Stewart, chief medical officer. In the ensuing decade, the pharma has reported on more than 5,500 studies.
Stewart agreed with Miller's contention that voluntary reporting of trial results includes three key factors: commitment from top management, a written policy and a system to operationalize those procedures. GSK's initiative incorporated all three.
"I wasn't surprised, in the findings, to see that GSK would be at 100 percent because that's what we're monitoring," Stewart told BioWorld Today. "It's not just by chance. Unless you actually make this a priority – develop a policy and put in metrics – the figures won't change."
Arriving at 100 percent compliance takes a few years, he acknowledged. But because GSK's top management has eyes on the trial reports, "a friendly rivalry" has emerged among development teams, which vie to ensure their reporting is on par with other units.
That's not to say biopharmas don't encounter barriers in seeking full transparency for trial data. One of the biggest, Stewart said, is the effort to get negative findings published, which he named as a key reason for the large discrepancy between reported and published trial data in the BMJ Open paper. Although GSK encourages investigators to try to publish studies with poor outcomes, the company puts a greater emphasis on transparent reporting.
"We will always try to get full publication," he said. "There may be a time lag because the results are negative or they're just not that exciting, but we will still make a commitment to publish on our web page. You a need a system where you're going to record the findings, irrespective of whether they get into a peer-reviewed journal."
Companies also need a culture of transparency.
"What I sometimes see is that once the headline is out, people don't care much," Stewart said, noting the tendency for teams to want to move onto the next project. "You need a culture that emphasizes completion of the study report to publication."
Instilling that culture requires resources, he admitted. GSK has a dedicated team that tracks and monitors the posting and publication of trial findings.
Similarly, J&J, through its subsidiary, Janssen Pharmaceutical Cos., has an agreement with the Yale University Open Data Access Project (YODA) to share data from its trials, according to Joanne Waldstreicher, J&J's chief medical officer.
"Our agreement with YODA ensures that every request for access to our pharmaceutical clinical data is reviewed objectively and independently," Waldstreicher told BioWorld Today. "Under this agreement, YODA serves as an independent body to review requests from investigators and physicians seeking access to clinical trials data from Janssen, and to make decisions on sharing data. YODA will independently interact with external data requesters and select and consult with an independent panel of scientific and medical experts to help with the review. The final decision to share or not share the requested data will be made by YODA."
That system ensures that J&J shares trial data in the form of clinical study reports as well as participant-level data "in a systematic and objective way that protects patient privacy and confidentiality," she added, noting that data are de-identified before they are shared.
GSK, too, is moving beyond reporting trial data in the scientific literature by making summaries available in lay language to patients and consumers and by providing more granular, patient-level data to the scientific community, Stewart said.
Gilead did not reply to requests for comment on the BMJ Open study.
Ultimately, Bioethics International aims to look at a broader set of ethics associated with drug development. In addition to data transparency, Miller said other issues include the design and conduct of clinical trials, marketing practices and accessibility.
"We want to do everything systematically and with the most rigor possible," she said. "We can't say, 'We don't have a problem here,' or with this drug or with this company. Now we have the data to have an informed conversation."
GSK's Stewart is hopeful that other biopharmas will improve voluntary reporting, and not just for self-serving reasons.
"This is the right thing to do because we will make advances in developing medicines and we'll get faster at developing medicines," he said. "I'm encouraged by the publication. It's a start. Principles and practices are always difficult, but we can get there. People shouldn't be put off by the challenge."