Clinical Trial Listings and Results Go Public

Sponsors react to pressure to register ongoing clinical trials and to provide access to drug studies.

After years of opposition, pharma companies are promising to better inform the public about opportunities to participate in clinical trials and the data resulting from those studies. Patient advocates, regulatory authorities, and the research community have been clamoring for this information for years and are now backing proposals for more extensive registries that can track sponsor activities and ensure broader access to both positive and negative study results. In an August speech at the Mayo Clinic, FDA acting commissioner Lester Crawford described underreporting of negative trial data as "both alarming and detrimental to the public health."

To head off more onerous disclosure requirements, industry has proposed to reveal data from unpublished studies, but skeptics say that this will not provide the "balanced flow of information" that Crawford urged.

Pushing trial registration
One factor driving disclosure initiatives is industry's fairly poor record in publicizing opportunities for patients to participate in clinical trials. The FDA Modernization Act of 1997 (FDAMA) directed FDA to establish a database of clinical trials to treat serious diseases and conditions. Sponsors are required to describe the purpose of each qualifying experimental drug entering Phase II efficacy studies plus eligibility criteria for participation, location of trial sites, and a point of contact; the prime aim is to make it easier for patients suffering from cancer, AIDS, and other life-threatening illnesses to enroll in studies of experimental treatments [see sidebar on FDA registration efforts].

Because many companies have failed to provide this data to ClinicalTrials.gov, critics are proposing more radical changes. One idea is to have FDA itself post information about clinical studies on the database by extracting relevant data from INDs. More proscriptive proposals are to put an IND on clinical hold or to delay approving a new drug application (NDA) for nonlisted products. However, neither FDA officials nor patient advocates support plans that would delay research or keep valuable therapies off the market, even when a manufacturer fails to meet legal requirements.

In September, the International Committee of Medical Journal Editors, a group of 11 publications including the Lancet, the New England Journal of Medicine, and the Journal of the American Medical Association, offered a new stick. The group said it would publish only articles that discuss studies previously listed on a public trials registry. To be eligible for publication, trials must register by the onset of patient enrollment on a public site such as ClinicalTrials.gov, beginning July 2005.

Drive for disclosure
By requiring sponsors to list trials on the FDA Web site, journal editors and other parties acknowledge that the the real aim is to track ongoing clinical research in order to ensure later disclosure of both negative as well as positive results. The clamor for such information erupted in June when New York attorney general Eliot Spitzer charged GlaxoSmithKline for allegedly concealing study data about the safety and efficacy in children of its antidepressant Paxil (paroxetine). Glaxo denied wrongdoing, but agreed in August to post online summaries of all clinical trials conducted after December 27, 2000.

Meanwhile, the American Medical Association House of Delegates called for the Department of Health and Human Services (HHS) to establish a registry for all U.S. clinical trials as a way to gain public access to study results. The AMA suggested that institutional review boards (IRBs) require public registration as a condition for approving a protocol, a task that IRBs hesitate to add to already overloaded portfolios.

The World Health Organization (WHO) similarly announced interest in creating an international clinical trials registry. The WHO proposal is slated to be discussed at a meeting of national health ministers in November.

Last month, PhRMA (Pharmaceutical Research and Manufacturers of America) announced that it would establish a new online database with summaries of most unpublished clinical trials sponsored by its members starting on October 1. The program is voluntary and limited to "hypothesis testing" studies-i.e., not early-stage exploratory trials. Companies will not have to list studies never completed or provide raw data, and the program applies only to drugs approved in the United States. And the PhRMA initiative intentionally steers clear of the debate over requiring registration of clinical trials enrolling patients.

In addition to Glaxo, Eli Lilly announced plans to post data from all company-sponsored trials for drugs recently approved by FDA, including early trials. Forest Laboratories, facing an investigation by Spitzer of New York, agreed to create an online registry of clinical trial results. These companies and others may shift their disclosure activities to the PhRMA site, where they may post information on earlier and/or successful studies if they so choose.

Congress concernedPhRMA's initiative was announced just two days before a House Energy and Commerce Committee hearing on whether sponsors of pediatric antidepressant studies tried to hide unfavorable study data. A House Government Reform committee hearing in May examined the lack of patient participation in cancer trials, prompting Rep. Henry Waxman (D-CA) to blast industry's poor compliance with the FDA database program. Several Democratic senators began to craft legislation to expand public information on study participation opportunities and research results.

Legislation may be needed to clarify sponsor disclosure requirements. The law requires pharma companies to submit clinical data from all studies to FDA in new drug applications (NDAs). Much of that is considered proprietary and never disclosed. While the Best Pharmaceutical for Children Act of 2002 gave FDA authority to post summaries of all pediatric trials-including studies that fail to lead to approved labeling-it does not apply to adult studies. FDA makes available its medical review when an NDA or supplement is approved, CDER acting director Steven Galson told ACT. But the agency does not post actual trial data, and does not have authority to disclose information on adult studies when an application is not approved.

Next step: Open accessEfforts to list all clinical trials and study results in a national database mesh with the related initiative to increase public access to government-funded research results. In July, the House Appropriations committee added language to the 2005 HHS funding bill calling for NIH to develop a system for public posting of articles about NIH-funded research. Such articles usually are published in expensive professional journals and may take months to appear.

Last month NIH responded by proposing to require researchers to submit to a new NIH database all papers based on NIH research that have been accepted for journal publication. NIH will post articles six months after publication, so publishers and medical societies have the opportunity to profit from their vetting and publishing activities.

Medical journal editors protest that the NIH plan will reduce subscriptions and income, and researchers are concerned that open access would translate into an "author-pays" business model. NIH is accepting comments on the open access plan through November.

Observers remain skeptical that industry's voluntary efforts will be enough. PhRMA intends to use an independent third party to manage and oversee its database, and says that it has "extremely strong commitment" to the program from top company executives, said Caroline Loew, vice president for scientific and regulatory affairs.

PhRMA's main message to legislators is that its initiative will provide access to data immediately, while a legislative route could take years. It remains to be seen if industry will continue to make public previously private information after the spotlight shifts to another medical research issue.

SIDEBAR: FDA Finds Poor Compliance with Trial Registration Program

To implement FDAMA, the National Library of Medicine (NLM) established the ClinicalTrials.gov database in February 2000, by incorporating cancer and AIDS databases established by the National Institutes of Health (NIH). The site now lists about 10,600 trials and attracts 16,000 visitors daily. But less than 10% of protocols come from sponsors of commercial studies.

To encourage industry to post trials on the Web site, FDA issued guidance in 2002 that clarifies which studies should be listed and procedures for doing so. FDA's Office of Special Health Issues (OSHI) launched an educational program and sent letters to some 3000 sponsors of INDs (investigational new drug applications) for drugs and biologics to determine which studies involved treatments for life-threatening diseases and thus should be listed.

As of June 2004, though, 246 pharmaceutical companies had listed only about 1200 protocols. Moreover, many of the listings are incomplete, often missing critical information on study sites or drug or company name. An OSHI review of all cancer protocols submitted to FDA's Center for Drug Evaluation and Research (CDER) during nine months in 2002 shows industry's poor compliance record: of 366 cancer protocols submitted to FDA, 127 were sponsored by drug companies and met the criteria for listing on ClinicalTrials.gov. However, less than half (48%) of the eligible industry studies were listed, compared to 91% of eligible NIH trials. And OSHI's analysis indicates that cancer and AIDS trials are the best-reported of all disease areas.

Pharma companies have claimed that trial registration could reveal proprietary information, such as inclusion/exclusion criteria and site locations. But FDA staffers maintain that such facts usually are well known by market analysts by the time a drug moves into Phase II or III studies. Conversely, regulators believe that registration provides an invaluable way for sponsors to accelerate patient enrollment in trials and to avoid duplicative research. OSHI is finalizing its report on registry compliance, which may lead to revisions in eligibility criteria and posting procedures.-J.W.