The fallout from the FDA investigation into the risk of suicide for adolescents taking antidepressants has included calls for publication of all drug study results. Pharmaceutical companies are pledging to do so on dedicated Web sites. Can the industry be trusted to police itself or is governmental oversight required?
One week before two U.S. Food and Drug Administration advisory committees in joint meeting determined that antidepressants can provoke suicidality in children with depression, a congressional committee concluded that such adverse effects, as well as the negative results of clinical trials, have not been adequately disclosed by either study sponsors or the FDA. (See related article discussing the reaction of the American Psychiatric Association on p48 of the print edition--Ed.)
During the Sept. 9 hearing on publication and disclosure issues in antidepressant medication trials, members of the U.S. House Oversight and Investigations Subcommittee of the Energy and Commerce Committee criticized the pharmaceutical industry for selectively releasing clinical trial data and publishing only studies with positive results. Committee members Edward Markey (D-Mass.) and Henry Waxman (D-Calif.) announced they would co-sponsor a bill to require timely posting of all clinical trial results and subsequently introduced the Fair Access to Clinical Trials Act (HR 5252) on Oct. 7.
The FDA was taken to task in the hearing by the Energy and Commerce Committee chairperson, Joe Barton (R-Texas), who accused the FDA of having a "spotty record" on sharing results from the pediatric clinical trial data with the public, despite this requirement in the Best Pharmaceuticals for Children Act. Janet Woodcock, M.D., FDA acting deputy commissioner for operations, responded to Barton's questions by insisting that the FDA has complied with the law.
"We have legal restraints on the amount of information that we can disclose and when we can disclose that information," Woodcock testified.
A follow-up hearing was held on Sept. 23 to consider whether the FDA had adequately fulfilled its role to protect public health in its review of pediatric antidepressant trials for safety and efficacy. An FDA epidemiologist, Andrew Mosholder, M.D., testified that he had concluded prior to the FDA's advisory committees that the data linked antidepressants to suicidality in children. He told the congressional panel that he had been prevented from presenting this analysis in an FDA advisory committee hearing in February by supervisors disagreeing with his assessment. In addition, Mosholder described an FDA internal probe into whether he had leaked his assessment to the press, and subsequent pressure to not disclose existence of the probe.
Robert Temple, M.D., director of the FDA Office of Medical Policy, explained that the FDA had been unable to conclude that antidepressants caused suicidality in children before the analysis of the clinical trials data by specialists at Columbia University, which was reported at the September advisory committee meeting. "The agency realizes its responsibility to the public to find the right answer to this question," Temple said.
Pressure to Report Results
Pressure for making all clinical trial results widely available was heightened by a wave of lawsuits against GlaxoSmithKline, the manufacturer of Paxil (paroxetine); these have followed the successful suit in June by New York State Attorney General Eliot Spitzer alleging suppression of clinical trial data on suicidality. GlaxoSmithKline agreed to the remedy of posting results of all of its drug trials. Several manufacturers quickly followed, announcing plans to post late-stage trials on their own Web-based sites or on the ClinicalTrials.gov registry maintained by the National Library of Medicine.
The Pharmaceutical Research and Manufacturers of America (PhRMA) announced on Sept. 7, two days before the congressional hearing, its planned Oct. 1 launch of the Web-based registry www.ClinicalStudyResults.org for manufacturers to voluntarily post trial results. Caroline Loew, PhRMA vice president for scientific and regulatory affairs, indicated in the announcement that this registry would be managed by a third party and reflect the industry's intent to make study results "as transparent and accessible as possible."
This splintered and voluntary approach to disclosing trial data has not been considered adequate, however, by the American Medical Association or by editors of major medical journals. At its annual meeting in June, the AMA passed a resolution calling for the U.S. Department of Health and Human Services (HHS) to establish a public registry for all clinical drug trail results. In September, JAMA, New England Journal of Medicine and other member journals of the International Committee of Medical Journal Editors (ICMJE) published an ICMJE statement announcing that the results from clinical studies starting enrollment after July 1, 2005, must be posted in a public registry to be considered for publication (JAMA 292:1363-1364).
The ICMJE statement asserted that research participant volunteers assume that their participation contributes to improved health for others and that researchers endeavor to minimize their risk. The statement further argued:
In return for the altruism and trust that make clinical research possible, the research enterprise has an obligation to conduct research ethically and to report it honestly. Honest reporting begins with revealing the existence of all clinical studies, even those that reflect unfavorably on a research sponsor's product.
The ICMJE statement acknowledged that public posting of trial results is only part of the means to an end, explaining, "that end is full transparency with respect to performance and reporting of clinical trials."
The ICMJE and other medical editors elaborated on the need for public disclosure of clinical trials in their September editorials. In a Sept. 16 Medscape posting, George D. Lundberg, M.D., editor of Medscape General Medicine and former editor of JAMA, criticized, as a long-standing practice, the selective submission of positive studies for publication.
"What if you are a researcher in the field and you want to test a treatment that has already been tested and shown not to work, or even found to be harmful?" Lundberg posed. "What if voluminous relevant data apply to a patient you are caring for and you can't find those data? Is that not a wasteful or even harmful situation?"
Drummon Rennie, M.D., deputy editor of JAMA, in his Sept. 15 editorial, lamented that it has taken 18 years after proof that registration of trials at inception eliminates publication bias for the current groundswell of support for a mandatory public registry (JAMA 292:1359-1362).
"The issue is fundamentally an ethical one," Rennie declared. "Should the results of trials be regarded as 'highly proprietary,' a view espoused by the manufacturers who pay for the trials?" Rennie's concerns coincide with Lundberg's, as he asked whether research volunteers "might have been less willing to participate had they known the results would be treated as trade secrets and often never made public?"
Rennie suggested that the difficulties the FDA advisory committees have had in evaluating suicidality in children receiving antidepressants demonstrates the dilemma for physicians wishing to practice evidence-based medicine without all the evidence. He cautioned against trial results being posted on the manufacturers' own sites, however, concerned that these could be "obscure, hard to access, idiosyncratic in the sorts of data they post, contain only what marketing departments allow them to contain, and are temporary and dependent on the company's institutional memory and will."
Rennie also rejected the idea that only late clinical Phase III and IV trials should be reported. He noted that Phase I trials are also used for gathering data on efficacy, as well as safety:
A system must be designed to register all trials (regardless of phase) at inception, so that reviewers can later ask for an accounting. Given the conflict of interest inherent in any decision by sponsors to enter trials into a register, any system must be free of the control of individual sponsors.
The resolution from the AMA 2004 Annual Meeting calling for a clinical trials registry was based on a report from the AMA Council on Scientific Affairs (CSA) which, in turn, had been prompted by a joint resolution of the American Psychiatric Association and the American Academy of Child and Adolescent Psychiatry. The CSA report considered the impact of funding sources on the outcome, validity and reliability of pharmaceutical research.
The CSA report acknowledged that without pharmaceutical industry sponsorship, important therapeutic advances would not have occurred. It noted, however, that a pattern of publication bias "distorts the medical literature, thereby affecting the validity and findings of systematic reviews and meta-analyses, the decisions of funding agencies, and ultimately the optimal practice of medicine."
In examining how clinical trial arrangements can confound disclosure of results, the CSA found that 53% of agreements in a sample of university-industry research centers allowed publication to be delayed; 35% allowed the sponsor to delete information from the publication, and 30% allowed both.
Although an Association of American Medical Colleges (AAMC) 2002 task force report provides guidance on conflicts of interest in clinical research, the CSA noted that sponsored clinical research is increasingly moving from academic research centers to contract research organizations (CROs). The CROs often use community physicians as a source for participant recruitment, according to the CSA, and enable less expensive trials with less cumbersome trial agreements.
In addition to recommending a public registry for all clinical trials, and preceding the ICMJE in recommending that registration be a required condition for publication, the CSA report cautioned physicians against entering into research contracts that permit the sponsoring company to "unduly delay or otherwise obstruct" the presentation or publication of results. The CSA report also urged institutions to adhere to AAMC 2002 guidelines in separating administrative responsibilities related to human participant research from those related to investment management and technology licensing.
The Fair Access to Clinical Trials Act would require clinical trials to be registered in a data bank maintained by the National Library of Medicine, under direction of the National Institutes of Health and authority of the HHS. It would also require a clinical trial to be registered before an Institutional Review Board could approve it to be conducted in the United States.
Trials would have 12 months after registering with the data bank to begin providing results, and then must update every six months. Some exceptions to this time frame are provided to accommodate publication considerations or "extraordinary circumstances" in the public interest. The bill specifies that trial data are to be reported in a form "that ensures that the information is accurate and not likely to mislead or distort the results of the trial."
The posted clinical trial results would include the percentage of individuals withdrawing from participation, and their reasons; as well as significant adverse events that may be associated with the tested product "including such events for which a causal relationship has not been established."