Gag Clauses in Clinical-Trial Agreements
Robert Steinbrook, M.D.
[editorial on prior post]
Gag clauses in clinical-trial agreements prevent investigators from examining the data independently or submitting a manuscript for publication without first obtaining the consent of the sponsor. Sponsors with a financial interest in the outcome of clinical research can suppress negative results. They can also interfere with the publication of unfavorable data on safety, as it was recently alleged that Merck officials did in the company-sponsored "Advantage" study of rofecoxib for the treatment of osteoarthritis published in 2003.1 Some of the hidden data, like those related to cyclooxygenase-2 inhibitors and the use of antidepressants in children, may eventually become public, but other studies are never reported.
In 2001, the International Committee of Medical Journal Editors (ICMJE), which then represented 11 general medical journals, including the New England Journal of Medicine, began to require that the responsible author of a study state in writing that he or she accepted full responsibility for the conduct of the trial, had access to the data, and controlled the decision to publish.2 Subsequently, researchers at Duke University (Durham, N.C.) surveyed the provisions in clinical-trial agreements between medical schools and industry sponsors and found that "academic institutions routinely engage in industry-sponsored research that fails to adhere to ICMJE guidelines regarding trial design, access to data, and publication rights."3 In response, the researchers and others proposed that there should be standard contract provisions for industry-sponsored research conducted at academic medical centers.4
In 2005, has anything changed? The short answer is no. Dr. Robert M. Califf, the director of the Duke Clinical Research Institute, noted recently, "I do not see any evidence that the average contract of a site participating in a multicenter study is any better now than when we wrote the article." According to Dr. Steven E. Nissen, the director of the Cleveland Clinic Cardiovascular Coordinating Center, "There has been little progress. The main reason is that commercial sponsors still exclusively control the database for most clinical trials."
There is model contractual language, such as that proposed by the Duke Clinical Research Institute (www.dcri.duke.edu/ccge/contracts), but there are no standardized clinical-trial agreements that have been adopted by multiple medical schools and academic medical centers. Elsewhere in this issue of the Journal, Mello, Clarridge, and Studdert (pages 2202–2210) report that standards for restrictive provisions in clinical-trial agreements with industry sponsors vary considerably among academic medical centers. For example, medical centers are split about whether sponsors can store the data and release portions to investigators, insert their own statistical analysis into manuscripts, and write up the results for publication, with the investigators' role limited to review and suggestions for revision. In a telling example of current contractual restrictions, medical school researchers who had conducted industry-sponsored studies of the use of antidepressants in children and adolescents were denied access to unpublished data from their own studies about the relation between the medications and the risk of suicide.5 In another example, Merck officials apparently did not inform academic authors about key safety data from the Advantage study of rofecoxib.1 The patient outcomes, including the number of deaths from cardiac causes, may have been incorrect in the medical journal report.
Progress against gag clauses, however, could be forthcoming, spurred by public concern about medication safety, distrust of the pharmaceutical industry, and advocacy within the medical community for greater openness in conducting and reporting clinical trials. In December 2004, the House of Delegates of the American Medical Association committed the organization to working with industry and other groups to eliminate gag clauses in clinical-trial agreements and to "take all appropriate action to protect the rights of physician researchers to present, publish and disseminate data from clinical trials." The Fair Access to Clinical Trials Act of 2005, as introduced in the Senate by Senators Grassley and Dodd in February, would ban contracts involving researchers or their institutions that prohibit, limit, or unreasonably delay the discussion of the results of a clinical trial at a scientific meeting or the publication of results.
The Association of American Medical Colleges, which has tried and failed to develop standardized contract provisions, plans to try again. The association, in conjunction with other groups, has launched a project to develop and promulgate a set of principles and recommendations for the interaction of academic medical institutions and faculty members with industry in clinical-trial research. The goal is to ensure integrity in the conduct and reporting of clinical trials by academic investigators.
A basic tenet of research ethics is that the data from clinical trials should be fully analyzed and published. If the knowledge gained from trials is not shared, subjects have been exposed to risk needlessly. Moreover, participants in future studies may be harmed because earlier results were not available. These principles are reflected in federal regulations regarding the protection of human subjects, which define research as "a systematic investigation designed to develop or contribute to generalizable knowledge."
Given their educational and public-service missions and their capability of conducting unbiased and rigorous clinical trials, academic medical centers should have a vested interest in working together to eliminate gag clauses. Academic institutions, however, compete for clinical-trial research contracts with other medical centers, as well as with private groups known as contract research organizations, or CROs. These organizations can often provide research services at lower cost and with fewer administrative burdens.2
Although industry is spending billions of dollars more for clinical trials than it did a decade ago, the proportion of its clinical research projects that go to academic medical centers continues to decrease substantially — from 63 percent in 1994 to 26 percent in 2004 (see graph). Industry spent $3.2 billion for investigator grants in 1994 and an estimated $10 billion in 2004, as compared with $1.6 billion for CRO services in 1994 and an estimated $7.6 billion in 2004, according to Thomson CenterWatch, a Boston-based clinical-trial listing service. An academic institution whose standards for access to data and publication rights are stricter than those of its competitors may lose trials — unless it has access to unique groups of patients or its researchers have some unique expertise.
View larger version (8K):
[in this window]
[in a new window]
Distribution of Sites of Clinical Research Sponsored by Industry in the United States, 1994–2004.
Data are from Thomson CenterWatch. Data for 2004 are projected.
Some research is already conducted without gag clauses — for example, large studies coordinated by well-established centers with considerable expertise, such as the Cleveland Clinic Cardiovascular Coordinating Center or the Duke Clinical Research Institute. An institutional example is the University of Toronto and its affiliated teaching hospitals, which have had comprehensive standards for all industry-sponsored clinical research since 2001. Their effort to institute standards was inspired in part by the intense controversy in the 1990s over the case of Dr. Nancy Olivieri, a Toronto hematologist. Olivieri entered into a research contract — which included a confidentiality agreement — with a company to study an orally active iron chelator. When she tried to publish findings about the lack of safety of the medication, the company sued her.
According to the University of Toronto standards, clinical-research contracts should not allow sponsors to suppress or censor research results, and researchers retain the right to disclose immediately any safety issues that arise during a study. In an interview, Dr. C. David Naylor, the dean of the University of Toronto Faculty of Medicine, said that some contracts had been lost as a result of the policy. "We believe there should be national guidelines in Canada that apply to all research sites. We know that some sponsors have moved trials elsewhere and bypassed academia. You are left with the sound of one hand clapping when you do this on your own, even in the country's largest medical school."
A key barrier to resolving the problem is the extensive financial connections that many academic investigators have with sponsors of their research. A related and contentious issue is the proper definition of the phrase "access to the data." Does the sponsor control the data, perform the statistical analysis, and provide a copy of the results to the investigators? Or is the trial database made available to the researchers, who independently confirm the findings and conduct their own analyses? According to the voluntary principles of the Pharmaceutical Research and Manufacturers of America, "As owners of the study database, sponsors have discretion to determine who will have access to the database. . . . Sponsors will make a summary of the study results available to the investigators. In addition, any investigators who participated in the conduct of a multi-site clinical trial will be able to review relevant statistical tables, figures and reports for the entire study at the sponsor's facilities, or other mutually agreeable location."
According to Nissen of the Cleveland Clinic, unless the investigators have direct and full access to the database, there is too much opportunity for sponsors to provide incomplete information and overemphasize the benefits or underestimate the risks of therapy. "At our coordinating center, we require the study contract to guarantee the transfer of the trial database to the investigators at the conclusion of the trial and to guarantee the independent right to publish," he said.
How might gag clauses in clinical-trial agreements be eliminated? Unless there is a federal ban, academic medical centers and other research organizations, in collaboration with leading clinical researchers, would have to develop strict policies about access to data and publication rights and adhere to them. If they did so, sponsors would be unable to seek more favorable terms from one institution than from another. Medical journals would need to strengthen their existing guidelines — for example, by explicitly defining minimum standards for authors' access to data and their right to examine the data independently, and then refusing to publish studies that do not meet these standards. In some instances, all the data might need to be made available to — and analyzed by — a statistical group that was independent of the industry sponsor. In the absence of a broad-based effort, however, the current opportunity for change may be lost.
Source Information
Dr. Steinbrook is a national correspondent for the Journal.
References
1. Berenson A. Evidence in Vioxx suits shows intervention by Merck officials. New York Times. April 24, 2005:A1.
2. Davidoff F, DeAngelis CD, Drazen JM, et al. Sponsorship, authorship, and accountability. N Engl J Med 2001;345:825-827. [Full Text]
3. Schulman KA, Seil DM, Timbie JW, et al. A national survey of provisions in clinical-trial agreements between medical schools and industry sponsors. N Engl J Med 2002;347:1335-1341. [Abstract/Full Text]
4. Drazen JM. Institutions, contracts and academic freedom. N Engl J Med 2002;347:1362-1363. [Full Text]
5. Meier B. Contracts keep drug research out of reach. New York Times. November 29, 2004:A1. |
