Special Report
Progress and Deficiencies in the Registration of Clinical Trials
N Engl J Med 2009; 360:824-830February 19, 2009
- Article
Clinical trials are essential to understanding the efficacy of medical interventions. The ethical underpinnings of this type of research involving human subjects, codified in the Belmont Report1 and the Declaration of Helsinki,2 require that the results be publicly available to inform medical practice as well as future research. In addition, basic principles of evidence-based practice require the analysis of all data on a given topic; the practice of publishing only some results, but not others,3,4 undermines our collective ability to make rational decisions about medical care.
In Section 801 of the Food and Drug Administration (FDA) Amendments Act,5 enacted in September 2007, Congress expanded the requirements for sponsors and investigators to post information about clinical trials, including selected aspects of trial results, on the U.S. government Web site ClinicalTrials.gov. These expanded requirements6,7 apply to all investigators and sponsors, whatever their source of funding. Section 801 also includes, for the first time, considerable penalties for noncompliance, including possible loss of grant monies from the National Institutes of Health (NIH) and civil monetary penalties of up to $10,000 per day. This mandate for registration and results reporting (Table 1Table 1
Registration and Reporting Requirements of the FDA Amendments Act, Section 801.) has important ethical, practical, and scientific implications for all who are involved in clinical research, as well as for all who rely on research results to guide their decisions on clinical care.Registration Requirements
Section 801 of the FDA Amendments Act requires the registration and public posting of all applicable trials of drugs (including biologics) and devices subject to FDA regulation (generally, this means having at least one trial site in the United States). Section 801 excludes phase 1 drug trials and “early feasibility device trials,” but it is important to note that the International Committee of Medical Journal Editors (ICMJE) and the World Health Organization (WHO) require the registration of all trials regardless of phase. In addition to broadening the scope of the trials that must be registered, the Act now mandates the posting of a number of data elements that had previously been optional on ClinicalTrials.gov but were required by the ICMJE7 and the WHO (Table 1). The law handles registration of device trials differently from drug trials, and device trials will not be discussed in detail in this article.
The requirements to submit registration data took effect on December 26, 2007, and ClinicalTrials.gov has been modified to accommodate compliance. However, the precise specification of mandatory data elements and delineation of compliance and enforcement provisions remain a work in progress and will be finalized at a later date by promulgating regulations. Sponsors and investigators are responsible for compliance with Section 801 even before the promulgation of formal regulations.
Results-Reporting Requirements
A novel aspect of Section 801 is the mandated incremental expansion of ClinicalTrials.gov to include information about results. This is being implemented by linking existing databases and creating a new results database at the National Library of Medicine (NLM). First, ClinicalTrials.gov registration records for trials of FDA-regulated drugs and devices will be linked to existing results information available on the FDA, NLM, and NIH Web sites. Examples include electronic drug labels, FDA public health advisories and drug-approval packages, and peer-reviewed publications indexed in Medline. Early versions of these links are now functional on ClinicalTrials.gov. Second, the NLM has created a “basic results” database to collect data elements specified by the law and display them on ClinicalTrials.gov. The creation of this database was completed on September 24, 2008, and it is now available for investigators to populate with data. Third, summary information on adverse events must be added by September 27, 2009, although a version of this information is now available for use by investigators. Finally, a more comprehensive results database — referred to in Section 801 as an “enhanced results” database — will be created after a public meeting is held and a rule-making process of up to 3 years has been completed.
Creating a comprehensive clinical trials database is very challenging because of the wide variety of trial designs that must be registered. Because the submitted data will not be peer-reviewed and will not be interpreted, qualified, or explained, the results reported at ClinicalTrials.gov will complement rather than replace the thoughtful presentation and discussion of results characteristic of the best peer-reviewed publications; journals represented on the ICMJE (www.icmje.org) do not consider the posting of results in this fashion in the database where the trial was registered as prior publication.
The reporting requirements are summarized in Table 1 and Table 2Table 2
Reporting Requirements for Drug Trial Results under the FDA Amendments Act, Section 801.. The results of all drug trials — other than phase 1 — that were ongoing on or after September 27, 2007 (the date of enactment), must be reported at ClinicalTrials.gov if the products under study have already been approved by the FDA. The results for drugs not yet approved by the FDA do not have to be posted until the drug receives such approval. In general, trial results should be reported no later than a year after the date on which the last measurement concerning the primary outcome has been completed. For trials of approved drugs in which a new, unapproved indication is being studied, reporting can be delayed by up to 2 additional years regardless of the ultimate regulatory decision made by the FDA on the indication. A deficiency of the current law is that the results of older trials of drugs that were approved before September 27, 2007, and that were no longer the subject of ongoing trials on or after September 27, 2007, do not need to be posted — and such drugs constitute the vast majority of the prescription drugs currently used by patients. Likewise, there is no requirement for posting the results of trials of drugs that were never approved.8 However, trials of previously approved drugs that have not been approved for a new indication will have to be posted, but not until 2 years after the completion date.An earlier law enacted by the State of Maine9 that required reporting of results from trials initiated after October 15, 2002, for drugs marketed in that state will be preempted once Section 801 is fully implemented. However, for the time that the Maine law remains in effect, public dissemination of results for some of the trials that are not covered under the FDA Amendments Act will be required by the State of Maine. The state recently issued guidelines for drug manufacturers reminding them of the need to comply with state law (the guidelines are available at www.maine.gov/dhhs/boh/documents/CT_reportingupdate9-08.pdf).
Clinical Trials Results Databases
ClinicalTrials.gov, the largest single registry of clinical trials, had more than 67,000 registered trials as of January 31, 2009. It contains trials representing a broad range of intervention types (e.g., drug, device, surgical intervention, and procedural intervention) from 157 countries. Under the terms of the FDA Amendments Act, ClinicalTrials.gov will substantially improve access to both registration and summary results information for new and ongoing trials of drugs and devices.
True public access to results depends on the existence of reports that are available to the public and on the convenience and accuracy with which the database can locate those reports. Right now there are three sources of results that meet these criteria: PubMed and other bibliographic databases (e.g., EMBASE) that provide indexes of publications, FDA analyses and documents that can be accessed through the FDA Web site (e.g., Drugs@FDA) and ClinicalTrials.gov, and existing industry results databases, which have limited search capabilities (e.g., GlaxoSmithKline Clinical Study Register [http://ctr.gsk.co.uk/welcome.asp] and the site sponsored by the Pharmaceutical Research and Manufacturers of America [PhRMA] [www.ClinicalStudyResults.org]).
The database of results from trials listed at ClinicalTrials.gov, searchable with the customized ClinicalTrials.gov search engine,10 will add another source of information about trial results. Ultimately, access to trial results is still insufficient if it is not possible to determine whether and how many sets of trial results may be missing. Unfortunately, none of these sources of results (i.e., publications, FDA reviews, industry-sponsored databases, or ClinicalTrials.gov entries) will necessarily contain a complete set of all clinical trial results for a given drug or device, either individually or among them.
The Need for Additional Legislation
Although incorporation of the NLM initiatives in the ClinicalTrials.gov database will be a great step forward, it is worth noting that the taxpayer already contributes to the maintenance of a treasure trove of clinical trial data at the FDA to which the public has very limited access. The FDA reviews and analyzes all trial results submitted in support of an application for approval to market drugs, biologics, or devices. According to the law, the FDA must treat much of the data on clinical trials it receives from sponsors, and its own analyses of these data, as “confidential commercial information” and is prevented from publicly disclosing this information. Therefore, any drug trial information released to the public by the FDA is often heavily redacted to prevent illegal disclosures.11 Such redacted disclosure often occurs only in response to one or multiple requests made under the Freedom of Information Act, and even then the disclosure is often made only after a prolonged delay. Section 916 of the FDA Amendments Act institutionalizes this limitation by mandating that the FDA publish on its Web site key information related to an approval of an application (the action package) for a drug that contains a previously approved active ingredient only after it has received three requests under the Freedom of Information Act.
Although FDA reviews of many approved products are posted online, the FDA Web site does not contain all data for all trials — or even most of the data for most of the trials — of approved products. In addition, because the law places limitations on the data the FDA is allowed to release, some of the review documents released are so heavily redacted that their content is largely incomprehensible. Even Section 916, which requires the posting of FDA reviews of trial results for approved drugs, does “not authorize the disclosure of any trade secret, confidential commercial or financial information.” Thus, mandatory disclosures under the FDA Amendments Act will still be heavily redacted because of current legislative restrictions.
The public health would be well served if the FDA were to post all reviews for all products after their approval.12,13 The FDA has committed to posting, in a timely manner, reviews of trials submitted with new drug applications for drugs approved since 1998, but it has rejected recommendations from the Institute of Medicine that it post all reviews for supplemental new drug applications (e.g., applications for use of approved products for a new indication), stating that doing so would be overly burdensome.13 Since the requirements under the FDA Amendments Act are not retroactive, the FDA may continue to refrain from posting older trials that contain data relevant to drugs on the market before September 27, 2007, which represent the vast majority of drugs currently used by patients and are therefore of great interest to patients and their physicians.
As mentioned, the FDA must currently consider data or analyses for certain trials to be confidential commercial information, including those concerning products under development, products that have been abandoned and for which approval is no longer being pursued, and products already on the market for which approval for a new indication is being sought. The balance of public interest in terms of protecting commercial interests during product development, avoiding unnecessary human experimentation, and informing current patients and potential research subjects about risks and benefits varies among these three groups. However, it is clear that research participants and patients have an interest in dissemination of these data to facilitate the ethical development and use of safe and effective drugs.
For example, there were apparently several suicides among participants who received the marketed selective serotonin-reuptake inhibitor duloxetine as part of a study of urinary incontinence.14 Despite the fact that the drug was already on the market for another indication (and that concerns had been raised about the risk of suicidality), the FDA review of these data was not made publicly available (nor were the data themselves) because the data were part of an application for a new indication that was never approved. The argument that such data are “confidential commercial information” and must be withheld so that potential competitors do not benefit runs counter to the principles of ethical human research, which require that risks to human volunteers be minimized and that human participation in research leads to generalizable knowledge. Knowledge cannot be generalized if the FDA is forced by legislation to withhold it from the public.
Although some protection of commercial interests is undoubtedly necessary to ensure ongoing drug development, the withholding of critical information about the safety or efficacy of marketed drugs from the public is unacceptable both ethically and scientifically. I believe that relevant FDA data on clinical trials and the agency's analyses of all these data should be made publicly available for all approved drugs and for drugs and indications no longer being actively pursued (Table 3Table 3
Proposed Legislative Changes to Section 801 of the FDA Amendments Act.).Under the FDA Amendments Act, the secretary of health and human services is specifically directed to consider, during the 3-year rule-making process, whether or not to extend the results-reporting requirements to unapproved products. It seems reasonable to propose that reporting should be extended to include drug trials to which the following circumstances apply: 2 years have elapsed since the FDA reached a “negative” decision on the new drug applications, or 2 years have elapsed since a company withdrew a new drug application, or a fixed period of time (e.g., 2 years) has elapsed after the last data were received by the FDA, without any additional studies submitted to the FDA, implying that clinical development has for practical purposes stopped. In addition, consideration should be given to the posting of phase 1 results (currently excluded under the FDA Amendments Act) after 2 years of inactivity as defined above. Safety data should be posted without a 2-year delay to rapidly inform institutional review boards and others reviewing similar drugs or devices.
Two years of inactivity would seem to be an adequate period for protection of legitimate concerns about maintaining confidentiality during drug development while also releasing information in a reasonably timely fashion to protect the public. However, safety data concerning patients should not be subject to such delays to protect commercial advantage. Posting of trial results would honor the pact between research participants and society that expects that data gathered through their participation in a trial will be made available in a timely manner to help future patients. Also relevant is knowledge of adverse outcomes associated with products similar to those currently on the market, even if the marketed product is for a different indication.
Many of the changes proposed to supplement Section 801 of the FDA Amendments Act (summarized in Table 3) can be accomplished relatively easily by enacting legislation to open up the FDA's database on clinical trials to the public — which has paid for the FDA's analysis. However, imposing a requirement on companies to post the results of clinical trials for drugs approved before September 2007 — especially results that might be decades old and not readily available — would be an unreasonable burden, however desirable the outcome. In addition, after patent protection expires and generic drugs enter the market, the company that developed the original drug may no longer market the drug. Access to old clinical trial data may also be problematic if substantial time has elapsed since completion of the trials and drug approval. In these and all other cases, removing the legislative provisions that define even ancient clinical trial data held by the FDA as “confidential commercial information” would allow the agency to post the data in its hands without imposing unreasonable retrospective demands on the companies concerned. Of course, to post such data, the FDA and the NLM will require additional funds from Congress (Table 3).
Conclusions
Ethical clinical research should contribute to generalizable knowledge and improve human health. The dedication of patients who take the risks to participate in clinical research is dishonored when their data remain secret. Section 801 of the FDA Amendments Act will greatly expand the type and amount of information available on clinical trials. The use of these data and the enhanced public access to the FDA's database on clinical trials proposed in this article will greatly improve the ability of investigators and others to assess the full set of results for a given intervention, whether FDA-approved or not, eventually leading to more accurate systematic reviews, better clinical decision making, improved patient care, and improved research efficiency and safety.
Dr. Wood reports serving on the National Library of Medicine Board of Regents Working Group on Clinical Trials, on the scientific advisory board of Sapphire Therapeutics, and on the board of directors of Oxigene and holding stock in Symphony Capital. No other potential conflict of interest relevant to this article was reported.
This article (10.1056/NEJMsr0806582) was updated on April 6, 2011, at NEJM.org.
I thank Deborah A. Zarin, M.D., and Tony Tse, Ph.D., for their substantial help, guidance, and support on earlier versions of this article.
Source Information
Dr. Wood is managing director of Symphony Capital and a Professor of Medicine and Pharmacology at Weill Medical College of Cornell University — both in New York.
- References
References
1
The Belmont Report: ethical principles and guidelines for the protection of human subjects of research. Washington, DC: Department of Health, Education, and Welfare, 1979.
2
World Medical Association. Declaration of Helsinki: ethical principles for medical research involving human subjects, as amended by the 48th World Medical Assembly, Somerset West, Republic of South Africa, October 1996. (Available at http://www.wma.net/e/policy/b3.htm.)
3
Nissen SE ,Wolski K . Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007;356:2457-2471[Erratum, N Engl J Med 2007;357:100.]
Full Text | Web of Science | Medline4
Turner EH ,Matthews AM ,Linardatos E ,Tell RA ,Rosenthal R . Selective publication of antidepressant trials and its influence on apparent efficacy. N Engl J Med 2008;358:252-260
Full Text | Web of Science | Medline5
Food and Drug Administration Amendments Act of 2007: Public Law No. 110-85 § 801; 2007.
6
Food and Drug Administration Modernization Act of 1997. Public Law No. 105-115 § 113.; 2007.
7
Laine C ,Horton R ,DeAngelis CD , et al. Clinical trial registration -- looking back and moving ahead. N Engl J Med 2007;356:2734-2736
Full Text | Web of Science | Medline8
Zarin D ,Tse T . Closing a loophole in the FDA Amendments Act. Science 2008;322:44-46
Web of Science9
An Act Regarding Advertising by Drug Manufacturers and Disclosure of Clinical Trials, Maine State Law, 22 MRSA c605, §2700-A (2005).
10
Ide NC ,Loane RF ,Demner-Fushman D . Essie: a concept-based search engine for structured biomedical text. J Am Med Inform Assoc 2007;14:253-263
CrossRef | Web of Science | Medline11
Kesselheim AS ,Mello MM . Confidentiality laws and secrecy in medical research: improving public access to data on drug safety. Health Aff (Millwood) 2007;26:483-491
CrossRef | Web of Science | Medline12
Turner EH . A taxpayer-funded clinical trials registry and results database. PLoS Med 2004;1:e60-e60
CrossRef | Web of Science | Medline13
Psaty BM ,Charo RA . FDA responds to Institute of Medicine drug safety recommendations -- in part. JAMA 2007;297:1917-1920
CrossRef | Web of Science | Medline14
Lenzer J ,Brownlee S . Antidepressants: an untold story? BMJ 2008;336:532-534
CrossRef | Web of Science | Medline
- Citing Articles (12)
Citing Articles
1
Joel Lexchin. (2012) Harmony in Drug Regulation, but Who's Calling the Tune? An Examination of Regulatory Harmonization in Health Canada. International Journal of Health Services 42:1, 119-136
CrossRef2
M. A. Rogawski, H. J. Federoff. (2011) Disclosure of Clinical Trial Results When Product Development Is Abandoned. Science Translational Medicine 3:102, 102cm29-102cm29
CrossRef3
Natalie S. Buchan, Deepak K. Rajpal, Yue Webster, Carlos Alatorre, Ranga Chandra Gudivada, Chengyi Zheng, Philippe Sanseau, Jacob Koehler. (2011) The role of translational bioinformatics in drug discovery. Drug Discovery Today 16:9-10, 426-434
CrossRef4
Joel Lexchin. (2011) Those Who Have the Gold Make the Evidence: How the Pharmaceutical Industry Biases the Outcomes of Clinical Trials of Medications. Science and Engineering Ethics
CrossRef5
Anna-Bettina Haidich, Charis Birtsou, Theodore Dardavessis, Ilias Tirodimos, Malamatenia Arvanitidou. (2011) The quality of safety reporting in trials is still suboptimal: Survey of major general medical journals. Journal of Clinical Epidemiology 64:2, 124-135
CrossRef6
Rafael Dal-Ré, Alejandro Pedromingo, Manuel García-Losa, Juan Lahuerta, Rafael Ortega. (2010) Are results from pharmaceutical-company-sponsored studies available to the public?. European Journal of Clinical Pharmacology 66:11, 1081-1089
CrossRef7
Matthias Dahm, Paula González, Nicolás Porteiro. (2010) Implementing registries and results databases of clinical trials: is there a side effect?. Expert Review of Pharmacoeconomics & Outcomes Research 10:3, 247-251
CrossRef8
Bernard Lo, Lindsay Parham. (2010) Resolving Ethical Issues in Stem Cell Clinical Trials: The Example of Parkinson Disease. The Journal of Law, Medicine & Ethics 38:2, 257-266
CrossRef9
A I Vitry. (2010) Reporting of Studies on New Medicines in Major Medical Journals: A Case Study in Breast Cancer. Clinical Pharmacology & Therapeutics 87:4, 398-400
CrossRef10
Z. H. Endre, J. W. Pickering. (2010) Outcome definitions in non-dialysis intervention and prevention trials in acute kidney injury (AKI). Nephrology Dialysis Transplantation 25:1, 107-118
CrossRef11
Califf, Robert M., Harrington, Robert A., Blazing, Michael A., . (2009) Premature Release of Data from Clinical Trials of Ezetimibe. New England Journal of Medicine 361:7, 712-717
Full Text12
Elske H. Weber, Ivo T.F. Timmermans, Martin Offringa. (2009) The Dutch Medicines for Children Research Network. Pediatric Drugs 11:3, 161-163
CrossRef
