|
While it is perhaps an extreme example, the use of science as a smoke screen for pharmaceutical marketing is old news. In the last decade, court documents related to other lawsuits against several drug manufacturers have given us undeniable evidence of the hiding of marketing behind a veil of respectable science, and of the involvement of the academic community in these practices. Publicity for a commercial health communication’s publication--available likely only to industry insiders considering the staggering $3,695 price tag--emphasizes how scientific publications are a crucial marketing tool. It seems to admit that selecting commercially interesting results was once common practice: “While picking and choosing favorable findings may have been acceptable a decade ago,” the publicity for this Best Practices publication states, “it is now considered unethical and potentially illegal.”[2]
What is new indeed, is that medical, academic and governmental organizations are putting the pressure on to curb some of these practices. In 2008, Joseph Ross and colleagues documented in an article in the Journal of the American Medical Association, one of the world’s leading medical journal, some other problematic publication practices associated with Vioxx.[3] The article discusses, for example, how employees of Merck and a communication agency working for Merck prepared final manuscripts on Vioxx for submission to top medical journals, but published the studies with academic researchers as the authors, often failing to reveal any link to Merck employees. Researchers received payment to borrow their name for these publications. David Healy and Dinah Cattell documented similar practices related to Pfizer’s drug Zoloft in a 2003 article in the British Journal of Psychiatry.[4] And a 2004 lawsuit of the Attorney General of New York against GlaxoSmithKline (GSK) for its off-label promotion of Paxil also revealed instances of ghost authorship, a term used to describe the practice of hiding the real authors behind publications.
When commercial sponsors design the protocol, select research subjects, conduct the research, collect the data, interpret the results, control publication of the outcome, pay academic authors to give credibility to ghost-authored publications, fund advocacy groups to lobby governments for approval or funding of the drug, and spend millions on commercial campaigns to promote a new drug, there is reason to worry about whether the process is generating unbiased results and unbiased communication of findings.
Hearings organized by US Senator Charles E. Grassley in 2008 to investigate financial ties between academic institutions, physicians, patient advocacy groups and industry, revealed examples of related practices and gave the public a sense of how much money is involved in the courting of academic researchers by industry. The hearings exposed how over the last decade, a large number of top-rated academic researchers in the United States earned sums of up to 2.8 million dollars for consultancy fees, advisory board membership, and other services for pharmaceutical sponsors, and often failed to report these payments to their university employer. Testimony and evidence revealed at the hearings also suggest that some were involved in ghost-authorship, in misrepresentation of data, in the delaying of publication of negative findings, or in tipping-off pharmaceutical sponsors about upcoming critical publications.
In earlier publications, I tried to sketch a picture of the various ways in which the financial interests associated with the marketing of blockbuster drugs have affected the design, organization, conduct and reporting of clinical trials.[5] I explored what various regulatory tools we have at our disposal to deal with the safety and effectiveness of pharmaceutical products and discussed why these have not prevented products with a very negative safety profile from entering the market. My current research, supported by a Social Sciences and Humanities Research Grant, focuses on an analysis and evaluation of various regulatory and policy tools to improve the integrity of medical research and the safety and effectiveness of pharmaceutical products. One of the questions I try to address is to what extent existing regulatory tools to improve safety and effectiveness of pharmaceuticals may in effect contribute to the growing control of industry over clinical research. The research also deals with how the growing entanglement of academic researchers with pharmaceutical industry interests may have an indirect impact on the legal system itself. Two research assistants, Sarah Jones and Sinziana Tugulea are directly involved in specific research topics that I will discuss further. With one faculty colleague, Simon Stern, I am also looking at the potential use of traditional tort law principles to address some of the concerns.
How Did Industry Gain so Much Control Over Clinical Research?
Financial interests in research are obviously nothing new. Since the birth of the pharmaceutical industry in the second part of the 19th century there have been significant commercial interests in research and drug development. But the ability of industry to control scientific development has significantly increased in the last three decades. The pharmaceutical and biotechnology industry has burgeoned into one of the most profitable industry listed on the Stock Exchange. With profit has come power. Backed by unprecedented financial means and pressured by stockholders’ demand for investment returns, pharmaceutical companies have taken a much larger stake in all stages of medical research. Industry has steadily extended its tentacles into all aspects of the development of scientific knowledge and into the use of this knowledge for the creation of health care products and the promotion of their use.
As a result, most of those involved in clinical research have some financial relationships with the industry. In the U.S.A., most funding for clinical research now comes from industry. Many academic researchers receive funding to be involved in clinical trials. Many of them also top off their salaries with stock options or fees for consultations, speaking arrangements, and membership on advisory boards of pharmaceutical companies. Financial rewards are also offered to consumer and advocacy groups, primary care physicians, research recruiters and bioethics consultants and centers. Even scientists at the renowned U.S. National Institutes of Health have come under fire for their significant financial connections to industry. Being on a company’s payroll has not precluded experts from functioning as members of influential governmental advisory panels involved in the approval of drugs and the determination of clinical practice guidelines.
Governmental policies and legislative initiatives that promote the commercialization of academic research have also contributed to a shift in institutional approaches to commercialization. Changes within the research culture have blurred the lines between public and private in research and fundamentally altered the landscape of medical research. Interaction between industry and academic researchers has become more direct and cozy. This has surely led to the development of many new health care products, as the rhetoric surrounding these partnerships consistently claims. But the courting of researchers, patients, health care workers, medical institutions, and regulators is also part of aggressive, orchestrated, and well-developed marketing campaigns which contribute little to the common good.
What are the concerns?
One of the larger and more general concerns, highlighted by the examples mentioned earlier, is the impact of the new trend on research integrity. When commercial sponsors design the protocol, select research subjects, conduct the research, collect the data, interpret the results, control publication of the outcome, pay academic authors to give credibility to ghost-authored publications, fund advocacy groups to lobby governments for approval or funding of the drug, and spend millions on commercial campaigns to promote a new drug, there is reason to worry about whether the process is generating unbiased results and unbiased communication of findings. A common response to such concerns is that a company’s desire to obtain results that will help it promote its products is counterbalanced by the culture of “organized skepticism” within science and also by the competing interests of other companies. But this view largely relies on the existence of a forum for scientific dialogue and for critical challenge of research findings. Commercialization has, however, also affected the environment where this debate is expected to take place. Pharmaceutical competitors often share core interests that outweigh the incentive to analyze critically a competitor’s research in certain respects. Pharmaceutical companies sell diseases to create markets for their drugs. Competitors will not challenge a company’s claim that anxiety, shyness, premenstrual dysphoric disorder, and post-traumatic stress disorder are endemic and have to be aggressively treated with medication. They will not question the claim that the best way to reduce your risk of heart disease or arthritis is to be on a life long regime of drugs, rather than to change your life-style and diet.
Second, there are increasing concerns about the integrity of the scientific literature itself. Academic journals have struggled for years with the question of how financial relations may influence the independence of the authors and reviewers of manuscripts. Disclosure of financial interests has now become the standard remedy in most of the major academic journals, but disclosure is clearly not sufficient. As the Grassley hearings show, not all scientists feel obliged to disclose their funding. Nor are financial interests simply parts of the research scene that can be laid on a scale alongside the scientific facts. They cannot be measured and evaluated in the same way as scientific data. Disclosure provides only a false security. It rings the alarm bell. But an alarm bell in and of itself is of little use if no effort is ever made to curb the negative impact of these conflicts.
And that financial interests have an impact is undeniable. Numerous studies have shown that there is a clear correlation between the source of funding and the outcome of clinical trials.[6] Trials funded by industry are much more likely to provide results that are favourable to the provider of the funding.
The growing phenomenon of ghostwriting, of which I mentioned some examples earlier, raises also concerns about integrity of the scientific literature. Pharmaceutical research is increasingly organized by specialized contract research organizations and specialized medical service agencies. Results of research are increasingly written up by ghost authors hired by industry. The papers written by these authors are then submitted for publication by established academics who are authorities in their field. Academic authorship gives credibility to the research and may help to get the studies published in the most respected medical journals. Yet the “authors” did not write the text and often did not participate in the analysis of data.[7]
The problem with these practices is not only that they involve some form of deception, run counter to established professional standards, and violate institutional rules of academic centers. What makes them particularly problematic is that they are part of a complex interwoven series of actions that have a very real impact on day-to-day medical practice. They aim to increase the consumption of pharmaceutical products to an extent that often seriously endangers the health and wellbeing of patients and consumers. The Vioxx controversy is the clearest evidence that this constitutes a serious public health threat.
If ghost authored publications were reflecting the accurate results of sufficiently independent studies, there would be no immediate concern for public health. But it is clear that ghost authorship is also associated with hiding of negative data (i.e. data revealing safety or effectiveness problems), and even with misrepresentation.
The 2005 lawsuit by the Attorney General of New York against GlaxoSmithKline for the promotion of off-label prescription of the anti-depressant and anti-anxiety drug Paxil highlights some of this.[8] This case was settled out of court after GSK accepted to pay US US $ 2,5 million and to make all of its clinical trials data publicly available on a website. The Attorney General accused GSK of conducting 5 different clinical trials involving Paxil for the treatment of depression in children and adolescents without disclosing the negative results of these trials. Internal GSK documents showed that company officials realized that the data were insufficiently robust to obtain a label change and to have the drug approved for the treatment of children and adolescents. The trials failed to show efficacy and even revealed an increased risk of suicidal ideation in the group taking the Paxil. Yet, the company decided, according to the statement of claim, to hide the negative data, and to selectively publish only the positive results of one of the trials. This publication, in a leading medical journal, was then distributed among physicians with the alleged goal of promoting further off-label prescription.
A 2008 study by Erick H. Turner and colleagues reveals more generally how problematic it can be to rely on the published peer-reviewed literature, because of the lack of reliability of published data.[9] They analyzed FDA reviews for studies of 12 anti-depressant agents and the published literature on these products and compared the outcomes of both. The researchers found that 31% of the FDA registered studies, nearly all of them negative ones, were not published. Of the negative studies, 22 were not published, while—even more problematically—11 were published in a way that conveyed a positive outcome, according to Turner and his colleagues. The overall message sent by the published literature stood also in stark contrast with the data submitted to the FDA: 94% of the published studies sent a clearly positive message about these drugs, while only 51% of the studies reported to the FDA had a positive outcome.
Clearly, pharmaceutical companies are in a powerful position when it comes to the collection and control of data from clinical trials. Although they can afford financially to study comprehensively a drug by studying in addition to its basic efficacy and safety in clinical trials also the potential long-term adverse effects of the drug after it enters the market, they generally don’t do so. The reason is that there is no regulatory obligation to continue studying a drug once it is approved, and that it can be convenient to use the absence of statistical evidence as a shield, particularly in the context of potential lawsuits.
The controversies highlight therefore also serious flaws in the reach of the drug regulatory system. Drug regulatory agencies have programs in place to monitor potential adverse effects of approved drugs. These adverse-reaction reporting systems can provide crucial information and have, for example, provided a basis for Health Canada to request the voluntary withdrawal of various drugs from the market. Health Canada has also, in the wake of safety issues associated with marketed drugs, tried to improve its adverse reporting system. Yet, the agency claims that it does not have the authority to impose additional safety studies once a drug is approved. And since industry is the main funder of expensive long-term clinical trials, industry can determine what is being researched. Recommendations for long-term safety studies were for example made in the context of the recent voluntary withdrawal of Vioxx from the market. Experts urged already in 2001 for appropriate clinical trials to assess cardiovascular risks; yet no such studies were undertaken. Because of these problems, new regulatory tools have been developed in the United States to improve post-marketing surveillance. The 2007 FDA Amendment Act has significantly improved the FDA’s ability to request specific post-marketing surveillance trials as a condition of regulatory approval.
Recent Initiatives in Canada to Strenghten Regulatory Review
In Canada, attempts have also been made to deal with some of these issues, but many of the proposed regulatory changes have yet to be implemented and it falls to be seen whether the measures will be sufficient and will really deal with all the problems that have been identified. In 2006-2007, Health Canada issued a Blueprint for Renewal, announcing its intention to switch to "life-cycle" drug regulation.[10] Whereas the current regulatory framework focuses on safety and efficacy at market-entry, a life-cycle approach would involve continuous evaluation of safety and efficacy, before and after market approval. Health Canada also intends to develop a Progressive Licensing Framework, which links a progression in drug approval status to ongoing drug surveillance.
Plans for legislative changes to the existing Food and Drug Act were introduced in parliament under the former government, but were stalled as a result of the recent elections. This type of legislative amendments and new regulations would give Health Canada the authority it is currently lacking to require post-market studies in response to reports of adverse reactions, or as a condition of drug licensing. The Blueprint also suggests that Health Canada would also work closely with drug manufacturers in developing appropriate clinical trial methods. Continuing market authorization of a drug could be more clearly linked to solid information gathering about adverse effects, long-term effects and drug interactions, which is often only becoming available post marketing.
There are, however, concerns about this proposal. Progressive licensing could be used to speed up initial market authorization, based on even more limited information, with consumers becoming the ultimate test subjects. Both the Blueprint and Health Canada's documents on progressive licensing refer to the pressures from consumer advocacy groups (often spurred on by drug manufacturers) for earlier access to new therapies. The progressive licensing framework as currently proposed would place the onus on the manufacturer to develop and carry out an ongoing drug monitoring plan. With the responsibility for both pre-market testing and post-market surveillance in drug companies' hands, the drug regulatory system would still be open to manipulation and suppression of evidence.
Another initiative, that has seen the light, is the Federal Government’s commitment to provide more than $30 million to develop a new Drug Safety and Effectiveness Network. This Network is currently being set up in a partnership between Health Canada and the Canadian Institutes for Health Research. The DSEN will link existing centres of excellence in post-market pharmaceutical research, provide coordination for post-marketing surveillance studies, and develop research priorities related to pharmaceutical products that are already on the market. While this is a valuable initiative, the number of post-marketing surveillance trials that can be undertaken with this level of funding remains limited, in light of the number of products that are already on or that are entering the market. It also pales in comparison with the money spent by industry in conducting clinical trials. Industry will clearly remain the main source of information on the safety and effectiveness of its products, even if the DSEN will provide a critical resource when doubts are raised about the safety of effectiveness of a drug already on the market.
An independent drug agency
One of the most significant problems remains the control of industry over the design of studies, the conduct of trials, and the publication of results. Several commentators have therefore argued that a fundamental change in the regulatory review of clinical trials is needed to separate those who design, conduct and peer review research from those who have financial interests in the outcome. In a 2006 article in the NEJM, Wayne A. Ray and C. Michael Stein recommended the establishment of three new independent national drug institutes in the United States.[11] These three independent centres would be responsible for: new drug approval, post-marketing studies and drug information. The companies would be funded by taxes on pharmaceutical products, but would not be directly funded by the industry. A company wishing to apply for approval of a new drug would have to negotiate an appropriate protocol with the new drug approval agency, which would organize the clinical trial, using qualified drug assessment centers.
This would also be a good model for Canada. In Canada, a drug testing agency would not only be important to improve the reliability of the data used to support applications for drug approval. The agency should also be involved in conducting comparative clinical studies to provide the necessary information to provincial governments for deciding which drugs should be on drug formularies supported by public funds. A national drug-testing agency would be an important cornerstone of an evidence-based rational public health care system. It could actually be based on the model of the Drug Safety and Effectiveness Network, to the extent that this network remains fully independent from industry and from the agencies that are involved in initial approval of the drug. The agency could contract out clinical trials for drug approval purposes to established clinical trials centres. This new drug testing structure could also help in restoring the independence of academic research, and public confidence in the research enterprise.
This idea has been voiced already in Canada. Royal Commissioner Roy Romanow supported an independent drug testing agency. More recently, Conservative MP Terency Young introduced a private member’s bill in federal parliament.[12] Terence Young’s proposal, motivated by his own experience as a father who has lost his 15-year old daughter due to complications resulting from the drug Prepulsid,[13] emphasizes the need to separate a national drug testing agency from Health Canada. The agency that has earlier given marketing approval for a drug can indeed be in a difficult position when it has to assess the potential serious side-effects of an earlier approved drug.
Incremental and Immediate Measures
While this more radical change may take time, incremental steps can immediately be taken to improve the transparency and reliability of the system. One common recommendation is to establish a registry of clinical trials. The International Committee of Medical Journal Editors, that represents the most prestigious medical journals, issued in 2004 a statement that its journals were going to publish from then on the clinical research findings of only those studies that were registered in such a registry prior to the commencement of enrollment of human research subjects.[14] A special working group of the World Health Organization has also recommended that all clinical research be registered prior to recruiting human subjects, and has provided guidance as to the items that have to be entered into publicly accessible databases in order to make these clinical trial registries meaningful and useful.[15] Other organizations have followed this lead. I am currently involved as a member of the Health Research Advisory Committee with the Pan American Health Organization’s efforts to promote clinical trials registration across the region and to create a clinical trial registration portal for all Latin-American countries and the Caribbean.
But access to information about a trial is in and of itself insufficient. The Ottawa Group, a group of international scientists who have been actively involved in the promotion of clinical trial registration, has always strongly argued that mandatory results reporting is an essential component of promoting real transparency and to avoid hiding or misrepresentation of data. This group recently met in Ottawa, with support from CIHR, to discuss further implementation of clinical trials result portals.[16] A WHO working group is currently also looking into the establishment of clear guidelines and rules about mandatory disclosure of trial results. Significant moral support for these initiatives has recently come from the World Medical Association, which has also introduced a requirement for clinical trials registration and results reporting in the 2008 revision of the Declaration of Helsinki “Ethical Principles for Medical Research Involving Human Subjects.”[17] The Declaration of Helsinki is the most influential international statement of ethical principles of medical research and this support for trial registration and results reporting may create a momentum for regulatory change.
Registration and result reporting should certainly be made obligatory, through national legislation or regulation. Recent amendments to the US Food and Drug Act are a source of inspiration in that context. The 2007 FDA Amendment Act introduced a strict obligation to register clinical trials and to report the results within a given period of time. It has accompanied these requirements with severe financial penalties. Those violating these provisions of the act risk penalties of up to $ 10,000 per violation and per day. Canada has not yet followed this lead, even though the federal funding agencies have confirmed their commitment to this requirement. The federal government should enact similar regulations with clear financial penalties to ensure that these requirements are also respected in Canada.
But a mandatory registration system and mandatory results reporting do not solve all problems. A registry falls short of promoting properly designed clinical studies which could challenge the findings in studies controlled by the pharmaceutical companies. And registration of results is pretty meaningless if the results cannot be relied upon. Thus, there will be a need for independent scrutiny of the methodology, analysis of data, and interpretation of findings in industry controlled clinical studies. This requires financial resources and availability of clinical researchers who are independent from the pharmaceutical industry. As a result of decades of handing out research funding, support to individuals and institutions, and largesses to researchers and physicians for membership of speaker bureaus and advisory committees and for providing lectures and consultation services, the number of truly independent researchers may be limited. More importantly, not even independent scrutiny can overcome manipulation of trial results by premature termination of the study or by analysis of only a part of the collected data. The problems of potential bias in industry-sponsored studies and the declining number of truly independent researchers remain a major challenge that requires structural reform. As Ray and Stein suggest in the New England Journal of Medicine, we need to move to a regulatory system which diminishes—rather than increases—the role of industry in establishing and publicizing scientific information.
In addition to developing a more independent, tightly regulated review sector, we must also reinvigorate traditional university values and promote an independent and unbiased publication culture in academia. This would involve countering the strong financial incentives offered by industry with equally strong deterrents to scientific misconduct. One on-going project, in which Sinziana Tugulea is the lead research assistant, examines the current mechanisms through which medical researchers can be penalized for failing to disclose conflicts of interest or for engaging in research fraud. These include reprimands for professional misconduct by disciplinary boards of hospitals and medical associations; academic sanctions imposed by universities on physician-researchers in affiliated hospitals; and criminal charges in egregious cases of fraud or negligence endangering human subjects. We are currently looking at how the Canadian emphasis on self-regulation by professional and academic institutions compares to a more pro-active regulatory review by federal regulatory agencies, such as exists in the United States. One emerging theme is that these self-regulated academic and professional bodies currently rarely investigate cases of scientific misconduct. They may have little incentive to make findings of misconduct against their own members and to inform the research community and the public. The sanctions they apply may not be commensurate to the harm done to the public trust in medical research, and to the potential tainting of the knowledge base, with safety implications for patients.
Although self-regulation has traditionally been the rule for the medical profession and academia alike, it may simply not be enough to regulate aggressive recruitment of researchers by the pharmaceutical industry. Nor is the honour system that pervades much of the medical publishing process enough to ensure researchers' integrity with such financial incentives at stake. We want to look at how academic institutions and professional colleges can become more actively involved in the sanctioning of professional misconduct. My colleague Simon Stern from the Faculty of Law is also taking the lead in exploring to what extent traditional tort law principles could be used to deal with some of the controversial practices associated with medical research and publication practices.
A second project we are working on, in which Sarah Jones is the lead research assistant, considers the effect that manipulation of evidence has in civil trials. More specifically, we examine the burden faced by a plaintiff who sues a drug company, given that the pharmaceutical industry largely controls the production and dissemination of evidence on the safety and efficacy of their drugs. We are interested in looking at how the regulatory system, with its current emphasis on limited pre-marketing safety and efficacy, in combination with the existing rules of liability, may create a negative incentive for companies to study real-world safety and efficacy of drugs.
Voluntary compliance and market mechanisms have not protected research subjects and consumers in the past. There is no reason to believe they will do so in the future. There are simply too many incentives and too many opportunities for the manipulation of research findings. We hope with our research project to contribute to new regulatory initiatives, some more immediate, other more drastic and long-term, that can strengthen research subject, consumer and patient protection and restore trust in the medical and research enterprise.
We are also interested in looking at how the courts’ emphasis on peer reviewed data and on academic credibility by expert witnesses as reflected in the number of publications, may affect the ability of people harmed by pharmaceutical products to obtain compensation.
Many maintain that the market creates the best system of accountability. The controversies surrounding drug development are a very strong indicator that—certainly when it comes to health care products—this is simply not true. Voluntary compliance and market mechanisms have not protected research subjects and consumers in the past. There is no reason to believe they will do so in the future. There are simply too many incentives and too many opportunities for the manipulation of research findings. We hope with our research project to contribute to new regulatory initiatives, some more immediate, other more drastic and long-term, that can strengthen research subject, consumer and patient protection and restore trust in the medical and research enterprise.
|
button (top left)