Credibility of industry-funded clinical trials

Abstract

If the pharmaceutical industry does not adhere to strict ethical standards both internally and externally when collaborating with academia, the public's trust will be further eroded.

In recent years, the credibility of medical research has been seriously compromised by both real and perceived conflicts of interest (COI). The consequence has been a public who are doubtful and dissatisfied with the biomedical enterprise. A recent survey reported that in the 7-yr period 2003–2010, trust in the oil, pharmaceutical, health insurance and tobacco industries are at the top of the list as those which the public believe need to be more closely regulated (Taylor, 2010). This mistrust is perhaps compounded by the public's discontent with the high cost of drugs and the fear that available drugs are ineffective or unsafe. Such fears are fuelled by media reports of scientific misconduct and fraud, such as those reporting the withdrawal of Vioxx (Garndener, 2004) and apparent misrepresentation of data of non-steroidal anti-inflammatory drugs (Ullman, 2007). Furthermore, there are also concerns that the scientific productivity of the pharmaceutical industry has declined in the last decade (Rafols et al., 2012).

Patients and payers want safer, more effective drugs and at lower prices. However, the cost of drug development continues to increase at an alarming rate which is a trend that is of major concern to the public. It is here that in the public's eye commercial and scientific goals become blurry. Such expensive pursuits require the kind of support that only industry can provide. Hence, the majority of drugs on the market have been developed by industry alone or in collaboration with academia. It is highly unlikely that industry-funded medical research will decline in the near future and therefore, it is in the public's best interest for industry and academia to address the public's concerns and restore the trust that has been lost. If these issues are not properly addressed, the credibility of the entire enterprise of academic and industry medical research will completely erode and the discovery and development of the next generation of drugs will no doubt slow down to the detriment of patients' best interests.

Conflicts of interest

COI, whether true, potential or apparent, often focus on financial gain, not because they are more prevalent but because they are more objective and easier to document. For industry scientists, disclosing potential COI is relatively straightforward; but it is often much more complex for academia or government scientists. Financial COI are not the sole source of conflict that investigators face when conducting clinical research. The publication of positive results will no doubt have an effect on the professional career of academic investigators, as federal or private funding is more likely to be awarded to those who have published in prestigious journals, which may disincline some investigators to publish negative reports especially if they are contrary to their own scientific agenda. Furthermore, clinical income as well as professional and social success may also benefit those with prestigious publications. Moreover, in some cases scientific misconduct may be solely ego-driven without any apparent financial motivation, as demonstrated by the search for fame demonstrated by investigators who authored the Korean cloning report (Bogner and Menz, 2006).

Given the complexity of these issues, the mere disclosure of COI is not enough (Thompson, 1993; Paul and Tohen, 2007; Tohen, 2007). In contrast, good clinical studies backed by sound methodology including choice of comparator and appropriate dosing will nullify the influence of any real or potential COI, regardless of the funding source or investigator affiliation. Additionally, bias in reporting and interpretation of medical research needs to be further evaluated by journal editors and peer-reviewers who have the responsibility to carefully review the interpretation of the results before they can be made public. Independent third party verification of study results also ensures fully transparent reporting of clinical trials from academia, industry or government. Many of the trials funded by industry have a strong financial interest in generating valid, reliable and reproducible data on new compounds to obtain regulatory approval or new indications for drugs already on the market. Results from industry studies are thoroughly scrutinized by regulatory agencies around the world. Poor methodology would likely lead to negative financial consequences. Therefore, efforts to eliminate COI should not solely focus on scrutinizing the source of funding, but also on employing measures that can reassure the public that any type of conflict will not have a voice.

Productive and meaningful collaborations among industry, government and academia are absolutely essential for the development of new therapeutics in psychiatry. The Bayh–Doyle Act of 1980, which enabled academic centres to retain the patents of inventions, has facilitated synergistic partnerships between industry and academia in the USA and has been considered a valuable step in the development of optimal pharmaceutical products (The Economist, 2002). However, the radical view that any scientific work conducted by industry scientists or supported by industry is not trustworthy threatens to reverse any benefits Bayh–Doyle may have bestowed by implying scientific endeavours and work for industry are inherently incompatible. Rothman (1991, 1993) suggested that such a position judges any investigator working with industry or any industry scientist as de facto fraudulent because of conscious or unconscious tendencies to favour the product of the sponsor or employer; additionally, it alleges an incompetence among industry scientists by insinuating they are unable to design a study that avoids bias. Considering industry's existing prominence and necessity in the drug development process, this continuing scepticism jeopardizes the very resources that make possible their continuing development about the proven safety and effectiveness of pharmaceutical products benefiting the public.

The government/industry/academia collaboration should also include data mining on existing large databases of industry-funded clinical trials. An example is a recent National Institute of Mental Health funded project where several pharmaceutical companies provided their databases to scientists from academia to develop new methodologies (Tohen et al., 2011).

Scientists working for industry have the obligation both to the public and to industry as a whole to restore the public's trust. Relying on goodwill and expecting ethical behaviour from all industry scientists may be non-intrusive but it is not practical because there is still the risk that a small number of individuals may engage in unethical practices. The pharmaceutical industry must ensure that its scientists follow rigorous methodology. Regulation is needed; ethical behaviour among industry scientists is expected, but law and ethics overlap. The involvement of government regulation enables the enforcement of law if needed. Academia is an ideal partner for industry to develop regulation considering that, while motivations may slightly differ at times, scientific pursuits are the same and interests in eliminating the risks of COI are mutually shared. Ethical scientific advice to industry from academic scientists should continue and be held to the highest standards as it benefits patients but as it may be perceived by the public as a potential COI, full disclosure is essential.

As suggested by Freedman et al. (2006) academia and the pharmaceutical industry need to jointly set their ethical boundaries and standards. Organizations such as the American Psychiatric Association or the American College of Neurospsychopharmacology, The European College of Neuropsychopharmacology or the Collegium International of Neuropsychopharmachology – which bring together industry, academia and government perspectives in an atmosphere of mutual respect for common scientific goals – have been the most effective sources to develop codes of conduct that can be embraced by all participants regardless of their interests (Paul and Tohen, 2007). Regulation of industry research by government agencies, however, rather than self-regulation would no doubt provide more reassurance to the public and also avoid COI.

It is also in the best interest of the public to have external monitoring processes applied to research conducted both by industry and academia. Unintentional bias is inherent in all scientific research. Good monitoring procedures detect all errors in design, implementation, analysis and perhaps most importantly the interpretation of the results of clinical trials, regardless of their intention. To dismiss the results of all industry-funded research is not in the public's best interest. Tohen (2007) has suggested that judging a study's value according to the investigator's affiliation or the source of funding is akin to prejudice based on ethnicity, gender, degree of physical attractiveness, or sexual preference of the investigator. The prejudicial interpretation of industry-sponsored research may lead readers attributing findings that favour the product of the funding source to COI to automatically discredit potentially valuable scientific findings.

Publication of industry-funded clinical trials

In the last couple of years, editorial boards of major medical journals, including psychiatric journals, have addressed the topic of collaboration between industry and academia. According to Lewis et al. (2006): ‘The collaboration between industry and academia is a useful part of the activity of many clinical investigators that enlists our best minds in the development of new therapeutics for our patients’. Furthermore, Freedman et al. (2006) wrote that ‘complete separation from the pharmaceutical industry is not the answer.’

Practices such as lead ghost authorship, which usually consists of having a recognized academic scientist named as lead author of a publication that reports an industry-funded study in cases where his or her participation was at best marginal, are unethical and should be avoided. Gotzsche et al. (2007) reported the incidence of ghost authorship of industry-funded studies to be as high as 75%. Another practice equally unethical but that has received less attention is ‘reversed-ghost authorship’ (Tohen, 2007) in which an industry scientist who designed a clinical trial, managed the implementation and played a major role in the statistical analysis plan, is not recognized as the lead author and is either delegated as one of the co-authors or not included as an author at all. This practice is clearly detrimental to ethical collaborations between academia and industry and should be avoided not only because it deprives industry scientists of well-deserved recognition, but also because it misleads readers of the actual authors and associations behind the trial. Other inappropriate practices that need to be eliminated are the use of industry marketing personnel who turn to external academics willing to author ‘promotional-driven’ manuscripts. Journal editors may actually favour a publication with an external lead author, rather than by an industry scientist. As stated by Wager (2007), the process of assigning authorship should preserve the purpose of informing readers about who contributed to the study, to not only grant credit, but also to clearly identify the influences behind the trial that may lead to bias in the design or interpretation of the results.

Manuscripts led by authors from industry, government or academia should be held to the same standard for publication. Regardless of funding or affiliation, all authors should follow the same guidelines, such as the Uniform Requirements for Manuscripts (URM) Submitted to Biomedical Journals, 5th edition, prepared by the International Committee of Medical Journal Editors (ICMJE, 1997). URM guidelines do not differ depending on the source of funding or affiliation. The use of professional medical writers has also become common-place in industry-funded research. Their role should be accurately described and should be limited to production of the manuscript and not in interpretation of the results. There may be, however, circumstances where medical writers make substantial contribution to a study and if medical writers satisfy the requirements of an authorship per the URM they should be included as such; otherwise, they should only be acknowledged in the proper section for their writing contribution.

Full access to industry-sponsored clinical trials is mandatory for regulatory agencies, such as the US Food and Drug Administration (FDA), the Japanese Pharmaceuticals and Medical Devices Agency, or the European Medicines Agency. The regulatory agencies have access to all relevant data on marketed compounds or those being reviewed for approval. More recently, at the request of the FDA, pharmaceutical companies have developed clinical trial registries where clinical trial data on all marketed products are routinely posted (http://www.clinicalstudyresults.org/home/) and are readily accessible by the public. Prestigious scientific journals such as the Journal of the American Medical Association have the option to require that analysis of industry-funded clinical trials be externally validated. Furthermore, the British Medical Journal recently indicated that, beginning in January 2013, it will no longer publish the results of industry- funded clinical trials unless the sponsors agree to provide detailed study data upon request (Thomas, 2012). Although regulatory agencies now require that all studies intended for registration be made public in clinical trial registries, there are still practices that need to be eliminated. As Colom and Vieta (2011) eloquently stated ‘History is written by winners’, alluding also to ‘winning drugs’ referring to the practice by industry of only publishing positive studies and either not publishing, delaying publication, or publishing negative studies in obscure journals. These practices need to be avoided by industry as well as the multiple repetitive publication of the same study or emphasis of post hoc finding without disclosure of the a priori primary outcome. Editors and reviewers are also at risk of publication bias by either only publishing positive studies or giving publication preference to unexpected findings. The latter practice could also include what appears to be the unexpected finding of the advantage of an older treatment over a new treatment which could make the publication more attractive to prove that new expensive treatments are not any better than older less costly treatments. Although this could be the case in many instances, reviewers and editors must make sure that the finding is not the product of selection bias or faulty interpretation of the results. Neither exaggerating nor minimizing the advantages or disadvantages of new treatments is in patients' best interests.

The recognition of the importance of collaborations between industry and academia was highlighted in a meeting held in March 2001 organized by The European College of Neuropsychopharmacology (Nutt and Goodwin, 2011) which was intended to address the consequences of what appears to be a decreased interest in neurosciences by major pharmaceutical companies. The summit was attended by leaders from both industry and academia. Recommendations from the group (Nutt and Goodwin, 2011) were comprehensive and included: (1) work on the ways to increase industry investment; (2) enhance research; (3) review regulatory process; (4) empower patients.

A strong and ethical collaboration between industry and academia is in the public's best interest. This collaboration is mutually beneficial because both parties share common goals, including the quest for scientific knowledge and better understanding of the pathophysiology and treatment of medical conditions. The relationship must, however, be a true scientific partnership rather one of sponsor–benefactor. Most importantly, in the interest of full disclosure, the partnership must be revealed. The industry should disclose publicly all medical research results that are significant to patients, health care providers or payers; whether these results are favourable or unfavourable to their specific products. The scientific report should be accurate, objective and balanced, so that caregivers can make informed decisions about industry products. Furthermore, a clear description of how clinical trials are developed and implemented by industry is essential to restoring trust.

Designing, implementing and analysing industry-funded clinical trials

Industry has the responsibility to ensure that the design, implementation, analysis and dissemination of its research are free of any bias about its integrity and commitment to patient safety. Industry-funded medical research should rigorously follow guidelines set forth by the Declaration of Helsinki's ethical principles and International Conference of Harmonization good clinical practices, and closely adhere to all applicable laws and regulations of the country or countries where a study is being conducted. The goal of clinical research conducted by the pharmaceutical industry should be to test the medications it produces either for the approved or intended indications. Medical research should never be used to promote pharmaceutical products. The pharmaceutical industry should only conduct medical research to answer scientific questions that are relevant to patient care.

Only the medical component of a pharmaceutical company should be accountable for the design, approval, implementation and results and the subsequent scientific dissemination of all clinical research funded by the company. Sales and marketing personnel should not be involved in the designing, approving or implementing any medical research or disseminating scientific results of research. Payment and other forms of remuneration to clinical trial investigators and institutions should reimburse amounts that are fair market value for legitimate, reasonable and necessary services.

Every step of clinical research must have the paramount emphasis on safety and well-being of study subjects. All pharmaceutical products developed by industry should have a safety surveillance plan based on non-clinical and clinical data. Once a clinical trial has been started, it must be terminated unless new safety or efficacy information suggests that the original scientific question cannot be reasonably answered. If the clinical trial is terminated, the sponsor must ensure the well-being of patients already enrolled.

Determining the countries in which to seek study sites to participate in industry-sponsored clinical trials should be based on the local prevalence of the disease being studied as well as other location-specific variables such as health needs, ethical practices and ethno-cultural variables. Ethical and scientific justification may not be universally applied. A reasonable degree of future availability of an unapproved drug or indication must be present in all countries from where patients will be recruited. Implementing clinical trials in developing countries that would not be approved in developed countries is highly unethical.

Pharmaceutical companies should have an independent well-structured adequately funded internal quality assurance group that reports to the senior medical leadership. The quality assurance group goals are to protect the rights and safety of subjects of investigation who participate in any medical research conducted by the company. Company employees should be able to report any suspected scientific misconduct committed by the company's investigators to the independent quality assurance group.

Importantly, the pharmaceutical industry needs to be committed to the balanced dissemination of scientific results from all clinical trials whether they prove favourable or unfavourable to their products. All results from industry-sponsored clinical trials need to be made available to the public in a manner consistent with industry-wide standards through publicly available clinical registries, such as clincaltrials.gov. In addition, clinical trial results must be submitted for peer-review publication regardless of whether the study results are positive, failed or negative.

Conclusions

If the pharmaceutical industry does not adhere to strict ethical standards both internally and externally when collaborating with academia, the public's trust will only be further eroded. Industry, academia and government, as the most knowledgeable and interested parties, must come together to establish the parameters of these partnerships. The pharmaceutical industry must understand because it is viewed as having the largest financial interests in research, it will have to adhere to the highest ethical principles and must avoid any appearance that it is using medical research as a means to circumvent laws that govern promotion of its products. Hence, the industry as a whole should adhere to practices that honour a respect for the goals and good clinical practices established by the ethical boards of professional bodies. Finally, it is responsible for creating a culture within its environment to dictate strict ethical codes and adherence from all its employees. Considering that the vast majority of pharmaceutical products are developed by the pharmaceutical industry, it is imperative to proactively participate in restoring the public's trust.

Acknowledgement

The author thanks Steven M. Paul and Alan Breier, who made important contributions to this article.

Statement of Interest

Dr Tohen was an employee of Lilly (1997–2008) and has received honoraria from or consulted for Abbott, AstraZeneca, Astrofarma, Bristol Myers Squibb, GlaxoSmithKline, Lilly, Corceps, Johnson & Johnson, Otsuka, Merck, Sunovion, Forest, Roche, Elan, Lundbeck, Wyeth and Wiley Publishing; his spouse is currently a Lilly full-time employee.

References