Abstract
Randomized placebo-controlled clinical trials have been the ‘golden standard’ during the last decades in the development of new drug therapies. This scientifically valid approach has recently been questioned in the fifth revised version of the Declaration of Helsinki, which states that the use of placebo-controlled clinical trials is only acceptable when no proven treatment exists for the studied disease. The World Medical Association further claims that no national ethical, legal or regulatory requirements should be allowed to reduce or eliminate any of the statements in the declaration. In spite of this, the document is not generally accepted as the world ethical standard, as demonstrated by its lack of adoption by many professional associations. In the evaluation process for a drug to be approved in many countries today, clinical investigators at the hospitals and researchers at the pharmaceutical companies are obliged to use study protocols that would be rejected if the new declaration were to be fully adopted. Adherence to the clinical trial guidelines of the International Headache Society could also mean violation of the new Helsinki declaration of ethics. Some ethics committees have already adopted the new declaration, which has caused concern among clinical investigators, who find this document to be vastly out of the line with common practice. At the moment, the situation is unclear and debated with increasing polarity concerning the scientific and ethical issues regarding the use of placebo in clinical trials.
Influential guidelines deeming placebo unethical
It is a common opinion, and also a finding in interviews, that patients do not join clinical trials for scientific reasons. Expectations of benefiting personally, rather than altruism, is more often their motive for participating (1–3). Patients generally rely heavily on the advice of their physician. Consent may therefore not always be ‘informed’, and the patient might not be aware that he or she risks sacrificing clinical benefit for the sake of helping others.
Some surveys indeed suggest that physicians underestimate patients’ need for information (1,4). The consent forms are sometimes not informative enough and may be signed without being understood. One reason for unclear information could be that physicians outside the trial context, in the usual clinical setting, may perform almost whatever procedure they choose without providing detailed information, seeking formal consent or asking a review committee for approval. This is a serious moral dilemma for the physician and may also cause concern. The physician may be experiencing a conflict between his role as the patient's personal healer and on the other hand his role as a modern scientist with a commitment to acquire knowledge. Furthermore, very comprehensive consent forms have been found to reduce willingness to participate in a study and increase anxiety (1,3–6). Nevertheless, if true informed consent is not sought for human experiments, the participant is treated as an object or thing, rather than as a person or co-adventurer. Needless to say, this is deception, which is unethical in a situation of trust and dependency (3–5). Medical ethicists, some researchers and patient groups have on these grounds questioned the use of no-treatment (placebo) in clinical trials (2,7–9).
In October 2000 the World Medical Association (WMA) voted to approve the fifth revised version of the 1964 version of the Declaration of Helsinki (10), which is often referred to as the cornerstone of clinical research ethics (5,8,11). According to the document, a researcher's role as a physician and healer must take precedence over his or her role as a scientist. The official main objective is to protect the patients. Much disagreement centres around the following controversial paragraph: ‘The benefits, risks, burdens and effectiveness of a new method should be tested against those of the best current prophylactic, diagnostic, and therapeutic methods. This does not exclude the use of placebo, or no treatment, in studies where no proven prophylactic, diagnostic or therapeutic method exists’ (10). In other words, if there is an appropriate drug already on the market, then a trial should compare any new treatment with the existing product. The use of placebo, even with informed consent, would be unethical. Placebos are only acceptable when no proven treatment exists (7,11–13).
It has been proposed by ethicists that the rules of the declaration should be relaxed to allow the use of placebos if they do not cause death or disability (7,14) This standpoint is suggested to be maintained even for trials where the disease is not life-threatening. Professor Kenneth Rothman at Boston University still claims that the declaration is essentially sound, and that no patient should have unnecessary pain or undertake unnecessary risks, no matter how small and no matter the impact on research. It is argued that placebo-controlled trials are unnecessary and result from ‘a misplaced emphasis on statistical significance testing’ (15).
The US National Bioethics Advisory Committee (NBAC) has also called for an end to placebo-controlled studies in diseases where an effective therapy already exists. Both the WMA and the NBAC are highly respected. Their new guiding principles for medical research push hundreds of ongoing clinical trials beyond the bounds of ethical acceptability and necessitate changes in the design of future drug trials (7,12).
In defence of placebo
The research community considers the active-control (also known as non-inferiority) approach to be, in many circumstances, ‘bad science’. The worst scenario for patients would be misleading publications and erroneous approval of drugs that do not work. The disagreement focuses on trials of therapy for non-life-threatening conditions, in which a delay in administration of the effective treatment is unlikely to cause permanent harm (11,12,14,15). Several authors believe that patients can indeed be properly informed about the options and asked to participate in a placebo-controlled study. An absolute condition is that the patient is not permanently harmed or put at significant risk during the trial (13,14,16). Concern about whether consent to participate in trials is as informed as we would like to believe is valid, but the problem is not unique to placebo-controlled trials (3,14). It is the belief of the authors that the possibility for the individual participating in a clinical trial to withdraw from taking part in the study at any stage without further ado is one of the best, if not the best, ethical arguments for keeping the randomized placebo-controlled clinical trial standard.
Comparing new drugs with placebo is often the only way to establish efficacy, especially in treating diseases where the effectiveness of the available treatment is modest and inconsistent (4,7,13–15,17,18). If a new drug and an active comparator evoke similar responses without placebo control, it cannot be categorically claimed that either was effective in improving the outcome over the natural history in this particular experiment conducted in this particular population as this was not tested (6,11,19). Clinical trials that seek to prove that a new agent and an active control have similar efficacy are inherently less reliable than trials that seek to prove the superiority to a comparator, whether active or inactive. Increasing the size of trials does not always reduce this problem (17). Indeed, it is in the patients’ interest to avoid or delay use of marginally or inconsistently effective standard medications and it may therefore be ethical to conduct placebo-controlled clinical trials in such situations (6,13,15,17).
The number of subjects who are exposed to unproven, albeit promising, investigational drugs should be kept to a minimum. Empirically obtained evidence suggests (20) (although seemingly paradoxical) that placebo controls can reduce the number of subjects who remain symptomatic in randomized controlled trials. This is based on the inverse relationship between sample size requirements and the detectable population effect size (6,20). Trials with active controls only may therefore expose more patients to harm than placebo-controlled trials (6). In addition, studies with active comparators tend to require larger numbers of patients than placebo-controlled studies to demonstrate equivalence or establish superiority (5,19,20).
The assumption that historically active controls always work better than placebo can not be generally extrapolated. There is normally considerable inter- and intraindividual variation in the study population, which renders comparison with historical controls impossible. Including a placebo-group, however, increases the scientific stringency by defining the study population. A placebo-response in the same range as in other similar studies suggests an acceptable randomization with an even recruitment of responders and non-responders. For these reasons, treatment with an active comparator may not be a better alternative to placebo. In addition, the effect and tolerability of the new drug to which the patients are exposed remain uncertain until the trial has been completed (4,12,13,21). Ideally from a scientific point of view, a new drug should be tested against both placebo and active control. In fact, the three-armed study-design provides information about the tested drug, the active control and placebo. However, such three-armed studies would be prohibited by a strict interpretation of the revised declaration (6,11,13).
The policy of the US Food and Drug Administration (FDA) is to require placebo-controlled trials before a new drug is approved. The FDA therefore disagrees strongly with the new Declaration of Helsinki (5,7,13,22). The European Agency for the Evaluation of Medicinal Products (EMEA) and the Committee for Proprietary Medicinal Products (CPMP) are responsible for providing scientific opinions to the European Commission for the granting of marketing authorizations for drugs within the European Union (EU) (17). It is the official position of the scientific committee of the EMEA/CPMP that forbidding placebo-controlled trials in therapeutic areas where there are proven therapeutic methods would be contrary to public health interest (8,17). The American Medical Association, the World Health Organization and the Council for International Organizations of Medical Sciences have also rejected the position that the declaration unequivocally bars placebo-controlled trials when proven therapy is available (14).
Ethics of conducting placebo-controlled trials in children and adolescents
Children are a vulnerable group in need of additional ethical safeguards, as they cannot protect their own interests through valid informed consent (22–24). Long-term effects on growth and development are always possible (4). The new Helsinki Declaration states that research should not include people who are legally incompetent, incapable of giving consent or under the age of majority unless necessary to promote the health of such a group and not otherwise attainable with legally competent people. This contravenes other guidelines, which mandate that children should be included in research unless there is good scientific and ethical cause for their exclusion (22). Whether or not a controlled trial in children can be ethically justified depends on such factors as the validity of control-treatment, the seriousness of the disorder, the likelihood of adverse consequences if treatment is delayed, the burden of the proposed intervention and value judgements by parents, working in collaboration with a caring physician (23,25,26).
Placebo-controlled trials raise particularly difficult ethical issues concerning the protection of children (24). It has been argued that a 50 : 50 chance of treatment in a placebo-controlled trial would be better for a child than no treatment at all, and the design would yield the strongest efficacy data in drug testing (5). However, a non-inferiority trial using active control would leave no child untreated (5). This argument is only valid if the studied drug has an effect. Good science sometimes dictates studies with placebo arms if populations differ from those previously studied (including children vs. adults) (11). The number of children who are exposed to unproven, albeit promising, investigational drugs should be kept to a minimum until preliminary placebo-controlled trials support the use of these drugs (20). Adolescents are another group in which constraints over participation in research arise from the requirements for informed consent (4).
Migraine research aspects
Representatives of the FDA and of Public Citizen, a consumer rights group in the United States of America (USA), do not believe that it is unethical to use placebos in short-term studies of drugs for migraine or other headaches or in fully informed patients (5,11,14). Bioethicists also argue that there is no meaningful harm that ethicists should worry about in letting a person who has given informed consent continue to suffer temporarily from a headache as part of a clinical trial (6). They also argue that placebo-controlled trials have a sound scientific rationale when the studied population has a high placebo-response rate; the condition is typically characterized by a waxing-and-waning course, frequent spontaneous remissions and existing therapies being only partly effective (6). All these criteria are fulfilled by migraine. Harm should be minimized by the administration of rescue medications (6).
According to the present guidelines of the International Headache Society (IHS), ‘placebo controls may and should be used to establish efficacy of a new drug, acute or prophylactic, that is expected to have some benefit over accepted therapies’ (19). However, the ethics subcommittee of the IHS, chaired by Dr Timothy J. Steiner, emphasize in their report, that the use of placebo in headache trials always hinges on informed consent with careful explanation of difficult terms and details of other treatment options that a patient could forego. The subcommittee rejects the argument that even informed patients may not be disinterested enough to decide rationally whether it is tolerable to be deprived of an accepted treatment. Although it has been argued that an informed patient may choose to accept pain or discomfort for a short time or to defer needed long-term therapy for a short time to participate in a placebo-controlled trial, this does not mean that indifference to patient discomfort is appropriate. It is possible to limit the duration of placebo exposure by an ‘early escape’ design that offers protection for the patient (13,14,19). All trials with new migraine drugs therefore allow rescue medication after 2 or 4 h. This reflects daily life, where patients may have no immediate access to their migraine treatment (27).
The unclear status of placebo in migraine research has led to differing doctrines between ethical committees. In the preparation of a three-armed multinational study, a 1 : 1 : 1 (triptan A : triptan B : placebo) randomization ratio was planned. The ethics committees of several countries objected to the inclusion of a placebo group, as an active comparator was available. They also argued that, despite the availability of rescue medication after 2 h, it was inappropriate to expose more than a limited number of patients to placebo. It was therefore decided to adjust the randomization ratio to 8 : 8 : 1. The patients were fully informed that the likelihood of receiving placebo was 1 : 16. Compared with earlier double-blind, placebo-controlled trials of the same triptans, the present study had an exaggerated placebo response. One reason could be that the unbalanced randomization created a high expectation of receiving active drug. Another shortcoming of the design was wider confidence intervals, requiring more patients to detect a difference between the treatments (27,28).
Conclusion and prospects for the future
Is it morally justifiable to sacrifice the patient's right to completely individualized treatment for the benefit of humanity and for the advancement of science? In a clinical trial, a moral dilemma exists at the outset between gaining knowledge that can be used in the longer term to benefit public health and the basic right of the individual patient to receive treatment (5,8,29). Randomized clinical trials involve a design of procedures that systematically manipulate subjects, using a control procedure for the purpose of gaining knowledge. When conflicts between ethical considerations and the trial's scientific quality occur, the ethical considerations must therefore be paramount. However, for clinical research to be ethical, it must be scientifically valid (6). The new Declaration of Helsinki is not accepted as the world ethical standard, as demonstrated by its lack of adoption by many professional associations. The process of revision of the declaration has been criticized. None of the four-member committee was recognized as having expertise in the conduct of clinical research or bioethics. The process was secretive and rushed, conflicting with the document's own spirit of transparency and disclosure (22).
In order to test a drug for approval on many markets (for example the USA), researchers and drug companies today choose to use protocols that are specifically rejected by the new declaration. Also, adhering to IHS guidelines could mean violating the new Helsinki declaration of ethics. The view of some ethics committees that a placebo group is not warranted in acute migraine trials is considered to be vastly out of line with common practice by leading scientists in the field representing a social welfare-centred rather than a patient-centred ethic.
It is claimed by the WMA that, ‘No national ethical, legal or regulatory requirement should be allowed to reduce … this Declaration’ (10). The declaration thereby claims priority over national laws and regulations. The authority and practical meaning of this self-proclaimed status is unclear (22). In fact, neither the Helsinki Declaration nor the NBAC report have the power of law (1,3–7,11–13,27,30). A working group has been appointed by the WMA to reconsider certain provisions. The General Assembly of the WMA is the only body with authority to adopt formal changes to the Declaration of Helsinki. This can not take place until their next conference (22,31,32). An extraordinary step has been necessitated in the form of a recently published note of clarification that can be found on the official WMA website. The Association:
affirms its position that extreme care must be taken in making use of a placebo-controlled trial and that in general this methodology should only be used in the absence of existing proven therapy. However, a placebo-controlled trial may be ethically acceptable, even if proven therapy is available, under the following circumstances:
• where for compelling and scientifically sound methodological reasons its use is necessary to determine the efficacy or safety of a prophylactic, diagnostic or therapeutic method; or
• where a prophylactic, diagnostic or therapeutic method is being investigated for a minor condition and the patients who receive placebo will not be subject to any additional risk of serious or irreversible harm (31).
The situation is confused at present, with increasing polarity about this issue in the research community.