ARTICLE IN BRIEF
Is it ethical to randomize study subjects to a placebo group if a standard effective therapy is available? Ethicists and clinical trial investigators discuss the pros and cons of the issue.
The best evidence for new treatments comes from randomized, placebo-controlled, double-blind studies. However, when an effective treatment for a disorder already exists, is it ethical to create a placebo group that will receive no treatment at all?
Multiple sclerosis (MS) provides a particularly thorny example, as a series of letters to the editor in the May 6 issue of the New England Journal of Medicine (NEJM) demonstrated. On the one hand, interferon beta and other treatments reduce recurrences of the myelin inflammation that coincides with numbness, paralysis, and other neurological symptoms. On the other hand, the treatments, which often cause fatigue and flu-like symptoms, require periodic self-injection or IV infusions, and they're only modestly effective.
The letter writers were responding to three Feb. 4 papers in the NEJM that tested the effectiveness of two new MS treatments — oral fingolimod and oral cladribine. Investigators compared data from 1,272 patients who completed the FREEDOMS study (FTY720 Research Evaluating Effects of Daily Oral Therapy in Multiple Sclerosis) on two doses of oral fingolimod (FTY720) to placebo, and found that the drug reduced relapses by more than 50 percent. The CLARITY study (Cladribine Tablets Treating Multiple Sclerosis Orally) randomized 1,326 patients into three groups — two received different doses of the drug, and the third received a placebo. Those who took the cladribine also reduced their relapses by more than half compared to the placebo group.
The third study, however, did not include a placebo group. Instead, the TRANSFORMS study (Trial Assessing Injectable Interferon versus FTY720 Oral in Relapsing-Remitting Multiple Sclerosis) compared two different doses of fingolimod to intramuscular interferon, a common treatment. Fingolimod reduced recurrences by up to half compared to the interferon group.
In his letter to the NEJM, Howard Mann, MD, professor in the Department of Radiology and program associate in the Division of Medical Ethics at the University of Utah, pointed out that the studies involving a placebo as the only control intervention denied an established effective therapy to patients allocated to that arm of the study.
“I asked them to justify the conduct of trials wherein the control intervention was not an established effective therapy, as was used in one of the three trials,” Dr. Mann said.
The letter illustrates in a vivid way the conflicting viewpoints about the use of placebo groups in studies involving diseases for which effective treatments already exist, said Jeffrey A. Cohen, MD, professor of neurology at the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University and the lead author of the TRANSFORMS study.
“Both my study and the FREEDOMS study, a placebo-controlled trial, showed benefit on relapses and on MRI, which are the more sensitive measures,” Dr. Cohen said. “But the placebo-controlled trial also showed benefit on disability, whereas my trial was unable to do so because disability is a less sensitive measure. That requires a placebo-controlled trial.”
Dr. Cohen coauthored a 2008 article in Neurology, which advocated restrictions on placebo-controlled MS trials. “For patients with relapsing MS for which established effective therapies exist, placebo-controlled trials should only be offered with rigorous informed consent if the subjects refuse to use these treatments, have not responded to them, or if these treatments are not available to them for other reasons (e.g., economics),” the authors wrote.
Dr. Cohen still agrees with that conclusion, but admits that placing MS patients in a placebo group can create tension between the best care of individual patients and the goals of research.
“In the case of placebo-controlled trials, there are two conflicting principles,” said Dr. Cohen. “One is non-maleficence, which means, basically, do no harm. When you have a potentially disabling disease like MS and you have effective treatments, it's not ethical to treat some with placebo. On the other hand there's an ethical principle called respect for persons, which means that informed patients should be allowed to have the autonomy to make informed decisions.”
Some MS patients, for example, dislike injecting themselves with interferon beta, or perhaps interferon hasn't worked for them. Such patients, according to Dr. Cohen, should be allowed to choose to participate in a placebo-controlled trial of a new drug.
“As long as you appropriately inform them of the potential risks of being in the trial, and incorporate appropriate measures so you don't let them get worse and worse, the respect-for-persons ethical principle would say it's up to them,” he said. “The hierarchy of ethical principles is still contentious among ethicists. Which of those two principles should take precedence — autonomy or non-maleficence? They're mutually contradictory.”
AN ETHICAL HIERARCHY
Clinical trials that include a placebo group can be ethical as long as the patients are selected in an ethical manner, said Winston Chiong, MD, PhD, a neurology fellow in the Department of Neurology at the University of California-San Francisco School of Medicine. In an editorial in the February Annals of Neurology, Dr. Chiong took issue with another article in the same issue that attempted to establish an ethical hierarchy for decision-making during medical emergencies.
The authors of the article argued that physicians treating an acute stroke, for example, should offer therapy that has been proven effective in randomized, blinded, controlled clinical trials. If no such treatment has met that standard, and if the patient meets the enrollment criteria for an open clinical trial, the patient should be enrolled. The next best option, according to the authors, would be to offer therapies as part of Phase 4 registries that would result in the collection of data on treatment outcomes.
Only in the absence of the above options should physicians use their “best judgment” based on their experience and on published case studies or anecdotes.
“Given the paucity of quality randomized clinical trial data for most medical decisions, the ‘best judgment’ option will be used most frequently,” the authors stated.
Such a hierarchy, in Dr. Chiong's view, applies a rigid rule to what should involve a flexible decision made by the doctor and the patient based on the particular features of the patient's condition. An MS patient who has an insurmountable phobia to self-injecting interferon, for example, or who has not responded to interferon, would not be sacrificing treatment by enrolling in a clinical trial of a new drug that included a placebo group.
The key, in Dr. Chiong's opinion, involves informed consent, but in his editorial he points to data showing that patients frequently misinterpret information presented to them because they believe their physician has their best interests in mind. This potential misinterpretation has been called the therapeutic misconception.
“It may be especially difficult for some patients to understand the use of placebo controls in a case when effective treatments are known to exist, since the physician is proposing to withhold an effective treatment for the sake of science,” Dr. Chiong said. “In these cases it's very important yet very challenging for investigators to ensure that consent is valid.”
In addition, the conflicting roles of scientist and clinician may create “cognitive dissonance” that may lead investigators to overestimate the value of the knowledge derived from a clinical trial, and underestimate the cost or burden to individual patients.
While physicians are responsible for the patients they treat, they're also responsible for future patients “who stand to benefit from medical knowledge gained by testing interventions in a controlled fashion,” Dr. Chiong wrote in his editorial. The challenge, then, involves finding a way “to strike an appropriate balance between the interests of present and future patients. It seems very unlikely that a blanket [ethical] hierarchy … will be a useful tool in helping physicians and clinical investigators to manage the moral conflicts that arise in clinical trial enrollment.”