The “gold standard” for evaluating a new approach to therapy, a novel medicine, or a new procedure is the “controlled, randomized, double-blind trial.” In this protocol, which is used in almost all trials of new drugs by the FDA, subjects are split, randomly, into two groups. This randomization should remove the influence of factors such as gender, age, race, and underlying medical conditions. Group A is given the real drug and group B is given an identically prepared inert drug, known as a placebo. In a variation of this protocol, both groups are given standard therapy; Group A gets the new drug and Group B a placebo.
The study is referred to as being “double blind” because the subjects don’t know which group they are in, nor does the investigator evaluating them. Certain outcome measures are selected before the study is started, and the subjects are followed for a chosen period of time. An outside group monitors the study for safety and efficacy. When the study is complete, the “blindfold is lifted” and the investigators compare the outcomes in the two groups.
In the article “Placebo Effect Stronger Than We Thought?” attention is directed toward a phenomenon that has been recognized for many years, but only recently characterized and evaluated. That is, when a subject, or patient, is given a placebo there can be responses similar to what we anticipate with the “real” medicine or procedure. The article by Beryl Lieff Benderly focuses on two areas: the placebo effect in patients with Parkinson’s disease and how the placebo effect results in difficulty in designing and interpreting clinical trials.
The placebo effect is in no way limited to Parkinson’s disease. It has been observed in studies of many other diseases of the brain, but particularly migraine, other forms of pain, depression, and even epilepsy. For those who wish to read more about this area, there is an excellent article by Fabrizio Benedetti and his colleagues, “Biological, clinical, and ethical advances of placebo effects” in the British journal Lancet. The article emphasizes that the placebo response is part of the overall therapeutic context. Thus when either a placebo or real drug is given, the response is the result of not only the effect of the treatment, but also the circumstances in which it is given. Further, the results can be thought of in terms of two mechanisms: the psychological and the neurobiological.
The psychological mechanism can, at least in part, be summarized as the “power of suggestion.” There is also considerable variation in how suggestible a person will be. Some are more likely to have a placebo response than others. There are a lot of components to the psychological response, but one of the leading factors is expectancy. What does the subject expect to happen? The circumstances and behavior of the physician play a big role here. If I tell you that I am giving you an injection to help your pain, your perception of pain will be less. On the other hand if I tell you that the injection is a blank, the pain will be perceived as not improved or even as getting worse. There are numerous examples in which this type of manipulation of results have been demonstrated—not just in pain, but also motor symptoms in Parkinson’s disease, depression, and addiction.
If you are a biologically-oriented neuroscientist, you may have a tendency to downplay the psychology aspect of things. But in the case of the placebo response, there is increasing evidence of demonstrated neurobiological mechanisms. Most of these responses are in the release or metabolism of neurotransmitters or hormones. This has been true of the release of dopamine in the basal ganglia in Parkinson’s patients and the release of endorphins (the brain’s endogenous opium-like compounds) in pain. Other examples are listed in the aforementioned article by Benedetti.
The placebo response has plagued clinical studies not only in Parkinson’s disease, but also in studies of depression, as mentioned by NIMH director Tom Insel (“Antidepressants: a Complicated Picture”). This is not an easy problem to solve. One attempt was devised by my colleague, Tony Ho, then at Merck, who faced varying placebo responses in a trial of a new drug for migraine. His approach was to start both groups, A and B, on placebo for a period of time. Those that showed a placebo response were then analyzed separately at the end of the study. In other words, he was trying to get the placebo responders out of the study.
I am sure there will be other modifications employed to minimize the placebo response effects in clinical trials which will require flexibility in the designs of clinical trials.