PDF is committed to using your funds efficiently to support our mission. To see how funds are spent, browse our financial information.
The Placebo Effect, How it Complicates Parkinsonís Disease Research
By Christopher Goetz, M.D.
To find out if a potential new treatment for Parkinson’s disease (PD) is effective, it is necessary to test it in people, through clinical trials. If you are a person who is touched by Parkinson’s disease, perhaps you or your loved one has participated in a clinical trial.
It is through clinical trials that scientists investigate the effects of a treatment in people, particularly its safety and efficacy. The advances we have made in understanding Parkinson’s, and the medications that you may be taking, are available only because people before you volunteered to participate in clinical trials.
In later stages of clinical trials, researchers usually use a methodology known as the “double blind” trial. In this type of a trial, the experience of a group of participants who are taking the treatment is compared with that of a group of participants who are taking an inactive substance called a “placebo” — typically, a simple sugar pill. Participants go into the trial knowing that they will receive either the active treatment or a placebo. But neither the participants nor the researchers know who is taking the real treatment and who is taking the placebo.
Researchers use placebos and blinded trials to better understand the true effectiveness of a new treatment. The placebo group is used as a “baseline” against which the treatment group can be measured. Theoretically, those participants who receive the placebo will not see any change in their Parkinson’s symptoms. But in nearly every Parkinson’s clinical trial, some people who take the placebo report that it eases their symptoms. We call this the placebo effect.
What causes the placebo effect? One cause may be a person’s expectations. Trial participants typically receive more medical attention than usual, including frequent visits to the clinic, laboratory tests, and frequent contact with doctors. Studies have found that these aspects of a trial may lead the participant to expect his or her symptoms to improve. This, often partnered with the hope that the treatment will work, feeds the placebo effect.
In the usual doctor-patient relationship, the placebo effect can be a wonderful thing because we want our patients to feel better. But in a clinical trial, it can confuse our results.
The Science Behind the Placebo Effect
It is important to note that the placebo effect is not a figment of the participant’s imagination. There are biochemical changes occurring in the brain. We think the placebo effect may be so prominent in Parkinson’s clinical trials because of the neurotransmitter called dopamine — the same neurotransmitter that is reduced
or lost in Parkinson’s. It turns out that dopamine also underlies the placebo effect. When a person is motivated to participate in a trial and anticipates a possible reward — for example, the easing of symptoms — these all boost dopamine activation in the brain.
Two studies have investigated how dopamine is involved in the placebo effect in people with Parkinson’s. In one of these, people with PD were treated with a placebo and then given brain scans measuring dopamine activation in the striatum (the brain area that exhibits reduced levels of dopamine in Parkinson’s).
Some of the people in the study, as measured by objective methods such as the Unified Parkinson’s Disease Rating Scale (UPDRS), experienced an improvement in their Parkinson’s symptoms. The brain scans showed that in these people, dopamine activation in the brain was markedly higher than in those whose symptoms did not improve while on the placebo.
Another study investigated dopamine activation at the cellular level. During the surgical procedure known as deep brain stimulation, people with Parkinson’s have an electrode placed to measure the activation of dopamine cells. After a placebo treatment in the operating room, the study leaders found that some participants experienced improvement in their rigidity, a hallmark symptom of Parkinson’s. They also found that in those individuals, the patterns of cellular firing in the brain had actually changed.
The Placebo Effect and PD Clinical Trials
In nearly every trial, some people who take a placebo will report that their symptoms have improved. And that means that the new treatment being tested has to be not just better than before enrollment in the study, but also better than the placebo in order for its effects to be considered significant. The placebo effect “raises the bar” for a new treatment to be declared effective. Because of this, my own research team at Rush University Medical Center in Chicago undertook a study to understand how the placebo effect impacts Parkinson’s trials.
We looked at data from placebo groups of 11 published trials, comprising a total of 858 participants, exhibiting a wide range of PD severity with a variety of treatments. We defined placebo response as an improvement of 50 percent on the UPDRS. This means “placebo responders” had a reduction by half in their disability simply by being involved in a clinical trial.
We found that 16 percent of people in placebo groups experienced improvement in Parkinson’s symptoms. The improvement was markedly greater for surgical trials, in which 42 percent of the study participants receiving placebo experienced improvement in symptoms, than for drug trials, in which the rate of placebo response ranged from zero to 27 percent.
We also tested an assumption common among researchers that the placebo effect is strongest at the beginning of a study, and wears off as it progresses. But we found no differences in placebo response rates at various stages of the studies. This means that we cannot filter out the placebo effect by ignoring data collected during an initial “placebo period.”
The Investigator’s Dilemma
Going back to my earlier point, the “gold standard” in establishing whether a new treatment works is whether it has a significant effect as compared to the placebo. When a clinical trial fails, one explanation is that the new treatment simply does not work. The other explanation is that the treatment does work, but that the evidence is overshadowed by the placebo effect. Our inability to “factor out” the placebo effect will make it difficult for treatments that have small, but important impacts on Parkinson’s symptoms to succeed.
On the one hand, we can say that these studies should fail, because we’re interested in treatments that can dramatically improve life for people with Parkinson’s. But for many people, even a modest improvement in one of their symptoms may be seen as a welcome result. Besides which, in clinical trials that show a placebo effect, evidence that a treatment will have a significant impact would require the enrollment of hundreds of participants. But this means asking people to volunteer just to overcome the placebo effect, not to get the data evaluating the true impact of the treatment. The time and money required for this will often be prohibitive.
Researchers are investigating other avenues to account for the placebo effect in Parkinson’s trials. One approach would be to find ways to identify people who respond to placebo treatment, the “placebo responders,” and screen them out of participation in the trial. Then, the number of participants could be kept to a minimum. On the other hand, it may be that the same people who will respond to placebos have a very active dopamine system, and may be the same ones who will respond to new treatments with small effects.
Many people ask, why not calculate the typical magnitude of the placebo effect and subtract it from our data to determine the “true” effect of the new treatment? For example, if we can say that on average, 16 percent of response is from the placebo effect, then anything more could be attributed to the treatment. The trouble with this idea is that the placebo response seems to be different in every trial, so we can’t construct a reliable formula.
Another approach would be to design trials with three groups of participants: those who receive active treatment, those who take a placebo and those who receive no treatment. But in this scenario, two thirds of participants would not receive the active treatment, meaning there is an even higher likelihood than in current trials of a participant receiving no treatment. The difficulty would be in finding hundreds of participants who are willing to go through the process to find that they are not receiving any treatment at all.
There is one other approach, used in some countries but not, for ethical reasons, in the United States. It is where study participants are not informed that they could be receiving a placebo drug. There is intentional deceit of the participants. To me, this breaches trust between doctor and patient, or investigator and study subject. But we will monitor scientific outcomes.
Researchers are working to understand the placebo effect — how it changes the brain, and how it affects the results of clinical trials.
If we can characterize it more precisely, then perhaps we can use it in two ways: in the clinic, to turn it to our advantage and help people with Parkinson’s feel better from their treatment and placebo effects, and in research, to decrease placebo effects so that we can design clinical trials that evaluate therapies more accurately.
Christopher G. Goetz, M.D., is Director of the Parkinson Disease and Movement Disorders Center at Rush University Medical Center, Chair of PDF’s Medical Policy Committee and a member of PDF's Scientific Advisory Board.