Monday, August 16, 2010

Sham Surgery

In clinical trials that test new drugs for Parkinson’s disease, the process is fairly straightforward: some people get the real stuff and others get a sugar or placebo pill. It looks identical to the pill containing the new drug, but doesn’t offer the new compound. Here everyone (the person with Parkinson’s and physician) enters the trial knowing that a certain amount of deception is necessary from the outset….no one knows which pill is which. That is a good thing scientifically.

There is an overwhelming consensus that double-blinded experiments, in which participants are randomly assigned to receive a placebo, result in the strongest and most reliable evidence – which means they give us the best indicator of which experimental treatments would truly help people living with PD and, just as importantly, help shield people from treatments that don’t.

This scenario gets more complicated when the new “drug” is actually a surgical intervention, in the case of PD, brain surgery. What is the surgical equivalent of a sugar pill? It’s called sham surgery and it’s not simple.

So the National Institutes of Health (NIH) held a two day conference, part of which is now available to view online via videocast, to discuss the scientific and ethical considerations underlying sham surgery for neurodegenerative diseases, chiefly Parkinson’s disease.

Why Sham Surgery?

Essentially, a clinical trial is an experiment where the results are unknown. Sure, previous data or testing indicates it is worth spending loads of money on it; but, really, no one knows how the experiment or trial will least that is how it is supposed to work. But because clinical trials are experiments involving humans, the process can quickly be biased by what people think or are expecting to happen: people with Parkinson’s believe they may be getting better and their physicians may too easily agree. Enter the very powerful and very real placebo effect in PD (see here and here for two studies highlighted by panelist Jon Stoessl, M.D.).

So if you have PD and the placebo effect can help you feel better, what is the big deal? Why not spend all this energy making the placebo effect better? The problem is the placebo effect works by convincing the individual something is real when it is in fact not. Before science entered medicine, placebo was mostly the only game in town. So while studying the mechanism underlying the placebo effect is a worthy vein of research, the placebo effect remains no substitute for medical therapy.

A solution has been to control for the placebo effect by keeping everyone in the dark (also called blinding or masking) about who is getting the experimental treatment and who is not. This way, designers of clinical trials try to eliminate the obvious forms of bias, like the placebo effect, as well as not so obvious forms of bias that are harder to predict.

Many PD therapeutics, however, start out as open-label trials where no secrets are kept and everyone knows who is getting the experimental treatment. It may not be surprising, as noted by meeting panelist Steven Piantadosi M.D., Ph.D., that these early trials often end in success —frequently demonstrating a 30 percent improvement in PD symptoms — yet are later thrown upon the bin heap of failure once the more rigorous experiment is conducted. With so many successes followed by failures, what is the price to the person with Parkinson’s?

Looking Ahead
There is no easy answer to this multi-faceted problem. However, several ideas emerged from the discussion at last month’s meeting.
  1. Be more rigorous in the preclinical stage. Without a true model of PD in animals, investigators need to be more thorough in understanding the biology behind a treatment before trying it in humans. (Mice are not people.)
  2. Early trials should be large enough to detect potential adverse events and should include some form of blinding to make them most informative about future directions.
  3. Sham controls should minimize the risk to the participant by using minimally invasive procedures.
  4. If the results of an open-label study are used to decide whether to go forward with more testing the treatment effect should be very large.

These are just a few of the suggestions that emerged to minimize the use of sham surgery while maximizing its scientific potential when employed. As the draft recommendations are refined by the panel participants, they will be available for public comment. We will let you know when that happens, so be certain to check back.

What Do You Think?

So what are your thoughts on how clinical trials are designed and conducted? When would it be appropriate to use sham surgery?


J said...

This type of thing has been talked about with regard to acupuncture, etc. Sham surgery isn't really the equivalent of a double-blind study: while the patient may not know if the surgery was real or not, the surgeon performing the operation does know.

To make things more parallel to a double-blind study, a surgeon who was not involved in the procedure would have to take over after the surgery was done and monitor recovery. He or she would then hand the patient over to the clinicians, who would be "in the dark" as to whether it was a sham surgery or not.

LindaH said...

I watched most of the first day’s webcast of the NIH conference. Maybe the following was discussed elsewhere, but it seems most of the researchers at the conference believe that any improvement noted by the trial participants must be due to the placebo effect, and couldn’t possibly be due to the experimental treatment working
You wrote:
“The problem is the placebo effect works by convincing the individual something is real when it is in fact not.”

Is it not possible that in open label trials , where improvement is often 30% or more that some of it at least is due to the treatment working? If a patient reports improvement in their symptoms, why should this not be considered “real” Many of us know PWP who have participated in trials (e.g. GDNF, Spheramine, fetal tissue transplant) and have experienced initial improvements and also lasting (5-10 years or more) benefits. There should be a way of including these patient reported outcomes in the clinical trial results. Currently they are just being ignored. They are real to us.

From the segments that I did view, there seemed to be very little opportunity for patient input at this conference. It could have been different. At one point, one of the researchers was talking about the Spheramine trials being terminated and the effect on the participants. There was a phase I Spheramine trial participant in the audience – what a great opportunity that could have been to incorporate her experiences into the discussion. But she was not acknowledged in any way.

I’m looking forward to seeing Day II. Hope it will be posted soon.

Peggy said...

Oh,My! We must watch our use of words when speaking of sham surgery and placebo effect. Quoting you:"The problem is the placebo effect works by convincing the individual something is real when it is in fact not... So while studying the mechanism underlying the placebo effect is a worthy vein of research, the placebo effect remains no substitute for medical therapy."

I attended the NIH Sham Neurosurgical Conference and heard evidence (albeit little) that there IS a chemical change in the brain associated with placebo effect. In an article found on PubMed regarding the effects of expectations on placebo effect, it states: "Significant dopamine release occurred when the declared probability of receiving active medication was 75%, but not at other probabilities." and in concluding: " The strength of belief of improvement can directly modulate dopamine release in patients with PD. Our findings demonstrate the importance of uncertainty and/or salience over and above a patient's prior treatment response in regulating the placebo effect and have important implications for the interpretation and design of clinical trials."
Arch Gen Psychiatry. 2010 Aug;67(8):857-65
So you see, the placebo effect is very real.

Dr. James Beck said...

@J - You bring up an excellent point - both the surgeon and surgical team do know who received a real vs. sham operation. Clearly, by the very nature of what is involved with surgery, they will have to know. This is part of what makes it difficult to perform rigorous assessment of surgical interventions. So what typically happens then is that an entirely separate team is involved in the post-surgical care and assessment of individuals who are in surgical trials. That separate team is blinded to who has received a sham vs. real operation and the original surgeon is no longer involved.

@LindaH and @Peggy - This is a great conversation. One of the meeting participants and director of PDF's Research Center at Rush University, Dr. Christopher Goetz, co-authored a review that delves into what they called placebo associated improvement. @LindaH points out that this improvement feels real to the person with PD, which is great for that individual. Indeed, as both @Peggy and I cite, there are measureable changes to brain chemistry. However, Dr. Goetz points out that the mechanism behind this improvement is likely not fundamental changes to the affected area, in this case the striatum, but involves a cortical or top-down change in neural activity. The result is that levels of many neurotransmitters, including dopamine, are altered. Dr. Goetz also noted that the more invasive the treatment (e.g., sham surgery) the more enhanced the placebo-associated improvement.

I think the quote @Peggy provided from the recent Stoessl paper is very pertinent: "The strength in belief of improvement can directly modulate dopamine release." That statement underscores the difficulty in designing clinical trials in the face of the expectations that people (both people with PD and physicians) have towards a new invasive treatment-everyone wants it to work. So while there are many hurdles to finding new treatments for PD, I agree with you that belief should not be a prerequisite for success.