The piece looks at experimental neurosurgical treatments for Parkinson's, such as "Spheramine". This consists of cultured human cells, which are implanted directly into the brain of the sufferer. The idea is that the cells will grow and help produce dopamine, which is deficient in Parkinson's.
Peggy Willocks, a 44 year old teacher, took part in a trial of the surgery in 2000. She says it helped stave off the symptoms for years, but the development of Spheramine was axed in 2008 after a controlled trial found it didn't work any better than a placebo.
The placebo was "sham surgery" i.e. putting the patient through a full surgical procedure, and making holes in their skull, but without doing anything to their brain.
It's cheap and easy to do a placebo controlled trial of a drug - all you need is a sugar pill. But with neurosurgery, it's clearly a lot more involved. A placebo has to be believable. Convincing sham surgery is expensive, time-consuming, and it has real risks, albeit small ones.
Is it ethical to put patients through that?
That, I think, can only be decided on a trial-by-trial basis. It depends on the likely benefits of the treatment, and whether the trial is scientifically sound. Obviously, it'd be wrong to do sham surgery as part of a flawed trial that won't tell us anything useful.
The Nature article, however, goes further than this, and suggests that placebo controlled trials may be unsuitable for testing these kinds of treatments, failing to detect a real benefit in some patients:
There are hints from some of the failed phase II trials that patients followed up beyond study endpoints might tell a more positive story. Some say, therefore, that sham controls are sinking the prospects of valuable drugs.A patient advocate agrees:
Anders Björklund, a neuroscientist at Lund University in Sweden who is collaborating with [Roger Barker of Cambridge], says that sham surgery can lead researchers to throw out a strategy prematurely if the trial fails because of technical or methodological glitches rather than a true lack of efficacy.
According to Perry Cohen, who leads a network of patient activists called the Parkinson Pipeline Project, that’s exactly what is happening. He had always questioned the need for sham surgery, he says, but after the string of phase II failures, “We started saying, ‘Hey, this is a problem. These trials failed, but we know they are working for some people.’”
...Cohen [says] that patients have different priorities and that researchers must take these into account. Researchers use placebo controls to weed out false positives. But for patients, the real ogre is the false negatives — which can sink a therapy before it has been optimized.I'm not sure about this. If I had Parkinson's, I would certainly hate to miss out on the genuine cure because a trial had failed to recognize that it worked. But equally, I would not be happy to be given a rubbish treatment that would have failed a placebo controlled trial, but never got one, because of arguments like this.
Placebo controlled trials can fail to detect benefits if they are too short, too small, methodologically flawed, or whatever. Certainly, a trial can be placebo controlled, and still crap. But the answer is surely to do better trials, not no trials.
It may well be that we shouldn't rush to do placebo controlled trials until later in the development process, when the technique has been properly refined. But the history of medicine is littered with treatments that "we know work for some people" - that didn't.
Katsnelson, A. (2011). Experimental therapies for Parkinson's disease: Why fake it? Nature, 476 (7359), 142-144 DOI: 10.1038/476142a
9 comments:
You know, I'm not aware of any sham surgery trials that showed surgery was superior to sham (but i could be wrong here). The whole act and association of cutting someone open is very powerful on an experiential level, and may be what drives a lot of successful surgeries (this is speculation).
It seems to me that we are all tapping the big wrinkly thing with appendages and missing it for what it is: the common denominator in the placebo effect is the patient's perception of receiving individual caring, compassion and acknowledgement of suffering and their sick role. It appears to have the capacity for powerful and permanent effect large enough number across diagnoses that it should be studied for what it is and what it can do. Give it a medicalized name, such as the Watson Effect (Jeanne Watson's Theory of Caring, Univ. of Colorado) or the Kolcaba Effect (Theory of Comfort, University of Akron).
But study it, use it (both professional nursing and medicine ascribe to the action and caring principles as a foundation of optimal treatment as usual), and elevate to a place of prime importance.
Thanks so much for this post - it brings up many excellent questions to ponder and challenges to meet.
The problem is that there is maybe a huge placebo effect in these trials, much more important than in other surgical cares.
- Doctors think that it must work because of nowadays easy-to-use knowledge of stem cells. (which must be much more complicated in reality)
- In an other hand, patients with Parkinson disease seems likely to answer to placebo effect (see very good video of Predrag Petrovic in english here : http://bit.ly/oKvTpg
An other main problem is I think to let a chance of improving a technique before considering it to be usefull or not. The first heart transplantation, pacemakers were not so successfull than now because the techniques improved.
We studied at university an article about verterbroplasty which deals about some of these problems http://bit.ly/pw5wEB
namely choosing experimented doctors which are logically beter than first users of the technique, making a "theater play".
I don't understand why we don't inject randomly choosed samples of placebo or corticoid/stemm cells.
AN other question is that, we really must understand what it placebo effect not pharmacologic speaking but cognitively speaking (Predrag Petrovic in english here : http://bit.ly/oKvTpg ) in order to improve patient-doctors relationship
According to Parry Cahin, who leads a network of patient activists called the Parkinson Placebo Project, that’s exactly what is happening. He had always questioned the need for expensive drug treatments, he says, but after the string of phase II sham surgery successes, “We started saying, ‘Hey, this is awesome. These trials failed, but we know the placebos are working for some people.’”
This is the key. Placebo trials are not meant to ask "does this treatment work?" They are meant to ask "Does this treatment work better than something cheaper?"
There are economics at play here.
If it takes years of expensive research to develop some drug and a sham surgery works just as well, rather than asking whether we should do the sham surgery, we should be asking whether it is even worth it to do the drug development.
He had always questioned the need for sham surgery, he says, but after the string of phase II failures, “We started saying, ‘Hey, this is a problem. These trials failed, but we know they are working for some people.’”
This just seems like your standard run-of-the-mill argument by anecdote, combined with a bit of special pleading. If RCTs using sham surgery as the placebo don't work the way other RCTs do, that would be a pretty significant result and would require better evidence than, "But many patients say it works for them!" (which is true of pretty much every failed treatment in the history of the world...)
If there's a problem here -- and I'm not saying there isn't, not least because I do question whether sham surgery is ethical at all -- it's going to take better evidence than that.
Since all placebos are fundamentally identical, what difference is there between sham surgery and a sham pill (that simulates side effects?) In either case, the intervention itself is irrelevant. The placebo effect is created by the patient's beliefs, not the actual cure.
As long as the control group and the experimental group never come into contact with one another, and thus never have the chance to compare treatments, any sham should be as good as any other.
"But the history of medicine is littered with treatments that "we know work for some people" - that didn't. "
Also the history of medicine is littered with treatments that 'we know work for some people" - that did.
Sham surgery is not riskless. Try to explain it to the family of a patient that dies from sham surgery (it happens, I saw at least one)
Stuart: That's a good point, but in practice it would be hard to compare surgery to a placebo pill - you would need to somehow convince people to sign up for a trial in which they might get an unknown treatment.
Because if you told them that they might get surgery, they would know whether it was real or not.
One limitation of using placebos is that the real effect of the drug or intervention can be 'masked'.
For example, let's say that a particular condition improves markedly in response to both (a) expectation and (b) an expectation / active drug combination. Even if no differences are found between them it would be premature to speculate that the drug has no physiological effect. That is, the condition might improves in response to the active drug in the absence of expectation (i.e., if administered without the person's knowledge).
Most placebo-controlled trials fail to manipulate levels of expectation (for ethical reasons) and most systematic reviews conclude that if superiority over placebo is not demonstrated, then the drug does not work. I wonder how many effective treatments have been cast aside as ineffective based on this reasoning?
Post a Comment