Placebos Are Getting More Effective. Drugmakers Are Desperate to Know Why. Wired

[jimwalsh]

This article (http://www.wired.com/medtech/drugs/magazine/17-09/ff_placebo_effect?currentPage=all#) Highlights the curious phenomenon of the placebo in clinical trials... it is interesting that the placebo effect is not a constant.

if according to the article it is now more difficult to actually beat the placebo in trials does that mean that all drugs that have been licensed in the past should be re-trialled or that the gold standard of RCT should be relaxed somewhat?

[Graham Lappin]

Depression, like many of the psychologically-based diseases (for want of a better phrase) are very difficult to measure in terms of effect. I work in more mechanistic-based medicine, where I want to see and measure some biochemical response - not always possible I grant you. In cases where the subjective response of a patient can be removed, then so the placebo effect is diminished.

The other point I would make is that it's all in the statistics. When is a therapeutic effect significantly greater than the placebo group? If the effects are statistically close then there is always a risk that it's all being lost in the noise. A lot of CAM trials suffer from this (or should I say that the proponents of CAM exploit minor differences).

Finally the end point being measured in clinical trials is also important. To measure the real effect (eg longevity) can be very difficult and so surrogate end points are often used http://www.annals.org/cgi/content/short/125/7/605. Such areas are the subject of great debate.

[JJM]

... if according to the article it is now more difficult to actually beat the placebo in trials ... the gold standard of RCT should be relaxed somewhat?Of course not. If your treatment cannot significantly rise above placebo, you don't get to sell it. One doesn't get to move the goalposts when one is losing. You cannot argue "This must be effective- so we'll design a procedure that proves it." We like to say- "That is so outrageous it is beyond wrong."

For example, homeopaths complain that the problem with scientific tests is that their magical water always fails. They know their nostrums work- so science doesn't. That is just bassackwards. We found out that anecdote misleads us, and science prevents our misconceptions. Claims to "other ways of knowing" are references to the Dark Ages, when ignorance and superstition ruled.

[dalriada]

Thus the headline might also have read:

"Drugs-Just Not as Good as They Used to Be" ?

Could this be true? Antibiotic resistance, New diseases, glut in the pharmaceutical market?

[Graham Lappin]

... or that the gold standard of RCT should be relaxed somewhat?

Don't forget some clinical trials are run against a comparator, that is a drug already on the market with known efficacy. It is sometimes necessary to show that a new drug has benefit over existing therapies in order to get a marketing authorisation.

[Floppit]

I work in more mechanistic-based medicine, where I want to see and measure some biochemical response - not always possible I grant you. In cases where the subjective response of a patient can be removed, then so the placebo effect is diminished.

Bunny, I noticed you used the word diminished rather than removed, therefore I'm assuming (maybe wrongly) that there are biochemical responses to the placebo effect. Are the placebo results ever used? Over time I imagine that it's possible to build quite an in depth picture of what biochemical aspects of disease are vulnerable to psychosomatic influence, especially a well researched area with multiple RCTs. Do people ever do lit reviews to gather the placebo information for particular diseases?

[jimwalsh]

Don't forget some clinical trials are run against a comparator, that is a drug already on the market with known efficacy. It is sometimes necessary to show that a new drug has benefit over existing therapies in order to get a marketing authorisation.

so could this apparent increase in placebo effect be more to do with increased efficacy of pharmaceuticals?

[Graham Lappin]

so could this apparent increase in placebo effect be more to do with increased efficacy of pharmaceuticals?

Sorry, don't see how the two go together, could you elaborate please?

Efficacy is one of the big three in pharmaceutical terms - dose, efficacy and toxicity. As drugs are developed that are more efficacious, then the dose that is needed to exhibit a therapeutic effect goes down. The trick is to get drugs with a wide therapeutic window - that is they provide a therapeutic effect over a wide dose range (more accurately target concentration) without resulting in toxicity (ie side effects). The efficacy of any given drug at any given dose regimen is independent of any placebo effect.

[Graham Lappin]

Bunny, I noticed you used the word diminished rather than removed, therefore I'm assuming (maybe wrongly) that there are biochemical responses to the placebo effect. Are the placebo results ever used? Over time I imagine that it's possible to build quite an in depth picture of what biochemical aspects of disease are vulnerable to psychosomatic influence, especially a well researched area with multiple RCTs. Do people ever do lit reviews to gather the placebo information for particular diseases?

My use of "diminished" was subconscious. I think I used it because "removed" is an absolute term and I am rarely comfortable with 100%. Nevertheless, you ask a very interesting question.

In mechanistic terms you are looking for some biochemical or pharmacological response to the active drug. This might be, for example, a change in some pathway or binding of the drug to some specific target (to use two fairly straightforward examples). My own area is more pharmacokinetics, where the drug concentration is monitored over time and then (sometimes wishful thinking) related to the pharmacodynamics (ie pharmacological response).

Under these descriptions, it is difficult to see how some mechanistic effect could be measured with a placebo. I can think of other circumstances however, say with MRI or PET scans of the brian how some actual effect could be monitored with a placebo. I have to say I am unaware of any such studies but the idea is quite fascinating. I wonder what the control would be?

The master of the literature - and it's interpretation is Pebble - any such trials I wonder?