Wednesday, September 24, 2008

News Flash: Hot Flashes Are Treated By...Everything!

...Well, I don't know that for certain, but that's what I thought when I read this headline "Acupuncture Reduces Side Effects Of Breast Cancer Treatment As Much As Conventional Drug Therapy, Study Suggests" (1). According to this "first-of-its-kind study," acupuncture is as "effective and longer-lasting in managing the common debilitating side effects of hot flashes, night sweats, and excessive sweating (vasomotor symptoms) associated with breast cancer treatment." What is the conventional drug therapy to which acupuncture was compared? Effexor of course (that's his slave name, he prefers to be called venlafaxine).

Are there any other scientifically validated treatments for hot flashes and other vasomotor symptoms? Well, according to these people (2), "Soy seems to have modest benefit for hot flashes, but studies are not conclusive. Isoflavone preparations seem to be less effective than soy foods. Black cohosh may be effective for menopausal symptoms, especially hot flashes, but the lack of adequate long-term safety data (mainly on estrogenic stimulation of the breast or endometrium) precludes recommending long-term use. Single clinical trials have found that dong quai, evening primrose oil, a Chinese herb mixture, vitamin E, and acupuncture do not affect hot flashes; two trials have shown that red clover has no benefit for treating hot flashes." [my emphasis]. Did I read that correctly? Acupuncture did not affect hot flashes? That must be some sort of fluke. Except that these people (3) and these people (4) both found that acupuncture was no better than sham treatment.

Assuming that my thought process is linear, acupuncture is as effects as venlafaxine, yet acupuncture is no better than sham treatment. So that would mean..., venlafaxine is no better than sham either. So what does the evidence say? Well, venlafaxine has these three open-label trials (5, 6, 7), all of which were...drum roll please...positive! (I'm shocked).

Science Lesson 1: Open-label studies are pointless. 98% of all open-label trials are positive. Sounds pretty high, right? That's probably because I just made that statistic up. But, when you have no comparison group, and all parties involved know about the treatment, positive results are the rule, not the exception.

Anyway, I was able to find this one randomized controlled study (8), which lasted for a staggering 4-weeks, and showed that venlafaxine was superior to placebo pill. Side-effects of the venlafaxine treatment included, "mouth dryness, decreased appetite, nausea, and constipation."

Science Lesson 2: When a drug has side effects, patients and doctors can accurately guess whether or not the patients were given placebo, thus breaking the blinding. Secondly, not all placebos are created equally. Pill placebo is less effective than capsule placebo, which is less effective than injection placebo. Also, if put a sticker price on the placebo, the more expensive placebo out performs the cheaper placebo (9). Lastly, another way to boost the placebo effect is to give a placebo that actually has side effects (10).

So quit jerking me around, does venlafaxine work or not? And while we're at it, does fluoxetine (11) and paroxetine (12) work too? Well, according to this meta-analysis (13) published in 2006, there are a total of 7 trials that compared either SSRIs or SNRIs to placebo. Only 3 out of those 7 trials were superior to placebo (43%). And those are only the published studies. Who knows what negative studies have not been published (14). Well, what about the science? There has to some sort of biological explanation, right? According to the former psychiatrist and current psychopharmacologist Stephen Stahl, "It may be that actions on both the serotonergic and noradrenergice systems are required to improve these (vasomotor) symptoms" (pg. 626). There's just one problem with that theory. In the controlled trial with venlafaxine, the highest dose given was 150mg. At that dose, venlafaxine is barely an SNRI. And since the 150mg dose was no better than the 75mg dose, that does not lend much support to Dr. Stahl's theory.

Science Lesson 3: A theory is suppose to lead to a hypothesis, which leads to data, which leads to revising the theory. In pharmacology, however, you discover that a drug has an effect by accident; then, you assume that the mechanism of that drug is why there was an effect. Then the thought process seems to stop there as these people tend to ignore contradictory evidence. Drug treatments for depression have these actions: SSRI, SNRI, DRI, NRI, Alpha2 antagonist, cortisol antagonist, CRH antagonist, blah, and blah. So the theory is that depression is caused by and treated by all these chemicals. Makes perfect sense.

So serotonergic drugs have some effect on hot flashes. But so does calcium channel blockers (15), CBT (16), and balancing your yin and your yang. And to be honest, I'm not even going to pretend that I know how black cohosh and soy work. So maybe my first thought was correct, hot flashes are treated by everything!

My long-winded point is this: When you have a condition, such as depression or hot flashes, which are highly subjective and have minimal to no reliable and objective identifiers, everything under the sun can be shown to have a positive effect. Everything that is, except magnets (17).

1 comment:

Marilyn Mann said...

Very interesting post. Breast cancer survivors are often advised to stay away from soy products, on the theory that soy may have an estrogenic effect.

There are ways to monitor hot flashes objectively, with temperature sensing equipment.

I am also skeptical that venlafaxine has much, if any, effect on hot flashes.

You may be aware of this already, but patients who are on tamoxifen are advised to avoid certain antidepressants (not including venlafaxine) because they are metabolized by the same enzyme as tamoxifen (or something like that; I don't remember the details).