Breast Cancer

December 14, 2006 6:02 PM

The incidence of breast cancer in the U.S. fell by 15% between August 2002 and December 2003. Why? Because starting in the summer of 2002 millions of menopausal women stopped taking hormone replacement therapy. How many women would be alive if they'd never started?
posted by alms (24 comments total)

15% more?
posted by Mach5 at 6:06 PM on December 14, 2006

Ha! Nice, Mach5.
posted by Dasein at 6:08 PM on December 14, 2006

“Epidemiology can never prove causality,” said Dr. Peter Ravdin, a medical oncologist at MD Anderson Cancer Center and one of the authors of the analysis.

O RLY?
posted by docgonzo at 6:18 PM on December 14, 2006

Mach5, clearly you are gunning for a job with Verizon.
posted by alms at 6:20 PM on December 14, 2006 [1 favorite]

So eventually I'll get to choose between crazy-bad menopause symptoms like my mother's getting now, or breast cancer. Awesome.
posted by heatherann at 6:21 PM on December 14, 2006

So stopping HRT had an immediate effect, on the order of days? (Summer 2002 vs. August 2002?) Cancers were immediately stopped, as if they were allergies with Benadryl? That would throw the 'Cancer requires multiple hits and many years' theory out the door.

Hard to believe this is true, praying it is. If a month-to-month graph of incidence shows the rate continually and slowly descending, maybe. Then the real payoff should be far better as the years go on.
posted by toma at 6:38 PM on December 14, 2006

The measured dropoff was between August 2002 and December 2003. That's 17 months for more and more women to stop taking the drugs and for the positive effects to show.

The hypothesis is that when women stopped taking menopausal hormones, tiny cancers already in their breasts were deprived of estrogen and stopped growing, never reaching a stage where they could been seen on mammograms. Other cancers may have regressed, making them undetectable. And, possibly, without hormones, cancers that would have gotten started may never have grown at all.
posted by alms at 6:46 PM on December 14, 2006

I did some similar research in the 90s, and I find it weird they picked the period immediately following HRT cessation to look at, and then took an average. I think I would have gotten laughed at if I had tried that.

The theory I mentioned is one of the oldest in the field, something virtually everybody accepts. If it's still holding, no real dropoff should have been seen in 8/02 (hence my allergy argument). So if there is a real cause-and-effect here, the actual benefits are far better than the 15% average over more than a year. That would be a jaw-dropper.

The only things I recall helping develop cancers that quickly (or prevent it, when withdrawn) are highly carcinogenic substances, like PCBs, tobacco smoke. If it's true, Estrogen is flat-out dangerous.
posted by toma at 7:18 PM on December 14, 2006

I second heatherann's Awesome comment up above...I'm supposed to start a round of hormone replacement therapy to deal with symptoms of early-onset menopause (I'm 39) on Monday. I've already got minor scars from three biopsies on "the twins" (fortunately all benign) and am unnerved by the possibility that the HRT might contribute to future malignant diagnosis.

I already hate Mondays...this just adds to the joy.
posted by squasha at 7:24 PM on December 14, 2006

Epidemiology may not be able to conclusively prove causation, but I'd call this a pretty goddamn strong correlation.
posted by chimaera at 7:34 PM on December 14, 2006

Rates of the most common form of breast cancer dropped a stunning 15 percent from August 2002 to December 2003, researchers reported yesterday.

Ah, yes, now that makes sense (rate/between vs rate/from). Same old jaw-dropping.

Whoever comes up with the molecule that alleviates menopausal symptoms without accelerating Cancer will be a hero.
posted by toma at 7:47 PM on December 14, 2006

Toma, can you describe the difference between rate/between and rate/from?
posted by alms at 7:50 PM on December 14, 2006

15% more?
posted by Mach5 at 6:06 PM PST on December 14

Ha! Nice, Mach5.
posted by Dasein at 6:08 PM PST on December 14


Not so fast. 100 women less 15% is 85 women.
85 women plus 15% is 97.75 women.
The answer is about 17% more.
posted by weapons-grade pandemonium at 7:51 PM on December 14, 2006

thanks wgp, way to dork up my first drunken thought upon seeing this thread. oh well, more boobies == better at any percentage point.
posted by Mach5 at 7:54 PM on December 14, 2006

Oh, yeah, I think it has to do with the traditional way science is parsed:

The GDP was up 2% between June and August, vs The GDP was up 2% from June to August.

One suggests the overall rate for 3 months, and the other compares the two rates. Regular English often uses between as a comparison, and Science uses it, I suppose, to include everything bounded by two points. Or something like that. Still is shocking and welcome news.

The guy I used to cancer research with is now at M.D. Andersen, they may be the best in the world. Cheers.
posted by toma at 8:06 PM on December 14, 2006

There was a report on NPR today that the Progesterone inhibiting drug RU486 prevented breast cancer in a sample of mice that were genetically engineered to be susceptible to breast cancer.

All of the 29 mice that took placebos developed breast cancer while the 14 mice that received RU486 were cancer free.

Apparently, women given estrogen/progesterone therapy at menopause have higher rates of breast cancer vs. women given estrogen therapy alone.
posted by tresbizzare at 10:26 PM on December 14, 2006

wgp and Mach5, maybe now you understand alms' joke.
posted by Kwantsar at 10:42 PM on December 14, 2006

Paging Suzanne Sommers ...
posted by maudlin at 10:58 PM on December 14, 2006

Oh, and this article is good, too. Hucksters like Sommers just make me see red.
posted by maudlin at 11:01 PM on December 14, 2006

It is worth noting that what is marketed to women as "hormone replacement therapy" is really "hormone substitution therapy." The hormones used in HRT are not the same as what normally occurs in a healthy pre-menopausal woman. Its like making "oatmeal" out of wheat and claiming it's the same thing. Hey, oats and wheat are both grains!

What Ms. Sommers is a proponent of is Bio-identical Hormone Replacement, which uses the same hormones a woman's body used to make. I hope I look as good as she and T.S. Wiley do when I am their age. Yes, Wiley (pic at bottom of page) looks that good in person, I met her last week. Boy howdy does her book have a lot of studies in the footnotes.
posted by ilsa at 9:51 AM on December 15, 2006

I'm curious about what happened to the rates of heart disease and osteoporosis.
posted by euphorb at 11:31 AM on December 15, 2006

ilsa, she claims that these are the same hormones our bodies make, but I wouldn't place much credence in that claim. From my second link:

The word bioidentical is a marketing term, not a scientific one, and it means different things to different people. To most doctors, bioidentical refers to a wide variety of FDA-approved drugs that are virtually identical to the hormones produced by women's ovaries. They come in many forms and doses, some of which have been used for years. Somers uses the term to refer to made-to-order treatments created by compounding pharmacies with dosages usually determined by the results of blood tests every two weeks (the method Somers herself uses), or regular saliva tests, a method most experts say is an unreliable way to measure a women's specific hormone needs. ...

Somers says these custom-made treatments are natural and not really drugs. That's just not true. Bioidenticals may start out as wild yams or soybeans, but by the time this plant matter has been converted into hormone therapy, it is in fact a drug. All of these products—whether or not they're approved by the FDA—are chemicals synthesized in a lab. Another thing you should know: there are only a few labs in the world that synthesize these hormones. Everyone—from small compounding pharmacies to big pharmaceutical companies—gets their ingredients from the same places.

Somers argues that bioidenticals are safer than FDA-approved hormones even though there are no high-quality studies to prove that assertion. In the absence of any reliable research to the contrary, most women's health experts say it's prudent to assume that all hormone products (FDA-approved or not) carry the same heart disease and cancer risks.

As far as her looks go, I'd credit good genes and plastic surgery.

And books with lots of footnotes mean dick. Ann Coulter's books have lots of footnotes, but they're still nasty, dishonest swill.
posted by maudlin at 7:24 PM on December 15, 2006

Ok, inasmuch as there are probably more medical degrees attached to my post than the rest of the tread, I'm going to authoritatively call shennanigans on everybody.

Statistics are just that. They prove nothing but can point to lots. Perhaps the decline in breast cancer is associated with rising atmospheric CO2 and has nothing to do with the HRT rates at all. Is it associated with an increase in consumption of antioxidant vitamins and fish oils? Is it associated with increased obesity?

The problem with all of this is the lack of actual research. Drug companies want only to sell the big patentable compounds. Nearly all real research anymore is drug company funded so there's no ( or very little) money to go to researching anything that can't be patented. Net net, we're left wondering about weird statistical anomalies and unable to determine causation because we can't control the variables.

As for the drug/not drug split that's a semantic game that many want to play but nobody really wants to sort out. Is water a drug? Why not? Is sugar a drug? Why or why not? Is a truly bio-identical hormone a drug? Why or why not?
posted by shagoth at 8:50 PM on December 15, 2006

Observational epidemiology can't, in fact, prove causation.

Not only that, the conclusion everyone seems to be jumping to - that HRT causes cancer - could be flawed. Just off the top of my head I can think of some other explanations:

1) HRT accelerates the development of cancers. Therefore, a large number of women stopping HRT means that a large number of cancers that would have been detected in that 17 month period are in fact simmering along more slowly, still under the bounds of detectibility. They will be detected later and the incidence of cancer in the target population will later spike to be higher than it was in the control period.

2) Doctors may ascertain people differently for cancer depending on whether or not they're on HRT. Perhaps these women are developing cancer at the same rate but docs who are stopping HRT are also ordering fewer mammograms. You want to be certain something like this isn't true before you draw any conclusions.

3) HRT may change the texture of the breast, rendering cancers less detectable, by physical exam, mammogram, or both. This explanation has what epidemiologists call "biological plausibility," and again you want to make sure that something like this isn't the true cause behind your explanation.

Epidemiology isn't a parlor sport, kids. Except when it is, and this isn't one of those times.
posted by ikkyu2 at 10:58 PM on December 15, 2006