> I think the women who suffer the side effects of "preventative  treatment" might differ with you. I suspect we are talking about two very different populations when we discuss risks and benefits.  Preventative treatment mandates that it first of all not cause unintended disease. And I am speaking *only* of preventative treatment in a healthy population. 

To a large extent I would agree with you. 

I might add that the major preventative treatments for osteoporosis in healthy populations are: sufficient dietary calcium and vitamin D, and exercise. These perhaps fall under the heading of "common sense" more than "treatments", but they do work.

Estrogen replacement therapy is a separate question, since it impacts several different aspects of aging disorders. Current [i.e.1998 -Tishy] opinion is that estrogen replacement decreases the risk of heart disease and osteoporosis, while having (when administered with progesterone, or in the absence of a uterus) no effect on uterine cancer, and *probably* little effect on breast cancer. (I have seen studies that ERT increase breast cancer risk, and others that say that it does not, and none saw a large effect.)

(ERT may have other effects as well, but these are currently under investigation, and are not established.)

The decision for hormone replacement therapy is one for post-menopausal women and their physicians, and should be based on both the objective alterations in disease risk (i.e. the result of population studies) and on factors related to the individual woman. Some women have high bone densities ( and low rates of bone loss) and low serum cholesterol levels, and few menopausal symptoms in the absence of HRT, and for them it is probably of little benefit. Some women object to HRT for personal reasons, and for them it is probably not worth doing. Some women suffer adverse reactions to the HRT, and should have either altered doses or cease the HRT.  On the other hand, some women who are currently healthy exhibit the "typical" 4-fold increase in bone loss associated with menopause, or the altered lipid profiles associated with menopause; for these women, the data suggest that HRT will significantly prolong their lives.

This prolongation of life takes into account any increase in breast cancer risk (note that this is an increase in *risk*; breast cancer occurs in women not on HRT also) and other problems. HRT (as with *all* "therapies", including ingesting food or exercise) is associated with risks and benefits.

--------------------------------------------------------- 
> > >The two published studies linking increased bone density with a doubling in risk of breast  cancer  should give pause to anyone jumping on the bandwagon of increased bone density to prevent osteoporosis

> > Which studies are these? And what exactly are their data? 
> 
> I have posted the cites below. Look up the data. 

Conclusions. Women in the highest quartile of bone mass are at higher risk for postmenopausal breast cancer than those in the lowest quartile. The mechanisms underlying this relation are not understood, but cumulative exposure to estrogen may play a part. (N Engl J Med 997;336:611-7.) 

Conclusions.--Bone mineral density predicts the risk of breast cancer in older women. The magnitude of the association is similar to that observed between BMD and all fractures. Our findings suggest a link between 2 of the most common conditions affecting a woman's health. Identifying a common denominator for these conditions should substantially improve our understanding of their etiology and prevention. JAMA. 1996;276:1404-1408 

On the other hand, increased bone density clearly does play a role in preventing osteoporosis, and far more individuals become osteoporotic than develop breast cancer. I will have to read the papers to decide how likely it is that this correlation is a real effect. 
------------------------------------ 
 > > As I pointed out in an earlier post, to a good approximation, the immune system and bone IL-6 are separate. The altered levels of IL-6 are only effective locally.
> 
> Can you explain this? Osteoporosis is systemic. I don't understand how something which is "effective locally" can create a systemic effect.

A few definitions for those lurking:
Osteoblasts -- cells that make bone
Osteoclasts -- cells that break down bone

Bone remodeling is a continuous process in which osteoclasts remove bone, and then osteoblasts replace it (the usual estimate is that in adults, about 0.5 gram of calcium is removed and replaced each day). Bone remodeling is necessary to maintain bone strength. In the *ideal* case in adults, the osteoblasts replace the same amount of bone removed by the osteoclasts. In order for this to happen, the osteoblasts and osteoclasts need to communicate. The two cell types communicate by releasing very small amounts of signaling molecules, of which IL-6 is one. These molecules regulate the *local* activity of the two cell types in order to control the amount of bone removed and replaced.

It is necessary that the IL-6 and other signaling molecules used by the osteoblasts and osteoclasts only be active locally, because the remodeling process will be at different stages both in different parts of the same bone and in different bones. If the IL-6 was active at significant distances, it would disrupt normal bone remodeling in other parts of the bone (or of the body); it would also interfere with immune system
regulation.

Now, for the cause of osteoporosis:

Unfortunately, the ideal case mentioned above occurs only briefly during most individual's lives. After the age of about 30, the average amount of bone removed is slightly higher than the average amount replaced. This is the result of local events (the result of the action of single cells) all over the bone structure. The *effect* is systemic, as a result of *local* actions of individual cells under the influence of *local* signaling molecules.

In women after cessation of ovarian function the decrease in levels of a systemic hormone (estrogen) causes an alteration in local levels of IL-6 (and some other signaling molecules). This change results in a marked increase in rate of bone loss; this increased rate of loss results in a much higher risk for osteoporosis. In males suffering from loss of testicular function, decreased levels of androgens also result in increased rates of bone loss.

(Rates of loss exhibit tremendous individual variation. The *average* woman in the US loses about 30 mg calcium per day (net deficit of excretion over intake). The effect is that calcium is lost from the bones, because the bones are the major storage location (roughly 99% of the total) for calcium. For the first 4 to 8 years after menopause, the rate of calcium loss rises about 4-fold, to over 100 mg of calcium per day (again, this is very variable, depending on the woman, and the day(s) measured). In general, women have a peak bone mass of less than 1000 grams of calcium (often much less; the average seems to be about 800 grams). The loss of 30 mg/day adds up to 11 grams per year; over 30 years (i.e. by age 60) this can result in a loss of more than 300 grams of the skeletal calcium, even without the increased losses due to menopause. The osteoporotic fracture threshold (i.e. the level at which bone fragility is such that fractures occur without major stress on the bone) seems to be (very approximately) 400 grams of total bone calcium. Individuals with low peak bone masses or average or above average rates of loss will reach the fracture threshold at earlier ages. Individuals with higher peak bone masses, or lower than average rates of loss will reach the threshold later.)

Some signaling molecules have systemic actions: estrogens, androgens, insulin, epinephrine, and many others. These compounds are released from one tissue and have effects throughout the body.   Other signaling molecules (IL-6 and many others) have local effects. They are released by cells in one tissue and responded to by cells in close physical proximity (centimeters or less). As a result, many of these signaling molecules are used by different parts of the body for different purposes. Thus, IL-6 released by the immune system cells has *in general* little, if any, effect on bone, and IL-6 released by the cells in bone has *in general* little, if any, effect on immune system function. (There are exceptions, some of which have been mentioned in this thread.)

> 
>  Breast cells are well separated  from bone.
> 
> So well separated that breast cancer most often metastasizes to bone. Yes indeed, the separation is quite distinct.

Metastases involve migration of single cells through the lymph and circulatory systems.

Breast cells *are* well separated from bone.  Breast cancer cells, on the other hand, are found within the same body as all of the other tissues of the body. All tissues are within a few cells of a blood vessel. Any metastatic cell that reaches the blood stream can go essentially anywhere in the body; the physical separation of the precursor tissue and target tissue is irrelevant.

On the other hand, IL-6 (and other local signaling molecules) are degraded or diluted rapidly as they travel through the blood or the tissues. A billion cells is rather a lot (a cubic centimeter; a readily palpable lump). A billion molecules of a signaling molecule in the same volume has *no* effect (the concentration is too low to result in significant activation of the relevant receptor). 
 

> > Actually, this is not quite true. You are partially correct in that  "only" 1 in 8 women develop breast cancer. However, more than half of women  over 75  will develop osteoporosis; in other words, "most" women who live long enough *will* develop osteoporosis.
> 
> Please define osteoporosis as you are using the term. When I read these kind of numbers I think of the many elderly women I know, all over 75 none of whom have any sign of bone disease. No visible vertebral fractures, no pain in their bones, no other fractures, And common sense then makes me doubt your numbers. Lowered bone density is *not* a clinical disease. 

Lowered bone density vastly increases the risk of fractures. When bone density is low enough, simply standing up can be enough stress to break bones.

I am happy for the women you know who are healthy. 

I have met a number of healthy women over 75. On the other hand, I also know (or knew) a number who have obvious signs of osteoporosis (or who died within a few months of breaking their hips).

> > >The long term safety of this drug is still very much open to question, and the experience with > > > tamoxifen should give researchers and doctors pause before haphazardly prescribing this drug.

> > Nearly everyone agrees that tamoxifen is pretty safe. In *rodents* tamoxifen is associated with liver cancer, but this does not the case  in  humans. Tamoxifen is only effective in a subset of breast cancer, but   this  is hardly an indictment of the drug.
> >
> 
> I'm quite certain the families of the women who died of endometrial cancer caused by tamoxifen would disagree. I also suspect that the women who survived their endometrial cancer but required hysterectomy and chemotherapy for it would disagree. I know  two women who stopped taking it as soon as the risks of endometrial cancer were made public, who remain at risk of endometrial cancer who disagree. And perhaps you know that new FDA regulations limit the use of tamoxifen to no more than five years and that they also require annual endometrial biopsies in intact women for each of those five years because the drug causes a particularly deadly form of the disease. Not quite a safe drug is it? As for my use of the word haphazard, it was not intended to apply to the  FDA process. It was intended to apply to the way physicians are prescribing it. 

1) Tamoxifen is an *anticancer* drug. It markedly reduces mortality in post-menopausal breast cancer, without the harsh side-effects of cytotoxic agents. (Adriamycin is a cardiac toxin; it is still used to treat certain types of cancer. Most, if not all, chemotherapeutic agents have *far* worse side effects than tamoxifen.)

2) The increased risk of endometrial cancer as a result of tamoxifen therapy is relatively minor (roughly 1 in 500). As with any cancer, this means tragedy for those affected, but significantly longer (and generally higher quality) lives for the majority of patients who respond. 

3) The increased risk of endometrial cancer is an *estrogenic* effect. Unfortunately, there are no clinically available compounds which are fully antiestrogenic in both breast cells and in uterus. Also, unfortunately, progesterone, while protective against uterine cancer when administered with estrogen, usually stimulates breast cancer growth; as a result, unlike combination HRT, progesterone (or synthetic progestins) is contraindicated in women being treated for breast cancer.

4) Regarding your statement about FDA limits on tamoxifen therapy: this 5 year limit is actually rather controversial. There is a clinical trial underway in Britain testing whether ten years of tamoxifen (or longer) may not actually be better. The limitation on treatment may actually be killing more women (due to recurrence of breast cancer they were already known to have) than would be affected by the slight increase in uterine cancer risk.

5) Most people are very bad at assessing relative risks. The risk of dying in a car accident is far higher than the risk of dying in a plane crash, and yet many people drive their cars without a thought but are afraid of flying. The risk of dying of breast cancer for all women is roughly 1 in 32 (and about 1 in 4 for women diagnosed with the disease). The risk of developing uterine cancer as a result of tamoxifen therapy is roughly 1 in 500 (the risk of dying of the disease is much less than that).

6) The risk/benefit ratio for prophylactic tamoxifen therapy (i.e. the current trial of women at high risk for breast cancer taking tamoxifen or placebo) is much more difficult to define (hence the trial). My suspicion is that, for women at high risk, tamoxifen is going to prove of overall benefit (I initially opposed the trial). For women not at high risk for breast cancer the risk/benefit ratio will probably be such that most will elect to avoid the drug. Excluding the clinical trial, very few women are taking tamoxifen unless previously diagnosed with breast cancer.
 

> > Raloxifene has been around for a while (it used to have a different  name).
> 
> What was that name? And what was it being used for? 

It used to be called keoxifene. It was tested as an anti-breast cancer drug, but found to have little effect on breast cancer. I don't think that it was ever approved as a drug, but it was used in various trials for years.  Both fosamax and estrogen replacement therapy have been shown to decrease fracture risk. By *my* definition, that is a real life benefit (YMMV).
> >
> > > Could you provide some cites, please.
> > Try:
> >

 Karpf DB; Shapiro DR; Seeman E; Ensrud KE; Johnston CC Jr; Adami S; Harris ST; Santora AC 2nd; Hirsch LJ; Oppenheimer L; et al. 
Prevention of nonvertebral fractures by alendronate. A meta-analysis. 
Alendronate Osteoporosis Treatment Study Groups. 
Journal of the American Medical Association, 1997 Apr 9,  277(14):1159-64.

I don't quite understand your quibble. Fosamax is not *used* in healthy post-menopausal women; obviously the studies have not been done. (Clinical trials are *very* expensive, and there is no reason to assume a benefit of fosamax in healthy women.) 

>Post menopausal women with clinical osteoporosis are not healthy  by definition. Further the study fails to distinguish between hip (dangerous) and wrist (relatively benign) fractures. And the difference is a rather anemic 12%, with actually very few fractures occurring in any of the groups. 
 

In all clinical trials of this sort, the number of fractures tends to be small. Unless you are going to begin physically assaulting the trial participants (which would tend to make recruitment difficult and/or cause problems with informed consent), the trial organizers have to wait for fractures to occur. Most people try to *avoid* fractures (adolescent boys being a slight exception), and older women tend not to be especially physically adventurous. 
 

>Actually I think lumping wrist and hip fractures together into the category of "non vertebral" fractures is quite disingenuous on  the part of the sponsor. Misleading but quite clever if you want to show a greater benefit than you have actually found. 
 

The problem with clinical trials is generating enough data to reach a conclusion of *any* sort. Trials require relatively large numbers of people who will put up with the drug and with reporting on the particulars of their lives. Rats are a great deal easier to work with, and you *can* break their bones. Unfortunately, after finding improved bone strength in rodents, you need to test the drug in humans.

This leads to an ethical dilemma:
Do you release an extremely promising drug (such as fosamax) with strong but not perfect supporting data, or do you wait for perfect proof and leave millions of individuals (in the US 1.3 million fractures are the result of osteoporosis each year) with osteoporosis without effective treatment?
  [refs and other stuff deleted]

> I don't subscribe to medline and can't get your other references on  line. However I suspect that, like the first, they discuss treatment of existing disease especially since one of them comes from Rheumatology.  As I have already said I was looking for cites showing decreased fractures in healthy post-menopausal women not those already diagnosed  with the disease. There is a huge group of *healthy* post menopausal women who are being urged to take lifelong drugs to "prevent" osteoporosis. I would like to see something showing that these drugs prevent the dangerous effects of the disease i.e. hip fractures in late life, while not causing disease themselves.

Of the drugs under discussion, fosamax is only recommended for individuals exhibiting evidence of markedly decreased bone density (the type of individual included in the trials), and raloxifene is also only recommended for women in selected groups. Neither is likely to be given to large numbers of healthy post-menopausal women. Estrogen is probably better for most post-menopausal women than raloxifene, especially since raloxifene does not prevent menopausal symptoms and is more expensive. Fosamax is sufficiently unpleasant as to be unlikely to be administered to anyone for whom it is unnecessary.

Thus, the only "life-long" drug being recommended for healthy post-menopausal women is HRT (i.e. estrogen, with progestin in women retaining their uterus). Estrogen replacement (as mentioned at the beginning) is protective against both cardiovascular disease (the leading cause of death in post-menopausal women) and osteoporosis (and probably other disorders, although the evidence is less clear).

Current evidence (from many trials) suggests that in women estrogen replacement *on average* extends life by several years. For some individuals, it is contraindicated, or it should be avoided as a matter of preference, or it is unlikely that the benefits will be greater than the risks. It is (and *should* be) a matter for the woman (in consultation with her doctor) to decide.

(FWIW, my mother is taking premarin. She discussed it with her physician, my father (also a biochemist), and me, and decided that it was a good idea, based on *current* knowledge. We won't know for 20+ years whether it *really* was a good idea, and possibly not then.)

-- 
Mark Brandt, Ph.D.
My opinions are my own, but I tend to give them away to anyone who fails to flee fast enough.