Category Archives: WHI

It’s not gonna’ kill you to take hormone replacement

It’s not going to kill you to take hormone replacement therapy.

That’s the take home message from the latest analysis of the Women’s Health Initiative, the largest and longest randomized trial of hormone replacement therapy (HRT) in menopausal women.

After almost 18 years of follow up in the WHI, there was no increase in overall mortality, including death rates from cancer, in women taking HRT for up to 5.6 years (estrogen plus progestin) or 7.2 years (estrogen alone). There was a non-significant reduction in mortality among those who started HRT between ages 50 and 59, the group most likely to be prescribed hormone therapy for menopausal symptoms.

I’ve blogged before about the results and limitations of the WHI, which found, on balance, that the health risks of HRT (breast cancer, blood clots, stroke) about equaled its health benefits (protection against colon cancer and osteoporosis). The study (and the US Preventive Services Task Force) concluded that there was no reason for women to take HRT for preventive health reasons.

The biggest criticism of the WHI was that 70% of its participants were 10 or more years post-menopausal. The study did not include women most likely to benefit from taking HRT – those with hot flashes, night sweats, insomnia and other menopausal symptoms – or enough women in the 50-59 year age group. These women start HRT at menopause, not a decade or more after it’s over. In this younger group, more recent research suggests there may be a reduction in heart disease risk not captured by the WHI. Or not. It’s an ongoing debate among health experts not likely to be decided very soon.

In the meantime, millions of women enter menopause every year, some of them with significant symptoms that impact the quality of their life (and sleep), who must decide whether or not to use HRT to treat their symptoms.

For these women, this newest update from the WHI is reassuring.

The breast cancer mortality data from the WHI are complicated, and based on a small number of cases in the study population. Let’s just say there was no statistical increase in deaths from breast cancer among users of estrogen and progestin. Since the breast cancers occuring in HRT users are not inherently less aggressive, its more likely the lack of increased breast cancer mortality is because most breast cancers are not lethal, while those that might become lethal are effectively treated. Paradoxically, there was a significant decrease in breast cancer mortality among users of estrogen alone, perhaps related to the fact that estrogen, when taken after a long hiatus, actually inhibits breast cancer growth. (I told you, it’s complicated. You can read more about this here.)

Will this new information matter to women deciding about using HRT?

The use of HRT has plummeted in the years since the WHI results were published in 2002.

In my practice, the reason for the decline is clear – women don’t want to increase their chances of getting breast cancer, however small that increased risk may be.

I tell my patients this : “If you take HRT for 20 years, your risk of breast cancer will be about 1% higher. Use it for 2-3 years for menopausal symptoms, and that increased risk is less than a quarter of a percent”.

After considering this, the majority of women thank me for the information, and decline to use estrogen. Their symptoms are just not bad enough to entertain even that small a breast cancer risk. (I’m happy to prescribe HRT or non-hormonal therapies for those who opt to use them.)

This new study won’t change that conversation much, but I will now add that taking HRT for a few years around menopause to alleviate its symptoms will not increase mortality. I’ll also tell them that the data on breast cancer mortality, while complicated, seem to suggest no overall impact. That’s most likely because the overwhelming majority of breast cancers are not lethal, either due to their inherent behavior or our improved treatments.

I’m curious to see if this changes the choices they make.


More Reading on Menopause and HRT 

HRT – Still No Place for Prevention

In a much-needed and thoughtful analysis, the United States Preventive Service Task Force has summarized what we have learned about HRT since the Women’s Health Initiative was published in 2002. (See summary chart above.)

They have also issued draft recommendations on the use of HRT for prevention of disease.

The U.S. Preventive Services Task Force (USPSTF) recommends against the use of combined estrogen and progestin for the prevention of chronic conditions in postmenopausal women.

The USPSTF recommends against the use of estrogen for the prevention of chronic conditions in postmenopausal women who have had a hysterectomy.

I wholeheartedly agree. 

The USPSTF Recommendations do not address the use of HRT for treatment of menopausal symptoms

That’s an important message for women who may want to consider using HRT for a few years around menopause. In that setting, the heart disease risks are probably minimal for an otherwise healthy woman, while the breast cancer risks, while real,  are quite small.

I tell my patients their breast cancer risk is about 1% higher for 20 years of HRT use, so HRT use for a few years conveys far less than even that risk. On balance, using HRT to get through the peri-menopause is not a terribly risky decision, and for women with severe symptoms, it is not an unreasonable choice.

Despite this reassurance, most of my patients with mild to moderate symptoms are choosing to forgo HRT and deal with their symptoms in other ways – exercise, reducing or eliminating alcohol, trying to lower stress. Those with more severe symptoms are still using hrt, but they are aiming for lower dosing and transdermal regimens, or considering non-hormonal alternatives such as SSRI’s instead.

It’s called informed choice. And when given that choice, these days,  more women than not are choosing against using HRT.

And just in case you’re interested, here are my rules for prescribing HRT.

The USPSTF Recommendations do not address premature menopause

Most practitioners agree that women with early menopause benefit from HRT. The risks for osteoporosis in this group is quite high, and symptoms are often quite severe.

In my practice, almost all of these women choose to use HRT until they reach the usual age of menopause, at which they begin to think about it the same way the rest of my patients do and usually wean off over time.

The USPSTF Recommendations do not address vaginal estrogen use

The one symptoms of menopause that does not get better over time is vaginal dryness. Most women, unless they are very frequently sexually active, will need to use something to treat dryness.

One treatment option is low dose vaginal estrogen.  Most practitioners feel comfortable prescribing vaginal estrogen, even in women at risk for blood clots and even in most women with breast cancer.

My patients, however, tend to want to avoid estrogen in any form and so they usually will try non-hormonal treatments first.  I’d say that about half of these women eventually end up using vaginal estrogen.

Those who might disagree with the USPSTF

The window theory believers

The WHI naysayers will argue that no one has properly studied HRT the way in which it is most often used, and in which it is most likely to prevent heart disease – namely, starting at menopause and continuing indefinitely.

They hypothesize that there is a “window of opportunity” during which estrogen will protect against heart disease, and after which starting estrogen will worsen pre-existing heart disease.

Subgroups analysis of WHI findings suggest that they may be correct in this regard. In the WHI, women starting HRT shortly after menopause had a lower risk of cardiovascular disease than those starting 20 years or more later.

Unfortunately, even if HRT were cardio-protective, the risks of breast cancer (and stroke and blood clots) with combination HRT use cannot be ignored, and mitigate against prescribing hormones for reason of heart disease prevention.

The Gap Theory Believers

There are those who argue that the estrogen-only arm of the WHI actually showed less breast cancers, indicating a potential protective effect of estrogen on breast cancer. They theorize something called the “gap theory”, which states that estrogen, when started 10 years of more after menopause, actually acts to inhibit breast cell growth. They are supported in this by in vitro data.

Unfortunately, while the gap time theory may explain the findings of the WHI estrogen-only arm, it is irrelevant to clinical practice, since the way most women use HRT is to start it at menopause. (ie., there is no gap)

They are both right – and both wrong

If both the gap and window theories are correct (and I suspect they are), when considered together they actually support the findings of the WHI and the recommendations of the USPSTF.  Starting estrogen at menopause may prevent heart disease, but it increases breast cancer risk.  Waiting to start HRT may decrease breast cancer risks, but it increases the risks of clotting and stroke and dementia. Or, as my mother used to say, “You’re damned if you do and damned if you don’t.”

The benefits of HRT as they exist in practicality and theory come at a price. That price is high enough to recommend against  the use of hormone replacement for the prevention of chronic disease.

If you have any issues with the USPSTF Recommendations on HRT –

You can submit a comment to the Task Force between now and June 26, 2012.


Chart above from USPSTF Draft Recommendations on HRT

The WHI through the Restrospectoscope

There is a concise, well-written discussion of the findings of the Women’s Health Initiative in this week’s Jama by Steven Hully and Deborah Grady. It’s the best short summary of the issues I have read to date, and I would encourage you to read it. I find that I agree with almost all of it.

As a practicing gynecologist caring for women in menopause, however, I find myseld a little more critical of the original study design, which enrolled primarily asymptomatic women on average 10 years post menopausal, which is not how HRT is generally used.

A residual loose end stems from the fact that the average age of women at enrollment for all 3 trials was the mid-60s, whereas the majority of women who use hormone therapy for treatment of vasomotor symptoms are in their early 50s. Hormone therapy among women aged 50 through 59 years in the WHI trials tended to have more favorable effects on CHD rates than in older women, but even if this finding is confirmed, the concerns about increased risk of stroke, venous thromboembolism, and breast or endometrial cancer remain.

This is a lot more than a “residual loose end” – it is a major flaw in the study design that understandably resulted from the good intentions of the investigators not to do harm by randomizing symptomatic women suffering from hot flashes to a placebo. I would agree that in younger women there remains a concern about thromboembolism and stroke, but the use of transdermal lower dose preparations may mitigate much of this risk. (Notice I say “may” – we still don’t have the clinical data we need.) The other big “loose end” is the issue of bioidentical hormones vs horse estrogens and medroxyprogesterone acetate, the drugs used in this study. Both these issues will require another large randomized clinical trial before we can say we have done this issue justice for menopausal women. Fortunately, such studies are now being conducted, and hopefully we will have some of the answers we need in a few more years.

I’m not sure why the risk for endometrial cancer are mentioned in the editorial – it did not appear in the WHI results, and should not be an issue if progesterone is taken according to recommended HRT protocols.

Still, the WHI answered many important questions about HRT. In particular, the decline in breast cancers that has resulted since hormone use dropped after the publication of its results more than supports this trial as one of the most important in the history of American healthcare.

In conclusion, the story of HERS and WHI is an excellent illustration of the evidence-based medicine tenet that practice guidelines should be based on rigorously designed research—preferably 2 or more randomized blinded trials with disease end points—even if consistent observational and mechanistic evidence suggests that such trials are not needed. Animal studies and clinical trials of surrogate outcomes can be misleading, and epidemiologic studies of preventive treatments are particularly susceptible to confounding because healthier individuals are more likely to seek and adhere to preventive measures. Weighing benefits and harms is especially important when considering the use of preventive interventions in healthy individuals, in whom there is a special obligation to do no harm. (italics mine)

It is so important to remember that before the WHI, we were prescribing hormone replacement to women just because they were menopausal. When a drug is being used in healthy individuals, the bar is raised to its highest in terms of safety.

Primum non nocere.

Hormone Replacement: Part 3

In part one and part two of this four part series, I reviewed the history and findings of the WHI, the impact of those findings on the medical establishment, and the newer data that have gotten us all muddled up again in our thinking about estrogen.

In the face of all that uncertainty, I have to practice medicine. To that end, I have created my own set of guidelines for prescribing HRT that I would like to share with you now. Feel free to comment, question, criticize or amend these guidelines, or god forbid, to use them yourself.

Please do not ask me if I am “pro-HRT” or “believe in HRT”. Health care is neither politics nor religion, despite that picture up there. It is, however, an uncertain science. Therefore, one must be wary of anyone expressing extremes of opinion about HRT, either for or against its use, and of anyone claiming to have the final word on HRT.

  1. I am willing to prescribe HRT for any of its approved indications (vasomotor symptoms, vaginal dryness, osteoporosis), provided you understand the risks as well as the benefits and know about alternative treatments for these conditions.
  2. I will give you the best data I can find that defines your personal risks and benefits from using HRT. Unfortunately, that data is imperfect, and may change in your lifetime. This will be frustrating for both of us.
  3. I am willing to prescribe HRT for reasons of well-being, mental or physical, provided you are willing to accept the risks and can describe the benefit for me as best as you can.
  4. If you are at increased risk for or have had breast cancer, I am willing to prescribe HRT provided you accept the risks and we have exhausted the non-HRT solutions to your problem, assuming, of course, that it is a problem that HRT can address. (I can count on less than one half of one hand the number of my patients with breast cancer who would fit this rule, but for them, I have it. )
  5. I will not prescribe estrogen without progesterone if you have an intact uterus. If you use anything less than standard progesterone regimens, you must undergo frequent monitoring of the endometrium.
  6. I am unwilling to prescribe estrogen if you are at increased risk for blood clot or stroke. I will do everything I can to find you an alternative that will address your symptoms or condition.
  7. I am not willing to prescribe HRT for cosmetic reasons alone.
  8. I am happy to prescribe bioidentical hormones, but you must assume they have the same risks as Prempro until there is data to prove otherwise.
  9. I consider vaginal estrogen to be safe in almost every woman. (I’ll let you know if I think you are the exception to this rule, and why.) If you feel otherwise, I completely understand.
  10. I will support your decision to use or not use HRT, and will work with you to find the optimal way to manage your menopause, whether it be through lifestyle changes, diet, exercise, hormones and/or medications. It is, after all, your body and your menopause.

Up next: A few words about bioidentical hormones.

Image: I mutilated The Ten Commandments Pressbook Cover. (I just ordered it from Ebay, and can’t wait to get it and see what other images are inside. It is one of my favorite movies of all time.)

Category: Second Opinions

Hormone Replacement – Part 2

See Part 1 here.

When we last left the WHI, the results of the estrogen-progesterone arm of the study had just been released. Unexpected results which showed the drug to increase the risks of heart disease, stroke and breast cancer…

The Immediate Reaction

Well, we all know what happened next.

Widespread confusion and not a little panic erupted, helped in no small part by the manner in which the WHI’s findings were released (via press release prior to publication of the paper). The investigators did little in their contact with the media to place their results in perspective or to address how they might relate to women using HRT for treatment of menopausal symptoms. Instead, they fanned the flames of hysteria by making statements to the press like “This is a dangerous drug.” (NY times) That was just irresponsible, in my opinion.

Critical analysis of the paper was lost in the melee. The fact that the WHI was never designed to study the use of HRT for treatment of menopausal symptoms was a fact that received cursory mention at most. Questions of statistical significance and relative risk were rarely mentioned. The WHI was accepted as gospel from day one. I don’t fault the media – most of the reporting I read was intelligent and well meaning. But critical analysis of the data itself was notoriously absent, because no one had time to do it before having to talk to the media.

Overall, about two thirds of women who were taking HRT at the time the WHI results were released stopped their therapy; those who stayed on the drugs tended to be those women who had the worst menopausal symptoms.

Physician reactions were just as strong. Some docs stopped prescribing HRT altogether. One doc I know told her patients, “You can kill yourself if you want to, but I am not going to prescribe the gun'”. One patient of mine was told by her former physician, “I won’t let my wife use it, and I’m not going to prescribe it for you.”

Perhaps they are worried about being sued. Given that there are more lawsuits out there against Wyeth than you can count, I can’t say as I blame them.

Still, it seems to me as if that approach merely exchanges the “one size fits all” approach to estrogen with the “no size fits anyone” approach, and is just more of the medical simple-mindedness that pre-dated the WHI. And when has the practice of medicine ever been simple?

The More Measured Response

In a relatively short time, the medical establishment recovered from the acute shock of the WHI. New guidelines for HRT were formulated by the US Preventive Services Task Force, recommending against the use of hormone replacement for the prevention of chronic conditions in women.

But what to do with all those women on HRT for treatment of menopausal symtoms? The WHI did not address these women, and yet they were the ones using most of the HRT. We were essentially left on own to figure this one out. Professional organizations like the American College of Obstetrics and Gynecology and the North American Menopause Society eventually came out with guidelines that were more detailed and helpful than the “just say no” guidelines of the USPSTF. In 2005, the FDA published HRT guidelines for women considering hormonal therapy.

The new guidelines can be summarized simply as: Talk to your doctor, take the lowest dose needed shortest shortest period of time, make sure you know the risks as well as the benefits, and consider alternative treatment when available.

Lower Doses, Shorter Use
The use of HRT for treatment of menopausal symptoms continues to be supported, but at the lowest effective dose for the shortest period of time necessary. Women are now encouraged to try to wean off of estrogen once it is no longer needed for the symptoms of the menopause transition, which last on average about 18 months. For women who continue to need estrogen past that time, the goal is to find the lowest effective dose.

Are lower doses of estrogen safer in terms of heart disease, blood clots, stroke and cancer? No one knows. We do know, however, that they are likely to be just as effective for prevention of osteoporosis, and are often enough to control symptoms. So why take more if less will do?

Vaginal Estrogen
Vaginal estrogen remains a viable option for the many women whose only real symptom is vaginal dryness. To date, I know of no adverse data associated with vaginal estrogen use, except that if the cream is used, one may occasionally get endometrial stimulation. Some experts still recommend some progesterone to balance that out. Newer vaginal estrogen formulations such as the ring or vaginal tablet appear to be quite safe for the endometrium. For women who are concerned, drawing blood levels can confirm that the absence of significant systemic absorption. Additionally, the endometrium can be sonogramed periodically.

Transdermal Estrogen
Since the WHI, some doctors have begun to preferentially prescribe transdermal estrogen, there being some data that it may impact clotting factors less than oral estrogen. Transdermal patches are a good option for women who want to stay with the so-called “bio-identical” estrogen, 17-beta estradiol. Taken orally, this molecule can have a shorter half life, and I have had more than a few patients need to take either higher doses of more frequent dosing to alleviate their hot flashes. For some women, however, patches can irritate the skin, so they are better off with oral.

Natural progesterone has become de rigeuer, although there are no really good data that it is any safer than Provera was. I do prescribe natural progesterone preferentially, because I find it is generally better tolerated than Provera. But I have no major objections to the so-called synthetic progestins, because natural progesterone can make some women very sleepy, even at low doses. (For most this is an advantage, insomnia being a symptom of menopause.)

Transdermal progesterone, while a very popular health food store item, is very variably absorbed through the skin, and not yet available in any reliable product for endometrial protection. As a result, it is not recommended to “balance” the estrogen part of HRT.

Some women cannot tolerate any progestin, oral, transdermal, natural or synthetic. The Mirena IUD might be an option for these women.

Non-Hormonal Options
For symptomatic women unwilling or unable to take HRT, newer non-hormonal options are being studied and offered off-label for treatment of hot flashes. The SSRI’s fluoxetine, paroxetine, venlafaxine and the anti-seizure medication gabapentin have been shown in small controlled trials to reduce hot flashes by about 65%. Not as good as estrogen, but for some women, enough. No large long term trials of these drugs have as yet been done.

Do you see what is happening? We are beginning to custom tailor the therapy to the patient – her needs, her symptoms, her concerns, her side effects. And her risks. Because, most importantly, doctors have began to weigh the risks and benefits of estrogen a little more carefully, patient by patient.

No more one size fits all. And that’s just better medicine, if you ask me.

The Plot Thickens…

Since 2002, the plot has taken a few twists. The first was the publication of data from the estrogen-only arm of the WHI. In that group, breast cancers were actually fewer in women who took Premarin alone, though not statistically so. (Of course, this data got much less attentionia atention than the combined hormone arm of the study.)

This difference between the data on estrogen-alone and combination HRT has led many to believe the the evil hormone is not estrogen, but progesterone. Or more specifically, medroxyprogesterone acetate, the progestin compound in Prempro. But is the so-called “natural” progesterone safer than Provera? Your guess is as good as mine. Again, no long term data.

The second twist has been an analysis of the younger women in the estogen-only arm of the WHI, those who began the drug in their 50’s, in a manner more typical to the way we gynecologists tend to prescribe it. In this group, there was not an increase in heart disease, but a suggestion of a decrease. This would support the hypothesis of a so-called “window of opportunity”, a period of as yet undefined time after menopause during which HRT, if started, might actually do all the heart-friendly things we all had thought it would do.

Recent findings from the Nurse’s Health Study also support this hypothesis, and suggest that for women who use HRT beginning at menopause, the risks of heart disease are reduced almost a third. A randomized, placebo-controlled trial to investigate this possibility is in the works, but don’t expect an answer anytime soon.

The relatively more favorable data from the estrogen-only arm has led science-saavy women to begin to ask for hormone regimens that use less progesterone. The problem with this approach is the ever-present risk of endometrial cancer from long term use of unopposed estrogen. I have some concerns about this – essentially treading off one low cancer risk (breast) for a higher one (uterine). Quarterly progesterone regimens are associated with higher rates of endometrial hyperplasia, and that has been my experience when my patients have tried these regimens. I have had better success with quarterly progestin regimens when they are combined with less than standard doses of estrogen. I have had one patient go to hysterectomy because she kept developing hyperplasia and was no longer willing to take progesterone of any kind, ever.

So, where are we now?

And so, four years later, we are practically back where we started. Maybe estrogen is good for the heart, as long as you start it early enough. Maybe the breast cancer risk isn’t from estrogen, but from progesterone. Maybe lower doses or transdermal regimens will prove to be safer. Maybe, maybe…

So how does a woman make a decision about HRT given all the unanswered questions? Her hot flashes will not wait for a better study to come along. And how does one practice medicine in this era of uncertainty? It’s not easy, I can tell you. You need to be clear in the face of fuzzy data. You must be internally consistent when externally, all is not. You have to be flexible, but not wishy-washy. It isn’t easy, but it can be done.

Up next…How I do it

Category: Second Opinions

Hormone Replacement Therapy – Part 1

I’ve been meaning for some time to pull all my thoughts together about hormone replacement therapy. The opportunity presented itself recently when I was asked to give a talk on the status of HRT four years after the WHI. After giving that talk, I realized that I had a lot more to say, so I am going to just say it all here.

This will be a four part, four day post. The first part is background on HRT use before the WHI, then discusses the WHI and its results. Part two addresses the response in the medical community to the findings of the WHI. Day Three will be TBTAM’s Rules for Prescribing HRT, and Day 4 will address the use of bioidentical hormones. So hop on board, the winds are fair and we’ve got a lot to cover…

Those Were the Days

Those were simpler days, the years prior to the publication of the WHI. “Simple” meaning that, for the most part, we had a “one size fits all” approach to managing menopause. You were a woman, you had menopausal symptoms, you had a heart — you got estrogen. It was that easy.

It seemed simple, but we were not stupid. There was more than a little scientific support for our approach to menopause. Dozens of well done, retrospective and cohort studies had all pointed to the same conclusion – women who took HRT had less heart disease than those who did not.

This conclusion made biologic sense. We knew that the onset of menopause was associated with an increase in the incidence of heart disease in women. Moreover, estrogen raised HDL, the good cholesterol, and lowered LDL, the bad cholesterol. Finally, there was indirect and animal evidence that estrogen caused vasodilation.

Turns out that estrogen also appeared to prevent osteoporosis. It was, and still is, is the most effective treatment for the symptoms of the menopause transition – hot flashes, night sweats, and vaginal dryness. Not to mention its effects on the skin, increasing collagen and decreasing wrinkles.

And the risks? Well, estrogen was known to increase the risk of thromboembolism. In addition, it had been known since the 70’s that it could lead to endometrial cancer, but that risk was alleviated if progesterone was added to the mix. Breast cancer, of course, was a persistent concern, showing up as a potential risk in a number of studies. But the results were inconsistent, even in meta-analyses.

Given that heart disease was the number one killer of women, and that the cardiac benefits seemed to be more consistently apparent than the breast cancer risks, there was a real hope in the medical community that the cardiopreventive properties of estrogen would outweigh the risks.

And yet, looking back at the review articles on estrogen written prior to the WHI studies, all the experts were clear that more evidence was needed before recommending HRT for heart disease prevention on a widespread scale. Almost every paper I read was careful to recommend individualized treatment, weighing the risks and benefits for each woman before prescribing.

As for the world of practicing docs, the primary indication for prescribing HRT in those days was still menopausal symptoms, but we had a much lower threshold for prescribing estrogen than we do now. It didn’t take much in the way of hot flashes for us to prescribe HRT, especially for women we thought might additionally benefit from the cardiac effects of estrogen. For some, a high LDL alone was enough reason to use estrogen. (Remember, we did not have the statins then). And more than a few physicians gave all their menopausal patients estrogen, regardless of symptoms. “My doctor said it was good for me.” is what some of my patients would tell me when I asked them at their first appointment why they were taking estrogen.

Overall, I would say that, prior to the WHI, estrogen was something we felt good about prescribing. We liked it and our patients liked it. It was one of those drugs that had a real and visible effect, which for the overwhelming majority of women was positive. Think about it – relief of hot flashes, insomnia, night sweats and vaginal dryness, prevention of osteoporosis, normalization of lipids, and the possibility of heart disease prevention. That’s one heck of a drug.

Big Pharma, as it turns out, also liked estrogen.

Wyeth and the WHI

Wyeth been the leader in the HRT market for years, spurred on in no small way by the publication in 1996 of “Feminine Forever“, a book which touted the wonders of estrogen in staving off the effects of aging. (The book, as it now turns out, was entirely underwritten by Wyeth. Why am I not surprised?)

Wyeth’s drug was Premarin, a proprietary mixture of conjugated equine estrogens. Equine – that’s a nice way of saying estrogens isolated from horses, specifically from the urine of pregnant mares. (I had read that if you crush the pill , soak it in water, and then sniff, it is clear where it comes from. I did not believe it until I tried it for myself. It’s true.)

I need to say at this point that, having prescribed Premarin extensively in the past and still occasionally now, that it is a very effective estrogen, and despite the smell, I have no real issue with its source. Many women do, however, and so I rarely prescribe it these days since there are numerous alternatives.

For the progesterone component of HRT, Wyeth had Provera, or medroxyprogesterone acetate. (Same stuff as in Depo-Provera, the birth control shot, but at a much lower dose). The two hormones are combined together as Prempro. I did not prescribe Prempro as much as I used Premarin, having moved towards micronized progesterone as soon as it became available, because I found it in general to be better tolerated than Provera.

By 1990, Premarin was one of the most prescribed drugs in the world, and was approved by the FDA not only for the treatment of both hot flashes and vaginal dryness, but also for the prevention of osteoporosis. When studies began to rack up showing that estrogen was also good for women’s hearts, Wyeth asked the FDA to approve Premarin for the prevention of heart disease. Of course, doctors everywhere were already prescribing the drug off-label for that reason. But with a formal FDA indication, Wyeth would be allowed to advertise for the indication, and we all know what Big Pharma advertising does to sales. It was sure to be a slam dunk, and a very profitable one at that, since the prevention market for a drug is absolutely enormous compared to the treatment market.

The problem was that all the data up to this point supporting the cardiac benefits of estrogen were from observational, retrospective or cohort studies. And there were those pesky little problems of breast cancer and blood clots that kept creeping into the study results. The risks appeared relatively small, but they were there nonetheless.

Despite this, the FDA’s advisory committee had actually recommended that Wyeth be given the cardiac indication based on the observational data. But in an unprecendented move, the FDA went against their own advisory commitee’s recommendation and asked for a randomized, placebo-controlled trial. The trial Wyeth would begin was the HERS trail, a study of HRT use in women with pre-existing heart disease, believing, I’m sure, that estrogen’s protective effect would show up quickly in this group at high risk. (It did not, as we found out later.)

At around the same time, feminists at the NIH, FDA and on the Hill were pushing for federal funding for research in Women’s Health. Congress responded by funding the Women’s Health Initiative, a large prospective study of women’s health enrolling over 60,000 women for 15 years. The jewel in the crown of the WHI was a randomized, prospective, placebo-controlled trial of hormone replacement in menopausal women. The study was designed to determine once and for all whether or not estrogen prevented heart disease and treated osteoporosis, and to answer the lurking question about breast cancer risks. Everyone, including the study investigators, expected favorable findings for estrogen in terms of heart disease.

But which estrogen to use? Well, Premarin and Prempro were the most prescribed hormones at that time, so it made sense (although maybe not biologic sense) to use them. And Wyeth? Well, they were more than happy to supply drug for the study. And that’s how Premarin and Prempro became the drugs used in the WHI.

There were two study arms to the WHI – one arm for women with a uterus, who received either Prempro or placebo, and a smaller arm for hysterectomized women, who recieved either Premarin or placebo.

The Bombshell Results

The first sign of trouble came when the HERS study results showed no benefit to HRT use in women with pre-existing heart disease, despite favorable effects on blood lipids. In that study, there was actually an increase in cardiac events in the first year of hormone use, probably due to clotting effects, that disappeared in the subsequent years such that overall, no difference was found in cardiac mortality among users and non-users of HRT.

Then came the results of the WHI, results that came earlier than expected when the Data Safety Monitoring Board decided to stop the study prematurely. Because as we all know now, Prempro did not to prevent heart disease in the WHI. Not only that, it actually appeared to increase that risk, along with that of blood clots and stroke. At least among the women in the Prempro arm of the study, who were on average 10 years post menopausal and 64 years of age at enrollment.

WHI results from the WHI Newsletter

To make things worse, breast cancers were also increased among users of Prempro in the WHI, to the tune of 8 additional cases for every 10,000 women for every year of use. Not a statistically significant increase if you want to get picky. But, since this was a prevention and not a treatment trial, the bar for safety was much higher since subjects were supposedly healthy to begin with. And the Data Safety Monitoring Board was not about to accept an increase in breast cancer if there was no cardiac benefit. Even the benefits in terms of bone and colon cancer were not enough to tip the balance in favor of HRT.

So, they stopped the trial. (The estrogen-only arm was allowed to continue till 2005, when increased stroke risks stopped that study.) Sure makes sense to me.

Whether the whole study design itself made sense in the first place is another story. The study population was deliberately weighted towards older women. Since most cardiac events occurred among older women, benefit would show up earlier in this group. Younger women with menopausal symptoms were specifically excluded from the WHI because it was felt to be unethical to give these women placebo. Of course, this is exactly the opposite of how HRT is prescribed in the real world, and remains the biggest limitation of the WHI in my humble opinion.

Which raises the question – Is an average age of 64 just too late to start estrogen? By that age, arterial plaque is likely to be long-established, and adding in a drug known to increase risk of blood clotting may not be such a great idea. Some experts still believe that if estrogen is given earlier in life, say at the menopause, it may still be preventive for heart disease. Think of it this way – Running is good for you, and can prevent heart disease, but if you are not a runner, have blocked coronaries and try to run a 10K, you might just trigger an MI, which would not be good. It’s an intruiging hypothesis, but at the time of the publication of the WHI, it was just that. An idea.

In 2002, like it or not, the bottom line was that HRT was associated with an increase in the risks of heart disease, stroke and breast cancer. In a disease prevention trial, that just doesn’t hold water, especially when the primary outcome, heart disease, was not prevented anyway.

And that, my dears, is why they stopped the study.

Up next…The Reaction

Category: Second Opinions

Calcium Confusion From the WHI

In yet another media “bombshell” from the Women’s Health Initiative and the New England Journal, we are now being told that calcium isn’t as good as we thought in preventing fractures. (See study here.) Moreover, women who take calcium supplements have higher rates of kidney stones than women who don’t. And once again, the media is off and running…

No problem. I can handle this one with my hands tied behind my back…aiebpqiwuy756435yh 46ty2w4… Oh, sorry, I forgot that I can’t type with my nose. Let me get my other shoe off, I’ll use my toes. Okay, are you ready? ‘Cause here we go…

First, let’s talk about the intervention studied by the WHI : Approximately 36,300 women nationwide were randomly assigned to placebo or 1000 mg/day elemental calcium combined with 400 IUs/day of vitamin D. The primary outcome was osteoporotic hip and other bone fractures, with colon cancer as a major secondary outcome. Overall, women assigned to take calcium supplements had no less osteoporotic fractures than those taking placebos.

What???? Oh, wait. I understand. When they looked at women who actually took the pills as advised, there was a 29% reduction in fractures.

So, what they seem to be saying is this – telling women to take calcium and vitamin D doesn’t prevent fractures. Sort of like telling my kids to brush their teeth to prevent cavities – it only works if they actually do it. OK, I’ll buy that.

The statisticians don’t like to hear this. They prefer the purity of an intent to treat analysis, where the only valid comparison is what goup study participants were assigned to, not what they actually did. As a clinician, I alswys have trouble with intent to treat analysis – I tend to want to drill down and find out why. I would argue that this is as valid as the intent to treat.

A benefit was also seen to calcium supplementation in the over-60-years-old crowd. Gee, that makes sense too. I wouldn’t expect to see a big difference in fractures in the under 60 crowd, because that group doesn’t get many fractures. And if someone in that age group is breaking her hip, I’d have a strong suspicion there’s something else going on other than just age. Like maybe anorexia or smoking or an overactive parathyroid gland or malabsorption, to name a few possibilities. No amount of calcium is going to fix that. Hmm…So far, it sounds like calcium is doing exactly what I’d expect it to do.

But that’s not what CBS News says. Their headline? “Calcium, Vitamin D Assumptions Shaken.” Why? Because the big groups analysis showed no benefit to calcium, that’s why. And do you know why? Because the women who received placebo were allowed to take calcium and vitamin D supplements on their own! That’s right. Here it is, right out of the WHI press release: “Since participants were not restricted from taking personal calcium or vitamin D supplements, they had a relatively high calcium and vitamin D intake at enrollment and intake rose even higher during the trial so the impact of study supplementation may have been muted.”

What the researchers are telling you is that since they didn’t control the very intervention they were studying, their results might not be right.

Oh, and one more confounding issue. They also did not restrict the use of Fosamax-type drugs in either group, so about 15% of women ended up on these fracture-preventing meds during the course of the study. This means that most likely the women at risk for fractures in both groups were already taking a more effective medication than just calcium and vitamin D.

Are you getting all this?

Now, let’s talk some more about the WHI’s intervention – namely, giving all women the same dose of calcium, regardless of what amounts they may have been getting in their diet.

You know what? This is not the way we doctors recommend that women take calcium.

Here’s the recommendation of the National Osteoporois Foundation and the American College of Obstetricians and Gynecolgists: “Food is the best source of calcium; however, most Americans do not have enough calcium in their diets. Fortunately, calcium-fortified foods and calcium supplements can fill the gap, ensuring that the daily calcium requirement is met. The amount needed from a supplement depends on how much calcium is consumed from food sources.”

Hmm…I don’t see any recommendation that every woman be given 1000 mg of calium as a supplement daily. Gee, I wonder what might happen if women were to take calcium supplements without taking into account their dietary sources? Maybe they’ll get too much calcium. And what can too much calcium cause? (Let’s shout it out together, shall we?) KIDNEY STONES!

Oh, wait, wait… I forgot to tell you about the vitamin D dose studied – 400 IU daily. Unfortunately, it’s too low. The current recommendation is 800 IU daily. (And not 600 IU as the NY Times says.)

Quick take

And there you have it, folks. The WHI researchers designed an intervention that doesn’t match current medical recommendations, failed to adequately control the intervention they were studying, and then used their results to question the very medical recommendations they didn’t follow. Yet despite this, they did manage to get some results that make sense if you understand what really happened.

Fortunately, Gina Kolata at the New York Times understands it. Here’s her headline: “Big Study Finds No Clear Benefit of Calcium Pills”. That just clears it all up for everyone, doesn’t it? And this from Forbes: “Calcium, Vitamin D Won’t Protect Older Women From Fracture.” The fact that it was the older women who actually did show a benefit is irrelevant to the headline. No wonder we’re all confused.

Here are the current recommendations for calcium and vitamin D (click here). I wouldn’t change a thing.

Mrs. O’Leary Milking Daisy by Norman Rockwell

Comment : define_me
February 16, 2006
Maybe I should aim to get my first publication with the New England Journal of Medicine…hehehe

Comment: Anonymous
February 17, 2006
Between the calcium and fat study, the headlines should read: Impossible to do prospective food study; massive waste of money and time. Conclusion: retrospective food studies, or even large population studies (Japanese on traditional diet v. American diet) are clearly the best information we’ll get—and far cheaper.) I’m not even going to think how many children could be fed and vaccinated, or how much prenatal nutrition could be paid for with the 415 million from the fat study…

Comment: Guinness_Girl
February 17, 2006
That’s all I have to say. Wow