Dr. Amsterdam: Here with me are two experts on heart disease in women: Dr. Kristin Newby, Professor of Medicine at Duke University School of Medicine and Co-Director of the Coronary Care Unit at that institution, and Dr. Jeffrey Borer, Chief of Cardiology and Chair of the Department of Medicine at Downstate Medical Center in Brooklyn, New York.
We're going to discuss an issue that has been in the headlines for more than a decade: heart disease in women. We’ll discuss what we know about it, where we've been in terms of problems that have been elucidated, and how we've addressed those problems. We all know that heart disease is the chief cause of mortality in women over their lifetime. Heart disease is actually the second highest cause of mortality in women until the age of 75; it is second to cancer until that age. But it is so great after the age of 75 that when averaged over the lifetime, heart disease becomes the chief cause of mortality in women. I think that may provide some perspective. Heart disease, particularly coronary heart disease, which is the major cause of mortality in women among the cardiovascular diseases, rises sharply after menopause. Coronary heart disease is unusual in women prior to menopause, unless there are outstanding risk factors such as familial hypercholesterolemia or diabetes. Finally, the risk of heart disease in women is always less than men throughout the lifetime.
Kristin, you authored the chapter with Pamela Douglas on heart disease in women in the latest edition of the Braunwald text on heart disease.1 Can you tell us about the situation as you see it now in terms of what we recognize as previous underdiagnosis and undertreatment of heart disease in women with a focus on coronary heart disease.
Dr. Newby: Thank you. I think that's an important question. We can look at a number of sources and I think most of them suggest that we have underappreciated the risk of heart disease in women. In part that's because women develop coronary artery disease later in life or the consequences of coronary artery disease are manifest later in life. I think every signal tells us that we are doing better, including physicians’ perception of risk and treatment. There was a recent study by Lori Mosca published maybe a couple of years ago that shows that women are increasingly aware of heart disease as a major risk factor.2 That said, however, only about half of women think that heart disease is their leading lifetime risk; many still believe it is cancer. So it's about 50/50 now, where it was about 30/70 maybe 5 to 10 years ago. We're making progress, and I think the Red Dress3 campaign and other awareness campaigns have been effective.
We still see gaps in treatment; use of evidence-based therapy in women is lower than it is in men, although it has improved. I think importantly on the side of clinical trials and testing new therapies we're seeing a better balance. But there is still underrepresentation of women in clinical trials that establish evidence for treatment.
Dr. Amsterdam: Your last point is key, and leads me to note that an age cutoff of 75 for enrollment in many clinical trials may be partly responsible for female under-representation, because it’s after 75 that cardiac disease really proliferates in women.
Jeff, what are your views on how much progress we’ve made?
Dr. Borer: Of course, I agree with everything Kristin said. I would emphasize something that we don't talk about so often, and that is that when women do manifest evidence of coronary disease, and particularly when they have a myocardial infarction (MI), they’re about twice as likely as men to die within a year of that first infarction. If only we were more attentive we would do something to modulate that bad outcome. Women are much less likely to get the needed drugs such as beta-blockers, angiotensin-converting enzyme inhibitors, aspirin, and other therapies known to improve survival after MI than are men. If we were more aggressive about that, maybe the outcome would be better. However, the problem is that men who have prognostically important coronary disease, which may be heralded by a first infarction, actually are best treated, or I should say their outcome is best improved, if they undergo coronary artery bypass grafting. One might think, therefore, that this strategy should then be applied to women. However, women are much more likely to die during a bypass grafting procedure than men.
Therefore, the appropriate therapy for women to reduce this extraordinarily higher risk after MI is not clear. I don't think we’re as good at giving medicines as we ought to be, but going the next step and using invasive therapy, that’s something else again. It’s not clear whether women benefit sufficiently to warrant that approach, and that's something we have to think about. We need a great deal more data on that subject.
Dr. Amsterdam: I would add a couple of short points. Kristin indicated that things have gotten better. In fact, in the past 10 years there's been a sharp decrease in the rate of death from cardiovascular disease in women. If you look at the curve of the decrease in mortality from heart disease for the past 50 years in this country, there is a steady downtrend. But when we break up the data, the curve for women was pretty flat, indicating that they did not share in this benefit until recently. It was mostly men until the past decade, when we saw a sharp drop in mortality in women. The death rate in women has fallen almost to that of men. This is attributable to programs such as the Red Dress campaign, as Kristin mentioned, and the “Get with the Guidelines” campaign. When women and men are risk-adjusted after an MI, the outlook is similar—women are older and have more comorbidities when they have their MI. So the question is, is there something intrinsic that puts women at higher risk, or are they just older, sicker, and with more comorbidities, which leads to a different outcome? Jeff?
Dr. Borer: I think it's a good point and I don't really know the answer, but the fact is that the disease is different in women than in men, whether that accounts for some of the difference in outcomes that we see or not. Typically, atherosclerosis in the coronary artery is much more diffuse in women than it is in men, so that angiographic findings often suggest narrow normal arteries when in fact they’re loaded with plaque. In contrast, men’s obstructions tend to be relatively discrete. That difference in the manifestations of atherosclerosis in the coronary arteries in women and in men may well have an important impact on outcome.So while it may be that risk adjustment makes things look more similar among men and women, the fact is the disease is different.
Dr. Amsterdam: Important point. Kristin, would you like to add something on that topic?
Dr. Newby: I think the points that you both have made are good in that looking at risk-adjusted outcomes is important and it gives us an idea of one group relative to another. Unfortunately, sometimes that masks the issue, which is that women are dying at a pretty high rate. Whether it's because they're older or not, it suggests that we've got a high-risk population in which we need to figure out what is different in the way they respond to therapy or whether there are other interventions that could be applied that might improve outcome.
Risk adjustment is important, but I don't think it is the whole answer and it doesn't do away with the challenge of the high event rates in women.
Dr. Borer: I agree with that and if you just look at patients younger than 50, just under age 50, a MI in a woman is twice as likely to be fatal as it is in a man.4
Dr. Amsterdam: On the other hand, of women under 35 who come to the emergency department (ED) with chest pain, only about 1 in 1000 turn out to have an MI. So there's a high frequency of chest pain in women and they generally tend to use the ED and medical system more than men for symptoms. This may be because there’s less denial in women, but there are lots of negative tests in women for symptoms like chest pain or chest discomfort. Would you agree with that, Kristin?
Dr. Newby: It is true. I think there are a number of other features, unfortunately, that may lead, perhaps not in young women but in older women, to missed or delayed diagnosis that may also contribute to their worse prognosis. That may explain some of the difference as well. I would just throw that out for discussion.
Dr. Amsterdam: A lot of the data that we’re talking about in underdiagnosis and undertreatment come from papers, many of which are 10 years and 20 years old. In your own institutions, are you seeing a persistence of these issues, because I'm not aware of any recent studies indicating that there's been an improvement in these trends, other than that we know that cardiovascular disease leading to mortality is decreasing in women.
In your own institutions, do you see a persistence of the woman who sits in the ED with chest pain and doesn't get an electrocardiogram (ECG) within 10 minutes, or a good workup, or a referral from a clinic or office for chest pain? Jeff, what’s your experience?
Dr. Borer: The issue, I suppose, is, what does the cardiologist do? I think over the past decade or more, maybe 2 decades, there's been a heightened sensitivity among cardiologists of the fact that women do in fact have atherosclerosis, they do have coronary disease, they do have MIs, and they do die from it. But I think the problem is a little bit different.
It's not so much now, I think, allowing a woman to sit in the ED with chest pain for too long. I don't think that happens terribly often, at least not from what I've seen. It's that the symptom pattern is different in women than in men, and while a lot of women have chest discomfort and often the chest discomfort isn't associated with coronary disease, a lot of women have coronary disease with symptoms that don't include chest discomfort. The difficulty in the sensitivity to that possibility is something that we still have to improve.
Dr. Amsterdam: Kristin, Jeff raised an important point about the difference of symptoms in women and men when they present with coronary disease. We know that symptoms in the elderly are different than those who are younger when presenting with ischemic heart disease. So is it that the symptoms are different in women per se, or is it that they comprise a large part of the elderly population with coronary artery disease, and if you compare elderly men with elderly women, would there be a difference in symptom presentation?
Dr. Newby: That's a good question. You know one of the better studies that's been done on this actually revealed that the majority of patients who were having an acute MI presented with some type of chest-related symptoms, such as tightness and pressure.1
My personal experience, and again I don't think this has been well studied specifically in the elderly, is that the issue with atypical symptoms is in part a phenomenon of age, where the elderly tend to present more frequently with just dyspnea or fatigue. One really has to be paying attention to that. That disproportionately affects women who present later in life with acute ischemic syndromes.
Dr. Amsterdam: I've asked our interventionists, what is the presenting symptom in a woman coming in with a an ST segment elevation MI? The response is that it is chest pain—and commonly crushing chest pain.
I want to move to the risk factors for coronary disease in men and women. Jeff, any differences or similarities—I will tell you my view right off the bat that the risk factors for coronary artery disease other than pregnancy and menopause, which act through the traditional risk factors, are the same in men and women. There are some differences, and in her chapter in Braunwald, Kristin pointed them out nicely with regard to diabetes from the INTERHEART study.5 What's your view on the risk factors in women? Do we need preventive cardiology clinics for women and for women with heart disease; or can women come to a good preventive cardiology clinic that men also attend?
Dr. Borer: I think the risk factors are pretty much the same. You know there may be some difference in outcomes associated with them, depending on age, with diabetes being the major issue. I think it has more of an impact on the risk of second MI in women than in men, so you have to be more cognizant of that particular risk factor in women.
But I think the risk factors are the same and the outcomes are qualitatively the same and so managing patients by trying to modulate risk factors could be done in a preventive cardiology unit. I don't think you need a separate one for women because I don't think we have sufficient data to suggest a different strategy for men and for women.
Dr. Amsterdam: Kristin?
Dr. Newby: I fully agree. I think that the risk factors, with the possible exception that diabetes may be a bit more potent in women, are still the same risk factors and we know the same treatments work in both groups. What I think is probably more interesting that came out of INTERHEART is the differential strength (stronger in women) of protective factors in women, such as exercise, eating a diet high in fruits and vegetables, and stress management. I found that to be interesting and potentially more informative than focusing on the proven risk factors.
Dr. Amsterdam: I think INTERHEART was a very good confirmation of what the Framingham Heart Study showed of the traditional risk factors.6
Would you agree, Kristin, that there is more stress in general, more psychosocial stress, in women than in men? Women work and take care of the kids—men do better now, but we don't do an equal share, and that seems to be having an impact on women and disease.
Dr. Newby: Yes, I think that's an interesting topic, and I think women's response to stress and the stress that they're dealing with may be different from that of men. I think there’s an interesting analysis that’s been done several times about support systems and how men do better when they have a spouse who’s taking care of them. Women on the other hand who have an MI who have a spouse actually do worse. So I'm not sure what that tells us, but I do think there's either some disproportionate amount of reaction to or ability to cope with stress among women. That is important for us to understand and work through.
Dr. Amsterdam: It's too bad we can't measure that the way we measure other data. Now let’s consider a stable patient who comes to the physician’s office with chest pain; it may be typical or atypical, but it's suspicious enough for the patient to require a careful evaluation.
Jeff, which tests do you use to detect objective evidence of myocardial ischemia and, by that factor, the likelihood of coronary disease? Exercise treadmill test has really been criticized for its value in women because of its repute for false-positive results. Some experts have advised going straight to a stress imaging study rather than a plain treadmill test.
Dr. Borer: Yes, well that's what I do, actually. The fact is that the specificity of every noninvasive test for coronary disease is lower in women than in men. The positive predictive value is less in women than in men, whether it be a radionuclide test, an echocardiography-based test, it's not just the exercise ECG. On the other hand, it seems to me that radionuclide based studies do provide greater accuracy in women than the exercise ECG, and that's what I use.
Dr. Amsterdam: This is good because thus far we’ve had little divergence of opinion and it’s good to have some differences. Thus, I am a strong proponent of the standard treadmill exercise test without imaging if the patient has a normal baseline ECG and the history indicates adequate exercise capacity, which is what the guidelines have advocated since 1998. So in these circumstances, my first test is always a plain treadmill study, and we train our fellows to follow this approach. Kristin, your method?
Dr. Newby: Yes, I think I fall a little bit more to Jeff's side of things. So there are two things I think about: one is if I see someone in the clinic who gives me a very good story for symptoms and they can't exercise and can’t do their usual activities. I either move to treat them with medicines and cardiac rehabilitation or with catheterization if I have a high index of suspicion that I'm dealing with coronary disease. So I think that's the first thing. Then the question is, how do we think about the individuals who aren't clear-cut? We want to rule out the group that has the highest likelihood to have a false positive—the ones with very atypical symptoms, which I know we all see in our clinics—and try to work with them without doing any kind of further workup if at all possible. But if I do a stress test in the clinic and I'm on the fence, I, like Jeff, add imaging.7 We tend to use more stress echocardiography than stress nuclear testing, but I don't think there's any evidence that one or the other, if you're adding an imaging study, is of a major advantage given the certain limitations of each study in terms of acquiring the images.
Dr. Amsterdam: I agree. They're very similar. Perhaps nuclear is a little more sensitive but less specific than echocardiography.
So I will say quickly, I think we all know the guidelines of the American Heart Association/American College of Cardiology (AHA/ACC)8—that I just referred to. They’re very clear in stating that the first test, even in a stable female patient with a normal ECG who is able to exercise should be a plain old treadmill test. I'll make a couple of points; one, we have in press a paper on how to improve the positive predictive value of the exercise ECG in women. The more risk factors the patient has, the deeper the ST depression, the longer the persistence of ST depression post-exercise—these factors can increase the positive predictive value up to about 80% in women. These ECG factors and non-ECG factors can help. We have also shown that if the patient, man or woman, has a positive treadmill ECG and can exercise to 10 metabolic equivalents (METs) (Stage III of the Bruce test),9,10 and then is referred for a stress imaging test, about a 94% likelihood that the stress imaging test will be negative.
So the key point for me is that assessment of the exercise test by ECG criteria alone is really obsolete. The poor positive predictive value and the high rate of false-positives are based solely on the ST segment when we know that the best prognostic indicator is functional capacity. So that's why we always do the treadmill test as the initial assessment unless there's a baseline abnormality or the patient can't exercise.
Do you have any comments on any of that, Jeff?
Dr. Borer: I would like to pick up on two things that Kristin said. She talked about symptoms, and I would say that if a person, and it doesn't matter whether it's a man or a woman, has typical symptoms, when you take a history and you put that person on a treadmill and you reproduce the symptoms and you do an ECG alone, the concordance of the ECG criteria and the symptoms has a very high predictive value. But that presumes that the person is going to get symptoms on the treadmill, and that may not happen.
The other point that Kristin made was that in certain patients with a compelling clinical presentation, she might go right to a catheterization. I would have to say that I don't do that. I guess it depends on how you define “compelling.” If somebody has a history that is consistent with crescendo angina, yes, I would send him or her to catheterization. But absent that, I would not because I believe that the noninvasive testing provides prognostic information even in the presence of true coronary disease.
The problem that I fear in sending someone to catheterization is that, in the absence of prognostically important disease, they'll come away with an angioplasty with no evidence at all that anything will be done to their natural history, only that their symptoms will be relieved and that could probably have been done with medication. So I'm a little more wary about sending people directly to catheterization. There really has to be what I would call a “compelling” history, and it may be the same that Kristin would call “compelling” or it may not.
Dr. Amsterdam: We’re into the art of medicine, and it’s legitimate to differ if we have good reasons to go down different paths. Kristin, your comments on this?
Dr. Newby: When I'm thinking of somebody whose symptoms are compelling enough that I would send them to catheterization, it's the woman who was active, doing her gardening, mowing her yard, whatever last summer, and has over the past several months progressively deteriorated in terms of activity levels. So you already know their functional capacity has declined by what they're telling you through their history. In those individuals, I don't think you gain a lot more information by documenting that on a treadmill.
You're right, maybe you take that person and you try medication or maybe you commit that they have coronary disease, you're going to take care of their angina, get them immediately back to functional capacity. I think those are style points, as you said, the art of medicine. To some extent it's driven by patient preferences and I think that's the thing that sometimes we also forget.
There are patients who, once they know what you're thinking, don't want to deal with multiple steps to get to an answer—they want to know. Maybe that's right, maybe that's wrong, maybe that's part of what's wrong with our healthcare system. We could debate those points for another hour probably, but I think those factors all play into how we manage patients.
Then you think about taking somebody when you're pretty sure you know what the diagnosis is, or maybe you're pretty sure you don't think it's the diagnosis, but you put them on a treadmill or a treadmill with imaging and it's positive, and now you've got them worried. They want to know what's going on, and they may or not may not tolerate medicines.
These are some practical issues in medicine, as well, that we have to think about, and again, as you said, that's the art of medicine.
Dr. Amsterdam: Kristin, can you tell us a little about the Duke treadmill score? Because it's a very good prognostic indicator.
Dr. Newby: Absolutely, and that's very much like the study that you described of Dr. Beller’s; if they could do 10 METs they did well. We look at our treadmill information, often we get it in conjunction with an echocardiogram, but we don't ignore the key physiologic information and functional information from the treadmill for that very reason; there's a high correlation between either METs done or various scores. The Duke treadmill score being one type of score that one can use to estimate risk of dying in the next 5 to 10 years. So it is critically important that we don't ignore the wealth of information that can come from a treadmill test.
Dr. Amsterdam: Jeff, you mentioned chest pain and it reminded me of something important. One of the values of stress testing is whether we can reproduce symptoms that the patient complains of. I notice in our chest pain unit that patients are admitted because of chest pain and they have a normal resting ECG and negative troponins. Our first test in these stable patients is, again, a treadmill test, symptom-limited. They do more activity on that stress test than they do in life but it's very rare that we reproduce chest pain. Even the patients with coronary disease who go to catheterization because of a very positive stress test infrequently have chest pain during treadmill testing. Does your experience differ?
Dr. Borer: No, my experience doesn't differ. When people come out of our chest pain unit they have an imaging study, they don't have a straight treadmill exercise ECG, but that's irrelevant. One thing that is perhaps not irrelevant is that in New York State, in the absence of unstable angina, you now cannot send a patient to catheterization unless you have demonstrated a positive exercise test.
If they have unstable symptoms certainly you can catheterize them, but in a patient who has relatively stable symptomatology, you can't catheterize (and expect to be paid for the procedure) unless you have evidence of ischemia by a noninvasive test. Those are called “appropriate use” criteria that have been set up here in New York, and if you don't follow them, the insurance companies don't pay. So there is sort of a brake here that may not be true in other states.
Dr. Amsterdam: Kristin, do you have such a proscription in North Carolina? What about your experience of chest pain on the treadmill?
Dr. Newby: First of all, we don't have such a proscription here in North Carolina, although I think we’re going to see more and more of that. New York State has historically been a leader in that kind of thinking through and monitoring and implementing criteria like these, and I don't think it's all a bad thing.
In our chest pain unit or in referrals from clinic, I agree that most of the time we don't reproduce symptoms on the treadmill and I find it very reassuring if we do reproduce symptoms and we don't see anything. Again we’re doing imaging, but imaging in conjunction with the ECG, I think, is actually helpful information to have, but in the majority of cases we don't reproduce the symptoms.
Dr. Amsterdam: It’s a really interesting phenomenon, and as Jeff has said, the symptoms on the treadmill are very important.
Let's make a couple of comments on heart failure. There are data that women present more typically with heart failure as an indication of underlying coronary disease and also have more diastolic dysfunction than men in terms of the underlying cause of heart failure. Your experience with this, Jeff?
Dr. Borer: I think that's true but, in fact, the truth also is that most heart failure trials include relatively few women. It was the same point you made earlier. I'm one of the principal investigators of the “Systolic Heart failure treatment with the If inhibitor ivabradine Trial” (SHIFT), the largest heart failure trial that was ever done, which used a heart rate–slowing strategy for therapy. We had a marked paucity of women compared with men, so it's hard to draw conclusions.11 I don't know how many women were excluded because they had diastolic dysfunction. It certainly is my perception and my bias that among those with heart failure, women commonly do have diastolic dysfunction, probably more commonly than men, but I think we lack a lot of information about heart failure.
Let me talk about the area on which I focus most closely, which is valve disease. The manifestations of valve disease are different in women than in men. That was demonstrated most clearly in the Placement of Aortic Transcatheter Valve Trial (PARTNER),12 which was the first randomized trial of any mechanical therapy for valve disease in patients with aortic stenosis. In fact, approximately 50% of the population in PARTNER comprised women. The average age was the mid-80s, and by that age there are a lot more women around than men. So perhaps that had something to do with the fact that many of those who entered the trial were women.
The extraordinary finding was that women did better with percutaneous valve therapy than men. Not only did they do better than men, the relationship of percutaneous therapy to conventional surgery was better in women than in men. We could speculate on why that is (smaller arteries, harder to do, etc.), and maybe we’d be correct. However, the finding does suggest that there is a difference, at least for that valve disease, in women compared to men.
Another piece of evidence is the affects of treadmill testing in patients with aortic stenosis. Asymptomatic patients with aortic stenosis are marked out as being at relatively high risk for the imminent development of symptoms or worse based on treadmill studies. One of the criteria that seems to separate those at high risk is ST-segment depression, but it’s 1 mm in men and 2 mm in women. We see the valve dysfunction more commonly in women than in men and the manifestations of the disease differ.
Dr. Amsterdam: These are fascinating data. We're doing an investigation in aortic stenosis and the overwhelming number of patients are men, but, in general, I agree with your point that because we’re seeing patients with critical aortic stenosis who are now in their 80s, women comprise a very large component of that group.
Kristin, can you comment on heart failure and whether there is more diastolic dysfunction, more hypertension in women, and so relative to their body size they have more left ventricular hypertrophy?
Dr. Newby: Yes I think that diastolic dysfunction is more common and more severe with aging. There are also the effects of hypertension. Both of those obviously are issues with older women. I do share Jeff's concern not just related to diastolic dysfunction, but related to any heart failure, of the underrepresentation of women in our studies that are designed to help us understand how to manage heart failure from any cause.
Dr. Amsterdam: Jeff; do you remember the age cutoff in the SHIFT trial?
Dr. Borer: There was no upper age limit. However, the average age was 60 years and 11% were at least 75 years old.
Dr. Newby: And if diastolic dysfunction is more prevalent in the older population than systolic dysfunction, the age of the patients will be a big factor in the etiology of the heart failure.
Dr. Amsterdam: In the past decade there has been a proliferation of publications on women with normal epicardial coronary arteries in whom myocardial ischemia and its complications are caused by (or associated with) coronary microvascular dysfunction (which is also presumed to be the underlying cause of the elusive Syndrome X). What are your comments on this issue?
Dr. Borer: Regarding microvascular dysfunction and Syndrome X, it has been my impression that some people, more often women than men, present with atypical chest pain and even typical angina pectoris in the absence of large-vessel coronary artery disease or any other definable cause, but that the problem of demonstrable microvascular dysfunction in these settings is relatively uncommon. Patients who present in this way may benefit from certain types of vasodilating drugs with relief of symptoms. However, I believe the outcomes for these people are substantially more benign than for those with large coronary artery obstruction. In other words, frank MI or death are relatively infrequently associated with this syndrome. Moreover, although it is plausible to associate such a clinical picture with endothelial dysfunction and small-vessel (arteriolar) hyperreactivity, I think the pathophysiology of the syndrome is not clear. Thus, while Syndrome X may contribute to debility and activity limitation particularly in women, I do not think the syndrome is very clearly understood and do not think it is an important cause of premature death or major morbidity.
Dr. Amsterdam: We're going to have concluding comments from each of you. So briefly, Jeff, are there any comments you'd like to make?
Dr. Borer: I would summarize that the situation with regard to sensitivity about heart disease in women has improved considerably over the past decade or two, but we still have a long way to go to understand the differences between women and men in terms of the manifestations of heart disease and how to manage them.
Dr. Amsterdam: And Kristin.
Dr. Newby: I completely agree with Jeff's summary. The only thing I would add is that of the therapies that we have available, they are all equally effective in men and women, and I think we just need to focus right now as we're trying to understand the underlying differences in pathophysiology and symptoms, among other things, on applying these therapies as appropriately as we can across both sexes.
Dr. Amsterdam: Thank you.
FoxP2 Media LLC is the publisher of The Medical Roundtable.