April 25th, 2010

Why I Don’t Recommend Statins for Primary Prevention in Women

For many treatment strategies, the highest-risk groups derive the most benefit. Prevention of cardiovascular disease is no exception: Secondary prevention is where the greatest benefits lie. With regard specifically to primary prevention of CVD in women, my review of the evidence has led me to conclude that statin use is not warranted as a broad-based strategy. Here are the most important reasons why:

1. No demonstrated benefit in preventing myocardial infarction
2. No demonstrated survival benefit
3. More known side effects in women than in men
4. The likelihood of as-yet unknown side effects of long-term treatment

Does that mean I would never find a woman in my practice whose clinical profile merits primary prevention of CVD with a statin drug? Of course not. But to prescribe a drug to an asymptomatic, healthy person, I want clear evidence that this drug will help her live longer. Without sufficient evidence in hand to show that the benefits outweigh the potential harms, I decline. What about you?

11 Responses to “Why I Don’t Recommend Statins for Primary Prevention in Women”

  1. kudos to Rita

    Kudos to Rita for articulating her views clearly — I just wonder if this is really about risk not sex – and the idea that low risk patients should not be treated with statins — the absolute benefit is just too small.

  2. Mandate for coronary calcium imaging

    There is no question but that if a woman has coronary artery disease, she will be benefited by the use of a statin. The problem is that conventional risk factors mis-classify most women. MESA has demonstrated the ability of coronary calcium imaging to accurately classify risk in women.

    Solution: Do heart scans on all women over the age of 50 to classify risk. Treat those with CAC scores over the 75th percentile for age and gender and all women with CAC>300, both of which are CAD equivalents. Prevent the heart attack before it happens.

    The fact that a woman is asymptomatic does not mean that she is healthy. The initial symptom for 60% of women who die from coronary disease is sudden death. Waiting for symptoms is not consistent with good practice.

    Competing Interests: I have a practice that emphasizes prevention and has evolved a protocol that has reduced heart attacks by 90% compared to national standards and has seen 0 heart attack deaths over 24,700 patient years. To this purpose, I have an investment interest in a facility that can do EBT heart imaging and carotid and aortic ultrasounds.

  3. A case example
    What would you do for a 65 year old woman with hypertension, but reasonably controlled on amlodipine and HCTZ; No history of diabetes, non-smoker, no family history of CAD. LDL 145, HDL 48?

    I personally would put her on simvastatin 40mg.

    Competing Interests: Christopher P Cannon reports receiving research/grant support from Accumetrics, AstraZeneca, GlaxoSmithKline, Merck, Intekrin Therapeutics, , and Takeda; He has served on advisory boards for Bristol-Myers Squibb/Sanofi-Aventis and Novartis (but funds donated to charity) and having equity ownership in Automedics Medical Systems.

  4. Two ways to look at this
    1) Her calculated 10-year Framingham risk is 9%. She is pretty close to intermediate risk and therefore we really need to figure out whether she is actually low (<6%), intermediate (6-19%), or high (20+%) risk for an event. 2) I would image her carotids. There is a good chance she already has an ACAS, and if not that, then at least 1.0 cm2 of total plaque area on her carotid/subclavian tree. 3) But if her carotid plaque is close to zero, I would not give her a statin (the Framingham Risk Score has an AUC-ROC of 0.78, meaning under or over classification in 1 in 5 people). If her carotid plaque area was high, I would treat her with a statin, and using the imaging to motivate adherence, aiming for plaque regression at follow-up.

    • Roshan Karki, MBBS says:

      Are we using statin in the above setting to prevent MI or stroke or combined cardiovascular events?

  5. Very intereting Rita…

    As always, you raise very interesting and provocative questions. I just wonder whether we can afford to adhere to the standards you set. That is, while it is true that no primary prevention trial has demonstrated a mortality benefit in women, it is also true that the question has not been adequately addressed. There is clearly a mortality benefit in secondary prevention trials in women, and there is clearly a benefit in high-risk primary prevention in men. Do you have reason to suspect that a West of Scotland-like trial in women would be negative? One way to think about it is, can you justify the cost of doing such a trial in women or any other subgroup for whom there is no current proven benefit? The CTT meta-analysis (Lancet 2005; 366: 1267–78) looked at over 90,000 patients randomized to statin or placebo across 14 primary and secondary prevention trials (including 21,000 women) and found that the risk reduction was proportional to the reduction in LDL and completely independent from age, sex, or pre-existing disease.

  6. Interesting Case, Chris

    However, in the absence of any data to show benefit for statins in this woman I would not recommend them. Her lifestyle habits were not included , but I would talk to her about the benefits of a heart healthy diet and regular physical activity, such as walking, compliment her on not smoking, and that’s it.

  7. The absence of proof is not the proof of absence
    Although I find Dr. Redberg’s position interesting and provocative, I disagree fairly strongly with it. While the absolute risk among women is of course lower than among men in the short term, there is not a biologically plausible reason to expect a qualititative difference in the effect of statins in primary vs secondary prevention in women. Moreover, epidemiological data do not suggest a lesser influence of cholesterol on the lifetime risk of CHD in women vs men nor do clinical trials suggest a difference in relative statin efficacy between men and women. Finally, this seems to me to be misuse of subgroup data in trials. Treatment interactions by sex have not been reported in most primary prevention statin trials, and the trials of course were not designed to have adequate statistical power to demonstrate hard endpoint reduction in a subgroup.

    If we are to set a standard for primary prevention that requires us to defer treatment until mortality benefit is demonstrated for the subgroup of women in a single trial, we may deny an entire generation of women a therapy that is highly likely to be effective based on the accumulation of epidemiological and clincial trial data. Such an argument seems to also require that we demonstrate similar benefit in all other subgroups (i,e defined by age, race/ethnicity, etc) before using statins in that subgroup.

    I would have an easier time accepting an argument against statin therapy for reasons of cost-effectiveness, given the low absolute event rates among women. However, the argument made by Dr. Redberg is based on risk-benefit, yet appears to overestimate presumed harms and underestimatie potential benefits.

  8. Primary Prevention – weighing risks and benefit
    Primary prevention means a healthy population that has no symptoms of heart disease. To recommend a lifetime of medication to this group there should be some benefit to them. What benefits can we advise women that they will accrue by taking a statin? It is hard to make asymptomatic people feel better, so the reason to recommend a statin must be to live longer. Yet no clinical trial in the last two decades of statin research, including the recent meta-analysis in Circulation of over 18,000 women, has shown any benefit of statins on mortality in primary prevention. Women are lower risk for heart disease than men. That is not a “subgroup” analysis. The problem with giving drugs with multiple side effects and risks to a low-risk group (like primary prevention in women) is that it cannot appreciably lower their risk further, but they will suffer the same harms from the medication (often more in women) as the high-risk groups. Finally, although there is strong epidemiologic data showing that populations which have lower cholesterol levels due to healthy lifestyle practices – heart healthy low-fat diets and regular physical activity – have lower rates of heart disease, we have never demonstrated that we can reproduce the benefits of a healthy lifestyle by taking a pill – including statins. Dr deLemos should remember that without a benefit, even small harms cannot be justified. I will continue advising women on healthy lifestyles as the best way to reduce their cardiac risk AND live longer.

  9. Primary prevention in women
    The problem with primary prevention in women is our inability to adequately measure risk with conventional risk factors and we therefore are not treating those most needing primary prevenion. MESA shows us that 2/3rds of women at highest risk are missed. On one hand we say that conventional risk factors are all we need (atherosclerosis imaging is a waste), on the other we acknowledge that conventional risk factors don’t accurately measure risk, especially in women.

    The argument that women are at lower risk loses credibility when you look at the statistics and realize that a larger percentage of women die from coronary disease than men, they just do it later in life. In addition, most women who die were asymptomatic until their initial fatal event.

    Yes, we should all live healthier, quit smoking and exercise but this is less of an evidence based argument than my argument to treat subjects based upon the presence of plaque by atherosclerosis imaging. We seem to have forgotten the message from Mr. Fit where the very small marginal benefit on cardiac mortality in the intensive intervention group was exactly compensated for with increased suicide and no net improvement in survival. If we are true to the principles of “evidence based medicine” we need to stop promoting healthy lifestyles.

    Competing Interests: I find asymptomatic significant vascular disease in otherwise low risk, untreated subjects by the use of carotid ultrasound and EBT calcium imaging.

  10. Dr. Redberg is one of my closest friends in the world. She has hosted my mom and me at her home for dinner and we agree on 95+% of things. My mom always says that Rita and Samia Mora are her favorite female Cardiologists. However, I feel that just because statin therapy has not been shown to reduce total mortality in women during 5 year followup, female patients with multiple risk factors should not be denied their proven benefits in reducing revascularizations, heart failure, and probably stroke. We prescribe bisphosphonates to women who have osteoporosis in their 50’s and 60’s but we do not have mortality benefits clearly seen in these types of women. Samia Mora’s superb Circulation article on the latest meta-analysis of statin therapy in women and the accompanying insightful editorial clearly outline the potential benefits of statin therapy in slowing the progression of atherosclerosis and markedly reducing cardiovascular morbidity. To maintain that we need mortality benefits for an intervention to be used in women with multiple risk factors is twisted logic. I haven’t seen anything as fractured as that logic since Lawrence Taylor mauled Joe Theismann’s right leg on Monday Night Football and fractured his tibia and fibula with blood spurting all over the place way back in 1985. I maintain that there is a clear role for judicious use of statin therapy in certain postmenopausal women with multiple risk factors. Happy holidays to my friends in beautiful San Fran and I hope to visit in 2011.

    Competing interests pertaining specifically to this post, comment, or both: