At the American College of Cardiology conference, another big trial showed that giving statins to people with some risk factors, but no evidence of cardiovascular disease, reduced major cardiovascular events including premature death.
We contacted both experts in cardiovascular prevention and primary care experts via email to ask:
Will you start prescribing statins more broadly than before, on the basis of these results?
If not, what additional evidence would persuade you to start doing so?
In general, do big trials like this reflect real-world patients well enough to significantly inform clinical practice? Or are the limitations of such trials -- such as the careful selection of patients and the often increased medical attention they receive from participation -- so great that the results can't be trusted.
The participants this week are:
Paul D. Thompson, MD, FACC, medical director, cardiology and The Athlete's Heart Program at Hartford Hospital in Connecticut
Marc Itskowitz, MD, associate professor of medicine at Philadelphia's Temple University School of Medicine and director of the Center for Perioperative Medicine for Allegheny Health Network in Pittsburgh
Joshua Septimus, MD, FACP, associate professor, clinical medicine at Houston Methodist Primary Care Group in Texas
Ronald Adler, MD, FAAFP, associate professor, family medicine and community health at University of Massachusetts Medical School in Worcester
Scott Grundy, MD, PhD, professor, internal medicine at UT Southwestern Medical Center in Dallas
Lee Green, MD, MPH, professor and chair, medicine and dentistry, Department of Family Medicine at the University of Alberta in Canada
Shilpi Agarwal, MD, a primary care physician at One Medical Group in Washington D.C.
No Surprise
Thompson: Statins reduce events in just about everyone. You find it faster and with fewer subjects in extremely high-risk folks. You find it slower and require more subjects in the intermediate risk folks. You will almost certainly find it as well in the low-risk groups, it just takes a lot longer. This study is critically important because it did show more muscle complaints in the statin users. The stain trialists have argued for years that statin muscle complaints don't exist and are created by the docs asking patients about it. Indeed, no prior trial has found increased muscle complaints probably because they did not ask. This one did find it probably because they asked!
Itskowitz: The results of the HOPE-3 trial confirm the benefits of statin therapy in primary prevention. I will continue to prescribe statins in primary prevention in patients who have risk factors for vascular disease. However, appropriate patient selection and a detailed discussion of the potential adverse effects remain critical strategies in successfully applying trial results to real world clinical practice.
Septimus: These results confirm my current practice and reinforce the American Heart Association guidelines: Treat at-risk patients with statin therapy based on risk assessment, not simply based on cholesterol readings.
More Research Needed
Adler: Widespread application of this approach would help some people avoid significant CV events, but these NNTs also mean that thousands of people would derive no benefit from taking daily medicine for years. Meanwhile, in the real world, many well-intentioned clinicians would likely deviate from the study protocol, checking lipids and transaminases and escalating doses. Should we therefore recommend this to all our patients with intermediate risk? No, but we can offer it, and patients can choose, but these decisions should be made after full discussion of alternative methods to reduce cardiovascular risk, especially smoking cessation, diet, and exercise, all of which produce numerous other benefits -- without over-medicalizing a risk factor and potentially undermining lifestyle improvement efforts.
Grundy: Current ACC/AHA guidelines recommend wide use of statins in the population. It is uncertain whether the HOPE-3 study would change these recommendations. There is not enough evidence from this trial to answer this question.
Green: I'll start offering them more broadly, but I don't know if that'll mean prescribing them more broadly. That I think is the key distinction, and one that is often lost in the discussion over findings like this. Statistically significant result does not equal compelling indication! With 5-year NNTs in the 75-100 range in the trial (almost certainly better than what would be achieved in community practice), statins for moderate risk patients are a reasonable option -- but a choice for patients to make. It would be quite inappropriate for me as a family physician to present this to my patients as a strong recommendation.
Agarwal: With any long-term medication, risks and benefits to the patient need to be considered on an individual basis. As a matter of practice, I tend to focus more on lifestyle changes before utilizing medications, especially because most -- including statins -- have some side effects. In certain populations, statins work well for patients, but I would like to see positive results in a longer-term setting before making statins part of general practice.
Limitations
Grundy: A major limitation of randomized controlled trials is that they do not represent the general population. Physicians should not be guided by individual trials to make clinical decisions, but should follow expert guidelines that take into account the totality of the available data.
Thompson: I think cholesterol treatment could wipe out atherosclerosis. Low dose, nearly life long treatment to slightly reduce everyone's cholesterol, with high doses for the very risky, would probably be a great thing. This is against the AHA guidelines, but they depend on "evidence based" studies and most don't last long enough ... 5 years max in most cases.
Agarwal: I wouldn't say the results can't be "trusted," but it is important to look at broader groups for longer periods of time and also take a closer look at all endpoints before making general statements suggesting a wider array of patients benefit from statins for primary prevention.
Qualified Trust in Results
Green: I think the results can be trusted, if taken with a suitably-sized grain of salt. First, these are self-selected volunteers. Second, the use of a run-in period means this study will continue the long-standing problem -- familiar to all family physicians -- of RCTs underestimating statin side effect rates in real-world practice. So realistically, the NNT will be higher in practice, and patient discontinuation rates will be higher too. The benefit is real though, it's a reasonable option for interested patients; we just have to resist the temptation to oversell it.
Septimus: Clinical trials are not the real world, but when done in a high-quality manner they can be applied in a reasonable way to the appropriate patient population. They are one of several critical tools to both learn about new therapies and about how to practice medicine in a cutting edge manner. The other important aspect of this discussion is that the HOPE-3 trial simply confirms what dozens of other trials have shown about statin therapy. One important thing about interpreting clinical trials is their consistency with previously published data.