Dr Paul Ridker

The study, presented by Dr Paul Ridker (Brigham and Women's Hospital, Boston, MA) and published online in the New England Journal of Medicine, showed that treating apparently healthy individuals with rosuvastatin (Crestor, AstraZeneca) 20 mg significantly reduced the primary end point—a composite of nonfatal MI, nonfatal stroke, hospitalization for unstable angina, revascularization, and confirmed death from cardiovascular causes—by 44% compared with individuals treated with placebo.

"We can no longer assume that patients with low cholesterol are at low risk," said Ridker. "This does not mean that cholesterol isn't important. We want our high-cholesterol patients to be treated very aggressively. But in this study, we found that patients with low cholesterol but high [C-reactive protein] CRP benefited greatly from statin therapy."

The findings, which dominated news the first day of the AHA meeting and were reported widely by the New York Times, the Associated Press, the Washington Post, CNN, and Time, among others, point to some particularly difficult choices for cardiologists as they assess whether or not to start healthy patients with low LDL-cholesterol levels on statin therapy.

We can no longer assume that patients with low cholesterol are at low risk.

"We need to start putting these data into the big picture," Dr Thomas Pearson (University of Rochester School of Medicine, NY) told heartwire. "We need to start converting this from an efficacy trial, which it was, and extrapolate it into the wider population. When we do that, we're going to start to come up with some very different perspectives."

Pearson said the evidence is in—patients at very low risk can benefit significantly from statin therapy—but cost-effectiveness remains the big issue. A quick back-of-the-envelope calculation by Pearson, based on treating 25 patients for five years to prevent one primary end point, showed that rosuvastatin, at $3.65 per pill, would cost more than $166 000 to prevent one MI, stroke, unstable angina, revascularization, or death from cardiovascular causes.

"None of this means that Crestor is a bad drug," said Pearson. "Even among patients with elevated CRP, it's still a very low-risk group, so cost-effectiveness becomes the issue." If, as Ridker and colleagues suggest, treating JUPITER-like patients could prevent 250 000 cardiovascular events in US per year, "you're starting to talk about some real money," said Pearson.

JUPITER is a large, multinational, long-term, double-blind, placebo-controlled, randomized clinical trial that included 17 802 healthy men and women with normal LDL cholesterol but elevated CRP (>2.0 mg/L) assigned to rosuvastatin 20 mg or placebo.

Stopped after 1.9 years of follow-up, treatment with rosuvastatin significantly reduced the primary composite end point 44% compared with placebo. This reduction was observed among nearly all of the individual end points, including a 55% reduction in nonfatal MI, a 48% reduction in the risk of nonfatal stroke, and a 47% reduction in the risk of MI, stroke, and death from cardiovascular causes.

In terms of absolute benefits, the proportion of patients who had an MI, stroke, revascularization, hospitalization for unstable angina, or died from cardiovascular causes was 1.6% in the rosuvastatin arm and 2.8% in the placebo arm, an absolute risk reduction of 1.2%. Similarly, the proportion of patients with hard cardiac events—cardiovascular death, MI, and stroke—was reduced from 1.8% in the placebo arm to 0.9% in the rosuvastatin arm, an absolute reduction of 0.9%.

"We have so many patients who continue to be nervous about taking these therapies, and we want to give out a message to continue with exercise, diet, and smoking cessation, but now we have very overwhelming evidence that this class of drug, this method of lowering surrogate end points, reduces hard clinical end points," said Ridker.

During the press conference, Ridker stressed that all subgroups, including large numbers of Hispanic and African-American patients, benefited from treatment with rosuvastatin. The study also showed that among the 6801 women included in JUPITER, rosuvastatin significantly reduced the primary composite end point 46%. Among patients with no other cardiovascular risk factors aside from elevated CRP levels, there was also a significant benefit with rosuvastatin therapy.

Ridker said the study is unable to discern how much the CRP or LDL lowering each contributed to the reduction in cardiovascular events. He pointed to the SEARCH trial, however, another statin study presented today, showing slightly larger reductions in LDL cholesterol than those in JUPITER but less cardiovascular risk reductions.

"The reduction [in risk in JUPITER] is larger than anticipated based on prior studies in this field," said Ridker. "That being said, these are very powerful drugs, rosuvastatin in particular, for LDL-cholesterol reductions and a very powerful drug for CRP reductions. We do have preliminary evidence. . . . getting LDL down is important, but getting CRP down on top of lowering LDL cholesterol appears to add further incremental benefit. That might suggest, and is certainly what I think, that both LDL and CRP are important."

Dr Subodh Verma (University of Toronto, ON), who was not affiliated with the study but who has performed research in the intersection of inflammation and heart disease, called the JUPITER trial an important study, one that should change the landscape of how physicians triage intermediate-risk patients.

We do a terrible job stratifying at-risk patients based on the Framingham risk score and need better ways to identify at-risk individuals.

"The results really are a giant step forward in the field of primary prevention," he said. "We do a terrible job stratifying at-risk patients based on the Framingham risk score and need better ways to identify at-risk individuals. . . . The guidelines will now need to reflect these tremendous results by recommending CRP testing to intermediate-risk patients so that we can better identify candidates for risk prevention."

Pearson, who is a cochair of the writing committee responsible for the National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP III) guidelines, told heartwire there have been recurrent requests to update the guidelines, but there had been no new advances in the field to justify changes.

"We had said that we were waiting for JUPITER and perhaps some other studies to reconsider that; that this would probably be the next big bit of information," said Pearson. "JUPITER, however, doesn't really say a lot about CRP. It is a randomized, controlled trial of two groups of very low-risk individuals treated with statin therapy. The efficacy benefit doesn't really speak to CRP."

Dr James Stein (University of Wisconsin Medical School, Madison) told heartwire that the use of more widespread screening of CRP values needs more investigation. The test has high variability and is elevated with infections and injuries, so abnormally high CRP levels do not always reflect arterial injury or cardiovascular disease risk.

"However, I suspect we should start using it more often," he said. "For starters, it makes communication easier. It summarizes a myriad of metabolic problems that individually are not bad enough to treat but collectively indicate increased risk. Increased CRP tells you that.  It is one number, and explaining and implementing it may be easier for physicians and patients than trying to explain why 'borderline' high blood pressure and being 'a little overweight' are bad. They damage and inflame blood vessels, and high CRP shows it."

Commenting on the study, Dr Andrew Tonkin (Monash University, Victoria, Australia) said the benefit of treatment to individuals as well as any changes to public-health policy depends on the absolute benefit and not the relative risk reduction.

"The reason we treat all patients who have coronary heart disease, who have had a stroke, or who have diabetes is because they have the highest risk of having future events," said Tonkin. "I think what we now need to see is the absolute risk reduction that occurred in the various subgroups in JUPITER."

Tonkin, as did others, pointed out that these patients were overweight, with the median body-mass index (BMI) of 28.3 kg/m², so it is unfair to say that these patients had no other risk factors. Moreover, very few individuals have ideal risk factors, including LDL-cholesterol levels. Many have suggested ideal LDL-cholesterol levels, based on hunter-gatherer societies and nonhuman primates, are likely in the range of 50 to 60 mg/dL.

To heartwire, Stein pointed out that despite being classified as healthy, many JUPITER patients had a number of cardiovascular risk factors.

"Many patients with heart attacks have normal LDL-cholesterol values," said Stein. "Thus, doctors and patients are lulled into a false sense of security when their LDL cholesterol is 'normal' and then are surprised when they have a heart attack. JUPITER illustrates this point perfectly. If you look at the median values—age 66 years, BMI 28.3 kg/m², systolic blood pressure 134 mm Hg, and 41% with metabolic syndrome—you know these people are going to have heart attacks and strokes and die."

Dr Colin Baigent (Oxford University, UK) told heartwire the JUPITER findings are broadly consistent with the evidence from all the statin trials. The larger the reduction in LDL cholesterol, the larger the reduction in major cardiovascular events, but the magnitude of benefit is going to be related to absolute risk.

"However you identify the absolute risk—by measuring CRP levels or however one does it—the trials are absolutely consistent in that regard," said Baigent. "If you are at high absolute risk, then your benefit is very clear. The lower the risk, the lower the absolute benefit. It's a massive public-health-policy task to try to decide when the absolute benefit justifies the cost."

Regarding cost, the elephant in the room with the JUPITER study, Stein and Dr Jon Keevil (University of Wisconsin, Madison) performed an analysis based on National Health and Nutrition Examination Survey data from 1999-2002.  These data include 171 million adult Americans between 20 and 79 years of age, and the two used these data and JUPITER results to evaluate its potential financial impact.

It's a massive public-health-policy task to try to decide when the absolute benefit justifies the cost.

There are 7.4 million adult Americans who meet the major JUPITER entry criteria. In other words, roughly 4.3% of all adult Americans would have qualified for JUPITER, said Stein. Prescribing rosuvastatin 20 mg daily vs not treating these patients would prevent 43 526 cardiovascular disease events per year, or 29 509 MIs, strokes, or deaths and 18 443 deaths.

According to their math, if statin therapy costs $1200 per year, treating this entire subpopulation would cost $8.9 billion per year and prevent a cardiovascular-disease event at a cost of $203 000 per event per year.  It would cost $480 000 to save a life.  If a generic statin is used and statin therapy costs $60 per year, the cost is $443 million per year. The cost of preventing a cardiovascular disease event is $10 200 per event per year and $24 000 to save one life.

The JUPITER trial was designed as a four-year study but was stopped by AstraZeneca after just 1.9 years based on recommendations from an independent data monitoring board and the JUPITER steering committee. Despite the benefits, Dr Sanjay Kaul (Cedars Sinai Medical Center, Los Angeles, CA) questioned why it was stopped so early, "especially when we have no idea about the long-term safety of very low LDL levels as achieved in this trial." Baigent, a member of the Cholesterol Trialists Collaboration (CTT), also took issue with trial being stopped after 1.9 years.

Will it change practice? . . . In some cases [doctors] might be using CRP appropriately, and in some cases they might not.

"This does have the effect, which is difficult to quantify, of inflating the estimated effects of treatment," said Baigent. "It's quite clear that rosuvastatin is quite effective, as are all statins when used in appropriate patients, in reducing major cardiovascular events. The size of the effect, however, is unknown because the trial was stopped early."

In terms of side effects, significantly more patients in the rosuvastatin arm developed new diabetes, and they also had significantly higher glycated hemoglobin levels, report investigators. Any reported serious adverse events were similar between the placebo and statin-therapy arms. There was also no risk of excess cancer among patients treated with rosuvastatin, a potential worry because of the low levels of LDL cholesterol achieved. However, follow-up was too short to assess any real risk, although most experts are unconcerned.

"When you look at the randomized evidence from the CTT, there is absolutely no evidence of cancer risk," said Baigent. "In fact, the data presented, quite likely by a play of chance, even went the other way. We know from the CTT, right down to the lowest levels studied, we don't see any evidence of cancer risk."

Gordon Tomaselli (Johns Hopkins University School of Medicine, Baltimore, MD) told heartwire that the study should reinforce the role of statins for reducing LDL-cholesterol levels. "I also think the findings are going to influence guidelines and clinical practice. Will it change practice? I think there are some things we don't have control over. Once these data get into the press, there are going to be some doctors who begin to measure CRP, and it's like anything else. In some cases they might be using CRP appropriately, and in some cases they might not."

Ironically, these new data, noted Pearson, might have provided some support for over-the-counter statin proponents, as the study showed that the drugs work in low-risk patients; if patients chose to buy them, the costs would not be borne by society.