JUPITER: How will it impact future statin guidelines?

January 21, 2009
Catherine Y. Campbell, MD

Roger S. Blumenthal, MD

From the Johns Hopkins Ciccarone Preventive Cardiology Center, Johns Hopkins University School of Medicine, Baltimore, MD and

Cardiology Review® Online, January 2009, Volume 26, Issue 1

One of the most talked about clinical trials in the field of preventive cardiology, the JUPITER (Justification for the Use of Statins in Prevention: An International Trial Evaluating Rosuvastatin) trial, was presented at the 2008 American Heart Association (AHA) Scientific Sessions and published in the New England Journal of Medicine.1 The investigators enrolled apparently healthy men, who were at least age 50, and women, who were at least age 60, with a high-sensitivity C-reactive protein (hs-CRP) level of 2.0 mg/dL or above. Individuals who had a history of previous cardiovascular disease or diabetes were excluded from the study. Nearly 30% of the participants were ethnic minorities, and the trial included more women (6801 of 17,802 subjects) than any other lipid-lowering trials to date. Study participants were enrolled from 1315 sites throughout 26 countries and randomized to receive 20 mg of rosuvastatin (Crestor) or placebo.

Low-density lipoprotein (LDL) cholesterol was lower in JUPITER than in other primary prevention trials. The mean LDL cholesterol level was 104 mg/dL and the median was 108 mg/dL. In other trials, the mean LDL cholesterol ranged from about 130 to 192 mg/dL.

The primary end point was a combination of myocardial infarction (MI), stroke, unstable angina requiring hospitalization, coronary artery revascularization, and cardiovascular death. Surprisingly, the study was stopped after a median follow-up of 1.9 years because of a striking 44% relative risk reduction in the primary end point (hazard ratio for rosuvastatin as compared with placebo was 0.56; 95% confidence interval [CI], 0.46-0.69). Women had the same relative risk as men, and those with a family history of premature coronary heart disease (CHD) had an even greater relative risk reduction of about 65%.

The event rate was at least twice as high as would have been predicted by the Framingham risk estimate, according to the investigators. This supports the idea that an elevated hs-CRP helped to identify high-risk individuals; however, JUPITER did not just look at the standard hard event rate of nonfatal MI/CHD death for their primary end point or MI/CHD death/unstable angina, as were looked at in AFCAPS/TexCAPS (Air Force/Texas Coronary Atherosclerosis Prevention Study), a prior large primary prevention study.2 To increase their statistical power to show a possible benefit with statin therapy, JUPITER also included stroke and revascularization end points.

The absolute event rate observed in JUPITER was 1.36 per 100 person-years versus 0.77 per 100 person-years, which could be extrapolated to indicate a cardiovascular event rate of 14% versus 7.5% at 10 years. All of the components of the primary end point were significantly reduced. Of note, all-cause mortality was reduced by 20% with rosuvastatin, with a decrease from 1.25 deaths per 100 person-years to 1.0 deaths per 100 person-years (12.5% vs 10% all-cause mortality with rosuvastatin over a decade).

No significant decline in total mortality has been observed in prior individual primary prevention trials. A recent meta-analysis of 19 primary prevention trials found a 7% decrease in total mortality with statins as compared with placebo (relative risk [RR], 0.89; 95% CI, 0.81-0.98; P = .02).3 This is in contrast to the 20% decrease in total mortality observed in JUPITER. The large decrease in total mortality with rosuvastatin may be because rosuvastatin lowered LDL cholesterol by about 50%, which was a much greater degree of lipid lowering than that obtained with pravastatin (Pravachol), lovastatin (Mevacor, Altocor), and low-dose atorvastatin (Lipitor) in prior primary prevention studies.

Implications of JUPITER

One challenge posed by JUPITER is whether an hs-CRP should be measured in all adults who are at least 50 years of age and in whom the decision to treat with lipid-lowering therapy is otherwise uncertain based on existing national guidelines. We know that there are ethnic and gender variations in hs-CRP levels. Women have higher median hs-CRP levels than men, and African Americans and persons of Hispanic descent tend to have higher hs-CRP levels than whites.4 Heavier adults also tend to have much higher levels of hs-CRP.

In the MESA (Multi-Ethnic Study of Atherosclerosis) study, 30% of participants classified as intermediate risk by the Framingham risk score (10%-20% risk of a hard CHD event [MI/CHD death]) had an hs-CRP above 3 mg/L, and 33% had moderate coronary calcium (an Agatston score >100). Only 9% of those at intermediate risk, however, had an elevated hs-CRP above 3 mg/L and a high coronary calcium score.5

In prior studies, the prognostic significance of moderate coronary calcium was much greater than that of an elevated hs-CRP. Yet from a clinical perspective, in many parts of the country, it is easier for a clinician to order an hs-CRP blood test than a coronary calcium scan. The finding of a high coronary calcium score has a more robust clinical database than an elevated hs-CRP in terms of increased cardiac event rates. Several studies, including MESA, have shown that a coronary calcium score above 100 is associated with a more than 8-fold increased risk of a future CHD event as compared with that of a person of comparable age and risk factor status, but with no coronary calcium. In contrast, the odds ratios for a high hs-CRP in most studies are in the range of about 1.5.

A low level of radiation is required with coronary calcium scans, and about 5% of patients have incidental findings that require further evaluation. For example, a noncalcified lung nodule found on a coronary calcium scan may require 1 or 2 additional chest computed tomography scans to document stability of the nodule. On the other hand, if a person has an elevated hs-CRP, it should be rechecked a few months later to see if it has returned to the normal range.

JUPITER subjects may be a higher risk group because of their high prevalence (41%) of the metabolic syndrome. It is likely that more than half of the subjects had at least 2 of the 5 components of the metabolic syndrome. Surprisingly, the relative risk reduction with rosuvastatin was a little less in those with the metabolic syndrome as compared to those without it.

Many clinicians have wondered if we would have seen similar benefits with aggressive lipid lowering if subjects with normal hs-CRP levels were included. The rationale for selecting persons with elevated hs-CRP was that in a post hoc analysis of the AFCAPS/TexCAPS trial, patients with normal LDL cholesterol levels and a low hs-CRP showed no trend for benefit with the less potent lovastatin therapy.6

JUPITER is most similar to the ASCOT (Anglo-Scandinavian Cardiac Outcomes Trial) study of treated hypertensive individuals with multiple risk factors and randomized to atorvastatin 10 mg or placebo. No data about the usefulness of atorvastatin in individuals with elevated hs-CRP versus those with lower hs-CRP have been published to date; however, at the AHA meeting, data from the CORONA (Controlled Rosuvastatin Multinational Study in Heart Failure) trial were presented. This study randomized persons with ischemic cardiomyopathy to rosuvastatin 10 mg versus placebo. The investigators showed that rosuvastatin did have a significant clinical benefit in those with an hs-CRP of 2 mg/L or more, but not in those with lower hs-CRP.

JUPITER showed that potent statin therapy with rosuvastatin can lower cardiovascular events by nearly 50% in persons with normal or near-normal LDL cholesterol levels and an elevated hs-CRP. The number of subjects who would need to be treated with rosuvastatin for 2 years to prevent the occurrence of 1 primary end point was estimated at 95, and the number needed to treat for 5 years was estimated to be 25.

It will be interesting to see how the new lipid-lowering guidelines that are scheduled to be made public in about a year will deal with the results of this trial. For the time being, many clinicians will use hs-CRP levels or coronary calcium scores as a tiebreaker to determine eligibility for aspirin or statin therapy in persons meeting the JUPITER age criteria. Others will be tempted to recommend a statin for any adult in this age group with at least 1 other risk factor. The investigators excluded adults with chronic inflammatory conditions, such as severe arthritis, lupus, and inflammatory bowel disease. On the basis of JUPITER, many clinicians may also be inclined to add a statin if the individual meets the entry criteria of the trial.

An elevated hs-CRP is most common in persons with increased visceral fat or insulin resistance. Before putting an adult with an elevated hs-CRP on life-long statin therapy, it would seem prudent to try to motivate them to significantly improve their dietary and exercise habits and then rechecking this blood test in about 6 months.7 In JUPITER, the relative risk reduction was the same in those with a body mass index below 25.


JUPITER shows that potent lipid lowering with statin therapy in persons with a mean age above 60 years and an elevated hs-CRP reduces cardiovascular events by nearly 50% in persons with elevated LDL cholesterol levels. It is likely that clinicians will now feel more comfortable recommending statin therapy for adults with LDL cholesterol levels below 130 mg/dL, but who have other risk factors. Many clinicians will start with a generic statin and then re-evaluate the lipid numbers before using a high dose of rosuvastatin or atorvastatin. Measurements of hs- CRP as a tiebreaker will undoubtedly increase in frequency if adults with at least 1 risk factor for cardiovascular disease do not otherwise qualify for aspirin or lipid-lowering therapy.