Three large trials of rosuvastatin to prevent cardiovascular events have been completed.1—3 Two of these, CORONA and GISSI-HF,1, 2 assessed 10 mg rosuvastatin daily. Substantial reductions in LDL cholesterol, a small increase in HDL cholesterol, and appreciable reductions in high-sensitivity C-reactive protein (hs-CRP) were reported. CORONA enrolled patients with ischaemic heart disease, whereas GISSI-HF included patients with ischaemic (40%), dilated (35%), and hypertensive (18%) causes of heart failure. Because such patients are probably at risk for future ischaemic vascular events, lowering of LDL cholesterol would be expected to reduce cardiovascular death, myocardial infarction, and stroke. Yet in neither trial was there a clear reduction in ischaemic vascular events or cardiovascular mortality.

Major trials of rosuvastatin in context of other major statin trials

By contrast, in the JUPITER trial,3 17 802 apparently healthy people, with LDL cholesterol less than 3·4 mmol/L and CRP concentrations above 2·0 mg/L, received rosuvastatin 20 mg daily. LDL cholesterol decreased by 50% and CRP by 37%. Over 1·9 years, there were substantial and significant reductions in ischaemic vascular events as well as total mortality, which were larger and more rapid than those in previous trials of rosuvastatin or other statins.4

How can we explain the apparently contradictory results of JUPITER compared with CORONA and GISSI-HF? In all three trials, CRP was raised, and substantial reductions in both LDL cholesterol and CRP occurred. The duration of the first two trials was much longer than that of JUPITER, and because the benefits of lipid lowering are enhanced by longer follow-up, the larger and more rapid benefit in JUPITER compared with previous statin trials is a surprise. For example, a 1 mmol/L lowering in LDL cholesterol reduces the risk of vascular events by 10% after 1 year, 25% after 2—3 years, and about 30% after 4 years.4 The 50% risk reduction in vascular events in JUPITER, after a 1·2 mmol/L drop in LDL cholesterol after 1·9 years, is unexpectedly large and rapid.

Although JUPITER explicitly chose healthy people with raised CRP, most people in CORONA and GISSI-HF were likely to also have raised CRP. Moreover, in all three trials, rosuvastatin lowered CRP; yet in CORONA and GISSI-HF little overall benefit was observed. A subgroup analysis from CORONA showed little benefit of rosuvastatin in those with high concentrations of CRP. But a re-analysis5 suggested a 13% relative risk reduction in those with a CRP concentration of 2·0 mg/L or more, compared with no benefit in those with CRP less than 2·0 mg/L. While subgroup results are intriguing, they suggest a far more modest effect than that observed in JUPITER and, being retrospective, require independent confirmation.

While advanced myocardial disease in individuals in the heart failure trials might explain a lack of effect of rosuvastatin on mortality, specific ischaemic events, such myocardial infarction or stroke, would be expected to be reduced. In CORONA, there were fewer myocardial infarctions and strokes (table), but the benefit was modest and not significant. In GISSI-HF, the reduction of fatal and non-fatal myocardial infarction was also modest and there was no reduction in strokes (table). The benefit of rosuvastatin for specific ischaemic events in the two trials seems to be less than what could be expected on the basis of secondary prevention trials, whereas the large effect in JUPITER was greater than that expected.6

The table compares rosuvastatin trials with a meta-analysis of previous primary and secondary prevention trials. This meta-analysis indicates that for about a 1 mmol/L reduction in LDL over 5 years, there is a 21% risk reduction in vascular events, and a 12% risk reduction in mortality. These estimates are similar in primary6 and secondary prevention trials,4 and are more modest than that in JUPITER. Furthermore, a recent trial of simvastatin and ezetemibe in mild-to-moderate aortic stenosis7 showed that a 50% reduction in LDL over 4·4 years led to a 17% decrease in vascular events and no reduction in mortality. Additionally, there was a reduction in non-cardiovascular deaths of about 20% in JUPITER (chiefly due to fewer cancer deaths) that was as large as the reduction in cardiovascular deaths. This finding is not biologically plausible (2 years is too short to influence cancer) and inconsistent with previous trials. Therefore, chance might have exaggerated the apparent benefits in JUPITER (compounded by the trial's early termination). In view of the results of all other statin trials, the real benefits of lowering LDL by 1·2 mmol/L for about 2 years (as in JUPITER) is unlikely to be larger than a 20—30% reduction in relative risk for ischaemic events, and a 10% reduction for total mortality. Both estimates are within the 95% CI of the estimates on specific events observed in JUPITER.

What are the clinical implications of the recent trials? First, rosuvastatin seems safe in the medium term. Whether the excess in diabetes observed in JUPITER is a chance finding or real is unclear, because it has not been observed in other trials of statins. Indeed one trial with pravastatin reported lower rates of diabetes,8 but it would be prudent to systematically explore this point in all trials. Second, one would expect benefits from rosuvastatin to be proportionate to the degree of lowering of LDL cholesterol and the duration of treatment, so the best estimate on specific ischaemic events might be the weighted average of all trials of rosuvastatin, when viewed in the context of all other previous statin trials.4, 6 Third, because the role of CRP as a predictor of risk remains controversial,9 its role as a marker of preferential benefits from statins should be evaluated in ongoing trials, and in a meta-analysis of all trials that have stored blood samples suitable for CRP measurement. Fourth, because in those without clinical cardiovascular disease the absolute benefits are likely to be modest in the short term (the absolute risk of events in JUPITER was low, despite use of CRP as a marker of risk), much longer-term trials than those currently completed are needed to discover the full benefits (which may increase over time) and safety of life-long use. Fifth, the effect of statins in several ethnic groups, such as Chinese people and south Asians, needs clarification. In view of the potentially large public-health and economic implications of widespread use of statins in apparently healthy individuals with average risk levels, confirmation of the long-term results of major lowering of LDL cholesterol (as can be safely achieved by rosuvastatin) is needed before potent statins are used widely in average-risk healthy people.

Substantial risk reductions in cardiovascular disease are theoretically possible by combined lowering of blood pressure and LDL cholesterol in those with average levels of both risk factors and no apparent vascular disease, but this promising hypothesis needs assessment.10, 11 Whether biomarkers such as CRP or N-terminal probrain natriuretic peptide7, 12 will be better than simple clinical risk factors in identifying individuals who would benefit from preventive strategies remains unclear. At present, tobacco avoidance, maintenance of optimum weight, a prudent diet, and regular exercise should remain the foundations for prevention of cardiovascular disease in apparently healthy individuals with average risk factors.

We thank Paul Ridker, Aldo Maggioni, John McMurray, and Colin Baigent for providing additional unpublished analyses from the JUPITER and the GISSI-HF trials, and from the Cholesterol Treatment Trialists' Collaboration; and Ann Raback for secretarial assistance. We are all investigators in the ongoing HOPE-3 trial that is evaluating combined lowering of blood pressure and lipids (with rosuvastatin) in primary prevention. This investigator-initiated trial is funded by the Canadian Institutes of Health Research through a peer-reviewed grant, and by AstraZeneca who manufacture rosuvastatin. SY has received research grants and honoraria from several companies that manufacture statins (AstraZeneca, Bristol Myers Squibb, and Merck), EL has received research grants and honoraria from several manufacturers of statins (AstraZeneca, Merck, Pfizer), and JB has received grants from AstraZeneca.

References

1 Kjekshus J, Apatrei E, Barrios V, et alfor the CORONA group. Rosuvastatin in older patients with systolic heart failure. N Engl J Med 2007; 357: 2248-2261.
2 GISSI-HF Investigators. Effect of rosuvastatin in patients with chronic heart failure (the GISSI-HF trial): a randomised, double-blind, placebo-controlled trial. Lancet 2008; 372: 1231-1239.
3 Ridker PM, Danielson E, Fonseca FAH, et alfor the JUPITER Study Group. Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein. N Engl J Med 2008; 359: 2195-2207.
4 Cholesterol Treatment Trialists' (CTT) Collaboration. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90 056 participants in 14 randomised trials of statins. Lancet 2005; 366: 1267-1278.
5 McMurray J, Kjekshus J, Hjialmarson A, Wedel H, Dunselman P, Wongstein F. LDL independent reduction in cardiovascular morbidity and mortality with rosuvastatin in heart failure patients with a raised C-reactive protein: a retrospective analysis of the Controlled Rosuvastatin Multinational Trial in heart failure (CORONA). Circulation 2008; 118 (suppl 2): S711.
6 Mills EJ, Rachlis B, Wu P, Devereaux PJ, Arora P, Perri D. Primary prevention of cardiovascular mortality and events with statin treatments: a network meta-analysis involving more than 65,000 patients. J Am Coll Cardiol 2008; 52: 1769-1781.
7 Rossebø AB, Pedersen TR, Boman K, et alfor the SEAS Investigators. Intensive lipid lowering with simvastatin and ezetimibe in aortic stenosis. N Engl J Med 2008; 359: 1343-1356.
8 Freeman DJ, Norrie J, Sattar N, et al. Pravastatin and the development of diabetes mellitus: evidence for a protective treatment effect in the West of Scotland Coronary Prevention Study. Circulation 2001; 103: 357-362.
9 Lloyd-Jones DM, Liu K, Tian L, Greenland P. Narrative review: assessment of C-reactive protein in risk prediction for cardiovascular disease. Ann Intern Med 2006; 145: 35-42.
10 Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80%. BMJ 2003; 326: 1419.
11 Xavier D, Pais P, Sigamani A, Pogue J, Afzal R, Yusuf Son behalf of The Indian Polycap Study (TIPS) Investigators. The need to test the theories behind the Polypill: rationale behind the Indian Polycap Study. Nature Clin Pract Cardiovasc Med 2009; 6: 96-97.
12 McKie PM, Rodeheffer RJ, Cataliotti A, et al. Amino-terminal pro-B type natriuretic peptide and B type natriuretic peptide: biomarkers for mortality in a large community-based cohort free of heart failure. Hypertension 2006; 47: 874-880.