6th April 2009
The AURORA study found rosuvastatin had no cardiovascular mortality or morbidity benefits in patients with end-stage renal disease undergoing haemodialysis.
Patients with chronic kidney disease (CKD) are at increased risk of cardiovascular (CV) disease. It remains uncertain whether the administration of statins to patients or subgroups of patients with CKD prevents CV events. Prescribers should continue to follow the recommendations in the NICE clinical guideline on chronic kidney disease. This recommends that statins should be offered to all people with CKD for the secondary prevention of CV disease, irrespective of baseline lipid levels. However, for the primary prevention of CV disease in people with CKD, the use of statin therapy should not differ from its use in people without CKD. Framingham-based risk tables underestimate CV risk in people with CKD and clinical assessment may be necessary as a basis for decision making.
What is the background to this?
People with CKD are at a greater risk of CV disease than the general population. However, the pattern of cardiovascular disease in such people differs from that in people without CKD; as does their dyslipidaemia, which varies with the stage of CKD the patient is at and the presence of diabetes and/or nephrotic syndrome. Some observational studies have suggested that statin therapy is associated with improved survival among patients with end-stage renal disease or those undergoing haemodialysis (study in end-stage renal disease, study in haemodialysis). However, others have suggested that hypercholesterolemia is actually a protective feature, associated with a greater survival among dialysis patients. Few randomised controlled trials (RCTs) have been conducted with statins in people with CKD.
The NICE clinical guideline on chronic kidney disease recommends that statins should be offered to all people with CKD for the secondary prevention of CV disease, irrespective of baseline lipid levels. However, for the primary prevention of CV disease in people with CKD, the use of statin therapy should not differ from its use in people without CKD, and should be based on existing risk tables for people with and without diabetes. For primary prevention in the general population, the NICE clinical guideline on lipid modification recommends statin therapy for people with a 20% or greater 10-year risk of developing CV disease. This level of risk should be estimated using an appropriate risk calculator, or by clinical assessment where an appropriate risk calculator is not available or appropriate. It should be noted that Framingham risk tables significantly underestimate risk in people with CKD.
What does this study claim?
The AURORA RCT in 2,776 patients with end-stage renal disease undergoing haemodialysis found rosuvastatin had no CV mortality or morbidity benefits. After a median of 3.8 years follow-up, rosuvastatin 10mg daily had no statistically significant effect on the combined primary endpoint of non-fatal myocardial infarction (MI), non-fatal stroke, or death from CV causes, compared with placebo. This endpoint was reached in 396 patients in the rosuvastatin group and 408 patients in the placebo group (9.2 and 9.5 events per 100 patient-years, respectively; hazard ratio [HR] 0.96; 95% confidence interval [CI] 0.84 to 1.11; P=0.59). Rosuvastatin had no effect on individual components of the primary end point or on all-cause mortality.
How does this relate to other studies?
In an earlier RCT, the 4D study, of 1,255 patients with type 2 diabetes mellitus undergoing haemodialysis, atorvastatin 20mg daily also showed no statistically significant effect on the composite outcome of CV death, non-fatal MI or non-fatal stroke. Further RCT evidence of statin therapy in people with CKD will become available when results of the ongoing SHARP study are published. This is a large multicentre RCT looking at the effects of simvastatin plus ezetimibe on outcomes in about 9,000 people with varying degrees of CKD but without established coronary heart disease.
A meta-analysis of data from RCTs or subgroups of RCTs in people with varying stages of CKD (pre-dialysis, dialysis or renal transplant) was published in 2008 looking at the benefits and harms of statin therapy, compared with placebo. However, the methodological quality of many of the included trials was poor. This found no significant reduction in the risk of all-cause mortality with statins in CKD overall, but there was a statistically significant reduction in fatal and non-fatal CV events. A substantial proportion of patients included in this meta-analysis already had established cardiovascular disease; and the authors concluded that their findings support a widespread use of statins for the secondary prevention of CV disease in people with CKD. However, their role in primary prevention remains to be defined.
An accompanying editorial to AURORA suggests that the disappointing results with rosuvastatin in this study add statins to the group of “promising but ineffective” interventions for improving survival and CV outcomes in patients undergoing dialysis. This is despite their beneficial effects on surrogate markers, in this case LDL-cholesterol. The lack of benefit of rosuvastatin in AURORA was observed despite a mean 43% reduction in LDL-cholesterol at three months.
Given the consistent benefits of statins seen in many large RCTs involving other patient groups, both the authors of the AURORA study and the author of the accompanying editorial discuss several limitations with the study which may have affected the results. One of these was the exclusion of patients already receiving statins, which may have excluded patients who were most likely to benefit from statin therapy, i.e. those with previous CV events or other evidence of increased risk. The study may also not have had sufficient statistical power, as event rates in the placebo group were lower than expected.
However, the most likely explanation given for the negative results with rosuvastatin in AURORA and atorvastatin in the 4D study is that CV disease in patients undergoing dialysis differs from that in other patients. Many patients undergoing dialysis who die from CV causes do so from sudden death or death due to arrhythmia, for which statins are not indicated, not from MI due to atheromas, as in the general population. The editorial suggests there is a broader message here, i.e. that the validity of surrogate markers, even well-established ones such as LDL-cholesterol, may vary unexpectedly according to the clinical setting.
Design: Randomised, double-blind, placebo-controlled, multicentre trial.
Patients: 2,776 patients aged 50 to 80 years of age who had end-stage renal disease and had been treated with regular haemodialysis or haemofiltration for at least three months. Patients who had received statin therapy within the previous six months were excluded.
Intervention: rosuvastatin 10mg daily
Outcomes: Primary endpoint was time to major CV event (non-fatal MI, non-fatal stroke, or death from CV causes). Secondary endpoints included all-cause mortality, CV event-free survival, vascular access procedures, coronary or peripheral revascularisation, death from CV causes and death from non-CV causes.
Results: Median length of follow-up was 3.8 years. Rosuvastatin had no significant effect on the primary endpoint, which was reached in 396 patients in the rosuvastatin group and 408 patients in the placebo group (9.2 and 9.5 events per 100 patient-years, respectively; HR 0.96; 95%CI 0.84 to 1.11; P=0.59). Rosuvastatin had no effect on individual components of the primary end point. There was also no significant effect on all-cause mortality (13.5 vs. 14.0 events per 100 patient-years; HR 0.96; 95%CI 0.86 to 1.07; P=0.51). The lack of an effect of rosuvastatin on the primary endpoint was consistent in all prespecified subgroups including patients with diabetes, pre-existing CV disease, hypertension, a high LDL-cholesterol level, or an elevated high-sensitivity C-reactive protein level. After three months, the mean reduction in LDL-cholesterol levels was 43% in patients receiving rosuvastatin, from a mean baseline level of 2.6mmol/L. There was a high incidence of adverse and serious adverse events, with no significant difference between treatment groups.
Sponsorship: AURORA was sponsored by AstraZeneca