14 May 2010
This meta-analysis found that while statins reduced hospitalisation for worsening heart failure, they did not significantly reduce all-cause mortality or cardiovascular mortality in people with heart failure. Limited data indicate that simvastatin or atorvastatin may have advantages over rosuvastatin for patients with heart failure, but these data are far from definitive.
Level of evidence:
Level 2 (limited quality patient-oriented evidence) according to the SORT criteria.
If statins are considered appropriate for a patient with heart failure (HF) (for example, in those with pre-existing cardiovascular (CV) disease), health professionals should refer to NICE guidance on lipid modification. NICE recommends that treatment is initiated with simvastatin 40mg daily. In line with NICE guidance, rosuvastatin should only be used in the minority of patients who cannot tolerate or who do not have a satisfactory response to the statins recommended first line.
What is the background to this?
Treatments such as ACE inhibitors and beta-blockers have been proven to improve the quality of life, reduce hospital admission and delay death in people with HF. CV disease is the commonest cause of HF, and many patients with HF also receive statins. However, their role is controversial as evidence for statins in this group of patients is limited and conflicting.
Observational studies and smaller prospective RCTs have demonstrated that statins are associated with improved outcomes in patients with HF. However, in two large RCTs of patients with HF (CORONA and GISSI-HF), adding rosuvastatin 10mg daily to established therapies did not reduce the risk of the combined primary outcome of death from CV causes, non-fatal myocardial infarction (MI) or non-fatal stroke, compared to placebo, despite a significant lowering of LDL-cholesterol.
The 2003 NICE clinical guideline on chronic heart failure states that “patients with the combination of heart failure and known atherosclerotic vascular disease should receive statins only in accordance with current indications”. This guideline is currently being partially updated and is due for publication in August 2010. However, the evidence on statins in HF is not being considered in this update.
What does this study claim?
The aim of this study was to systematically review RCTs comparing statins to placebo in people with HF, and compare the impact of different statins. The meta-analysis of 10 RCTs (n=10,192) found that statins did not statistically significantly affect all-cause mortality (odds ratio [OR] 0.89, 95% confidence interval [CI] 0.72 to 1.10) or CV mortality (OR 0.89, 95%CI 0.71 to 1.13), compared with placebo. However, statins did statistically significantly decrease hospitalisation for worsening HF (OR 0.67, 95%CI 0.50 to 0.90, P=0.008) as well as increasing the left ventricular ejection fraction (LVEF; 4.2% 95%CI 1.3 to 7.1%, P=0.004), although the latter outcome was based on small numbers of patients. Statins did not significantly increase serious adverse drug reactions (OR 1.04, 95%CI 0.89 to 1.21).
The role of statins in HF remains controversial, but this meta-analysis does provide some reassurance on the safety of statins in HF. This study found that statins were not associated with an increased risk of adverse events. However, updated safety information on statins has recently been published by the MHRA, and other important adverse effects such as myopathy are well recognised. This should be factored into discussions with both new and existing patients on the likely benefits and risks of treatment.
Although no mortality benefits were observed in this meta-analysis, this does not provide sufficient evidence to discourage prescribers from continuing to use evidence-based statins in people with heart failure and co-existing CV disease (such as angina or history of MI). The choice and dose of statin should be based on the underlying type of CV disease in line with the NICE guideline on lipid modification. Overall, statins significantly reduced hospitalisation for worsening HF during follow-up, and significantly increased LVEF, which may have benefits in terms of cost and patient morbidity.
Most patients included in this meta-analysis were from the CORONA and GISSI-HF studies which randomised patients to rosuvastatin 10mg daily — and no improvements in any of the primary efficacy endpoints were observed in these studies. It is therefore not surprising that post hoc analyses showed heterogeneity among different statins and demonstrated that randomisation to atorvastatin significantly decreased all-cause mortality (OR 0.39, P=0.004), and decreased hospitalisation for worsening HF (OR 0.30, P<0.00001). Randomisation to atorvastatin and simvastatin led to a significant improvement in LVEF, whereas these benefits were not observed in patients randomised to rosuvastatin. Meta-regression analysis found that the type of statin was the only factor that had a significant influence on all-cause mortality and hospitalisation for worsening HF. However, this finding may not be reliable for a number of reasons such as the small numbers of patients taking simvastatin or atorvastatin (n=523); the wide variation in the patient inclusion criteria and because the analysis was conducted post hoc.
Despite the limitations of this meta-analysis, this study suggests that rosuvastatin has no advantages and may have some disadvantages compared with some other statins for patients with HF, and for whom statin treatment is appropriate. In line with NICE guidance, rosuvastatin should only be used in the minority of patients who cannot tolerate or who do not have a satisfactory response to the statins recommended first line.
Systematic review and meta-analysis of 10 randomised placebo-controlled trials with follow-up from 3 to 47 months.
10,192 patients with HF (weighted mean age 69.2 years). Inclusion criteria varied between studies. Patients from the two large RCTs (CORONA and GISSI-HF) had HF NYHA class II to IV.
Intervention and comparison
Patients were randomised to statin or placebo with weighted mean follow-up of 38 months. The majority of included patients were from studies of rosuvastatin (3 studies, n=9,680). One study randomised to simvastatin (n=51), the remaining six studies randomised to atorvastatin (n=472). Rosuvastatin dosages were 10–40mg daily, simvastatin dosages were 5–10mg daily, and atorvastatin dosages were 10–40mg daily.
Outcomes and results
|Primary endpoint||n||Statin vs. placebo (95%CI)||P value||Heterogeneitya I2|
|All-cause mortality||10,192||OR 0.89 (0.72 to 1.10)||0.27 NS||50%|
|CV mortality||9,881||OR 0.89 (0.71 to 1.13)||0.35 NS||65%|
|Hospitalisation for worsening HF||10,193||OR 0.67 (0.50 to 0.90)||0.008||64%|
|Serious adverse drug events||9,828||OR 1.04 (0.89 to 1.21)||0.64 NS||0%|
|Changes in LVEF||485||–4.20% (–7.08% to –1.31%)||0.004||94%|
a Heterogeneity is described here. The higher the I2 value, the greater the heterogeneity
NS Not statistically significant
Funding for this meta-analysis is not stated. However, the authors “do not wish to imply support for any particular statin over any other statin”.
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you.