Two aldosterone inhibitors are currently licensed for heart failure (HF) in the UK: spironolactone and eplerenone. Recent clinical guidelines recommend eplerenone after an acute myocardial infarction (MI) for patients with symptoms and/or signs of HF and left ventricular dysfunction.
The primary objective was to evaluate relative clinical effectiveness and cost-effectiveness of spironolactone and eplerenone in patients with postMI HF and explore the possibility of conducting an indirect comparison of spironolactone and eplerenone. A second objective was to undertake value-of-information (VOI) analyses to determine the need for further research to identify research questions critical to decision-making and to help inform the design of future studies.
Relevant databases including MEDLINE, EMBASE and CENTRAL were searched between September and December 2008. Randomised controlled trials (RCTs) of spironolactone, eplerenone, canrenone or potassium canrenoate were included if conducted in a postMI HF population. Trials of general HF patients with a subgroup of postMI HF patients were considered if they had at least 100 ischaemic participants per arm and the authors provided subgroup data when contacted. Adverse events summary data were sought from recognised reference sources and RCTs or observational studies in any population that recruited more than 100 participants.
The comparative clinical effectiveness and cost-effectiveness of spironolactone and eplerenone was derived using Bayesian meta-regression drawing on a wider 'network' of aldosterone trials to those considered in the main clinical effectiveness review. An alternative scenario was also considered assuming a 'class effect' for the aldosterone antagonists in terms of major clinical events, but allowing for potential differences in side effect profiles. Cost-effectiveness was assessed using incremental cost-effectiveness ratios (ICERs) where appropriate. Uncertainty in cost-effectiveness results was also presented and used to inform future research priorities using VOI analyses based on expected value of perfect information (EVPI). A probabilistic decision analytic model was developed to estimate cost-effectiveness of spironolactone, eplerenone and standard care for management of postMI HF, provide estimates relevant to the NHS and explore alternative approaches to an indirect comparison between spironolactone and eplerenone. The model incorporated a lifetime horizon to estimate outcomes in terms of quality-adjusted life-years (QALYs) and costs from the NHS persepctive. In the base-case analysis, 2-year treatment duration was assumed, consistent with the follow-up in the main RCTs. Other scenarios were explored to examine the robustness of alternative assumptions including impact of different treatment durations.
Searches yielded five RCTs: two spironolactone trials of poor methodological quality and three trials of which only one (of eplerenone) specifically examined postMI HF (Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study, EPHESUS). One trial of spironolactone (Randomised Aldactone Evaluation Study, RALES) and one of canrenone (Antiremodelling Effect of Aldosterone receptors blockade with canrenone In mild Chronic Heart Failure, AREA IN-CHF) comprised general HF, but data were available for an ischaemic subgroup. Structural similarity of spironolactone and eplerenone suggests that they may be interchangeable, but formal indirect comparison between the three trials was severely limited by trial differences. Relative safety data were limited from RCTs and observational sources. Hyperkalaemia rates varied, but were generally higher than for placebo; data were insufficient to assess discontinuation because of hyperkalaemia.Gynaecomastia rates were higher with spironolactone. Adverse event data were sparse. Systematic review of economic evidence identified three main published studies but none used a UK perspective or attempted to compare cost-effectiveness in postMI HF. The new decision model indicated that eplerenone was the most cost-effective strategy for postMI HF (ICER of eplerenone compared with standard care was 4457 pounds per QALY, increasing to 7893 pounds per QALY if treatment continued over the patient's lifetime); in neither scenario did spironolactone appear cost-effective. The ICER of eplerenone was consistently under the 20,000-30,000 pounds per QALY threshold used to establish value for money in the NHS. Uncertainty resulted in EVPI estimates between 820M pounds (base-case) and 1265M pounds (lifetime treatment duration scenario). When class effect for mortality and hospitalisations was assumed spironolactone emerged as the most cost-effective treatment and EVPI estimates were negligible. If class effect is considered more plausible than the results of the evidence synthesis model then there would be limited value in additional research.
Exchangeability between trials was poor and there was a lack of robust data in RCTs.
Only two good-quality trials of aldosterone inhibitors in the postMI HF population were found, but lack of exchangeability with respect to study populations, meant that a comparison between these drugs could not be done. It consistently emerged that, compared with usual care, use of an aldosterone antagonist appears to be a highly cost-effective strategy for the management of postMI HF patients in the NHS. An adequately powered, well-conducted RCT that directly compares spironolactone and eplerenone is required to provide more robust evidence on the optimal management of postMI HF patients.