Health Technology Assessment (Jul 2009)
Early high-dose lipid-lowering therapy to avoid cardiac events: a systematic review and economic evaluation
Abstract
Objective: To evaluate the cost-effectiveness of high-dose statins (atorvastatin 80 mg/day, rosuvastatin 40 mg/day and simvastatin 80 mg/day) versus simvastatin 40 mg/day in individuals with acute coronary syndrome (ACS). Data sources: Eleven bibliographic databases, including MEDLINE, CINAHL, EMBASE, Cochrane Database of Systematic Reviews, CENTRAL, DARE and NHS EED, were searched from inception to 2008. Review methods: Data relating to study design, baseline patient characteristics, clinical or surrogate outcome, and adverse events were abstracted, and methodological quality was assessed according to standard methods. A synthesis of the available evidence was performed using a Bayesian mixed treatment meta-analysis using both direct and indirect evidence. An existing Markov model was modified to explore the costs and benefits associated with a lifetime of the differing treatment regimens. Results: A total of 3345 titles and abstracts were screened for inclusion in the review of clinical effectiveness and 125 full papers retrieved and assessed in detail. Of these, 30 papers met the inclusion criteria for the review, describing 28 trials. The Bayesian mixed treatment meta-analysis demonstrated a clear dose–response relationship in terms of reductions in low-density lipoprotein cholesterol (LDL-c), with rosuvastatin 40 mg/day achieving the greatest percentage reduction (56%) from baseline, followed by atorvastatin 80 mg/day (52%), simvastatin 80 mg/day (45%) and simvastatin 40 mg/day (37%). Although serious adverse events with statins are rare, their incidence is likely to be greater with higher doses. Several clinical scenarios were used to explore the effect of adherence on the cost-effectiveness of the treatment regimens. Using a threshold of £20,000 per quality-adjusted life-year (QALY) and assuming that the benefits and adherence rates observed in the clinical trials are generalisable to a clinical setting and that individuals who do not tolerate the higher-dose statins are prescribed simvastatin 40 mg/day, then simvastatin 80 mg/day, atorvastatin 80 mg/day and rosuvastatin 40 mg/day would be considered cost-effective compared with simvastatin 40 mg/day in individuals with ACS. Simvastatin 80 mg/day is not well tolerated because of the high incidence rates of less severe adverse events such as myopathy (26-fold higher than rates in those receiving simvastatin 20 mg/day), which are likely to affect adherence levels in clinical practice. The reference case shows that rosuvastatin is the optimal treatment for individuals with a recent history of ACS using a threshold of £20,000 per QALY. However, this is based on the assumption that the additional incremental reductions in LDL-c observed in patients treated with rosuvastatin 40 mg/day compared with atorvastatin will transfer into corresponding changes in relative risks of cardiovascular events. Conclusions: Simvastatin 80 mg/day cannot be recommended because of the high incidence rates of adverse events. If the cost of atorvastatin decreases in line with that observed for simvastatin when the patent ends in 2011, atorvastatin 80 mg/day will be the most cost-effective treatment for all thresholds; if the cost reduces to 25% of the current value, atorvastatin 80 mg/day will be the most cost-effective treatment for thresholds between £5000 and £30,000 per QALY. Large long-term RCTs reporting effects in terms of clinical events are required to determine the optimum statin use for subgroups.
Keywords