Answers are grounded in this paper and cite their sources.
Across Europe, an aging population and improved survival after acute coronary syndrome are increasing the prevalence of stable coronary artery disease. Large numbers of people therefore require treatment with long-term anti-anginal and secondary prevention medication, and may be offered expensive cardiac investigations and myocardial revascularization procedures. This use of resources places a substantial economic burden on healthcare systems, and in the United Kingdom it has been estimated that the management of angina consumes 1% of the total National Health Service budget.1 More recently, in the USA, the direct and indirect costs of coronary heart disease were estimated at $177.1 billion and a substantial proportion of these costs were attributable to myocardial revascularization procedures.2 Evidence that this investment represents value for money is limited, and the review of the published evidence on the cost-effectiveness of myocardial revascularization procedures for the treatment of multi-vessel coronary artery disease by Ariyaratne and colleagues in this issue of the Journal is therefore timely. Cost-effectiveness is typically assessed with the incremental cost-effectiveness ratio (ICER), which incorporates differences in clinical efficacy and economic cost between alternative treatments into a single metric, expressed as the incremental cost per additional unit of efficacy. The standard metric for assessment of clinical efficacy is the quality adjusted life year (QALY), which combines the impact of treatment on quantity and quality of life. A treatment that is less effective and more expensive than an alternative treatment is not cost-effective and would not be recommended. Conversely, treatment that is more effective and less expensive has a lower ICER, dominates the alternative treatment, and is generally considered to be the preferred strategy. Health economic analysis is particularly useful, however, when a treatment increases clinical effectiveness but incurs a higher cost than an alternative treatment, and in these circumstances the ICER is judged against the amount that the relevant society is willing to pay per QALY. When Andreas Grüntzig introduced percutaneous transluminal coronary balloon angioplasty into routine practice nearly 40 years ago, the early clinical results were marred by high rates of procedure-related coronary occlusion and target lesion restenosis during subsequent follow-up. Over the ensuing decades, these problems have been addressed by iterative improvements in interventional practice, including the development of bare-metal and drug-eluting coronary stents, combined with dual antiplatelet therapy to minimize the risk of stent thrombosis. As a result, percutaneous coronary intervention is now used to treat many patients with multi-vessel coronary artery disease for whom coronary artery bypass graft surgery was previously mandated. Nevertheless, the optimal long-term revascularization strategy for most patients with stable multi-vessel coronary artery disease remains controversial.3,4 Numerous randomised trials of coronary artery bypass graft surgery versus percutaneous coronary intervention suggest that both methods of myocardial revascularization relieve angina, but have reported inconsistent effects on other major clinical outcomes. Meta-analyses of the trials suggest that coronary artery bypass surgery confers a survival advantage, particularly in patients with diabetes,5 but the relevance of this finding to contemporary practice is limited by wide variation in the percutaneous revascularization techniques that were studied in the individual trials. Two recent randomized trials that assessed the efficacy of percutaneous coronary intervention using drug-eluting stents are therefore of particular interest. The SYNTAX trial randomized 1800 patients with multi-vessel coronary disease to coronary artery bypass surgery or to percutaneous revascularization using ‘first-generation’ paclitaxel-eluting stents. Although SYNTAX was designed as a non-inferiority trial, at 5 years there was a 10% absolute difference in the rate of the primary end point (a composite of all-cause mortality, stroke, myocardial infarction, and repeat revascularization), which favoured surgical treatment and exceeded the pre-specified non-inferiority margin. This difference was due to higher rates of repeat revascularization and myocardial infarction in the percutaneous coronary intervention group with no differences in mortality or stroke.6 The mechanism for the difference in the rate of myocardial infarction is unclear, but it is plausible that a coronary bypass graft provides greater protection against the effects of future plaque disruption than focal treatment with a stent. On the other hand, the difference in the rate of myocardial infarction might also be explained by ascertainment bias in an open study, and the clinical impact of myocardial infarction related to interventional coronary procedures varies substantially, depending on the size of the infarct.7 Hence, the clinical relevance of the differences in the rates of myocardial infarction and the primary end point in SYNTAX are open to debate. Nevertheless, multiple exploratory subgroup analyses of the primary endpoint by angiographic risk score and pattern of coronary disease have received considerable emphasis in SYNTAX trial publications and have influenced guideline recommendations,8 even though such analyses have limited statistical power. The FREEDOM trial randomized 1900 patients with diabetes and multi-vessel coronary disease to coronary artery bypass surgery or percutaneous revascularization using ‘first-generation’ sirolimus-eluting coronary stents. At 5 years, the surgical strategy was associated with a 7.9% absolute reduction in risk of the primary outcome (a composite of death from any cause, nonfatal myocardial infarction, or nonfatal stroke) and a 5.4% absolute reduction in mortality. The mortality benefit only appeared after 2 years and for the average patient, surgery only extended life by a few months. Moreover, stroke was more frequent in the surgical group, with a 2.8% absolute difference at 5 years.9 These randomized trials enrolled selected patients who were considered suitable for both myocardial revascularization strategies, and the results may not be generalizable to the larger population of patients with multi-vessel coronary disease undergoing myocardial revascularization in routine practice. Large observational registries may provide additional information, and the ASCERT study of patients with multi-vessel disease reported outcomes after coronary artery surgery in 86 244 patients and percutaneous revascularization in 103 549 patients, 78% of whom received a drug-eluting stent. After 4 years propensity-score adjusted mortality was lower after surgical revascularization (absolute difference 4.4%),10 but residual confounding by variables that were not included in multivariable adjustment is an inherent limitation of this type of analysis. In another registry of 1013 patients undergoing percutaneous coronary intervention for multi-vessel disease, ineligibility for coronary bypass surgery in 22% of patients was associated with a six-fold increase in the odds of in-hospital, and a three-fold increase in the hazard of longer-term mortality. This effect could not be explained by differences in other known risk factors for adverse outcome.11 Healthcare costs vary widely between different countries and healthcare systems, reflecting differences in health policy, arrangements for clinical care, and the price of consumables. Long-term healthcare costs of individual patients with stable coronary artery disease may also be influenced by cardiovascular risk profile, and low-risk patients with longer life expectancy may incur substantial future healthcare costs.12 The costs of myocardial revascularization procedures have evolved over time as new technologies have become available. In the pre-stent era the cost of a coronary artery bypass operation was estimated to be nearly twice the cost of a coronary balloon angioplasty procedure, but this difference attenuated during follow-up, reflecting a greater requirement for additional revascularization procedures and medications after percutaneous treatment.13 The introduction of bare-metal stents narrowed the difference in costs between the two revascularization strategies because of the requirement for additional equipment and medication, and the need to treat late complications including in-stent restenosis.14 The development of drug-eluting stents further improved the clinical outcome but also increased the costs of percutaneous revascularization, leading to early restriction in the use of this technology in many countries. In the SYNTAX trial, the initial procedural costs were lower for coronary artery surgery than for percutaneous revascularization, reflecting the high cost of drug-eluting stents, but the total cost of the initial hospitalization and estimated lifetime costs were higher in the surgical group.15 In the FREEDOM trial, the total costs of the index hospitalization were also higher for patients assigned to surgery, but this difference attenuated over 5 years as patients assigned to percutaneous intervention required more repeat revascularization procedures and incurred higher medication costs.16 Although there is evidence from some recent randomized trials that myocardial revascularization by coronary artery bypass surgery confers a modest survival benefit over percutaneous revascularization, uncertainty about the optimal choice of revascularization strategy for patients with multi-vessel coronary artery disease persists. Moreover, the estimated costs of revascularization have varied over time, reflecting rapid advances in interventional cardiology with the introduction of effective but expensive new technologies. Against this background reliable estimation of cost-effectiveness of myocardial revascularization strategies is challenging. Ariyaratne et al. are therefore to be congratulated on their systematic review of studies comparing surgical and percutaneous revascularization procedures in subjects with multi-vessel disease that also reported health economic analyses. The review identified 16 relevant health economic studies, including 12 that were conducted alongside 11 randomized trials, and 4 that were based on comparison of non-randomized patient cohorts. This selection process excludes studies comparing coronary bypass surgery and percutaneous coronary intervention that did not conduct cost-effectiveness analyses, which potentially biases assessment of the efficacy of the two revascularization strategies. Patient characteristics varied between the selected studies, with mean patient ages ranging from 53 to 74 years. Moreover, the studies recruited patients over 18 years (1989 to 2007) and there was considerable heterogeneity in the revascularization techniques employed over this time period. Although most patients were treated by conventional coronary artery bypass surgery, some patients underwent minimally invasive or ‘off-pump’ surgery. Only five of the studies included patients treated by percutaneous coronary intervention using drug-eluting stents. The studies also adopted different analytic approaches with differences in time horizon, health-related utility weights, and modelling methods. The reported ICERs varied widely and no formal pooled analysis of the data from the selected studies was carried out. Nevertheless, studies with short-time horizons mainly favoured percutaneous revascularization, but three recent studies with longer time horizons (SYNTAX, FREEDOM, and ASCERT) reported lifetime ICERs that favoured coronary artery bypass surgery. On this basis, the authors concluded that for patients with multi-vessel disease coronary bypass surgery is economically more attractive than percutaneous revascularization. Other health economic analyses of the same problem have reached different conclusions. In an economic analysis conducted alongside the development of a clinical guideline on stable angina, the National Institute of Health and Care Excellence concluded that percutaneous coronary intervention in patients with multi-vessel disease is likely to be cost-effective when compared with surgical revascularization.17 In another study, 15 randomized trials of treatment strategies for stable angina that reported cost-effectiveness analyses were included in a network meta-analysis. Medical, surgical, and percutaneous interventional strategies had comparable effects on the 3-year rate of death or myocardial infarction, but coronary artery bypass surgery was the most costly, and by inference the least cost-effective strategy.18 Recent developments in the design of drug-eluting stents, including the use of novel alloys for the metal scaffold and the use of biocompatible and bioabsorbable polymers to deliver optimal doses of drug, have improved procedural success rates and reduced rates of stent thrombosis and in-stent restenosis.19 As yet, no large randomized trials have compared percutaneous revascularization using such ‘second-generation’ drug-eluting stents with surgical revascularization, although use of everolimus-eluting stents in patients with stable coronary artery disease has been associated with improved survival.20 Over time, the price of drug-eluting stents has also declined, partly as a result of competition between manufacturers, with a beneficial effect on cost-effectiveness. Hence, the cost-effectiveness of treatments for patients with multi-vessel coronary artery disease is likely to remain a moving target. Perhaps the most important message from the review by Ariyaratne et al. is that there are important gaps in our knowledge about the relative clinical and cost-effectiveness of alternative myocardial revascularization strategies for multi-vessel coronary artery disease. Evidence that one myocardial revascularization strategy is clinically and economically superior is lacking, and in this context the choice of strategy for individual patients will depend on the clinical circumstances, angiographic pattern of disease, local policy and patient preference. In discussion with patients, some clinicians may emphasize potential survival benefits over other effects of revascularization but patient expectations of treatment vary widely. Indeed, some fully informed patients may be willing to trade any long-term survival advantage from surgical revascularization for the lower procedural morbidity (especially stroke) and better short-term quality of life associated with percutaneous revascularization. In the SYNTAX trial, 50 patients assigned to surgical revascularization but only 11 patients assigned to percutaneous revascularization withdrew consent to the randomised treatment, suggesting that patient preference influenced patient acceptance of assigned treatment.6 Further randomized trials of contemporary revascularization strategies in patients with multi-vessel coronary artery disease are required and should take account of the patient perspective, and include prospective health economic analyses. Such trials should also address patients who have been under-represented in previous studies, including women and older patients, and those with particular anatomical patterns of coronary artery disease. In the interim, clinicians, commissioners, and purchasers of healthcare should be aware that the available studies of cost-effectiveness of different myocardial revascularization strategies in patients with multi-vessel coronary disease have important limitations, have reported inconsistent results, and cannot easily be used to support particular health policies. Conflict of interest: Dr. Henderson was Clinical Advisor and Guideline Development Group member, NICE Stable Angina guideline 2009-11.
No takes yet. Share an insight, caveat, or question.
Robert Henderson (2016) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: