Coronary revascularisation

Abstract

Surgery is effective on clinical and economic grounds, but stenting does not seem to be cost effective

This week, the BMJ publishes three studies dealing with revascularisation in ischaemic heart disease.^{1 2 3} Two of the studies compare the clinical effectiveness¹ and cost effectiveness² of revascularisation of isolated left anterior descending coronary disease by stenting or surgery, while the third examines the cost effectiveness of medical treatment, stenting, and surgery in multivessel disease.³ The studies raise key issues not only about the decision making process for intervention in the individual patient but also how to obtain maximum value from limited health service resources.

Isolated left anterior descending coronary artery disease

Because the left anterior descending coronary artery supplies more myocardium than the circumflex or right coronary arteries, disease in its proximal portion carries a worse prognosis. When ischaemia is present, revascularisation improves survival⁴ even in asymptomatic patients.⁵ For more than two decades, the most durable and effective option for revascularisation has been an internal mammary artery graft which, unlike vein grafts, is almost immune to the development of atherosclerosis.⁶ This strategy significantly reduces the risk of death, subsequent myocardial infarction, recurrent angina, and the need for repeat intervention.⁶ However, because surgery has conventionally required a median sternotomy incision and cardiopulmonary bypass, many cardiologists have favoured the less invasive option of percutaneous revascularisation with stents, unless this is contraindicated by certain anatomical or pathological complexities.

Two studies in this issue, one a systematic review and meta-analysis,¹ the other a cost effectiveness analysis,² report that internal mammary artery grafting using a less invasive surgical approach (through a small thoracotomy on the beating heart) is clinically at least as effective¹ and probably more cost effective² than stenting over the medium to long term. Compared with surgery, stenting resulted in an almost threefold increase in recurrent angina (odds ratio 2.62, 95% confidence interval 1.32 to 5.21) and an almost fourfold increase in the need for reintervention (4.63, 2.52 to 8.51).¹

While mortality did not differ significantly between the interventions, the survival benefit of surgery may have been underestimated because follow-up was limited to less than four years in most of the studies (the survival benefit of surgery may not appear until later). More importantly, patients with severe and complex lesions that were not suitable for stenting were excluded from the trials, even though they would still have had a survival benefit with surgery.^{4 5 6}Despite the findings, practice is unlikely to change as many patients with left anterior descending coronary artery disease—which is amenable to either intervention—may still favour the less invasive approach of stenting (even at a higher risk of reintervention) in the absence of a definitive survival advantage.

Multivessel coronary artery disease

Several trials of stents versus surgery in patients with multivessel coronary artery disease have reported no survival benefit from surgery. However, as the trials randomised less than 5% of all potentially eligible patients and included only low risk patients (patients with more severe disease who would have a survival benefit from surgery were actively excluded). In effect, therefore, these trials were inherently biased against the prognostic benefit of surgery in most patients with multivessel disease.⁷ Consequently, the use of these trials to justify the widespread practice of inserting stents is not evidence based and denies some patients the survival advantage of surgery. Several large “real life” registries show that most patients with multivessel disease survive significantly longer after coronary artery bypass grafting rather than stenting,^{8 9 10} and the benefit is even greater in patients with diabetes, who usually have more severe coronary artery disease.¹¹ For example, in the New York Registry database of almost 60 000 patients, after risk matching for cardiac and non-cardiac comorbidity, the three year mortality was 15.6% for stenting compared with 10.7% for coronary artery bypass grafting (P<0.01), and reintervention (35% v 5%) was seven times higher in patients receiving stents rather than surgery.⁷ Nevertheless, despite the benefits of surgery on survival and freedom from reintervention, the number of stent procedures has increased dramatically in most industrialised countries in the past few years, so that this intervention now outnumbers surgery at least fourfold.

In contrast to the studies of the clinical effectiveness of surgery and stenting, the study by Griffin and colleagues³ examines the cost effectiveness of medical treatment, stenting, and surgery in patients deemed suitable for each treatment by a panel of experts. It concludes that both medical treatment and surgery (but not stents) are cost effective at a conventional National Health Service quality adjusted life year threshold of £30 000 (€44 000; $58 000) and that the additional benefit of percutaneous coronary intervention over medical treatment is “too small to justify its additional costs.”³ While these findings are unlikely to be welcomed by the stent industry, valued at around $6bn each year, they echo the concerns of a previous report questioning the clinical effectiveness and cost effectiveness of stents compared with medical or surgical treatment.¹²

Will the findings remain robust in the era of drug eluting stents?

The key to answering the question of whether the findings will remain robust in the era of drug eluting stents is based on understanding why surgery has a survival advantage over stents in multivessel disease.^{8 9 10} Firstly, because bypass grafts are placed to the midcoronary vessels, surgery protects whole zones of vulnerable proximal myocardium against the “culprit” lesion (of any complexity) and against new lesions in diffusely diseased endothelium. In contrast, stents deal only with “suitable” culprit lesions and offer no protection against new disease. Secondly, the failure of stenting to achieve complete revascularisation in most patients with multivessel disease reduces survival proportional to the degree of incomplete revascularisation.¹³

For these reasons, even drug eluting stents are unlikely to match the results of surgery for most patients with multivessel disease. These same reasons probably explain the findings of several meta-analyses, which report that although drug eluting stents reduce the risk of restenosis in low risk coronary lesions, they do not reduce mortality or the risk of subsequent myocardial infarction.¹⁴

And disquiet about the lack of improved clinical outcome with drug eluting stents, despite their increased costs,¹² has recently been superseded by concerns about the increased risk of late thrombosis and its high associated mortality.¹⁵ These clinical concerns are compounded by cost implications. Drug eluting stents are much more expensive than bare metal stents, and new recommendations that patients remain on clopidogrel for at least a year¹⁵ and possibly indefinitely—despite the increased risks of bleeding and at a minimum cost of £500-£1000 a year—will add greatly to costs.

Implications for health services and for patients

Griffin and colleagues highlight the tension between the adverse economic implications of the phenomenal growth in stent procedures and the absence of an appropriate evidence base to support such a policy. More importantly, this strategy has denied many patients with multivessel disease the prospect of a better long term outcome in terms of survival and freedom from reintervention offered by surgery. This highlights the dangers of individual practitioners rather than multidisciplinary teams making recommendations for stenting in patients with multivessel disease. Such teams should include a non-interventional cardiologist and surgeon and are likely to offer more balanced advice.⁷ Griffin and colleagues have laid the challenge for “physicians, providers, and payers to prove that clinically appropriate treatments are also cost effective.” To this we should add the challenge that a multidisciplinary approach should be a minimum mandatory “standard of care” to ensure that patients are offered the most clinically appropriate treatment.