Editorials

Chronic coronary artery disease: drugs, angioplasty, or surgery?

The cardiovascular community has been studying the effectiveness of percutaneous transluminal coronary angioplasty for about 15 years and coronary artery bypass graft surgery for about 25; yet important questions remain about the relative roles of these two revascularisation procedures in patients with chronic coronary artery disease. A recent meta-analysis comparing angioplasty with bypass surgery¹ is most appropriately interpreted in the context of previous trials comparing angioplasty or bypass surgery with medical treatment.

An overview of randomised trials published in 1994 included data on about 2500 patients; it found that bypass surgery resulted in a highly significant overall reduction in mortality compared to initial medical treatment in patients with chronic coronary artery disease.² This difference in mortality was clearly evident only after four to five years of follow up and was confined to patients at high and moderate risk of death (those with five year mortality during medical treatment ranging from 25% to about 12%, in whom the risk reduction with bypass surgery was 50% and about 37%, respectively.) Patients at low risk of death (those with five year mortality rates of about 5%) were not likely to live longer after bypass surgery.

Two trials have compared medical treatment with angioplasty in patients with single vessel disease.^{3 4} Neither trial was large enough to detect a significant difference in the major clinical outcomes of death or myocardial infarction. The ACME trial, which included patients with angina and exercise induced ischaemia, found that angioplasty led to greater relief of angina and exercise induced ischaemia and better psychological wellbeing.³ The second trial, which was limited to patients with asymptomatic single vessel disease, found no difference in any outcome.⁴

In their meta-analysis, Pocock and colleagues identified eight prospective, randomised trials comparing angioplasty and bypass surgery, enrolling a total of 3371 patients with chronic coronary artery disease.¹ Using summary data, they evaluated the overall effects on mortality, myocardial infarction, and subsequent revascularisation procedures over a mean follow up period of 2.7 years. Despite variations in study design and patient characteristics (such as number of diseased vessels and severity of angina), the results from the eight trials were broadly consistent. Seventy nine (4.6%) patients undergoing angioplasty and 73 (4.4%) undergoing bypass surgery died during follow up (relative risk 1.10; 95% confidence interval 0.79 to 1.50). Cardiac death or non-fatal myocardial infarction occurred in 169 (9.9%) patients undergoing angioplasty and 154 (9.3%) undergoing bypass surgery (1.10; 0.89 to 1.37). At one year, bypass surgery achieved complete relief of angina more often than did angioplasty, although this difference diminished after three years of follow up. Thirty four per cent of patients undergoing angioplasty required at least one other revascularisation procedure during the first year of follow up, including 18% who "crossed over" to bypass surgery. Clearly, restenosis after angioplasty remains a challenge.

Hard to generalise to high risk patients

These data are extended by recently presented results from the Bypass Angioplasty Revascularisation Investigation (BARI) study (RL Frye, American Heart Association, Anaheim, California, 16 November 1995). The study randomly assigned 1829 patients with multivessel disease to either angioplasty or bypass surgery. Overall five year mortality was 13.7% in patients undergoing angioplasty and 10.7% in those assigned to bypass surgery; cumulative rates of myocardial infarction were 21.1% and 19.6% respectively. The observed 3.0% difference in absolute mortality represents a risk reduction of 22% but does not reach statistical significance (P=0.17; -0.2% to 6.0%). Adding these data to those in the recent meta-analysis achieves a combined sample size of 5200 patients; 7.8% of patients randomised to angioplasty and 6.6% of those randomised to bypass surgery died during follow up (odds ratio 1.20; 0.97 to 1.48), a difference in favour of bypass surgery that approaches nominal levels of statistical significance.

How should clinicians interpret these data in practice? Most of the patients included in the meta-analysis and in the BARI trial were at low or moderate risk of death; all had normal mean ejection fractions, less than 10% had significant impairment of left ventricular function, and 70% had one or two vessel disease. In the meta-analysis, the overall mortality was 2.6% in the first year and only 1.9% thereafter. This means that a large proportion of patients who entered these trials are not among those for whom a clear mortality benefit from bypass surgery (in comparison to medical treatment) had previously been found. These data cannot therefore be generalised to patients at moderate or high risk of death.

A 20% difference in the risk of death between bypass surgery or angioplasty could be reliably shown only if at least 600 deaths occurred in the study population. This would mean enrolling about 16000 patients at low risk or 8000 at moderate risk, and following them for five years. A larger difference in mortality (40-50%) can be reliably excluded by the currently available data, but we would not anyway expect to find differences between two forms of revascularisation on the same scale as those observed between bypass surgery and medical treatment. In the trials comparing bypass surgery and medical treatment, important differences emerged after about three years of follow up and became statistically significant at five years. In the trials comparing angioplasty with bypass surgery, the high rates of bypass surgery (20-30%) among patients initially treated with angioplasty further diminish the possibility of detecting differences. Given the moderate power of even the combined data and the relatively short follow up in most trials, it remains possible that differences between the two procedures of about 20-25% reduction in the relative risk of death may yet emerge and the difference become statistically significant with further follow up.

The significantly lower mortality associated with bypass surgery among patients with single vessel disease, which was observed in the meta-analysis by Pocock et al¹ and in the subgroup of patients with diabetes in the BARI trial, was unexpected and may well have been due to chance. These observations should be re-examined by combining the relevant subgroups (preferably using individual patient data) when data from at least five years of follow up are available from all trials. Equally important would be an analysis of subgroups of patients in whom bypass surgery has previously been shown to improve survival when compared to medical treatment.

Perhaps the relevant clinical question is not which mode of treatment is best, but which combinations of treatment, in what sequence, are appropriate for a specific patient at a specific point in their clinical course. Aggressive and tailored modification of risk factors (smoking cessation, lipid lowering, moderate exercise, and control of blood pressure and blood glucose) is the mainstay of treatment in all patients with coronary artery disease. In addition, patients with angina but at low risk of death should initially be given antianginal drugs and aspirin. If their symptoms continue, a revascularisation procedure should be considered, the choice being made on the basis of suitability of coronary anatomy, operator expertise, and patient preference.^{5 6} In patients at moderate or high risk of death, bypass surgery should generally be preferred, while angioplasty may be preferred in lower risk patients. These recommendations should be reassessed periodically as technology advances (especially if the problem of restenosis after angioplasty can be mitigated) and as we develop more effective approaches to preventing the progression of atherosclerosis or its thrombotic complications.

Assistant professor of medicine Professor of medicine Division of Cardiology, McMaster University and Preventive Cardiology and Therapeutics, Hamilton Civic Hospitals Research Center, Hamilton General Hospital, Hamilton, Ontario, Canada L8L 2X2

Charanjit S Rihal, Salim Yusuf 

1. Pocock SJ, Henderson RA, Rickards AF, Hampton JR, King SB, Hamm CW, et al. Meta-analysis of randomised trials comparing coronary angioplasty with bypass surgery. Lancet 1995;346:1148-9.
2. Yusuf S, Zucker D, Peduzzi P, Fisher LD, Takaro T, Kennedy JW, et al. Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Triallists Collaboration. Lancet 1994;344:563-70. [Medline]
3. Parisi AF, Folland ED, Hartigan P for the Veterans Affairs ACME Investigators. A comparison of angioplasty with medical therapy in the treatment of single vessel coronary artery disease. N Engl J Med 1992;326:10. [Abstract]
4. Sievers B, Hamm C, Herzner AE. Medical therapy versus PTCA: a prospective, randomized trial in patients with asymptomatic coronary single vessel disease. Circulation 1993;88 (part II):1-297.
5. Raco D, Rihal CS, Yusuf S. Overview of randomized trials of percutaneous transluminal coronary angioplasty: comparison with medical and surgical therapy for chronic coronary artery disease. In: Grech E, Ramsdale A, eds. Practical interventional cardiology. London: Martin Dunitz (in press.)
6. White HD. Angioplasty versus bypass surgery. Lancet 1995;346:1174-5. [Medline]