Determining prognosis after acute myocardial infarction in the thrombolytic eraBMJ 1997; 315 doi: https://doi.org/10.1136/bmj.315.7111.761 (Published 27 September 1997) Cite this as: BMJ 1997;315:761
Non-invasive investigations still have a place
- George A Beller, Chiefa
- a Cardiovascular Division, Department of Internal Medicine, University of Virginia Health Sciences Center, Charlottesville, VA 22908, USA
The most appropriate and cost effective approach to assessing prognosis in patients who survive an acute myocardial infarction in the thrombolytic era remains controversial. Prognosis is determined mainly by the degree of left ventricular dysfunction and the extent of residual jeopardised myocardium, both generally and in the distribution of the infarct related artery. The extent of myocardial damage and inducible ischaemia can be assessed with non-invasive stress imaging and the extent of coronary artery disease with angiography. The issue is whether “routine” coronary angiography performed soon after infarction in patients reperfused early and with an uncomplicated course yields better prognostic information than exercise or pharmacological stress perfusion imaging.
Outcome studies in America and Canada have yielded unexpected findings about the value of routine invasive investigations in patients with uncomplicated courses. Rouleau et al reported that, although coronary angiography was more often performed in America than in Canada (68% v 35%), as was revascularisation after infarction (31% v 12%), no difference in mortality (23 v 22%) or rate of reinfarction (13% v 14%) was observed at a mean follow up of 42 months.1 In the GUSTO-1 trial, despite a much higher rate of angiography (72% v 25%) and angioplasty (43% v 14%) in America than in Canada, there was no difference in survival at 1 year (90.7% v 90.3%).2 The GUSTO-2 trial and a more recent one have confirmed the same picture,3 4 and other studies in American centres suggest that merely the presence of a cardiac catheterisation laboratory in a hospital is one of the strongest predictors of catheterisation in patients with acute infarction.
Studies in the 1980s showed that submaximal exercise or pharmacological stress myocardial perfusion imaging performed before discharge successfully distinguished patients at high risk of subsequent cardiac events and did so better than exercise electrocardiographic testing alone. Perfusion imaging was better at detecting and localising ischaemia at submaximal exercise heart rates; identifying multivessel coronary artery disease and residual ischaemia within the zone of infarction; and measuring infarct size. Gibson et al reported that about half of patients who were 65 or younger with an uncomplicated myocardial infarction who showed major defects on imaging subsequently experienced cardiac death, recurrent infarction, or class III-IV angina requiring admission to hospital.5 These defects were either multiple perfusion defects in more than one coronary vascular supply region on submaximal exercise scintigraphy with thallium-201, reversible 201Tl defects (ischaemia) within or outside the infarct zone, or abnormal lung thallium uptake. The cardiac event rate was only 6% in patients with normal scans or only persistent defects in the supply region of the infarct related artery. Studies published in the prethrombolytic era showed that stress perfusion had a sensitivity of about 70% and a specificity of 85% for detecting patients with multivessel disease.6 Similar findings were reported with dipyridamole or adenosine stress in conjunction with perfusion imaging.7 8
In the thrombolytic era, however, it is hard to show the worth of stress perfusion. Patients who are eligible for thrombolysis comprise a relatively low risk group, and many asymptomatic patients undergo routine angiography before stress testing. Those with multivessel disease or a residual high grade infarct related stenosis are often referred straight for revascularisation. Most of these high risk patients would have been identified by stress imaging but underwent the invasive strategy first. Also, many patients who do initially undergo stress testing are referred for coronary angiography and revascularisation as a result. This reduces the future cardiac event rate in these cohorts, resulting in a low positive predictive value of non-invasive imaging variables for predicting cardiac death or reinfarction (post test referral bias).9
Nevertheless, data are now emerging which show the continuing value of non-invasive stress imaging for risk stratification after acute myocardial infarction. Dakik et al recently reported that quantitative exercise 201Tl imaging performed in patients who had received thrombolysis provided extra prognostic information over and above clinical findings and ejection fraction data; coronary angiographic variables did not further improve prognostic information.10 Again, other studies have shown similar findings, with both exercise and pharmacological stress.11 12 13
Although the negative predictive value of a low risk stress perfusion scan for predicting a low event rate is excellent, the positive predictive value of an ischaemic response for predicting subsequent cardiac death or infarction is only 40-50%. A cost effective approach therefore may be to categorise patients clinically into high risk, intermediate risk, or low risk groups before determining which investigations to use. Clinically high risk patients (those with postinfarction angina, history of infarction, rales in over a third of the lung field on admission, hypotension and sinus tachycardia on admission) could go directly to coronary angiography with a view to coronary revascularisation. Patients at intermediate or low risk could undergo an initial non-invasive investigation with angiography performed in those with significant ischaemia or a scan pattern on perfusion imaging that suggests multivessel disease. Patients with a depressed left ventricular ejection fraction but without clinical manifestations of heart failure could also undergo a non-invasive investigation to determine both myocardial viability within the zone of dysfunction and extent of inducible ischaemia. Those whose depression of left ventricular function is caused by jeopardised myocardium could be referred for angiography followed by revascularisation of arteries with important stenoses; those in which it is caused predominantly by myocardial scar would be treated with angiotensin converting enzyme inhibitors, aspirin, and ß blockers. To validate the worth of this approach, however, we need a clinical trial comparing it with “routine” angiography.