ABSTRACT & COMMENTARY
Current Utility of Exercise ECG Testing
By Michael H. Crawford, MD
Professor of Medicine, Lucie Stern Chair in Cardiology, Director, Cardiology Fellowship Program, Chief of Clinical Cardiology, University of California, San Francisco
Dr. Crawford reports no financial relationships relevant to this field of study. This article originally appeared in the June 2014 issue of Clinical Cardiology Alert.
SYNOPSIS: The authors concluded that these findings can be used to identify which patients would benefit from further testing after an initial exercise ECG test to diagnose coronary artery disease.
SOURCES: Christman MP, et al. Yield of downstream tests after exercise treadmill testing. J Am Coll Cardiol 2014;63:1264-1274.
Sinusas AJ, Spatz ES. Reframing the interpretation and application of exercise electrocardiography. J Am Coll Cardiol 2014;63:1275-1277.
Most current guidelines recommend exercise electrocardiographic (ECG) testing for suspected coronary artery disease (CAD) in patients who can exercise and have a normal resting ECG. If the results are inconclusive, often another stress test with non-invasive imaging is done. These investigators sought to analyze the results of this downstream testing and identify characteristics that would make such testing valuable or not. The patient sample was collected over 2 years and excluded patients with known CAD. The Bruce treadmill protocol was used and standard criteria for positive, negative, and inconclusive ECGs was used. They analyzed any subsequent testing done by the ordering physician for 6 months after the initial test. Also, the patients were followed for subsequent cardiac events (death, myocardial infarction, or revascularization). The study population included 3656 patients, of whom 90% (3270) had complete follow-up for a mean of 2.5 years. Negative tests were most common (68%), inconclusive next (28%), and positive least (4%). The most common reasons for an inconclusive test were suboptimal exercise (57%), rapid resolution of ECG changes (13%), and test cessation for typical angina (10%). Further testing was performed in 11% (9% noninvasive imagining, 2% invasive angiography). Subsequent noninvasive imaging included stress nuclear perfusion (81%), stress echo (12%), coronary CT angiography (5%), and cardiac MRI (2%). The combined outcome endpoint occurred in those with a negative initial test was 0.2%, inconclusive 1.3%, and positive 12%. Multivariate analysis showed that younger age, female sex, higher exercise performance, and rapid recovery of any ECG changes predicted negative further testing and event free survival. The development of typical angina during the initial test predicted positive downstream testing and a worse prognosis. The authors concluded that these findings can be used to identify which patients would benefit from further testing after an initial exercise ECG test to diagnose CAD.
The major limitation of this trial is that it is an observational study done at one center. However, since randomized trials are unlikely to be conducted on this topic, the data can help inform our decisions in this complex and controversial area. Treadmill exercise ECG testing is commonly used as the initial diagnostic test in patients with symptoms that could represent myocardial ischemia, with the caveat that they can exercise near maximally and have a normal resting ECG. Prior observational studies have shown that these requirements are only present in about one-third of patients referred for stress testing. Most undiagnosed patients either have abnormal ECGs, can’t fully exercise, or have unstable angina. Even in this academic center series, more than half of the patients with an inconclusive test were not able to exercise fully. In my practice, anyone > 80 years of age or obese, I automatically eliminate exercise testing.
Interpreting this study is challenging because some of the downstream testing was obviously indicated, such as the patients with typical angina, but a normal ECG. In such patients, it is not inappropriate to do an imaging study or invasive angiography. In this study, 100% of those with obvious angina on stress testing had a significant coronary lesion that was stented. On the other side, a negative stress test would rarely indicate a need for further testing and in this study, the incidence of the combined endpoint was 0.2% in such patients. However, someone with angina symptoms and a negative stress test may have vasospasm or small vessel disease and further sophisticated imaging could be appropriate. There are nuanced areas in ischemic heart disease that may trump usual thinking.
Perhaps the most interesting data from this study are the subgroup with an inconclusive ECG exercise test because of rapid reversal of ST changes in recovery (< 60 seconds). There is considerable prior literature that suggests this may be characteristic of a false-positive test. In this study, such patients had an excellent prognosis. They had no positive downstream tests and no deaths or myocardial infarctions. Their data support the notion that this finding represents a false-positive test.
Another major limitation of this study is that we do not know the pre-test probability of disease. According to Bayes’ Theorem, this markedly affects the outcome of testing. For example, not everyone in the study had further testing, so there may be a section bias toward the more likely to have CAD patients. This would make further testing seem more valuable.
My conclusion is that in the small group of patients who meet criteria for an exercise ECG test, it is still a reasonable test. If typical angina is provoked despite non-diagnostic ECG changes, further testing is indicated. Lacking angina, if there are ECG changes suggestive of ischemia that resolve in < 60 seconds of recovery, this is likely a false-positive test.