Stress tests were not reliable in predicting coronary obstructions in patients with chest pain but no known coronary artery disease (CAD) referred for invasive angiography, and performance varied by angina status, a single-center observational study suggested.
Patients with typical angina were actually more likely to have obstructive CAD if they had a negative stress test than if they had a positive stress test (74.3% versus less than 60%).
And a positive stress test in atypical angina was confirmed as obstructive CAD upon invasive angiography in only 24.6% of cases, John P. Vavalle, MD, MHS, of the University of North Carolina at Chapel Hill School of Medicine, and colleagues reported.
“We believe that this study highlights the importance of clinical judgment and a detailed understanding of patients’ clinical symptoms in identifying those at highest risk for obstructive CAD,” they concluded online in JAMA Cardiology.
Armin Arbab-Zadeh, MD, PhD, MPH, of the Johns Hopkins Hospital in Baltimore, said the “remarkable” study confirmed his group’s previous findings of “poor sensitivity of stress testing for identifying patients with obstructive CAD.”
“These data document that the stress test results correlate poorly with both, patient symptom and outcome. It is particularly concerning to note the high rates of nonobstructive and obstructive CAD findings in patients with normal stress test findings,” he explained.
Arbab-Zadeh, who was not involved in Vavalle’s investigation, wrote in an email to MedPage Today that “the take home message — in my mind — is to reconsider our approach to patients with stable chest pain. We want to establish or rule out the diagnosis of CAD and establish the patient’s prognosis.”
“There is strong data now suggesting that CT coronary angiography is superior to stress testing for both purposes. It’s time to have practice guidelines reflect these data,” he concluded.
Vavalle’s study included 15,888 patients with no history of CAD who were referred to the cath lab between 1996 and 2010. Data was collected from the Duke Databank for Cardiovascular Disease.
Ten-year rates of myocardial infarction (MI) did not differ between patients with a positive stress test versus a negative one (P=0.10), although patients with typical angina had the highest rate of MI (6.7%).
Revascularization was less likely for patients with positive stress test results (35.2% versus 47.9% for negative stress test results and 40.3% for no stress test). In addition, revascularization was more likely for patients with typical angina (64.0% versus 30.0% for atypical angina, 26.9% for no chest pain) by 10 years.
Arbab-Zadeh pointed out that “more than 60% of patients who are being referred for elective cardiac cardiac catheterization have either no or only atypical chest pain.”
“This is important because the main benefit of performing percutaneous coronary intervention in patients is symptom relief. If only about one-third of patients have typical angina (and most of these have only mild angina according to the PROMISE trial), why are we sending these patients for cardiac catheterization in the first place?” he questioned.
The PROMISE trial had shown that CT angiography and exercise or stress testing came out similar for clinical outcomes as initial screening tests for suspected stable coronary artery disease, although initial CT angiography led to fewer catheterizations that turned up no obstructive CAD.
In any case, Vavalle and colleagues acknowledged that their investigation suffered from several limitations, not least of all its nature as a single-center observational study. “It is possible that we did not capture nuances of clinical presentation, such as symptom severity and frequency, that may critically affect a physician’s evaluation and decision making,” they wrote.
Kavitha M. Chinnaiyan, MD, of Beaumont Health System in Royal Oak, Mich., told MedPage Today in an email, however, that “this study demonstrates that the strongest correlation lies between patients’ symptoms and invasive coronary angiography findings. From these results, it would appear that the clinical acumen of physicians was superior to stress test findings in predicting coronary angiography results.”
“Trust your clinical intuition when it comes to patients’ symptoms! No test has 100% accuracy, which is why medicine is an art as much as it is a science,” urged Chinnaiyan, who was also not part of the investigation.
Vavalle, Arbab-Zadeh, and Chinnaiyan disclosed no relevant conflicts of interest.
Vavalle JP, et al “Effect of the presence and type of angina on cardiovascular events in patients without known coronary artery disease referred for elective coronary angiography” JAMA Cardiol 2016; DOI: 10.1001/jamacardio.2016.0076.