CT Angiography and Stress Testing: Prognostic Value Differs Between Women and Men
(UPDATED) Chicago, IL—Whether coronary CT angiography (CTA) or stress testing offers the most prognostic value for individuals with suspected coronary artery disease varies by patient sex, according to a fresh analysis of the PROMISE trial. In women, a positive CTA more closely correlates with subsequent clinical events than does a positive stress test, whereas in men, the two tests are on more equal footing.
“We know that whether a patient is a woman or a man affects the presentation, pathophysiology, and outcomes of coronary artery disease—and it affects the diagnostic performance of noninvasive testing,” said lead author Neha J. Pagidipati, MD (Duke Clinical Research Institute, Durham, NC), during a featured clinical research session at the American College of Cardiology 2016 Scientific Sessions.
The findings also were published online ahead of print in the Journal of the American College of Cardiology.
PROMISE randomized 10,003 symptomatic patients with suspected CAD to an initial diagnostic strategy of coronary CTA or functional testing (exercise ECG, nuclear stress testing, or stress echocardiography). Reported in March 2015, the main results showed that rates of the primary composite endpoint (death, MI, hospitalization for unstable angina, or major procedural complication) were similar in the CTA and functional testing groups (3.3% vs 3.0%) through a mean follow-up of 25 months.
Interestingly, in the overall trial, there was no difference in clinical event rates by sex between patients randomized to CTA versus stress testing.
Separating the Sexes
Pagidipati et al sought to compare the prognostic value of the two tests separately in men and in women, who made up 53% of the PROMISE cohort. Their analysis was restricted to the 90% of patients in the trial who were tested as randomized and had interpretable test results. Their main outcome of interest was clinical events, defined as the composite of death from any cause, MI, or hospitalization for unstable angina.
Randomization in the original trial was not stratified by sex; however, in the current analysis, baseline characteristics were well balanced between the CTA and stress testing arms for both men and women.
In women, a positive CTA was less common than a positive stress test (8% vs 12%), though “event rates for each type of test were low and nearly identical,” Pagidipati said. “This means that a larger proportion of women with a positive stress test did not go on to have a clinical event compared to those women with a positive CTA.”
In men, the pattern was reversed, with a positive result more frequently seen for CTA than for stress testing (16% vs 14%). Again, event rates were low and similar between test types, she noted. Compared to men having a positive CTA, Pagidipati explained, “those with a positive stress test had a higher clinical event rate, but this difference in men was not statistically significant.”
The interplay between test type and sex on test positivity was “highly significant,” she stressed.
Compared with negative tests, a positive CTA in women was more predictive than a positive stress test of subsequent clinical events (adjusted P = 0.028). Among men, a positive CTA was slightly but not significantly less informative of risk compared with a positive stress test (adjusted P = 0.168).
According to the researchers, both sex and noninvasive testing type appear to “jointly influence the relationship between test result and clinical events.”
Prospective Study Must Come First
“I think one major point to take away from this study is that it adds to the growing evidence that men and women do experience coronary artery disease differently, and that the patient’s sex should be taken into account in clinical decision making. This may extend to which test is ordered and how that test is interpreted,” Pagidipati told TCTMD.
But before these findings make their way into a clinical recommendation, the issue must be looked at prospectively, she urged. The patterns seen here are “something that probably clinicians should be aware of, . . . but to recommend a change in clinical practice at this point would be a bit premature.”
It is important to keep in mind, Pagidipati said, that these observations were made in the setting of what was a neutral trial. However, she predicted that the likelihood of a prospective randomized trial being done to study “this specific question of what tests should be ordered in men and in women . . . realistically is relatively low, because as you saw in the PROMISE trial, there were over 10,000 patients followed for over 2 years, and still there was only an event rate of around 2-3%.”
But “there are other ways of getting at this question, for instance looking prospectively through registry-based data,” perhaps via electronic health records, Pagidipati suggested. “It would be fascinating . . . , especially in this era of growing CT angiography use, [to see] in real-world settings, in community practices across the country and across the globe, what are people ordering in men and women? How is that changing over time? And how do men and women do with these different tests?”
False Positives and Microvascular Dysfunction
Following the presentation, panelist Prediman K. Shah (Cedars-Sinai Medical Center, Los Angeles, CA), said, “The message that I derive from this study is reconfirmation of the long-standing view that false-positive stress tests are more common in women than in men.”
This is indeed possible, Pagidipati agreed. “Yes, I do think that this would indicate that is most likely the case. That would likely explain what we’re seeing in women. . . . That may also be related to microvascular dysfunction, which we know is common in women, that may be picked up by stress testing and not by CTA. That may also help to explain the differences in prognostic value.”
The event rate associated with microvascular dysfunction is lower than that associated with obstructive disease, Pagidipati said.
Panelist Allen J. Taylor, MD (MedStar Health, Washington, DC), asked for further details on how positive tests were defined in the analysis and whether alternate definitions might narrow the gap between men and women.
Pagidipati et al defined positive CTA as epicardial stenosis ≥ 70% or left main stenosis ≥ 50%. A stress echo or stress nuclear test was considered positive if there was inducible ischemia in at least one coronary territory and/or early termination of exercise stress (< 3 minutes) due to ST changes consistent with ischemia, symptom reproduction, arrhythmia, and/or hypotension.
“Your point is very well taken about looking at the different levels of positivity, . . . because not all positive tests are the same,” she said. “We didn’t do that in this study for a couple of reasons. One is that I think there are only so many ways that we can slice a handful of events before our statisticians start trying to rein us in.” Also, the study looked more at a population level than at an individual level—admittedly, for individual patients, the degree of positivity would be more important for clinical decision making, Pagidipati said.
But with the definitions used here, already a “distinction between positive and negative results is very nicely splayed out in two curves. That may tell us all we need to know,” she suggested.
Pagidipati NJ, Hemal K, Coles A, et al. Sex differences in functional stress test versus CT angiography in symptomatic patients with suspected CAD: insights from PROMISE. J Am Coll Cardiol. 2016;Epub ahead of print.
- Pagidipati reports no relevant conflicts of interest.