COURAGE Substudy: Anatomy Trumps Ischemia for Predicting Outcomes

Download this article's Factoid (PDF & PPT for Gold Subscribers) 

Among patients with stable ischemic heart disease, baseline assessment of anatomic burden predicts future events, but evaluation of ischemic burden does not, according to a substudy of the COURAGE trial published online January 15, 2014, ahead of print in JACC: Cardiovascular Interventions. Moreover, neither risk-stratification strategy alone, nor their combination, identifies those most likely to benefit from revascularization.

In the main COURAGE trial, after undergoing angiography, 2,287 patients with myocardial ischemia and significant CAD were randomized to optimal medical therapy (OMT) with or without PCI. After a median follow-up of 4.6 years, there was no difference between the groups in the primary outcome of all-cause death and nonfatal MI.

For the current analysis, investigators led by G.B. John Mancini, MD, of the University of British Columbia (Vancouver, Canada), looked at 621 COURAGE patients who had baseline QCA and quantitative nuclear SPECT data, 313 from the OMT plus PCI group and 308 from the OMT alone group.

At follow-up, the overall rate for the composite endpoint of death, MI, and NSTE-ACS was 30.2%.

No Treatment Interactions

The researchers performed 3 multiple logistic regression analyses, 1 containing only the primary factors of interest and 2 including interaction terms. Assignment to OMT + PCI or OMT alone did not predict outcome in any model, nor did ischemic burden. In contrast, LVEF and anatomic disease burden were consistent predictors of outcome, as was a history of hypertension and CABG surgery (table 1).

Table 1. Multiple Logistic Survival Analysis to Determine Outcome Predictors

Model 1
(No Interactions)


95% CI

P Value

Treatment Assignment




Ischemic Burden








Atherosclerotic Burden




There was no interaction between treatment and LVEF (P = 0.51), anatomic burden (P = 0.07), ischemic burden (P = 0.07), or the combination of anatomic and ischemic burden (P = 0.24). The only interaction approaching statistical significance (defined as P < 0.01) was that between anatomic and ischemic burden (P = 0.03). The basis for this interaction emerges from an analysis showing progressively higher event rates within the higher atherosclerotic burden group. On the other hand, event rates were progressively higher based on the degree of anatomic burden irrespective of the amount of ischemic myocardium.

The current analysis is discordant with observational studies implying that the severity of ischemia can identify patients who would benefit from an invasive strategy, the authors note. Although the discrepancy may be explained in part by the strict adherence to OMT in COURAGE or inadequate adjustment for confounders, there may also be a physiological basis for the findings, they say.

Although PCI reduces ischemic burden more rapidly than does OMT, in the long run medical therapy may achieve a comparable reduction in angina, Dr. Mancini and colleagues observe. “Thus, the overlapping anti-ischemic nature of both strategies may contribute over the long term to the abrogation of the prognostic importance of baseline, pretreatment inducible ischemia,” they write.

Prelude to ISCHEMIA

“Our data suggesting that ischemia may be most important in those with more severe atherosclerotic burden provide an important justification for the ongoing ISCHEMIA …trial, which is using coronary CTA to exclude patients with minimal atherosclerotic disease,” the authors write. “Due to its randomization strategy [no prior catheterization], it will be possible to determine whether ischemic burden at baseline retains prognostic power in those receiving OMT.”

In an accompanying editorial, Spencer B. King III, MD, of Emory University School of Medicine (Atlanta, GA), writes that the study findings provide both good news and bad news for ISCHEMIA. The good news is that while the COURAGE population consisted mostly of patients with less than 10% ischemia, the ISCHEMIA trial will include patients with a greater anatomic burden who will therefore have a greater chance of benefiting from revascularization. The bad news, he adds, is that because there is no anatomic criterion for participation, some patients with the greatest predictive risk of events may be excluded, while others with extensive ischemia but less compelling anatomic burden may be included.

Importantly, Dr. King observes, the mandated baseline CTA—to which investigators are blinded—“will provide a post hoc opportunity to evaluate the prognostic importance of the anatomic burden of disease, and we may learn whether anatomy, once again, trumps ischemia in this higher-risk group.”

The Perils of Ignoring Anatomy

In a telephone interview with TCTMD, Ajay J. Kirtane, MD, SM, of Columbia University Medical Center (New York, NY), called attention to the clinical consequences of misinterpretation of the COURAGE trial results.

“A lot of people have interpreted COURAGE [to imply that] if you do a stress test and a patient is at low or intermediate risk, you don’t have to cath them because revascularization is not going to make a difference,” he said. “But that is not consistent with COURAGE, BARI 2D, and several other studies because they mandated [prerandomization] catheterization specifically to exclude patients with high-risk anatomy who would get revascularized. So the authors’ assertion that there is no treatment effect has simply not been demonstrated in a high anatomic risk population.”

In fact, ignoring the anatomy can sometimes be dangerous, Dr. Kirtane added, explaining, “A clinician who relies only on stress testing may end up medically managing left main disease, which is not only not standard of care, it can harm the patient.”

Dr. Kirtane also faulted the study’s low 50% threshold for stenosis. “I wouldn’t expect treatment of [that kind of] lesion to affect outcome, so I don’t see [these data] as being relevant to the revascularization vs. OMT argument.”

There should be some filter for deciding which patients warrant catheterization, Dr. Kirtane acknowledged. “But stress testing is just not that good,” he said. “You have to use clinical judgment, but if there is a question, it’s important to define the anatomy, because clearly anatomy predicts outcomes—even in a low-risk population like COURAGE.”

Study Details

There were no baseline differences between patients assigned to OMT + PCI or OMT alone. Compared with the overall COURAGE population, patients in the substudy were more likely to be white, hypertensive, and diabetic; have had previous CABG surgery; and have triple-vessel disease and lower HDL levels; they were also less likely to be in Canadian Cardiovascular Society class II and III. On the other hand, their angina duration was shorter.  


1. Mancini GBJ, Hartigan PM, Shaw LJ, et al. Predicting Outcome in the COURAGE trial: Coronary anatomy vs. ischemia. J Am Coll Cardiol Intv. 2013;Epub ahead of print.

2. King SB III. Is it form or function? The “COURAGE” to ask. J Am Coll Cardiol Intv. 2013;Epub ahead of print.



Related Stories:

COURAGE Substudy: Anatomy Trumps Ischemia for Predicting Outcomes

Among patients with stable ischemic heart disease, baseline assessment of anatomic burden predicts future events, but evaluation of ischemic burden does not, according to a substudy of the COURAGE trial published online January 15, 2014, ahead of print in JACC Cardiovascular
  • The study was funded by the US Department of Veterans Affairs in collaboration with the Canadian Institutes of Health Research and unrestricted research grants from multiple pharmaceutical and other health companies.
  • Dr. Mancini reports receiving honoraria from AstraZeneca and Merck Canada and consulting fees from Amgen, GlaxoSmithKline, and Sanofi-Aventis.
  • Dr. King reports receiving honoraria for serving on the data safety monitoring boards of Merck and Wyeth Pharmaceuticals and serving as a consultant to and holding equity in Celonova Biosciences.
  • Dr. Kirtane reports participating in studies through his institution, which receives institutional research support from Abbott Vascular, Abiomed, Boston Scientific, Medtronic, and St. Jude Medical.