COURAGE: Medical Therapy ‘Crossovers’ Have More Symptoms, not More Hard Events

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

Approximately 1 in 6 stable patients who start off with optimal medical therapy will require revascularization within one year, largely due to unstable angina, according to a subanalysis of the COURAGE trial published online July 9, 2013, ahead of print in Circulation: Cardiovascular Quality and Outcomes. But crossing over to percutaneous coronary intervention (PCI) was not associated with a clearly higher rate of irreversible ischemic events and thus, the authors say, the trial’s medical therapy-first strategy remains reasonable. 

In the original trial, 2,287 stable patients with ischemia and significant CAD were randomized to optimal medical therapy with (n = 1,149) or without (n = 1,138) PCI. At a median 4.6 years, there was no difference between the PCI and the medical therapy-only groups for the primary endpoint of all-cause death and nonfatal MI (P = 0.62). 

For the subanalysis, investigators led by John A. Spertus, MD, MPH, of St. Luke’s Mid America Heart Institute (Kansas City, MO), looked at the frequency and predictors of crossing over to PCI among the medical therapy group, and assessed the clinical consequences of the initial medical therapy strategy for those patients. 

One in 6 Medical Therapy Patients Crossed Over

Among the 1,148 patients assigned to medical therapy, 185 (16.1%) underwent revascularization (90% receiving PCI, 10% CABG) within the first year after randomization. Of the 40 patients who crossed over to revascularization within the first 2 months, 21 were hospitalized for an adjudicated MI or unstable angina, while 70% of the remainder experienced increasing angina.

In a multivariable model, the strongest predictors of early revascularization were patients’ baseline health status, assessed by the angina frequency, stability, and treatment satisfaction scales of the Seattle Angina Questionnaire (SAQ), as well as the health-care system in which they were treated (Veterans Affairs [VA] vs. non-VA or Canadian; table 1).

Table 1. Likelihood of OMT Patients Requiring Revascularization within 1 Year


Adjusted HR

95% CI

Baseline SAQ
Angina Frequency, per -10 points
Angina Stability, per -25 points
Treatment Satisfaction, per -12.5 points



Health-Care System
Canada vs. VA
US non-VA vs. VA



P ≤ 0.001 for all comparisons.

No interactions were seen between the health-care system and any of the SAQ scales (P ≥ 0.50 for all comparisons).

To evaluate the potential clinical consequences of delayed revascularization, researchers matched 156 medical therapy patients who crossed over to early PCI with an equal number of patients randomized to upfront PCI. The pairs differed only in that crossovers had higher baseline LDL (P = 0.003) and use of nitrates (P = 0.004) and statins (P = 0.03), and were more likely to be Canadian Cardiovascular Society class III/IV (P = 0.003); these characteristics were included in the multivariable model.

Over the first year, mortality rates were similar between crossover patients and those with upfront PCI. Although there was no difference between paired patients in risk of nonfatal MI, 51% of events in crossover patients occurred before revascularization. There was a suggestion that crossover patients were more likely to be admitted for unstable angina. In addition, angina control and quality of life (QOL) were worse for crossover patients (table 2).

Table 2. Outcomes: Upfront PCI vs. Medical Therapy Requiring Revascularization Within 1 Year


OR or Difference (95% CI)

P Value


0.51 (0.13-2.10)


Nonfatal MI

1.86 (0.75-4.63)


Unstable Angina

2.78 (1.05-7.47)


SAQ Angina Frequency over First Year

-7.46 (-10.7 to -4.2)

< 0.001

SAQ QOL over First Year

-7.88 (-11.65 to -4.11)

< 0.001

a Statistical significance was defined as P = 0.01.

“These data support the principal findings of the COURAGE trial that it is reasonable to attempt [optimal medical therapy] in all patients meeting the entry criteria for the trial because there were no clear adverse consequences at 1 year after attempting to treat with [optimal medical therapy] alone,” the authors write.

Dr. Spertus and colleagues admit that crossover patients tended to need more frequent hospitalization for unstable angina and suffered worse quality of life during the first year. Thus, they say, “[i]dentifying such patients who are likely to need revascularization at an earlier time point could potentially lead to improved overall health status for the [optimal medical therapy] treatment strategy and prevent a transient period of suboptimal health status by offering early revascularization.”

Nonetheless, the authors conclude, “our data suggest that there are no irreversible consequences from starting with a more conservative management strategy. Given the unfavorable cost-effectiveness of upfront PCI from a societal perspective, these data, in aggregate, support an initial trial of [optimal medical therapy] in symptomatic patients with stable ischemic heart disease and close follow-up of the most symptomatic patients so that early revascularization can be offered if their symptoms are not adequately controlled with [optimal medical therapy] alone.”

Analysis not Patient Centered

“What bothers me [about this analysis] is that it’s the opposite of patient centric,” Jeffrey W. Moses, MD, of Columbia University Medical Center/Weill Cornell Medical Center (New York, NY), told TCTMD in a telephone interview. Instead, it adopts the perspective of the health-care system, he noted.

“We know from FAME II that if patients have ischemia they seem to do much better with PCI from the standpoint of symptoms, rehospitalization, and cost,” Dr. Moses observed. “Now from COURAGE we see that if they have a moderate amount of symptoms they tend to cross over,” And revealingly, the crossover rate was much higher in the more “empowered” non-VA health-care systems, he added.

With regard to the mortality risk among crossover patients, Dr. Moses commented, “Of course it’s not statistically significant: you can’t get a mortality statistic out of 300 patients. But it should at least be disturbing, especially against a background of several meta-analyses showing strong trends toward a mortality reduction with a PCI-first strategy.” Moreover, crossover patients had more MIs, which, in fact, led to many instances of crossover, he added.

Why Push Medical Therapy for All?

“What we should be doing is asking, is there a group of patients who should be getting PCI upfront?” Dr. Moses said. He noted that the overall COURAGE cohort “was not very symptomatic and not very ischemic. When you look at a population like that in the aggregate, medical therapy will do fine,” he said. “But once you start culling out the groups with ischemia or symptoms, it seems you’re better off with PCI first.”

Countering the authors’ claim for a medical therapy-first approach across the board, Dr. Moses emphasized that its cost-effectiveness has not been tested in at least moderately symptomatic patients in a contemporary setting. “Certainly PCI looked like it was cost-effective in the FAME II [ischemic] group,” he remarked.

In the end, Dr. Moses said, “Why push medical therapy for all patients until they hurt, so to speak? My attitude is, if a patient has moderate symptoms, knowing [what we do now], why not start with PCI?


Spertus JA, Maron DJ, Cohen DJ, et al. Frequency, predictors, and consequences of crossing over to revascularization within 12 months of randomization to optimal medical therapy in the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial. Circ Cardiovasc Qual Outcomes. 2013;Epub ahead of print.

  • Dr. Spertus reports holding the copyright to the Seattle Angina Questionnaire, receiving research grants from Eli Lilly and Genentech, and serving as a consultant to St. Jude Medical, United Healthcare, and the American Heart Association, where he serves as an editor for Circulation and Circulation: Cardiovascular Quality and Outcomes.
  • Dr. Moses reports serving as a consultant to Boston Scientific.


Related Stories:

COURAGE: Medical Therapy ‘Crossovers’ Have More Symptoms, not More Hard Events

Approximately 1 in 6 stable patients who start off with optimal medical therapy will require revascularization within one year, largely due to unstable angina, according to a subanalysis of the COURAGE trial published online July 9, 2013, ahead of print in Circulation
  • The study was supported by unrestricted research grants from AstraZeneca, Bristol-Myers Squibb, Datascope, First Horizon, Fujisawa, GE Healthcare, Key Pharmaceutical, Kos Pharmaceuticals, Merck, Pfizer, and Sanofi-Aventis.