The COURAGE Trial: A Mid-Course Correction
It's fair to say that there was perhaps an initial overreaction when the study first came out, and at least a superficial unrest within the interventional community. Now that people have had an opportunity to read the study and digest the results, more people are embracing the trial results. Interventional cardiologists have indicated that perhaps the trial represents a much-needed mid-course correction and that it underscores the benefit of optimal medical therapy in reducing events. There's an acknowledgement even from within the interventional community that similar to the improvement and evolution of technology in catheter-based revascularization, so too have there been significant advances in pharmacotherapy over the last decade. COURAGE represents a kind of balancing, if you will, of the importance of medical therapy as well as the continued importance of angioplasty and stenting in certain subsets.
The most immediate criticism that came out in the aftermath of the study was that we didn't use drug-eluting stents (DES). Of course, DES were not available until the last six months of patient enrollment. They were not FDA-approved until April of 2003, and by that time, almost all the patients had been randomized. However, if you look at the published data from a variety of sources, one can see that there are actually no data that support the benefit of DES in terms of reducing death or myocardial infarction (MI) compared to bare-metal stents. In fact, even data derived from a large Swedish registry, published in early March in The New England Journal of Medicine, indicated that if anything, there was a significant increase in, or excess, mortality in patients who received DES compared to bare-metal stents. So I do not think that had we used DES it would have altered the primary outcome of death or MI. Perhaps it may have lessened the incidence of a repeat revascularization in the percutaneous coronary intervention (PCI) arm, which in the trial was 21%; it's possible that number may have been lower. Other than that, I don't think that the study would have measurably changed had we used DES.
Should angina relief have been a primary trial endpoint in addition to death and MI?
No. I absolutely disagree with that, because angina is a soft endpoint. All the stent trials have, almost across the board, incorporated soft endpoints. All of the stent trials have used angiographic surrogates or have used angina as part of a surrogate composite. If one looks back 20 years ago at the classic trials which evaluated bypass surgery versus medical therapy, they either used death or MI or death alone as the primary endpoint. Angina and quality of life were pre-specified secondary endpoints, and if anything, it would have weakened and degraded the science of the study had we used a soft endpoint. COURAGE had a robust endpoint and it was the correct decision to use death and MI for the primary outcome measure of this trial. Frankly, there wasn't a great deal of difference in angina, to tell you the truth. I'm not sure that even if we had used death, MI and angina, that it would have been significantly different. The differences between the two groups were modest. They were statistically significant and perhaps numerically significant, but there were not large differences in the percentage of patients who were angina-free. Of course, by the end of the study there were no differences.
Can you talk about the intention to treat analysis in the trial, where some patients eventually crossed over into the PCI group, yet were still counted in medical therapy group?
In any randomized trial, particularly in a strategy trial, inevitably there will be certain patients who, during the course of the study, particularly if assigned to medical therapy, will presumably have a recurrence or worsening of angina, or the development of ischemia, and will require, for clinically appropriate reasons, a revascularization. However, one cannot really preserve the science of the trial by utilizing the outcomes as a function of those who crossed over. The intention-to-treat principle is a sacrosanct principle in clinical trial design and reporting. It is the only way that one can, in a completely unbiased fashion, assess the comparative value of two strategies, two devices or two drugs by using the strategy to which the patient was originally assigned. The crossover rate of 32.6% in this trial was low, relative to the symptomatic status of the patients and their co-morbidity, as well as the fact that they had significant ischemia and coronary disease.
Some have said COURAGE had a low-risk group of patients. I would say, read the paper, because you can't possibly conclude that this is a low-risk group. It is an intermediate-to-high risk group of patients. These patients had angina for an average of 26 months, an average of 10 episodes of angina per week and the median values were 5 months and 3 episodes per week. Two-thirds were hypertensive, one-third were diabetic, close to 40% had a prior infarction. Seventy-one percent of patients had dyslipidemia on treatment at the time of randomization. Two-thirds of the patients had multivessel coronary disease. You needed a proximal 70% coronary stenosis in a major epicardial coronary artery to qualify, and 95% of the patients had demonstrable evidence of myocardial ischemia. Two-thirds of the patients who had undergone myocardial perfusion imaging had multiple reversible perfusion defects. These are incredibly symptomatic patients with documented ischemia and high-grade coronary disease, and I think it is remarkable that only one-third of them crossed over to the PCI group.
One thing the interventional community has played down is that during the 7 years of follow-up, among the patients originally randomized to the medical therapy-only group, two-thirds of patients never, at any point, required even a first revascularization. If you met the rigorous inclusion criteria for enrollment in COURAGE, chances were 2 in 3 that you never needed a procedure. I think that is incredibly good news for patients and physicians. What it says is that aggressive medical therapy and lifestyle intervention works. If you can avoid the procedure to begin with, then restenosis and target lesion revascularization, and all these other things that happen when you do an angioplasty, are taken off the table.
How much of this trial do you think could be extrapolated to peripheral arterial or cerebrovascular disease?
A good question. These are the next vascular frontiers. Quite honestly, I think we are undertaking an awful lot of carotid stenting and intervention, and certainly a fair amount of peripheral intervention. There is a great deal of fervor associated with computerized tomographic angiography (CTA) and non-invasive ways of detecting vascular disease. One of the fears I have about that technology is that physicians will be actually adopting it for the wrong reasons, which is to troll for occult vascular disease. Doing procedures on asymptomatic or low-risk patients is of troubling concern. If you can't demonstrate benefit in a highly symptomatic group, as we could not in COURAGE, then I would be concerned about the value of doing any kind of catheter-based procedures in patients with carotid disease or peripheral arterial disease, if they didn't have rigorous criteria met for intervention.
The implications of COURAGE are such that it may create potential research opportunities to look at this issue in other vascular beds. Maybe there needs to be a carotid artery study where patients are treated with optimal medical therapy versus revascularization, or a similar kind of trial designed for patients with peripheral arterial disease. Randomized, optimal medical therapy versus revascularization these studies have just never been done.
The hypothesis of COURAGE was that PCI plus optimal medical therapy would be superior to optimal medical therapy alone. Some have stated that the results, showing equivalency of optimal medical therapy (OMT) to PCI plus OMT, were something that was well-known in this group of patients, and that the trial was a good reminder of that fact. Why did you choose to set up the trial with that particular hypothesis?
To say that the results are not a surprise, or that COURAGE just tells us what we already knew, is patent hypocrisy. Many in the interventional community are trying to downplay the results of COURAGE because they themselves and many others have been purporting to perform angioplasty as a prognosis-altering strategy. Patients undergoing angioplasty don't do this with the mere expectation that they are doing it only to take a few less nitroglycerin tablets or that they are going to walk a few feet longer on the treadmill or walk an extra block without getting chest discomfort. I believe most patients who have undergone angioplasty felt that it is a procedure that will somehow confer a long-lasting, durable clinical benefit. That is to say, they think it will improve their survival or reduce the likelihood of having a heart attack. To interventionalists who have been saying, we never said that angioplasty alters outcomes, I would say, then why even bother to do the procedure if you haven't first tried an aggressive course of medical therapy to see if it will, in fact, reduce the ischemia? In fact, if you look at the guidelines promulgated by our two organizations, the American Heart Association and the American College of Cardiology, they consistently say to start with medical therapy, and uptitrate. Only if you have been unable to control the symptoms of angina should you consider revascularization. I would say, that is just not the way we practice in the United States. We flip that algorithm upside-down and start with revascularization.
Sadly, one of the things that has happened in the last 10-15 years is similar to the following example. An internist who has a long-standing relationship with a patient does a stress test on that patient. It is positive. They refer the patient to an interventional cardiologist, who does a cath. The coronary anatomy is defined in the cath lab, and right then and there while the sheaths are in place, the interventionalist tells the patient, Mr. Jones, you've got a 70% blockage in this artery, you have an 80% blockage in that artery, and we need to fix it right away. The poor patient, who is now hit between the eyes with that information and unable to process it, certainly is in no position to challenge the physician's judgment. They essentially have no choice but to acquiesce in that decision, which is to proceed with the angioplasty.
There has been a preemptive type of decision-making going on, without the patient having the opportunity to weigh the pros and cons, and the risks and benefits, of other treatment choices. Nor has the referring internist or the physician who has a long-standing relationship with that patient had an opportunity to weigh in on the terms of that decision-making, either. In the old days, what would happen is that the patient would have an angiogram and then would go home, and then would come back and have a subsequent procedure. At least in that context, there would be an opportunity for a time-out, for the patient to perhaps consider the pros and cons, and the risks and benefits, of bypass surgery, PCI or optimal medical therapy. Maybe what we need to do is go back to that and reestablish a pathway of decision-making that now includes the family practitioner, the internist, or whomever, a physician I think is systematically cut out of the decision-making process.
Ultimately, I think COURAGE is a big win for medical therapy, and as I said, great news for patients and physicians. It tells us that optimal medical therapy, to a degree, has leveled the playing field. We now have a viable third option for patients and physicians to consider. The biggest surprise, to my way of thinking, is the degree to which optimal medical therapy was in fact able to relieve angina. That's something no one would have predicted. I would have thought the differences between PCI and optimal medical therapy would have been significantly greater than they really turned out to be, but they were actually quite modest. The improvement from baseline to follow-up in the medical therapy arm alone was much greater than the differences between the two groups. Again, roughly two-thirds of the patients in the medical therapy arm were completely symptom-free at three years. Close to three-quarters were symptom-free at five years. I think that was an unexpected outcome.
Finally, I want to reemphasize the fact that in the medical therapy group, two-thirds of patients at no time ever required even a first revascularization procedure. It is crucial to note that now, if you decide to defer angioplasty for the chronic stable angina patient, you know you are not putting that patient in harm's way. I think one of the fears has been, in the minds of many cardiologists, now that I am confronted with this diseased coronary anatomy, how can I turn my back on the angiogram and not proceed immediately to fix what is there? I do think there has been an implicit fear and concern that if the physician did not perform angioplasty, that he might be potentially vulnerable to litigation if something were to happen. The results of COURAGE should give everyone confidence that you are no more or less at risk of dying or having a heart attack if you decide to defer PCI, as opposed to proceeding with PCI. It is important news.
Dr. Boden can be contacted at wboden (at) kaleidahealth. org.
