The COURAGE Trial: Looking past the controversy to patient care

Do you think drug-eluting stents (DES) would have made a major difference? No, not a major difference. Everything always tends to balance out. Obviously, everyone is well aware of the recent concerns about late and very late stent thrombosis with DES. Our current best guess as to the rate of late stent thrombosis is somewhere in the range of 1 in 200 to 1 in 500 patients. If used in the COURAGE trial, DES may have added the adverse consequences of late stent thrombosis to the PCI group. In contrast, the benefit from DES would have been to lower the rate of restenosis and thus to lower the rate of subsequent revascularization procedures in the PCI group. So, just to speculate, there were about 1,000 patients in the PCI arm that received at least one stent. That would add 2-5 late adverse events due to DES use. There were 228 patients in the PCI group that underwent revascularization during follow-up; 77 were coronary bypass procedures. The restenosis rate in COURAGE was roughly 15%, which is consistent with what we expect with bare metal stents. Assuming a 5% restenosis rate had DES been used, there would have been about 100 fewer repeat PCIs in the PCI group. However, it is important to emphasize that the primary endpoint in COURAGE was all-cause death and MI; repeat revascularization was only a secondary endpoint. Adding a few more events to the primary outcome variable in the PCI group would not likely change the overall outcome. More importantly, the practical application of the trial to daily practice would not be changed. Whether you think the COURAGE trial is an absolutely fantastic trial or whether you think it's a terribly flawed trial, it provides a good, solid message from a contemporary evaluation of medical therapy versus PCI in patients with chronic stable angina. You can quibble about individual subgroups and dissect the trial in many ways, but the same message is still there for the patient with chronic stable angina, two very good treatment options exist. If the patient can take medications and if their symptoms are well-controlled, if they do not have troublesome side effects and they are happy with the lifestyle they experience while on medical therapy, then it is a perfectly good therapy and patients can proceed with confidence. Medical therapy does not in any way affect their mortality, which is very low, and it is not affecting their risk of heart attack compared with PCI. The alternative, if for some reason patients cannot take their medications perhaps they are in the group that has some side effects, or because of the dosing necessary to control their symptoms, they have a level of physical limitations not suitable for their lifestyle a PCI procedure is a good alternative. Patients can also choose this strategy with confidence. There's no question that putting in a stent makes the artery look better on the angiogram than with medical therapy. We are making the picture look better and relieving angina, but not particularly affecting the outcome for the patient in terms of mortality and the risk of MI. It has been known for at least 10 years now that a patient's chance of having a myocardial infarction is not directly proportional to the numerical severity of the blockage in their artery. A 90% stenosis is not more likely to lead to an MI than a 50% stenosis. In fact, it is frequently the opposite. Treating just the severity of the stenosis is a good way to relieve symptoms of chronic stable angina, but we have known for a long time that it is not going to dramatically affect someone’s chance of having a MI or ultimately, their prognosis. An extended clopidogrel regimen was incorporated midway into the trial. What if this extended regimen had been used from the beginning of the trial? The prolonged clopidogrel regimen would have caused a small shift, but I do not think it would have had a dramatic effect. Does the trial design mean that some patients received stents that were not necessary? In a truly randomized trial, yes, that’s a possibility. I don’t view that as being a bad thing. There were probably some patients that, outside of the trial, could have been treated very well with medicines, but because of the randomized design of the trial, wound up getting a stent. The opposite is also true; some patients randomized to medical could have been well-treated by a stent. That is the very nature of a randomized trial. Can you discuss some aspects of the COURAGE trial that have undergone criticism? There were about 35,000 patients screened for the trial, and subsequently about 32,000 patients were excluded. When the number excluded is so high, you always have to become concerned. In my own opinion, this doesn’t deflect the validity of the study; the findings still remain valid. What it does do is call into question how generally applicable the study is going to be to a large number of patients in the real world if so many were excluded. It is important to understand the specific patient cohort that actually underwent treatment in COURAGE. Critics have also noted that this trial included a large number of patients from Veterans Administration (VA) hospitals, which for obvious reasons, tend to have more males than females. In fact, 85% of the patients randomized were men. Then there is the fact that many of these hospitals were in Canada, and the healthcare delivery system is different in Canada. It is interesting to look at the difference in outcome between Canadian and U.S. non-VA hospitals. To be fair, the confidence intervals overlap, so this difference is not statistically significant. The composite event rate for the primary outcome variable (all-cause death and MI) was 3% lower with medical therapy compared with PCI in Canadian hospitals (hazard ratio 1.27). In contrast, the composite event rate for the primary outcome variable was 6% lower with PCI in U.S. non-VA hospitals compared with Canadian hospitals (hazard ratio 0.71). This 9% swing leads to several provocative questions and merits further evaluation. Finally, this was an intention-to-treat analysis, meaning that the patient (and their outcome) remained in the group to which the patient was originally randomized. In COURAGE, there were 46 patients in the PCI group who never got a PCI. In comparison, during the follow-up period almost one-third of those in the medical therapy group crossed over and underwent revascularization. The intention-to-treat analysis is well-accepted and avoids the known pitfalls that occur if patients are shifted from one group to another after randomization. When substantial crossover occurs, it becomes more challenging to understand the application of the study results to real-world situations. Even though I am an interventional cardiologist, I would say that it’s important to not let the desire to be critical of the study cloud over what I believe is a very important message from the trial. Over the past 20 years, there have been substantial advances in medical therapy and in the use of PCI to treat patients. PCI procedures now are much safer than 20 years ago, and medical therapy is also much better. The trial’s message is clear: for patients with chronic stable angina, there are two very good choices, and they are just about equivalent. The trial showed that for symptom relief, there was a slight favoring of PCI, but it was not dramatic. The message should be very clear for physicians as well as patients. The message I take home is that physicians should always have a complete discussion with their patients before recommending either course of therapy, whether it be medicines or PCI, and make sure that the patients understand the goals of therapy. Whichever strategy is chosen in a patient with chronic stable angina, the goal is to improve symptoms. The goals are not to prolong life, because life expectancy with chronic stable angina is already very good, and not to prevent a heart attack, because neither therapy can do that. The incidence of death and MI was equivalent between the two treatments. It’s a matter of the patient and the physician deciding what is best. How easy is it to anticipate and monitor patient compliance? How should this issue impact the physician’s choice of strategy? Compliance is a big problem. There have been a number of reports recently showing that even when given instructions, patients will not always stay on their dual anti-platelet therapy after having a stent. There are multiple different reasons why they stop: sometimes they can’t afford the medications, sometimes it’s socioeconomic or educational status. There are complicated reasons why patients may not adhere to medical therapy regimens. We see this all the time in treating high cholesterol. There are no symptoms of a high cholesterol. You don’t feel bad if your cholesterol is 350 as compared to 200. Yet we know that lowering it is certainly much better; that is unquestionable. Still, it is hard to get patients to take their medications when no symptoms exist. Some patients with chronic stable angina will have trouble taking their medication. If they are not having angina 10 or 15 times a day, it is easy to forget to take your medications. In addition, all medications have side effects. The beta blockers are very effective medicines for controlling symptoms, but at high doses they do have side effects that are troublesome for some patients. What should the interventionalist tell his or her patients when asked about their need for stents? Always have an open and factual discussion with your patient about the goal of therapy. Stents have vastly improved interventional cardiology, but they are not perfect devices. Neither is bypass surgery perfect. If you are treating a patient with chronic stable angina, make sure they understand that stenting is expected to improve their symptoms, but not prolong their life or prevent a MI. Use this as an opportunity to emphasize those therapies that can reduce the risk of MI, such as medical therapy for hyperlipidemia, or lifestyle changes that make a difference, like stopping smoking and daily exercise. Make sure the patient understands the need for dual antiplatelet therapy and the duration of such therapy after stent placement. Make sure they understand that despite receiving a stent they will likely require continued medical therapy for the treatment of hyperlipidemia, hypertension and diabetes. These are really life-long commitments to therapy. Has the general media helped or hurt patient care by how it has reported the COURAGE trial? I had the privilege of being at the press conference when the COURAGE investigators presented the trial results, and to their credit, the investigators went out of their way to emphasize the patient cohort that was evaluated in the COURAGE trial, as well as which groups of patients were not evaluated in the trial. Despite these efforts, subsequent headlines in many of the newspapers read something like: Most angioplasties unnecessary. It was a very unfortunate way to have such important information play out in the lay press. These headlines really confused a lot of patients and did not put this trial in a proper context. This is not the fault of the investigators, who tried to put the trial results in the proper context. Unfortunately, some of the media reports confused and worried patients unnecessarily. Some elements of the ongoing discussion about the COURAGE trial have centered on the nature of interventional fellow training, which has been very hospital and device-centered, rather than what might be called patient-centric. It’s a very reasonable thing to question. In a general sense, it applies to many other areas besides cardiology. It applies to surgery, gastroenterology, all sorts of different areas. Physicians have become very interested in all the new technologies available for the diagnosis of disease, and all the wonderful treatments, whether they are surgical or other less invasive treatments. At some point, there is reason that we should all go back to the basics. Within cardiology, echocardiography has become so sophisticated that some people have suggested that stethoscopes are no longer necessary. That will be a sad day in terms of doctors losing these core diagnostic skills that were demonstrated years ago by some of the giants of cardiology, such as W. Proctor Harvey and other individuals of that stature. I would agree that too much emphasis has been placed on the procedural aspects of these newer diagnostic techniques. Sometimes just talking and listening to the patient is the best therapy you can offer. The New England Journal of Medicine published an editorial discussing the COURAGE Trial results on March 26, 2007, entitled Does Preventive PCI Work? by Judith S. Hochman, MD, and P. Gabriel Steg, MD. The editorial concludes, Patients whose condition is clinically unstable, who have left main coronary artery disease, or in whom medical therapy has failed to control symptoms remain candidates for revascularization, but PCI should not play a major role as part of a secondary prevention strategy.* Do you agree? I understand what the writers were trying to address, but it is rather provocative to lead with title like Does Preventative PCI work? That said, editorials are often meant to be a bit provocative. Overall, I agree with the concluding remarks of the editorialists because they place the COURAGE trial in proper perspective. Specifically, they note that for patients who are clinically unstable, such as those with acute STEMI or those with other types of acute coronary syndromes, invasive evaluation followed by an appropriate revascularization procedure (which could be bypass surgery) is best. I also agree that for chronic stable angina, PCI does not prevent MI or death and thus does not prevent such secondary events. However, being an interventional cardiologist, I likely would have ended on a more positive note, that PCI is an effective procedure for relieving symptoms and a good option for many patients. Dr. Dehmer can be contacted at kbdavid@scai.org

* Accessed April 10th, 2007 at http://content.nejm.org/cgi/content/full/NEJMe078036v1