Revascularization for Stable Coronary Artery Disease

Abstract & Commentary

By Jonathan Abrams, MDProfessor of Medicine, Division of Cardiology, University of New Mexico, AlbuquerqueAbstract originally printed in November, 2007 Clinical Cardiology Alert. Dr. Abrams serves on the speaker's bureau for Merck, Pfizer, and Parke-Davis.

Sources: Lin GA, et al. Cardiologist's use of percutaneous coronary interventions for stable coronary artery disease. Arch Intern Med. 2007;167:1604-1609; Moscucci M. Behavioral factors, bias, and practice guidelines in the decision to use percutaneous coronary interventions for stable coronary artery disease. Arch Intern Med. 2007;167:1573-1575; Boden WE, et al. Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med. 2007;356:1503-1516.

Exceptional attention has been paid to the recent publication of the COURAGE Trial (April 2007) regarding the efficacy of percutaneous coronary intervention (PCI) in stable coronary disease patients. Whereas in acute coronary syndromes, including STEMI and unstable angina, there is reasonable consensus as to an interventional approach in patients with either PCI or CABG as opposed to medical therapy, a highly visible dialogue is now taking place regarding the appropriate approach in stable CAD. This unusual study from University of California-San Francisco reports on a series of focus groups comprised of interventional and non-interventional cardiologists in California in an effort to assess their views of PCI in stable CAD patients vs high-quality medical therapy without revascularization.

The results of COURAGE confirm that PCI for stable coronary disease does not offer a mortality or MI benefit vs aggressive medical therapy. Previous studies and metaanalyses have also shown that PCI does not improve survival or non-fatal MI reduction with a PCI strategy in stable angina patients when compared with medical therapy. The data do support earlier and greater relief of angina for PCI compared to medical therapy, but after one or more years, chest pain symptoms tend to be comparable between the strategies of PCI or medical therapy. In the United States, there is considerable geographic variation in the use of angioplasty that suggests multiple factors other than coronary anatomy influence, whether an invasive approach is utilized. In order to explore reasons for the variation in decision making, physicians from UCSF designed a qualitative study using 3 focus groups comprised of 4 to 9 cardiologists and asked the participants to discuss the issues regarding their decisions whether or not to utilize PCI in stable angina patients.

Overall, the physicians acknowledged that PCI has not been shown to reduce the hard end points of death or acute MI in chronic stable CAD patients. The "oculostenotic" reflex was cited by the physicians, who more often than not voted for PCI, even in asymptomatic patients. The 3 focus groups were similarly structured, with a moderator and 3 hypothetical case scenarios. The fictional patients all had stable CAD with either no symptoms or atypical complaints, and were felt by the study authors to represent individuals who clearly have not been shown to derive benefit from PCI. The focus groups lasted for 90 minutes and were comprised of invasive and noninvasive cardiologists, cardiology fellows, and individuals from both rural and urban environments. A systematic and detailed approach to discussion was used. New methods of triangulation were utilized to reduce bias. A summary of the major systematic discussions of each focus group was provided; the participants agreed with the summary record. Themes discussed were related to physician's medical-legal concerns and technical advances that relate to PCI. The majority believed that PCI "would benefit the patients described in our case scenarios by preventing cardiac events, even in asymptomatic patients." The participants did acknowledge that PCI was less likely than medical therapy to provide benefit in terms of preventing MI or death; however, they stressed the benefit of patients leaving the hospital with an open artery. It was also believed to be important to perform PCI for equivocal stress testing results or lesion-ischemic mismatch on stress test. Concern about not intervening on an obstructive lesion was common; the likelihood of a subsequent event was felt to be unacceptable. In general cardiologists believed that even in low-risk patients, complications due to catheterization would be more meaningful than the "potential consequences of not performing PCI". Decreased anxiety following a procedure in asymptomatic patients was commonly cited, especially in individuals who were self referred. Patients who reached the cath lab generally underwent PCI, regardless of why the patient was sent for the angiogram. Thus " ... by the time one is this far along, the die is cast. The cath lab staff probably wouldn't leave the lab unless we did something with the lesion ... " Medical-legal concerns were also an important motivation; not performing an intervention was believed by some to be a setup for a potential law suit after an event occurred.

Cardiologists agreed that available technological advances, such as electron beam CT and CT angiography, are and will be increasing the number of asymptomatic patients who are referred for studies. They believe that these patients should be treated aggressively, although no evidence is available that such an approach would prevent death or myocardial infarction. The availability of drug-eluting stents seems to have been important in supporting catheterization decisions. The authors comment, " ... the current practice of cardiologists in our sample is to recommend PCI for almost all significant lesions seen at cardiac catheterization." The cardiologists believed that they were benefiting even asymptomatic patients by performing PCI." Emotional and psychological factors were important determinants of physician judgment, suggesting an overly positive belief in PCI efficacy. The authors discuss a variety of reasons why cardiologists favor an invasive approach more than being guideline driven. Physician and patient anxiety about abnormal test results, as well as the ease in doing PCI following a diagnostic catheterization "has made PCI almost inevitable in anything with a significant lesion."

The experience of the cardiologists was comparable whether they came from rural, suburban, or urban areas. The conclusion of the authors is that "cardiologists may believe they are benefiting their stable patients ... but this belief (for PCI) appears to be based on emotional or psychological factors rather than on evidenced clinical benefits." Physicians need to find a "greater balance between emotion and beliefs and clinical evidence to provide the best treatment for patients."

In an accompanying editorial, Dr. Moscucci from the division of cardiology at the University of Michigan, discusses deviation from treatment guidelines and the potential impact of increased CT coronary angiogram availability, which will provide even more asymptomatic patients for a decision of PCI or not. Moscucci calls for the development of appropriate criteria and improved physician and patient communication.


It is likely that this subject will engender considerable comment, and hopefully, alterations in the way physicians deal with decision making. The increasing dialogue among health professionals, industry, and patients regarding invasive vs noninvasive procedures for patients with CAD hopefully will be helpful and informative for the thousands of individuals for whom decisions for or against PCI hold great importance.

The results of the University of California focus groups in dealing with PCI are of considerable interest. While the database is quite small, the authors found that there was substantial agreement among the physicians regarding various factors pro and con for the role of PCI in stable CAD. A much larger and more specific study is needed; it is likely that providers and those working in health policy will be looking carefully into this matter with great interest. One needs to expand the hypothetical case scenarios made available to the participants; the cases appear to be somewhat simplistic, and perhaps, overly amenable to a quick decision of go vs no go. The more dialogue about this matter, the better it is for our patients, physicians, and healthcare providers. It is widely acknowledged that PCI rates are extremely disparate around the United States, with 2- to 3-fold differences in utilization of angioplasty among cities. Canadian use of angioplasty and PCI is lower than in the United States on a per capita basis, yet overall survival rates do not appear to be different.

One important area that is not discussed in these two articles is the current emphasis on noninvasive or conservative therapy. Thus, the COURAGE trial has recently caused somewhat of a furor because of the finding that individuals who had active ischemia on stress testing, and angina with stable symptoms, had comparable morbidity and mortality at 5 years to those randomized to PCI. The impact of this study indicates that aggressive medical therapy and lifestyle changes without PCI resulted in equal outcomes of MI or cardiac death in individuals randomized to no angiogram vs those randomized to PCI. Of note, in COURAGE, the large diabetic cohort showed no difference in cardiac death or myocardial infarction when compared to the non-diabetics, thus emphasizing the outstanding outcomes achievable with aggressive medical therapy. Other trials are also persuasive in that optimal medical therapy in stable patients may be appropriate, at least until symptoms arise or worsen, without jeopardizing the patient. Street talk suggests a current 10-20% decrease in recent cardiac catheterization since COURAGE was reported. The drug-eluting stent controversy regarding late thrombosis may have also had a dampening effect on performing PCI. It appears that an algorithm is taking shape that provides a better fit for the patient and cardiologist with respect to how to deal with significant CAD in stable individuals.