Guidelines on Diagnosis and Treatment of Stable Ischemic Heart Disease: Keeping Up With a Constantly Evolving Evidence Base FREE

In 1999, the American Heart Association, American College of Cardiology, and American College of Physicians (ACP)–American Society of Internal Medicine published joint guidelines on the management of patients with chronic stable angina (1). Updates in 2002 and 2007 reflected advances in diagnostic imaging, medical therapies, and revascularization techniques (2–3). Five years later, we have new guidelines from the initial organizations plus additional colleagues in nursing and cardiovascular surgery (4). The new guidelines are broad, covering both diagnosis and management of patients with this condition, now called stable ischemic heart disease (SIHD). Reading, absorbing, and applying the numerous resulting recommendations, with their nuances and caveats, will surely challenge clinicians.

To help clinicians identify the most clinically consequential recommendations, the ACP Clinical Guidelines Committee synthesized 2 summary documents, appearing in this issue (5–6), based on the corresponding multispecialty SIHD guideline. These summaries emphasize the recommendations identified as class I (strongly recommended) or class III (strongly recommended against). Concentrating on class I and III recommendations allows focus on areas with the strongest evidence but risks neglecting many common situations where patient care decisions are necessary despite the lack of high-quality evidence. For example, there is no class I or III recommendation regarding stress testing in a patient who has chest pain but whose symptoms and risk factors suggest a low risk for ischemic heart disease.

Digesting the 28 recommendations for diagnosis and 48 for management is daunting. Fortunately, the guidelines also contain algorithm figures that are clear, usable, and comprehensive. The algorithms incorporate almost every typical clinical scenario, including those for which we must rely on class IIa and IIb recommendations.

Take the example of a middle-aged man with left bundle branch block on electrocardiography (ECG). The algorithm directs us to begin by deciding whether he has unstable angina, which would warrant an expedited evaluation. If the pattern is not unstable, the algorithm guides us to collect historical, clinical, and laboratory data to identify the appropriate form of stress testing. If stress testing is contraindicated, we have the interesting options of performing coronary computed tomography angiography (CCTA) first or initiating empirical medical therapy, which, if successful, would lead to monitoring without testing.

If the patient had no contraindications to stress testing or history of coronary revascularization (that is, bypass surgery or stenting), can exercise, and has a resting ECG that would be interpretable during exercise, the algorithm leads to a treadmill ECG stress test without imaging. However, the example patient has left bundle branch block, which would make treadmill testing uninterpretable, so the algorithm recommends exercise stress testing with either echocardiographic or multiplanar nuclear scintigraphic imaging (also the appropriate test if the patient had a previous coronary procedure, even if the resting ECG is normal). If the patient cannot exercise, pharmacologic vasodilator stress with echocardiographic, nuclear, or magnetic resonance imaging or CCTA are options.

However performed, stress testing results should indicate whether ischemic disease is present and whether its characteristics are high risk. If high risk, the advice is to consider revascularization. If the cardiac catheterization laboratory will not be the patient's next destination, medical therapy is undertaken with the goal of “success” (about which, more later).

Algorithms for risk assessment, medical therapy, and revascularization are similarly useful despite necessarily including many branch points. In line with the COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial (7) and the strong evidence that optimal medical therapy is a good thing, the therapy algorithm advises reserving revascularization (and cardiac catheterization) for when “success” is not achieved despite at least 2 classes of antiangina drugs and appropriate management of diabetes, hypertension, hyperlipidemia, physical activity, and smoking. My advice: Go straight to the algorithms.

Several features of the guidelines are surprising. First, there is a recommendation for treadmill ECG stress testing without imaging in women with intermediate-probability symptoms. This may not match common clinical practice, but it probably should (8). Nuclear scintigraphic imaging is costly, may be overused (9), has been a target of governmental cost-reduction efforts (10), and has radiation-related risks (11–12). Second, there is a recommendation for initial testing with stress imaging for all patients with prior revascularization. Again, this may not match current practice but makes good clinical sense, because it tells us not only if the chest pain is “real,” but also how large a region of myocardium is at risk. Third, the discussion of CCTA and cardiac stress magnetic resonance imaging in the guidelines will please many and cause concern in others. Both methods have strong evidence regarding their test performance characteristics (13–15), but neither is widely available (especially stress magnetic resonance) and both are expensive. The relative risks of CCTA are debated, given that the patient's exposure to both radiation and iodinated contrast are considerable and only marginally less than that in invasive angiography (13). The availability of several types of stress imaging is itself a cost problem, if newer tests are done in addition to—rather than instead of—older ones. In addition, I was surprised that CT coronary calcium scoring was absent from the diagnostic algorithm. This test involves less radiation than CCTA and no contrast. A coronary calcium score of zero drastically reduces the probability of having significant coronary disease (16). Perhaps future guidelines will include this method.

The revascularization algorithms separately consider “revascularization to improve survival” and “revascularization to improve symptoms.” If, for example, a patient has significant left main stenosis, good evidence supports a recommendation to proceed with revascularization regardless of symptom status. Of course, we would not know about the left main stenosis without imaging of the coronary arteries (invasive or noninvasive). But the guideline does not clearly specify the timing of coronary angiography in the diagnostic or therapeutic process. Perhaps it should, because details of coronary anatomy can significantly influence treatment decisions. On the cost-conscious hand, however, a patient on the catheterization table who is found to have any severe stenosis may well wind up with 1 or more expensive drug-eluting stents, thereby initiating a new cascade of follow-up tests, drug therapies, and further revascularizations. Angiographers refer to this as the “oculostenotic reflex”: see a lesion, place a stent (17). Thoughtfully constructed, evidence-based guidelines, such as those summarized in this issue, may mitigate such practices. Yet, there is something compelling about knowing the anatomy.

Prevention of death and myocardial infarction are clearly necessary goals in the management of patients with coronary artery disease and define the major part of “success.” What about symptoms? Patients dislike having chest pain and taking medications. Physicians want patients to feel better. These pressures often lead patients to request nonsurgical percutaneous coronary intervention (usually involving stents) and lead invasive-interventional physicians to gladly comply. Why live with symptomatic stenosis, when you can get rid of it?

The answer, of course, lies in balancing risks, benefits, and costs—a concept useful intellectually and in policymaking but often elusive in practice. Angioplasty has both risks and costs. In patients with SIHD, angioplasty does not reduce long-term mortality and may not reduce the overall rate of myocardial infarction (18). Indeed, periprocedural myocardial enzyme release may numerically counterbalance future prevention of spontaneous infarction, although their clinical equivalence is debated. This leaves symptom relief as the only likely benefit of percutaneous coronary intervention in SIHD. But isn't the absence of angina a worthwhile goal?

These guidelines provide sound guidance for the diagnosis and management of SIHD. Practitioners can apply these with confidence but also with a caveat: Things will change. For example, the recent FAME 2 (Fractional Flow Reserve versus Angiography for Multivessel Evaluation 2) trial (19) has already ignited discussion about the value of percutaneous coronary intervention in patients who have stable coronary artery disease.

Stay tuned for more guideline updates.