Making the case
When evaluating patients in the office for the workup of chest pain, it's essential to estimate the probability that they have coronary artery disease (CAD). This can be done initially through a complete clinical assessment.

Three descriptors make a patient's chest pain more likely to be angina:

• substernal chest discomfort with a characteristic quality -- pressure/squeezing/heaviness -- and duration
• provoked by exertion or emotional stress
• relieved by rest or nitroglycerin

Chest pain can be described as typical angina if it has all three of these characteristics, atypical angina if it has two, and noncardiac chest pain if there's only one or none of them.

Predicting CAD risk
In 1979, Diamond and Forester combined data from several clinical trials on a total of 4,952 patients who'd all had a coronary angiogram. They found that having typical angina, atypical angina or non-cardiac chest pain could predict the likelihood of CAD. For those with typical angina, an estimated 88.9% had CAD, compared to 49.9% with atypical angina and only 16.0% with noncardiac chest pain. This estimate can be further stratified by age and gender.1 Mr. X., who has atypical angina and is 45 years old, has a pretest likelihood of CAD of 46.1%.

This probability must be adjusted on a patient-by-patient basis. Other historic factors that may alter the risk of CAD include having vascular risk factors, particularly smoking (> 10 cigarettes/day in the past five years, or a total > 25 pack-years), diabetes or hyperlipidemia. Physical examination should note the presence of bruits, abnormal cardiac apex or murmurs, all of which can raise the pretest probability of CAD. The presence of Q-waves or ST-T changes on the ECG also increases the likelihood. Following a full assessment, the pretest probability of CAD can be classified as low (< 10%), intermediate (10-90%) or high (> 90%).

Selecting the best test
Current guidelines suggest that patients with low CAD probability be investigated by the most appropriate test according to their symptoms. An exercise stress test in these cases isn't useful, as the post-test likelihood of CAD is low, whether the result is negative or positive. Similarly, exercise stress tests aren't helpful in diagnosing CAD in those with a high pretest probability, since their post-test likelihood is high regardless of outcome. On the other hand, they can provide useful prognostic information in these high-risk patients. For people with an intermediate (10-90%) probability of CAD, an exercise stress test is recommended to diagnose CAD (provided they're able to exercise and have a normal ECG). Patients who can't exercise, or have an abnormal ECG, should have a pharmacologic stress test plus imaging.2

Cardiac computed tomography (CT) angiograms are now suitable for diagnosing CAD in certain patients. The quality of images from 64-slice CT scanners allows imaging of coronary arteries and detection of intra-arterial stenosis. The images are timed to the ECG and obtained within a single breath-hold. A recent study analyzed the coronary arteries of 70 patients by both coronary angiography and cardiac CT angiography. Compared to coronary angiography, CT had a sensitivity and specificity of 95% and 86% in detecting stenoses of 50% or greater.3

Diagnosis
Mr. X. underwent an exercise treadmill test, where he was able to complete 8 min 30 sec of the Bruce protocol and had a normal BP response to exertion. At peak exercise, he developed 1 mm of upsloping ST depression in the inferolateral leads. The test was interpreted as being equivocal for ischemia.

He was referred for a cardiac CT angiogram, which revealed patent coronary arteries with no evidence of CAD (see figure on facing page). Consequently, his physician was able to reassure him that his symptoms weren't due to CAD.

This case illustrates the importance of estimating the pretest probability of CAD in patients presenting with chest pain. It allows the physician to choose the most appropriate test for further investigation. Cardiac CT angiography is useful in patients with an intermediate probability of CAD who can't exercise, have an abnormal resting ECG, or have uninterpretable or equivocal stress tests.4 In selected patients, this test is more sensitive and specific than other noninvasive tests, compared to coronary angiography. Ultimately, it may provide more specific information on the presence and/or extent of CAD.