Using radiopharmaceuticals to visualize the regional distribution of myocardial perfusion during rest and stress is a well-established modality for the evaluation of known or suspected coronary artery disease (CAD). In 1964, the first scintigraphic images of myocardial perfusion were acquired by Carrea et al.,¹ while Zaret et al. were the first to demonstrate exercise-induced myocardial ischemia using radioactive potassium in 1973.² Since then, the field of nuclear cardiology has grown dramatically, and numerous studies have validated the utility of both exercise and pharmacologic stress myocardial perfusion imaging (MPI) for risk assessment and the prediction of future cardiac events. With >8 million such studies being performed yearly in the United States alone, understanding the logistics of and options available for radionuclide stress testing is critical.³

Whenever possible, exercise is the preferred modality for stress testing, because it allows for a physiologic assessment of functional capacity, hemodynamics, and symptoms. In addition, when compared to pharmacologic stress testing, exercise is associated with less extensive hepatic and gastrointestinal tracer uptake, which significantly improves image quality.^4,5

MPI in conjunction with exercise stress testing enhances diagnostic sensitivity and specificity, particularly among patients with resting electrocardiographic (ECG) abnormalities that preclude the interpretation of ST-segment deviation. Similarly, MPI can differentiate true-positive from false-positive ST-segment depression (STD), which is helpful, because among patients referred for exercise ECG testing with a low to intermediate pretest probability of CAD, approximately 40% of those who develop STD will not have CAD.⁶ When compared to ECG interpretation in isolation, MPI not only provides a more accurate assessment of the extent and severity of disease, but it can also localize ischemia to a particular vascular distribution. MPI is also useful when patients fail to achieve their target heart rate during exercise, because myocardial perfusion abnormalities in response to stress occur earlier than ECG changes.⁷ Finally, when combined with exercise, MPI not only improves diagnostic capability, but is also predictive of short- and long-term cardiac events.⁸ This important prognostic ability does not apply to ECG interpretation without concurrent use of the Duke treadmill score (Table 8-1) or the presence of significant ischemic changes, such as ST-segment elevation (STE). Despite the clear advantages of MPI in conjunction with exercise testing, an important consideration should be made for patients who are able to achieve ≥10 METS on exercise stress testing. A recent study by Bourque et al.⁹ showed that individuals who are at intermediate risk for CAD or those who have known CAD and are able to accomplish ≥10 METS have an excellent cardiovascular prognosis regardless of peak heart rate achieved, and thus the addition and utility of MPI in this population is questionable.

Logistics and Procedures

The indications and contraindications for exercise MPI are listed in Tables 8-2, 8-3, and 8-4. Of note, the appropriate use of exercise and pharmacologic radionuclide imaging is specifically addressed in Chapter 13. Among patients who are capable of physical exercise, MPI is usually performed using one of the several standardized treadmill protocols. Individuals who are generally healthy should perform treadmill exercise using the Bruce protocol, which calls for 3-minute stages of gradually increasing speed and grade. Older individuals, or those with limited exercise capacity, can be evaluated with a modified Bruce protocol that incorporates two warm-up stages. Other protocols, such as the Naughton or Weber, use 1- or 2-minute stages with incremental 1-MET increases, and are appropriate for patients with significantly limited exercise tolerance. Cycle ergometers, while less expensive than treadmills, are infrequently used in the United States. They are unfamiliar to many patients, and often preclude maximum levels of exercise due to muscle fatigue. Regardless of the particular protocol or equipment, 6 to 12 minutes of continuous and progressive exercise produces maximal myocardial metabolic demand and is optimal for diagnostic and prognostic purposes.^6,10–12

From a procedural standpoint, patients should be instructed not to eat, drink, or smoke for 8 hours prior to exercise testing, and to wear comfortable shoes and loose-fitting clothing. Antihypertensive and antianginal medications can limit the development of ischemia and blunt the physiologic heart rate response during exercise, resulting in a lower level of sensitivity for detecting CAD. Thus, whenever clinically feasible, these medications (beta blockers, calcium channel blockers, and long-acting nitrates) should be tapered and discontinued at least 12 hours prior to exercise testing,^5,13 especially if the study is being performed for diagnostic purposes.

After explaining the logistics of the test as well as the potential risks and benefits, informed consent is obtained, after which a brief history and physical examination should be conducted in order to elicit any medical issues that may limit or preclude exercise. An intravenous line is then placed for injection of the radiopharmaceutical agent. During exercise, the heart rate, blood pressure, and ECG should be recorded toward the end of each stage, along with symptoms should they occur. When the endpoint of exercise is reached, the radiopharmaceutical is injected rapidly and followed by a saline flush, and the patient is encouraged to continue exercising for at least 1 to 2 more minutes. Continuation of exercise is crucial because it allows for myocardial extraction of radiopharmaceutical during peak blood flow and maximum ischemic stress. If necessary, the speed and grade of the treadmill can be decreased to allow for continuation of exercise. Following completion of the test, monitoring should continue for at least 5 to 10 minutes or until any symptoms resolve. The indications for terminating exercise MPI, aside from patient fatigue, are listed in Table 8-5.

ECG Interpretation

Prior to stress testing, a standard 12-lead ECG should be recorded along with blood pressure and heart rate in both the supine and standing positions, since postural changes can elicit ST–T-wave abnormalities. Hyperventilation, which can also produce nonspecific ST-segment changes, is no longer recommended as a routine prior to stress testing.¹⁴ If false-positive ST-segment changes due to hyperventilation are suspected, a hyperventilation ECG can always be obtained after the test is complete,⁶ and then compared with the maximal ST-segment abnormalities observed during exercise.

Multiple ECG changes can occur as part of the normal physiologic response to exercise, including PR, QRS, and QT interval shortening, and J point, or junctional, depression with rapid, upsloping ST segments. In the presence of underlying ischemia, the ST segment classically becomes horizontal during exercise, a finding that may be associated with angina or become more pronounced with increasing workload (Fig. 8-1). Abnormal ECG findings during exercise include ≥1 mm of horizontal or downsloping STD, and 1.5 mm of upsloping STD, all measured 60 to 80 ms after the J point.^15–17 Of these criteria, downsloping STD is the strongest predictor of underlying CAD. Precordial STD, especially in lead V₅, is more reliable for detecting CAD in patients without prior myocardial infarction (MI) and normal resting ECGs than is STD in the inferior leads, which carries a high false-positive rate.

STD may first appear after exercise is complete or persist during recovery,^11,12 emphasizing the need for continuous monitoring throughout the procedure. STD that begins after the cessation of exercise is not more significant than STD that occurs during exercise,^18–20 but when ST changes during exercise are equivocal, downsloping STD during recovery indicates a significant ischemic burden, and portends a poor long-term prognosis.²¹ In general, true ischemic STD tends to coincide with the termination of exercise, and frequently persists or intensifies for at least 2 to 3 minutes during recovery. If STD does not occur until >2 or 3 minutes into recovery, or if it occurs near peak exertion and resolves rapidly in early recovery, a false-positive response is likely.^16–22

Overall, the diagnostic accuracy of the exercise ECG in isolation is extremely variable, with an overall mean sensitivity and specificity of ~70%, based on meta-analyses of 147 consecutively published reports involving >20,000 patients who underwent both coronary angiography and exercise testing.¹⁴ The sensitivity of exercise-induced ECG changes is higher in populations with a greater prevalence of disease, such as the elderly or those with multiple cardiac risk factors. Likewise, specificity is reduced when false-positive results are likely, as when resting ECG abnormalities or confounding clinical conditions are present.

The location of the exercise-induced STD (leads) does not reliably predict the location of coronary stenoses.²³ In contrast, the rare finding of exercise-induced STE without prior MI implies a high-grade coronary lesion,^23,24 and localizes myocardial ischemia quite accurately. Exercise-induced STE in the presence of a previous transmural MI is relatively common and of debatable significance. The mechanism is unclear, but has been ascribed to wall-motion abnormalities or residual viability in the infarcted area.^25–29

Confounders of Stress ECG Interpretation

The clinical conditions associated with false-positive ST-segment responses to exercise are listed in Table 8-6. In addition to postural changes and hyperventilation, several additional factors deserve further comment. Resting STD of <1 mm in the absence of other abnormal findings is nonspecific, but it is also a relatively sensitive marker for significant CAD, and has been shown to be associated with adverse outcomes.^30–34 Thus, exercise-induced ST-segment changes can be reasonably well interpreted in this cohort,¹⁴ with the potential exception of patients taking digoxin and those with left ventricular hypertrophy (LVH). Digoxin produces abnormal ventricular repolarization and STD in response to exercise.^35–37 Similarly, the repolarization abnormalities associated with LVH decrease the specificity of any ECG changes during exercise. Thus, current consensus guidelines for stress testing recommend imaging modalities in combination with exercise among patients taking digoxin, or for those with LVH.¹⁴

Exercise-induced STD usually occurs in the presence of left bundle branch block (LBBB), and has no diagnostic significance or association with ischemia.³⁸ However, during exercise, both increased heart rate and augmented myocardial workload decrease septal blood flow in the setting of LBBB, which often results in falsely abnormal MPI. Therefore, vasodilator pharmacologic stress MPI is the preferred modality when baseline LBBB or a ventricular-paced rhythm is present. Exercise in the setting of right bundle branch block (RBBB) is associated with nonischemic anterior STD (in leads V₁–V₃) secondary to abnormal repolarization.³⁹ Nonetheless, ischemic changes can still be interpreted in the presence of RBBB using the left chest leads (V₅ and V₆) and inferior leads (II, III, and aVF) without reduced sensitivity, specificity, or predictive value. The induction of complete RBBB or LBBB during exercise is a nonspecific finding in isolation, but it may suggest myocardial ischemia if noted in conjunction with hemodynamic or clinical symptoms.¹⁶

Additional Stress Testing Parameters

Aside from exercise-induced STD, other modalities may provide supplementary clinical and prognostic information during and after stress testing. Michaelides et al.⁴⁰ evaluated the utility of right-sided precordial leads during exercise testing in a group of 245 patients, and found that the sensitivity and specificity of the ECG to detect CAD were enhanced when compared to standard leads, yielding comparable results to those obtained with stress MPI. However, there was a high pretest prevalence of CAD in their patient population, and without confirmatory studies involving larger groups, the routine use of right-sided leads during stress testing is not currently recommended.¹⁴

The degree of ST-segment displacement relative to the maximum heart rate achieved during exercise, or ST/HR index, has also been suggested as a means to enhance the detection of CAD. This measurement can be derived manually or generated by computer, although its use among symptomatic patients has been limited.^41–46 Thus, the ST/HR index has not been validated for routine use during stress testing, but it may be helpful in certain situations, such as when there is equivocal STD associated with a high exercise heart rate.^14,47 Also of note, computer processing of exercise ECG data to calculate STD is part of most standard software programs, but a significant number of false-positive findings can result.⁴⁷ Computerized scores or measurements of ST-segment deviation, while useful, should always be preceded by and compared to the raw ECG data, and never used in isolation.

The increase in heart rate during exercise is a function of parasympathetic withdrawal and sympathetic activation, while heart rate recovery immediately after exercise is mediated by reactivation of the parasympathetic nervous system. Chronotropic incompetence during exercise in the absence of rate-limiting medications, although variably defined, generally signifies significant cardiac disease, and among patients with known or suspected CAD is independently associated with higher all-cause mortality.⁴⁸ Heart rate recovery after exercise can also provide prognostic information. In a study involving 2428 consecutive patients undergoing exercise MPI, Cole et al.⁴⁹ were the first to demonstrate that a delayed decrease in the heart rate during the first minute after exercise was an independent predictor of overall mortality regardless of workload, changes in heart rate during exercise, or perfusion defects. Their findings have been independently confirmed in several subsequent studies.^50–53 Adverse outcomes are also more frequent with persistently elevated systolic blood pressure after exercise⁵⁴ or an inability to adequately augment systolic blood pressure in response to exercise.

Finally, the Duke treadmill score (Table 8-1) is a valuable clinical tool that is used for diagnosis as well as risk assessment and prognosis. It was originally devised by Mark et al.⁵⁵ using clinical and ECG data from almost 3000 inpatients with known or suspected CAD who underwent exercise stress testing prior to coronary angiography. Briefly, the formula incorporates exercise time, ECG changes, and angina to calculate a score from –11 to +5 that has been shown to be a powerful predictor of mortality. It works equally well with men and women, and has subsequently been validated in outpatients, patients at other centers, and in those with resting nonspecific ST–T-wave changes.^55–58

Even when MPI is combined with exercise stress testing, failure to attain an adequate heart rate response during exercise reduces the sensitivity of stress MPI for detecting CAD, diminishes the extent of defects seen on perfusion scintigraphy,^59–63 and can result in a substantial number (up to 25%) of false-negative results.^64–66 Many individuals referred for stress testing are elderly, and are unable to perform maximum exercise because of chronic obstructive pulmonary disease (COPD), physical deconditioning, musculoskeletal and peripheral vascular disease, previous stroke, extremity amputation, unfamiliarity with the treadmill, or simply poor motivation. Heart rate responsiveness is also frequently affected by the use of medications, particularly beta blockers. In addition, as noted above, complete LBBB or a ventricular paced rhythm can produce artifactual septal perfusion defects with exercise.

Pharmacologic stress testing is increasingly being utilized for stress perfusion imaging, and currently accounts for nearly 50% of all nuclear stress tests in the United States.⁶⁷ Pharmacologic stress MPI has been well validated when compared to exercise MPI in terms of diagnostic sensitivity, specificity, and risk stratification, and can safely be accomplished with vasodilating agents such as adenosine and dipyridamole, and recently with selective A2_a adenosine agonists, such as regadenoson. Alternatively, catecholamines, such as dobutamine, may be used, primarily in the setting of contraindications to the use of vasodilators. In general, pharmacologic stress testing should be reserved for patients who are unable to exercise adequately or who possess contraindications for exercise; the indications for pharmacologic stress MPI are listed in Table 8-7. A listing of general and specific contraindications for pharmacologic testing is noted in Table 8-8.