Myocardial perfusion imaging (MPI) allows for noninvasive evaluation of coronary artery disease (CAD) by detecting flow-limiting disease and providing risk stratification.1-3 Despite a recent decline in popularity likely due to changing physician behavior, it remains the most frequently ordered nuclear medicine exam.4 It is important for any provider reading nuclear medicine studies to have a firm understanding of MPI acquisition and interpretation. This review of single photon emission computed tomography (SPECT) MPI, includes its indications, sensitivity and specificity, radiopharmaceuticals, protocols, and an approach to image interpretation with illustrative cases.
MPI is performed to evaluate for the presence and severity of CAD by detecting flow-limiting disease, risk-stratifying patients, and assessing and quantifying patient risk. It is most frequently ordered for perioperative cardiovascular evaluation and chest pain.
A meta-analysis has shown that a moderate to large defect on MPI is a positive predictor for postoperative cardiac events, while a normal MPI has high negative predictive value for postoperative myocardial infarction (MI) or cardiac death.5 The 2014 American College of Cardiology (ACC)/American Heart Association (AHA) guidelines for perioperative cardiovascular evaluation and management of patients undergoing noncardiac surgery is a good reference for perioperative indications.6 According to these guidelines, exercise testing with cardiac imaging is indicated for preoperative patients with elevated risk and poor or unknown functional capacity, or noninvasive pharmacologic stress testing (dobutamine stress echocardiogram or MPI) in patients with elevated risk and poor (< 4 METs) functional capacity, if the results will change surgical management.6 MPI also may be indicated in patients with suspected valvular heart disease, as valvular disease increases the cardiac risk associated with noncardiac surgery.6 Overall, a reversible defect on MPI is a more concerning predictor for perioperative cardiac events, and a fixed defect is a predictor for long-term cardiac events.6
The American College of Radiology (ACR) appropriateness criteria is another useful resource for reviewing appropriate study utilization. According to these criteria, SPECT MPI is usually appropriate in the clinical setting of chest pain suggestive of acute coronary syndrome in stable patients with no ischemic electrocardiogram (ECG) changes and negative lab markers.7 A rest-only SPECT has been shown to have a high negative predictive value for acute coronary syndrome during active chest pain, although it is less sensitive than stress SPECT after the pain has resolved.7,8 SPECT MPI is also usually appropriate for use in chronic chest pain with a high probability of CAD,9 chronic chest pain with a low to intermediate probability of CAD,10 and in dyspnea with a suspected cardiac origin.11 The ACR has given a may be appropriate grading to the use of SPECT MPI in acute nonspecific chest pain in a setting of low probability of CAD.12 This is most appropriate when there is a need to exclude CAD as a potential source in the setting of negative cardiac enzymes and negative or equivocal ECG findings. A usually not appropriate grade applies to the use in asymptomatic patients with a low to intermediate risk of CAD but a may be appropriate grade applies in asymptomatic patients with a high risk of CAD.13 Table 1 is a brief review of these guidelines.
The ACC Foundation Appropriate Use Criteria Task Force, along with other subspecialty groups, performed an appropriate use review for multiple categories of stable ischemic heart disease, which is another helpful resource but out of the scope of this review article.14
Coronary angiography remains the gold standard for CAD evaluation. A large meta-analysis study published in 2012 compiling data from 2000-2010 that compared SPECT MPI, positron emission tomography (PET) and cardiac MR (CMR) found MPI has overall sensitivity of 88% and specificity of 61% for diagnosing obstructive CAD.15 Sensitivity was similar to PET and CMR but MPI has lower specificity. MPI remains more widely used than PET and CMR, and can be used in patients in whom other modalities are limited.16
Quantification software was developed to assist with MPI interpretation that allows for quantified risk stratification. Such software, discussed more below, compares patient perfusion to a database of normals and provides quantification of the differences expressed as summed stress score (SSS), summed rest score (SRS) and summed difference score (SDS). According to one study, SSS ≤ 3 carries a 0.9% annual risk for cardiac mortality vs. 5% for SSS > 12.17
One area of low sensitivity for SPECT MPI is balanced or multivessel ischemia. This is due to the assumption that the best perfused areas are normal, leading to diffuse normalization of diffuse decreased perfusion. Increased pulmonary uptake, right ventricular (RV) uptake, decline in left ventricular (LV) ejection fraction (EF) with stress, and transient ischemic dilation (TID) are indicators of possible balanced multivessel disease.16,17
Historically, thallium-201 (Tl-201) was used in nuclear medicine to image the myocardium. Tl-201 is a potassium analogue that decays primarily via electron capture to mercury-201 emitting 69-81 KeV x-rays.18 It is actively transported into myocardial cells by sodium-potassium pumps. Thallium uptake is directly proportional to perfusion of the myocardial cells. After initial distribution, Tl-201 redistributes throughout the myocardium as a result of cellular washout and continued cellular uptake. Ischemic regions of myocardium demonstrate delayed washout, but eventually equalize with the remainder of the myocardium. Regions that do not show redistribution are considered fixed defects. With a relatively long 73-hour half-life, Tl-201 is administered in low dosages, 1.5-2.0 mCi at both rest and stress.19 With such low dosage and low energy photons, Tl-201-based imaging produces counting statistics inferior to those provided by newer agents, which have mostly replaced thallium for MPI (Figure 1). Tl-201 remains useful in viability imaging. Twenty-four-hour delayed redistribution into a fixed defect suggests viable, hibernating myocardium, which may respond to revascularization.
Technetium 99m (Tc99m)-based agents are primarily used for SPECT MPI now. It is well-known that the 140 KeV photopeak from Tc-99m isomeric transition is ideal for gamma camera imaging. For perfusion studies, the primary agents bound to Tc99m are tetrofosmin and sestamibi. Both agents are lipophilic and, therefore, enter myocardial cells by passive diffusion in a quantity directly proportional to blood flow. There they bind to proteins within mitochondria via electrostatic interactions.20 When a myocyte is damaged, these electrostatic interactions occur in much smaller numbers, resulting in far lower accumulation and reduced counts.20
Both sestamibi and tetrofosmin demonstrate efficient myocardial uptake without significant physiologic redistribution.18 Both also underestimate myocardial blood flow at stress, beginning when blood flow reaches approximately double baseline.18 Because a radiopharmaceutical is primarily removed from the myocardium by blood flow, a larger dose is required at stress when blood flow is higher to maintain adequate counts.18,21 Tetrofosmin has less hepatic uptake and faster hepatic clearance.18 Tetrofosmin is thus preferred with pharmacologic stress, as hepatic uptake is greater than with exercise.22
Recommended dosage for sestamibi is 10 mCi at rest imaging and 30 mCi at stress, and 5-12 mCi at rest, 15-33 mCi at stress with tetrofosmin. There are no provided dose adjustments for either agent in pediatric or geriatric patients, or in patients with metabolic impairment.23,24 Estimated effective radiation dose for myocardial perfusion imaging studies is 9-11 mSv for a 1-day Tc99m-based study compared to 20 mSv for a Tl-201 study.25
Before any stress test, the patient should be NPO for 4 to 6 hours to reduce blood flow to the bowel and liver. Calcium channel blockers and beta-blockers should be discontinued for 48-72 hours, and long-acting nitrates discontinued for 12-24 hours. Caffeine should not be consumed for 1 day prior to the exam.18 Myocardial stress is achieved either with an exercise treadmill test (ETT) or pharmacologic agents.
ETT involves a multistage exercise test based on a Bruce or modified Bruce protocol, during which the speed and grade of the treadmill are gradually increased. The modified Bruce protocol starts with no slope, vs. 10% for the Bruce protocol, and the speed and slope are increased at a slower rate. The patient must achieve a heart rate of at least 85% of their predicted maximum heart rate (220 beats/minute [min] minus the patient’s age) for an ETT to be adequate. The radiopharmaceutical is injected at peak stress and the patient should continue to exercise for an additional minute.
Pharmacologic stress can be used if the patient is unable to perform exercise or an ETT must be terminated before the patient’s heart rate reaches 85% of the maximum predicted. Other indications for a pharmacologic stress test include acute coronary syndrome, hypotension, a left bundle branch block (LBBB), or severe congestive heart failure. A patient with LBBB may have false positive results from an exercise stress test, typically an artifactual septal defect.
Non-nitrate vasodilators such as dipyridamole, adenosine, and regadenoson can be used for pharmacologic stress. The resultant vasodilation, typically 3 to 5 times baseline, tests coronary flow reserve.26 Minor side effects are common and include flushing, headache and dyspnea. Regadenoson (Lexiscan) is the most frequently used agent. It is a selective A2A receptor agonist that results in coronary vasodilation with fewer side effects, injected as a 400µg bolus over 10 seconds.26 Non-nitrate vasodilators should not be used in patients with second- or third-degree heart block. Caution should also be used in patients with reactive airway disease as these agents can cause severe bronchospasm.
Dobutamine has also been used as a pharmacologic stress agent. Dobutamine is a synthetic catecholamine that acts on alpha- and beta-adrenergic receptors, which increase cardiac work through positive inotropic and chronotropic effects. Dobutamine can be used as an alternative to vasodilator therapy in cases of severe asthma. Initial infusion dose is 5 ug/kg/min over 3 min. The infusion is then increased to 10 µg/kg/min for another 3 minutes. The infusion rate can be increased to a maximum of 40 µg/kg/min every 10 minutes.26 The radiopharmaceutical is then injected 1 minute after the maximum tolerable dose. Frequent side effects include chest pain and arrhythmias.
Due to the physiologic redistribution of Tl-201, imaging starts with the stress portion of the examination. While Tl-201 can be injected at both stress and rest, many will use a single injection protocol. This entails an injection of around 4 mCi of Tl-201 at peak myocardial stress with the poststress images obtained as soon as heart rate returns to normal. If the stress images are abnormal, redistribution images are obtained around the 4-hour mark. Viability images can be obtained at 24 hours if there is a fixed defect, and determination of viability will assist clinical management.
With the Tc99m-based mitochondrial imaging agents, the primary options are using either a 1 or 2-day protocol. The 1-day protocol is primarily used, as it is more convenient for patients and more efficient. The 2-day protocol allows for higher administered dosage at rest, which can be advantageous particularly in obese patients, and offers the possibility of stress-only imaging.
The 1-day protocol involves obtaining a rest study and a stress study on the same day. The patient receives 2 doses of radiopharmaceutical approximately 1.5 to 2 hours apart.26 The rest study is performed first with images obtained generally 30 to 60 minutes after injection of 8-10 mCi of Tc-99m sestamibi or tetrofosmin. After 1.5 to 2 hours, allowing sufficient clearance of the rest dose, the patient undergoes the stress portion of the study. During peak stress a second dose of radiopharmaceutical is administered, typically 25 to 30 mCi. The patient is then imaged quickly (15-20 min) after radiopharmaceutical administration. The larger second dose provides higher quality images, generally favored with the stress portion of the examination, especially for gated images.
In large patients, a small dose of radiopharmaceutical can result in poor image quality and nondiagnostic images due to increased attenuation. In these cases, a 2-day protocol can be used to allow for higher administered activity for both the rest and stress portions.26 The larger doses must be separated by a day to ensure counts originate from the desired injection. For a 2-day protocol, the stress portion is performed on the first day. The Tc-99m radiopharmaceutical is injected at peak stress and the patient is imaged quickly (15-20 min). If the stress portion of the study is normal, no further imaging is necessary. If the stress portion of the study is abnormal, the patient returns the next day to receive of a second dose of Tc-99m radiopharmaceutical at rest with images obtained 30-60 min after injection.
SPECT imaging can utilize either a 180° or 360° acquisition, with 32 or 64 stops or continuous acquisition. For nonattenuation-corrected imaging, 180° is preferred. The patient is usually supine, unless using a dedicated cardiac chair, with arms up to position outside the field of acquisition. Prone imaging can be helpful for patients with a lot of attenuation. The use of attenuation correction is occasionally essential, such as in patients with recent pacemaker placement who cannot put their arm up out of the field of view. If attenuation correction is used, it is imperative to evaluate both the corrected and uncorrected data as the correction software can create artifactual perfusion defects.
The interpretation of SPECT MPI includes four broad categories: raw data, qualitative analysis, quantitative analysis, and evaluation of motion and function.
Raw data images are generally viewed as rotating cinematic (cine) images. These images allow the interpreter to evaluate radiotracer distribution, and assess for attenuation, motion or interfering splanchnic activity. Normal physiologic distribution of the Tc-99m-based mitochondrial imaging agents includes activity within the heart, liver, gallbladder, small bowel, kidneys, bladder and sometimes the stomach. Mild uptake may be seen within the lungs and skeletal muscle as well, particularly if an exercise stress protocol is used. Uptake within the RV is normal, especially on stress imaging; however, if RV activity approaches the left, it is abnormal and may indicate RV strain. Any uptake outside of these normal structures requires attention. Particular attention should be paid to the axillary nodes and breast tissue, as the mitochondrial agents are nonspecific tumor imaging agents. In fact, Tc99m-sestamibi is used for breast-specific gamma imaging. Any breast uptake should be correlated with mammography. Uptake in the axillary nodes is often from radiopharmaceutical extravasation at the injection site, but can be seen with metastatic disease and may warrant further workup. A right upper quadrant ultrasound should be recommended if there is no uptake within the gallbladder in a patient without prior cholecystectomy (Figure 2).
Identifying attenuation on the raw cine images may aid problem-solving for defects noted during qualitative analysis. It is particularly helpful to note whether a shift in attenuation occurs between the rest and stress images as this may result in an artifactual reversible defect. Breast tissue leads to attenuation on most female patients, usually resulting in an anteroseptal defect on qualitative review. The diaphragm and lateral chest wall also frequently attenuate, resulting in defects involving the inferior base of the heart. Also, the interpreter should be aware of external devices such as pacemakers that may overlie portions of the heart and mimic perfusion defects. SPECT computed tomography (CT) can be helpful as the CT component allows for attenuation correction of the SPECT data (Figure 3). At our institution, we perform the examination as a SPECT CT if the patient has a high BMI or is unable to raise both arms above the head. Repeating the stress images in the prone position is also sometimes helpful, as a true perfusion defect will not shift.
Motion correction should be applied during image processing if there is significant motion. Assessment after correction should show no significant cardiac motion.
Splanchnic activity projecting adjacent to the heart—usually bowel or hepatobiliary activity—may limit evaluation. This “gut crosstalk” (Figure 4) usually affects the inferior wall. The high activity within the adjacent organs may either artificially add or subtract counts from the myocardium. Falsely elevated counts can mask a true defect, while falsely decreased counts can create a false defect.3 Prone positioning may increase distance between abdominal viscera and the heart, possibly limiting gut crosstalk, and adding a small exercise component to pharmacologic stress may decrease splanchnic activity. The most important way to limit gut crosstalk is to allow sufficient time before imaging to allow activity to clear.3 At some facilities, patients drink a carbonated beverage prior to imaging to help clear gastric activity.
The raw SPECT data is processed to create image slices that are parallel and perpendicular to the long axis of the heart for qualitative analysis. The horizontal long axis (HLA) is parallel to the inferior wall giving the best view of the septum and lateral wall; the vertical long axis (VLA) is parallel to the septum giving the best view of the anterior and inferior walls; the short axis (SA) is perpendicular to the long axes and parallel to the base (Figure 5). Evaluating these images for perfusion abnormalities provides the primary information clinicians need. The poststress data is the most important. If the stress data is normal, there is no evidence of ischemia or infarct. The primary purpose of rest images is to provide specificity to a stress defect. The images can be evaluated using multiple color schemes, each of which comes with an inherent associated sensitivity and specificity. Determining the color scheme is a personal preference, but requires an understanding of the statistics associated with the choice.
If present, stress perfusion defects are described by size, severity, location, extent and reversibility. Size is the amount of LV myocardium involved, described as small, moderate or large, equating to < 10%, 10% to 20%, or > 20% of myocardium. Severity is described by the percentage of counts compared to the most active area of myocardium. As the images are normalized to the “hottest” pixel, this is one of the ways “gut crosstalk” can affect interpretation. Masking the data to exclude extra-cardiac activity can help, but it is impossible to exclude activity immediately adjacent to the heart without losing real data. At our institution, anything visibly within 75% to 100% of the hottest pixel is considered normal. Mild decrease is 50% to 75%, moderate is 25% to 50%, and severe is between 0% to 25%, with 0% considered complete. Location refers to the ventricular segments/walls involved, and extent describes which slices within the segment are involved, typically broken down into base, mid slices, distal/apical slices and apex. Rest images are needed to determine reversibility, which in turn, determines the primary differential. When evaluating for reversibility, it is imperative to ensure that the rest and stress data are similarly aligned.
An example description of a perfusion defect could be, “a small, mild severity anterior wall perfusion defect extending from mid slices through the apex that is reversible.”
A fixed defect is a perfusion defect present at stress and rest. Primary differential for a fixed defect includes scarring from infarction, chronically ischemic areas called hibernating myocardium, or attenuation. Attenuation defects generally are worse at rest than stress (because approximately 3 times the dosage is given at stress with a 1-day protocol), are in a distribution that matches known areas of attenuation, should be seen on the raw images, and will have normal motion. Infarction and hibernating myocardium will look identical on SPECT MPI (Figure 6). A separate viability study can differentiate, if it is clinically necessary, as hibernating myocardium will respond to revascularization (Table 2).
A poststress perfusion defect that improves at rest is called reversible, and is concerning for stress-induced ischemia (Figure 7). As discussed, shifting attenuation leading to an area of myocardium being attenuated on the poststress images but not the rest images is a possible cause for a false positive. Ensuring the defect matches a vascular distribution and thorough evaluation of the raw data can help prevent this pitfall. SPECT-CT with attenuation correction is also an option. Areas of reversibility within a mostly fixed defect should be reported as such. An example is a fixed defect with reversibility along the margins representing an infarct with peri-infarct ischemia (Figure 8).
Computer software can help provide quantitative analysis of myocardial perfusion. Most programs use a 17-segment polar map of the LV to help identify specific areas of perfusion abnormality. The images are compared to a normalized control database. A 0-4 scoring system is used for each segment, with 0 being normal and 1-4 given for mildly decreased to completely absent counts. This scoring applies to both stress and rest data, and the numbers in each polar map are added to provide the SSS and SRS. SSS can help assess patient risk of a significant cardiac event. SSS of ≤ 3 is normal, 4-7 is low risk, 8-12 is intermediate risk, and ≥ 13 is high risk.21 The SDS is assessed by subtracting the SRS from the SSS. Any segment with an SDS of ≥ 2 is considered reversible and concerning for ischemia (Figure 9). We typically give more weight to our qualitative analysis if the quantitative results are discordant, as the quantitative data relies on appropriate contouring and is easily manipulated.
As previously noted, TID may be a sign of balanced multivessel ischemia. The term is somewhat of a misnomer as there is no actual dilation, especially as the stress images are usually not obtained until the patient has returned to resting physiology. The apparent dilation is likely due to diffuse subendocardial hypoperfusion, which can be seen with multivessel CAD, hypertensive heart disease and dilated cardiomyopathy (Figure 10). The TID ratio compares the apparent stress chamber size to the rest chamber. We use a cutoff of 1.22 for ETT and 1.32 for pharmacologic stress, at or above which is abnormal.27
With the addition of gated images, LV motion and function can be analyzed. We only obtain gated images poststress, as the statistics are better due to the higher administered activity, and the patient is already physiologically at rest at the time of acquisition. Normal LV motion results in at least 20% shortening of the long axis and 40% shortening along the short axis, with the anterior wall moving the most and the apex moving the least.18 The septum should thicken and move toward the center. Cine analysis of slices from the 3 axes of the gated images should show all walls thickening and brightening with systole. The basal one-third of the septum is membranous and should not be mistaken for a defect or motion abnormality. Apical thinning, which is anatomically normal, may also mimic a defect. If wall motion is normal, the ventricle thickens normally, and a small defect is evident on rest and stress views, it is likely normal.21 Most software will also create a 3D map of the left ventricle, highlighting the extrapolated chamber throughout the cardiac cycle.
Wall motion is either normal, hypokinetic, akinetic, or dyskinetic. Hypokinesis can be seen with areas of ischemia if the patient has not fully recovered from stress. Akinesis is usually associated with infarction or hibernating myocardium. Dyskinesis, outward movement during systole, suggests aneurysm formation at the site of transmural infarction.
After coronary artery bypass grafting (CABG), anteromedial deviation of the heart—commonly referred to as post-CABG kinetics—can be seen due to interruption of the normal fascial bands. The 3D renderings of the chamber will resemble septal hypo- or akinesis, but review of the slices will show the septum normally thickening and increasing in counts (Figure 11). A “septal bounce,” or paradoxical movement of the septum inward during diastole, is nonspecific but may appear with LV restriction (constrictive pericarditis, tamponade, or restrictive cardiomyopathy) or with LBBB (Figure 12). If no LBBB is seen on ECG, further analysis is in order. Calculated LVEF is usually provided, with 50% to 70% being normal. This is not as accurate as radionuclide ventriculography.
SPECT MPI remains a widely used modality to analyze myocardial perfusion. It is particularly useful in identifying flow-limiting stenoses and for risk stratification. Use of appropriate protocols and radiopharmaceutical dosage is paramount in obtaining diagnostic quality images. An ideal approach to image interpretation includes evaluation of raw data, qualitative analysis, quantitative analysis, and cardiac motion and function. Accurate interpretation requires a working knowledge of radiopharmaceutical biodistribution, cardiac physiology, as well as common pathologies, pitfalls, and artifacts.
Disclaimer: The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Department of the Air Force, Department of Defense, or the U.S. Government.Back To Top
Holder L, Lewis S, Abrames E, Wolin EA. Review of SPECT Myocardial Perfusion Imaging. J Am Osteopath Coll Radiol. 2016;5(3):5-13.