Version July 2025
INTRODUCTION —
Radionuclide myocardial perfusion imaging (MPI) involves the visualization of a radiopharmaceutical distributed throughout the myocardium in proportion to coronary blood flow, thereby permitting the determination of relative blood flow in various heart regions. Regional coronary blood flow (delivery) determines the amount of tracer activity within a specific area; a close correlation between flow and activity has been demonstrated with available radiopharmaceuticals over a physiologic range of coronary blood flow.
Perfusion imaging depends upon the radiolabeled tracer's physical properties, delivery, and extraction and retention by the myocyte. Both cell membrane integrity and energy utilization are necessary for intracellular extraction and retention of tracer. Thus, retained tracer activity reflects myocyte viability. Revascularization of such segments can lead to improvement in left ventricular function. (See "Evaluation of hibernating myocardium" and "Treatment of ischemic cardiomyopathy".)
The ideal perfusion agent would have the following characteristics:
●High first-pass myocardial extraction
●Linear relationship between uptake and flow
●Uptake independent of metabolic state
●Stable distribution during imaging
Several radiolabeled myocardial perfusion agents are available. Understanding these agents' basic properties is necessary to choose the proper agent for each particular clinical setting and to accurately interpret radionuclide myocardial perfusion images (table 1). (See "Assessment of myocardial viability by radionuclide imaging in coronary heart disease".)
SPECT TRACERS —
Single photon emission computed tomography (SPECT) is the most commonly used myocardial perfusion imaging technique. Thallium-201 and technetium-99 tracers can be used in SPECT MPI.
Thallium-201 — Thallium-201 (Tl-201) is a radioactive element similar to potassium analogs first used in perfusion imaging but with superior imaging characteristics.
General characteristics — Tl-201 has the following general characteristics:
●It is cyclotron-produced and, therefore, requires off-site manufacturing.
●The principal photopeaks are gamma rays at 135 keV (2.7 percent) and 167 keV (10 percent), and mercury X-rays of 69 to 83 keV (85 to 90 percent) [1].
●The physical half-life of thallium-201 is prolonged (73 hours), limiting the overall amount that can be administered to 2 to 4 mCi.
●Thallium uptake is partly an active process involving the Na+/K+ ATPase pump. Due to the relatively small contribution of active transport, extraction and uptake of thallium is relatively unaffected by ischemia, hypoxia, or digoxin, and is directly proportional to coronary blood flow [2-5].
●The relationship between flow, as measured by microspheres, and myocardial thallium activity is linear and proportional up to at least twice the resting flow rates [5,6].
●Thallium uptake may overestimate flow within regions with low flow rates, and underestimate flow with high flow rates [6,7].
Thallium-201 has the following advantages:
●Myocardial uptake is proportional to flow.
●Redistribution enables a single injection for both stress and rest images.
●It is considered the "gold standard" for viability assessment among single photon agents. (See "Assessment of myocardial viability by radionuclide imaging in coronary heart disease".)
However, it also has two disadvantages:
●Low photon energy results in more scatter and soft tissue attenuation
●Longer physical half-life limits the allowable dose, which limits image quality
Redistribution — Following thallium's initial extraction, there is a continuous exchange between the myocyte and the extracellular compartment, resulting in redistribution. Intake of thallium into the cell continues via additional extraction of thallium that remains in the blood (arterial input) and recirculation of tracer that has already been washed out of the intracellular compartment. Thallium may leak out of various regions within the myocardium at different rates based on coronary blood flow. Thus, an area with higher coronary flow may permit the egress of thallium at a faster rate than a region of low flow, demonstrating a "differential washout" of thallium.
Redistribution often begins as early as 20 minutes following thallium administration (figure 1) and may result in the partial or total resolution of perfusion defects noted shortly after stress imaging [8]. Thus, poststress imaging should begin within 15 minutes after the initial injection. The redistribution process enables thallium to assess the relative regional coronary blood flow following stress and under "resting" conditions after a single infusion of tracer [7].
However, the perfusion estimate is different at three to four hours after thallium delivery than under actual resting conditions, and it can overestimate the extent of myocardial necrosis. This potential limitation may be overcome by a second injection of a smaller dose of thallium immediately following the redistribution images ("reinjection"). (See "Assessment of myocardial viability by radionuclide imaging in coronary heart disease".)
Thallium redistribution can be affected by several factors:
●Changes in coronary blood flow can substantially impact redistribution, although tracer may be redistributed even without a significant alteration in flow.
●The administration of nitrates appears to facilitate the appearance of thallium redistribution and, consequently, the recognition of reversible myocardial ischemia [9].
●Redistribution may be affected by metabolic interventions that alter the transport of thallium (eg, ribose administration), potentially permitting earlier and more obvious determination of redistribution after an ischemic event [10].
Reverse redistribution — Many authors have commented on a finding called "reverse redistribution," in which the perfusion defect appears worse on the delayed redistribution perfusion images (three to four hours after thallium injection) than on the initial images. This phenomenon is thought to be related to hyperemic blood flow that causes enhanced thallium uptake on initial poststress images [11], and a more rapid clearance of thallium, producing the appearance of a perfusion abnormality.
Reverse redistribution is consistent with viable myocardium [12]. However, its presence in a patient with a low likelihood of ischemia and without other evidence of ischemia is felt to represent an artifact.
Technetium-99 labeled agents — There are several physical advantages of technetium-99m (Tc-99m) perfusion tracers (table 2):
●The higher photon energy (140 keV) is well suited for gamma camera imaging and may result in less photon attenuation and scatter due to soft tissue compared with thallium-201. (See "Artifacts in SPECT radionuclide myocardial perfusion imaging".)
●A half-life of six hours and favorable dosimetry permits the administration of substantially more activity, resulting in a high number of emitted photons and improved image resolution. The increased photon flux also enables functional imaging with gated SPECT or first-pass techniques.
●Tc-99m is a generator-produced product, so it is readily available (on site) at most institutions.
Two 99m Tc-labeled myocardial perfusion agents are routinely used in clinical practice (sestamibi and tetrofosmin). Each compound has unique properties that make it suitable for certain types of imaging (table 1).
Sestamibi — Tc-99m sestamibi (Cardiolite) is in a class of compounds known as isonitriles. Sestamibi has the following general characteristics:
●It is a lipophilic monovalent cation with transient hepatic uptake, minimal lung uptake, and an excellent target-to-background ratio [13].
●The exact mechanism of myocardial uptake of sestamibi is unclear. Uptake appears to be passive across the plasma and mitochondrial membranes with sestamibi being sequestered in the mitochondria by the large negative membrane potential [14]. In experimental models, sestamibi uptake is not blocked by ouabain. Thus, unlike thallium uptake, sestamibi uptake is not dependent upon the Na+/K+ ATPase pump [3,15].
●The fraction of sestamibi extracted on first pass by the heart is lower than that for thallium [16].
●Distribution within the myocardium is proportional to blood flow until approximately two to three times the resting flow levels when uptake plateaus [17].
●Animal models show that sestamibi myocardial activity overestimates coronary flow at low flow rates, as determined by radiolabeled microspheres [17]. This finding may be attributed to increased extraction of tracers at low flow and, as mentioned above, has also been seen with Tl-201.
Tc-99m sestamibi has the following advantages:
●Myocardial uptake is proportional to flow
●There is stable retention of tracer (minimal redistribution)
●Simultaneous perfusion and function assessment is possible
Sestamibi undergoes minimal redistribution over time [17,18]. This stable distribution has numerous ramifications for the use of this agent:
●Sestamibi is well suited to the prolonged acquisition times associated with tomographic imaging.
●The absence of significant sestamibi redistribution necessitates two separate injections of the radiopharmaceutical, one during peak stress and a second while at rest.
●The lack of sestamibi redistribution permits greater flexibility in scheduling since imaging is not mandated immediately after injection.
●Sestamibi can be used to quantify the area of risk in a patient suffering an acute myocardial infarction. As an example, sestamibi can be injected at the time of presentation, and the patient imaged after stabilization in the coronary care unit. Since sestamibi activity reflects myocardial perfusion during injection, defects on these early images indicate the "area at risk." Subsequent injections and imaging after interventions (eg, thrombolysis) can then be used to demonstrate the amount of salvaged myocardium [19].
●Sestamibi can be given to patients with chest pain and nondiagnostic findings on ECG, and images taken later, even after treatment of the chest pain (image 1). This method permits assessment of coronary blood flow during chest pain, enabling improved triage and management. As an example, a patient with a normal perfusion study after being injected with sestamibi during an episode of chest pain may be triaged to a less intensive care setting or even home, depending upon the clinical circumstances.
●The stable retention and high-count rates of Tc-99m sestamibi permits the evaluation of left ventricular function as part of a perfusion imaging protocol. Both gated SPECT and first-pass imaging have been performed successfully with this agent [20]. This evaluation may be one of sestamibi scintigraphy's most valuable clinical attributes.
Although its lack of redistribution can limit the use of Tc-99m sestamibi for evaluating myocardial viability and detecting hibernating myocardium [21-24], data suggest that quantitative analysis of sestamibi activity is helpful in predicting functional recovery of dysfunctional ventricular segments after revascularization [21]. In addition, nitrate administration before tracer injection improves the detection of viable myocardium by reducing the perfusion defect size and intensity in most viable regions [22-24]. (See "Assessment of myocardial viability by radionuclide imaging in coronary heart disease".)
Tetrofosmin — Tetrofosmin (Myoview) is a lipophilic, cationic diphosphine compound. It has good myocardial uptake with rapid clearance from the liver, lungs, and blood [25,26]. The mechanism of uptake into the myocytes is not clear but appears to involve potential driven diffusion across the sarcolemmal and mitochondrial membranes [27].
Important features of Tc-99m tetrofosmin include:
●Myocardial uptake is proportional to flow. Myocardial uptake of tetrofosmin is proportional to microsphere confirmed blood flow until uptake plateaus at flow rates greater than 2.0 mL/min/g [28].
●There is minimal tetrofosmin redistribution. As with sestamibi, there is little myocardial washout over time. Thus, high quality myocardial images can be obtained from five minutes to several hours after injection.
●There may be more rapid washout from the liver than with sestamibi, permitting earlier imaging after injection. The more rapid clearance from the liver with tetrofosmin may lead to less artifact from subdiaphragmatic activity, which is a common problem with Tc-99m sestamibi [29].
Although there is less widespread experience with tetrofosmin compared with the other available agents, the diagnostic accuracy of stress Tc-99m tetrofosmin imaging appears to be similar to that of Tl-201 and Tc-99m sestamibi [29-31]. Given the many similarities between tetrofosmin and sestamibi, these two agents can be used in clinically similar situations:
●Exercise tetrofosmin imaging is of prognostic value in patients with coronary heart disease, adding incremental information to that provided by clinical and exercise data [32,33]. In a multicenter review of 4278 patients, the mortality rate in those with a normal exercise or adenosine tetrofosmin SPECT study was 0.6 percent per year, similar to published rates for normal thallium and sestamibi studies [32].
●Some studies utilizing vasodilator stress have shown that, compared with thallium and sestamibi, tetrofosmin detected fewer ischemic segments, and the extent and severity of reversible defects were smaller. As a result, tetrofosmin may be less sensitive in identifying mild coronary stenoses [34,35].
●Tetrofosmin appears to be useful for the detection of viable myocardium [36-38]. In one report, its ability to predict functional recovery after percutaneous revascularization in patients with a prior myocardial infarction was comparable to that of thallium and sestamibi [38].
●First pass imaging is feasible, and gated SPECT tetrofosmin imaging can also be performed [39].
●Preliminary studies have revealed that tetrofosmin is of clinical utility in patients who present to emergency departments with chest pain and nondiagnostic electrocardiograms [40].
PET TRACERS —
Positron emission tomography (PET) imaging differs from single photon emission computed tomography (SPECT) in that PET uses radionuclides that decay by positron emission. PET imaging relies on the detection of two 511 keV gamma photons. Typical PET tracers have short half-lives; they must be synthesized and imaged in a short period of time, which usually requires the presence of an on-site cyclotron or a generator close to the scanner. The half-lives of the two commercially available radiotracers limit the ability to use these tracers for exercise-based stress protocols. Like SPECT tracers, PET tracers must be easily accessible, cost effective, and have a linear extraction over a wide range of myocardial flow.
Compared with SPECT, PET has several advantages that contribute to superior image quality and higher diagnostic accuracy.
●High spatial, temporal, and contrast resolution.
●Lower radiation (2 to 3 milliSieverts for PET radiotracers compared with 11 to 22 milliSieverts for 99m-technetium and thallium-201, respectively).
●Attenuation correction (attenuation correction is optional with SPECT imaging, and it is mandatory with PET) results in improved image quality by accounting for soft tissue-related attenuation.
●Quantitative assessment of absolute myocardial blood flow and myocardial flow reserve. SPECT radiotracers, with the exception of thallium, plateau at typical normal stress coronary blood flow values.
●Assessment of left ventricular function closer to peak stress (especially Rb-82).
The most commonly used cardiac PET tracers are rubidium-82 (Rb-82), nitrogen-13 (N-13), oxygen-15 (O-15), and fluorine-18 (F-18). PET imaging is based on two concepts: myocardial flow and myocardial metabolism. Rb-82, N-13 ammonia, O-15 water, and flurpiridaz F-18 all assess myocardial flow, while fluorodeoxyglucose (FDG) assesses myocardial metabolism.
Quantification of myocardial blood flow at rest and stress enables assessment of myocardial flow reserve (MFR). In both PET and SPECT imaging, perfusion imaging may underestimate the degree of coronary artery disease due to the challenge of "balanced ischemia" in patients with multivessel disease. MFR calculation, globally low in multivessel disease, increases the sensitivity of these studies.
Tracers such as F-18 flurpiridaz and O-15 have the closest to linear tracer uptake relative to blood flow, making them especially attractive in determining MFR. However, the more commercially available radiotracers Rb-82 and N-13 accurately assess MFR, aiding in diagnosing ischemia, microvascular dysfunction, and multivessel coronary artery disease. (See "Overview of stress radionuclide myocardial perfusion imaging", section on 'PET imaging'.)
Rubidium-82 — Rubidium-82, like thallium, is a potassium analog. It depends on the Na+/K+ ATPase pump for cellular uptake, and its myocardial extraction is similar to thallium. The following characteristics make Rb-82 a clinically useful cardiac PET imaging tracer:
●It is generator produced
●Its short half-life enables high throughput for clinical pharmacologic stress testing but does not allow exercise testing
Unlike thallium, rubidium is generator-produced, allowing it to be manufactured on site. It is eluted from the decay of strontium-82 in a generator, which must be replaced approximately every four weeks. Rb-82 decays by positron emission with a half-life of only 78 seconds. Although this short half-life makes the logistics of imaging challenging, it also enables rapid completion of a series of rest and stress imaging studies. The cost of the generator is often a fixed monthly expense, making it economically feasible for larger volume centers and those without cyclotron availability.
Rb-82 based PET imaging has been compared with SPECT imaging with all of the various SPECT tracers in numerous studies and several systematic reviews [41-45]. Generally, Rb-82 based PET imaging has shown equivalent, if not greater, overall sensitivity, specificity, and accuracy for the detection of coronary disease [43,45]:
●Sensitivity – 84 to 93 percent for PET versus 88 percent for SPECT
●Specificity – 81 percent for PET versus 61 to 76 percent for SPECT
In a small series, Rb-82 has been shown to identify multivessel disease [46]. When used to measure myocardial flow reserve, Rb-82 PET can predict the likelihood of multivessel disease or microvascular disease [47]. Several studies in both patient volunteers and animals have compared Rb-82 with N-13 ammonia and shown that Rb-82 can be useful in quantifying myocardial blood flow [48,49].
Fluorine-18
F-18 fluorodeoxyglucose — Using F-18, one can create FDG, which is a radiolabeled glucose analog used for metabolic and perfusion imaging, enabling one to distinguish ischemic myocardium from nonviable myocardium [50]. The half-life of F-18 is 110 minutes, and it must be produced in a cyclotron.
The use of FDG was initially reported in 1986 for the assessment of myocardial viability [51]. In patients with poor ventricular function, FDG-PET showed an increased benefit over thallium or sestamibi imaging in predicting improvement of ventricular function [52]. Since that time, there have been increasing data on the prognostic utility of FDG after coronary revascularization [53].
F-18 flurpiridaz is a novel perfusion radiotracer that shows promise in phase 3 clinical trials [54]. Compared with SPECT imaging, the sensitivity and specificity were higher, especially in females and patients with obesity.
F-18 flurpiridaz — F-18 flurpiridaz has drawn attention due to its longer half-life (110 minutes) compared with tracers using Rb-82, N-13, or O-15, making it more suitable for cardiac perfusion imaging. With its longer half-life, there is more flexibility with exercise-based stress protocols, and it can be distributed from a centrally located cyclotron center rather than requiring an on-site cyclotron.
Additionally, due to its physical properties, such as a shorter positron range, it is notable for producing higher spatial resolution images compared with other PET imaging agents, such as Rb-82 and O-15 [55].
F-18 flurpiridaz enters the cell by passive distribution and is concentrated in the mitochondria. Based on animal studies, the heart-to-liver ratio at one hour was high, making it a favorable myocardial imaging agent in its ability to produce high-quality images [56]. F-18 flurpiridaz also exhibits features important for an effective myocardial perfusion agent: flow-independent first-pass extraction and high heart-to-surrounding-tissue uptake. Flow-independent first-pass extraction enables assessment of quantitative myocardial blood flow, a distinct advantage over relative blood flow in detecting balanced ischemia and microvascular dysfunction.
Several studies have shown the safety and diagnostic performance of this agent. A small study of 12 patients showed high myocardial-to-background uptake and a favorable safety profile [57]. In phase 2 and 3 studies, its side-effect profile was found to be noninferior if not superior in comparison with SPECT tracers [54,58,59].
Nitrogen-13 — As nitrogen is an essential component of basic amino acids, N-13 enables labeling amino acids without altering their physiologic properties. These amino acids are important components of myocardial protein synthesis and metabolism in both normal and diseased states. The neutral component diffuses across the cell membranes, reequilibrating with ammonium and becoming trapped intracellularly. In vitro studies have shown that in ischemic myocardium, more N-13 becomes trapped within myocytes with a diminished tracer washout [60].
N-13 enables the gating of both rest and stress images in addition to the quantification of myocardial blood flow. In small studies, quantitative myocardial blood flow with N-13 was reproducible and accurate compared with estimates using Rb-82 [61]. Its use is limited by its short half-life (10 minutes) and required cyclotron production; thus, coordination between a cyclotron operator, a radiochemist, and a stress testing technologist is required. However, its short half-life results in decreased radiation exposure to the patient.
N-13 PET perfusion and flow reserve have long-term prognostic value. Individuals with abnormal coronary flow reserve measured by N-13 PET had significantly higher rates of major cardiac events over a five-year follow-up [62].
Oxygen-15 — O-15 has a short half-life (2.1 minutes), requiring the use of an on-site cyclotron. Due to its ability to freely diffuse across the cell membrane, it is considered the gold standard for myocardial flow quantification [63]. This same characteristic is what limits its clinical utility. Image quality is less than ideal due to low-count statistics arising from its rapid clearance from the myocardium. It is not approved for clinical use in the United States.
SUMMARY AND RECOMMENDATIONS
●General considerations – Although an ideal myocardial perfusion agent has not been discovered, several single photon agents are available with differing physical properties and applications (table 1). By understanding the characteristics of these radiopharmaceuticals, appropriate selection of the agent may be accomplished and imaging protocols optimized.
●Thallium-201 – This agent has the following advantages and disadvantages:
•Myocardial uptake is proportional to flow.
•Redistribution enables a single injection for both stress and rest images.
•It is considered the "gold standard" for viability assessment among single photon agents. (See "Assessment of myocardial viability by radionuclide imaging in coronary heart disease".)
•Low photon energy results in more scatter and soft tissue attenuation.
•Longer physical half-life limits the allowable dose which reduces image quality.
•Newer agents, including single photon emission computed tomography (SPECT) and positron emission tomography (PET) radiotracers, have largely replaced the use of thallium, although thallium still plays a role at small centers for viability assessment when PET imaging is not available.
●Technetium-99m (Tc-99m) – These perfusion tracers have several advantages (table 2):
•The higher photon energy (140 keV) is well suited for gamma camera imaging and may result in less photon attenuation and scatter due to soft tissue compared with thallium-201. (See "Artifacts in SPECT radionuclide myocardial perfusion imaging".)
•A half-life of six hours and favorable dosimetry permits the administration of substantially more activity, thereby resulting in a high number of emitted photons and improved image resolution. The increased photon flux also permits functional imaging with gated SPECT or first-pass techniques.
•Tc-99m is a generator produced product, and as a result, it is readily available (on site) at most institutions.
●PET – PET imaging differs from SPECT in that PET uses radionuclides that decay by positron emission. PET imaging relies on the detection of two 511 keV gamma photons. Typical PET tracers have short half-lives; they must be synthesized and imaged in a short period of time. The most commonly used cardiac PET tracers are rubidium-82 (Rb-82), nitrogen-13 (N-13), oxygen-15 (O-15), and fluorine-18 (F-18). PET imaging is based on two concepts: myocardial flow and myocardial metabolism. Rb-82, N-13 ammonia, F-18 flurpiridaz, and O-15 water all enable an assessment of myocardial flow, while F-18 fluorodeoxyglucose (FDG) enables an assessment of myocardial metabolism.
●Comparison of agents – For the routine detection and evaluation of coronary artery disease, thallium and technetium-99m agents have proven efficacy and are comparable in terms of diagnostic accuracy [64]. The technetium-99m agents provide improved image quality and more easily permit simultaneous assessment of ventricular function at a lower radiation dose. However, the choice of agent and protocol will depend upon the needs and requirements of the particular nuclear cardiology laboratory (eg, scheduling, availability).
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Thomas Holly, MD, who contributed to earlier versions of this topic review.
