Review

Developments in Nuclear Cardiology: Transition from Single Photon Emission Computed Tomography to Positron Emission Tomography/C

*Martin A. Lodge, PhD, *Henning Braess, PhD, *Faaiza Mahmoud, MD, *Jongdae Suh, MD, *Nancy Englar, RN, *Sandra Geyser-Stoops, BCNP, *Jason Jenkins, CNMT, ¥Stephen L. Bacharach, PhD, *Vasken Dilsizian, MD
*Martin A. Lodge, PhD, *Henning Braess, PhD, *Faaiza Mahmoud, MD, *Jongdae Suh, MD, *Nancy Englar, RN, *Sandra Geyser-Stoops, BCNP, *Jason Jenkins, CNMT, ¥Stephen L. Bacharach, PhD, *Vasken Dilsizian, MD
The advent of single photon emission computed tomography (SPECT) in the late 1970s and PET in the 1980s dramatically changed the clinical utility of radiotracer techniques for the assessment of myocardial ischemia and viability. Recent advances in instrumentation of multichannel spiral CT permit detailed (noninvasive) visualization of the coronary arteries as an adjunct to SPECT and PET techniques. The hybrid PET/CT technology has already been adopted in the clinical practice of oncology. Anatomic lesions identified by CT can be further characterized by PET metabolism, which can differentiate benign or inflammatory processes from malignancy. Similar to the widespread use in oncology, functional assessment of myocardial blood flow (SPECT or PET) in the setting of coronary artery lesions (multichannel CT) has the potential to be equally essential in the practice of cardiology. The ability to determine coronary artery disease (CAD), myocardial perfusion, viability and ventricular function from a single hybrid PET/CT study will become a powerful diagnostic and prognostic tool. Similarities and Differences Between SPECT and PET Both SPECT and PET utilize radionuclide tracer techniques that produce images of the in vivo radionuclide distribution using measurements made with an external detector system. Like CT, the resulting images represent cross-sectional slices through the patient, but with SPECT and PET, the image intensity reflects organ function rather than anatomy. The functional information depicted in SPECT and PET images depends upon the radiopharmaceutical employed for that particular study. SPECT allows for the noninvasive evaluation of myocardial blood flow by extractable tracers such as thallium-201 and technetium-99m (Tc-99m) labeled perfusion tracers.1 PET, on the other hand, allows for the noninvasive assessment of regional blood flow, function and metabolism using physiological substrates prepared with positron-emitting isotopes of such elements as carbon, oxygen, nitrogen and fluorine. These isotopes have half-lives that are considerably shorter than those used in SPECT and typically must be produced in close proximity to the scanner. Most medically important positron-emitting isotopes (Table 1) are produced using a cyclotron, and the requirement for such a machine has been a significant limitation of PET. Sites without a cyclotron are generally limited to fluorine-18 (F-18) tracers, as this has a 110-minute half-life and can be shipped from a central production facility to remote imaging sites. Exceptions to this include Rb-82, which is produced using a generator in a similar manner as Tc-99m. Radioisotopes commonly used with SPECT emit gamma rays of varying energies and have relatively long physical half-lives. Localization of gamma rays emitted by single photon-emitting radiotracers in the heart is accomplished by an Anger scintillation camera (gamma camera), which converts the gamma rays to light photons via sodium iodide scintillation detectors. Radioisotopes used for SPECT are limited to those that emit gamma rays with an energy range that is suitable for the gamma camera such as thallium-201, Tc-99m and iodine-123. The spatial resolution of SPECT systems is in the range of 10–14 mm. Although clinically useful, estimates of relative myocardial blood flow by SPECT are significantly affected by attenuation artifacts. The characteristic property of radioisotopes used in PET is that they can reach a more stable configuration by the emission of a positron (Figure 1). Positrons are positively charged particles with the same rest mass as electrons and are emitted from the parent nucleus with a range of energies. They travel a short distance before they interact with other atoms in the body, losing energy and changing direction. These scattering interactions are repeated until the positron comes to rest and collides with an electron. The result is complete annihilation of both the positron and the electron and conversion of the combined mass to 2 gamma rays, 511 keV each, that are emitted almost exactly 180? apart. Conservation of momentum, which is almost zero immediately before annihilation, ensures that the photons are emitted in opposite directions and conservation of energy ensures that they always have 511 keV (the rest masses of the positron and electron). Because the gamma rays are perfectly collinear and travel in opposite directions, the PET detectors can be configured to register only photon pairs that strike opposing detectors at approximately the same time (coincidence detection). This results in improved spatial (5–7 mm) and temporal resolution. In the most common design, no information is assumed about the position of the annihilation event along the line of response, and information is only measured in one direction. Over the course of a typical scan, thousands of coincidence events are recorded and projections of the activity distribution are measured at all angles around the patient. These projections can subsequently be used to reconstruct an image of the in vivo radionuclide distribution using the same algorithms as those used in X-ray CT. Advantages of PET Over SPECT While the basic principles of PET are similar to those of SPECT, PET systems are generally more sensitive than SPECT systems, have better spatial resolution and provide the possibility of more accurate attenuation correction. The consequence of these advantages with PET is the possibility for quantification of tracer concentration in absolute units. Attenuation correction. Nonuniform attenuation of photons between their point of emission within the patient and their escape from the body gives rise to characteristic artifacts which are particularly problematic in myocardial perfusion SPECT.2,3 Although similar artifacts can occur in PET, the back-to-back nature of the emissions means that they can be accurately corrected using a separate measurement of tissue attenuation. This transmission scan uses data from an external photon source, which rotates about the patient and can be routinely performed as part of the PET scan. As long as the patient does not move during the scanning procedure, cardiac PET images will be free from attenuation artifacts, removing a major limitation of SPECT. Similar approaches have been attempted to correct attenuation artifacts in SPECT, but these have not been widely adopted because the problem of attenuation correction is fundamentally more challenging in SPECT than in PET. In PET, the total attenuation factor is the same along any given line of response, irrespective of where the annihilation event took place, because both back-to-back photons have to escape the body. In SPECT, photons originating from different depths within the patient will experience different levels of attenuation depending on their position relative to the surface of the patient. Even with gamma cameras that have the ability to acquire transmission data, the advantage of current attenuation correction implementations has not been consistently demonstrated.4 Spatial resolution. The spatial resolution of PET is significantly better than that of SPECT (typically 5–7 mm with PET and 10–14 mm with SPECT). Furthermore, the resolution of SPECT detectors degrades rapidly with distance. This means that when the heart is viewed from one angle, the data can be recorded with much poorer resolution than from another leading to image distortion and nonuniformity. Although the spatial resolution of PET is generally better than SPECT, it is dependent on the distance traveled by the positron between its point of emission and annihilation in tissue. This distance is a function of the isotope concerned and is one of the reasons why Rb-82 (PET myocardial blood flow tracer) images do not usually exhibit significantly higher resolution than corresponding SPECT images. Sensitivity. In SPECT, a collimator is situated immediately in front of the detector and is essential to provide positional information, although it dramatically reduces photon sensitivity. With PET, no such collimator is required and the resulting sensitivity gains mean that PET images can typically be acquired in a shorter period of time. Exact scan durations vary according to institution and protocol, but typical N-13 ammonia (PET myocardial perfusion tracer) acquisitions may last 5–15 minutes, whereas typical Tc-99m sestamibi (SPECT myocardial perfusion tracer) acquisitions may last at least 15 minutes, and many more counts are usually acquired in the PET study, resulting in less noisy images. The advantages discussed above mean that PET images are not only of higher quality than SPECT, but can be accurately quantified in terms of local activity concentration (MBq/ml). In addition, most clinical PET scanners are full-ring devices that simultaneously measure projections at all angles, whereas SPECT cameras require time to rotate the detector heads about the patient. This is significant because it enables PET to image rapidly-changing processes, which, in turn, allows the possibility to quantify such things as myocardial perfusion in absolute as opposed to relative terms. PET Instrumentation Although some dedicated PET scanners employ sodium iodide detectors (the same material used in gamma cameras), most tomographs use detector materials with a higher stopping power for high-energy 511 keV photons. Bismuth germanate (BGO) has been widely validated for cardiac applications, although many scanner designs have adopted lutetium oxyorthosilicate (LSO) and gadolinium oxyorthosilicate (GSO). LSO and GSO sacrifice a lower stopping power for 511 keV photons for greater energy and timing resolution. When combined with 3-D acquisition geometry, these crystals can result in less noisy data and shorter scan durations, particularly in whole-body oncology applications. Work is ongoing to determine whether similar benefits translate to cardiac studies where count rates, and thus detector deadtime, can be higher. While the trend in the design of dedicated PET tomographs is toward scanners that operate without inter-plane septa, certain manufacturers support the option to scan either with (2-D) or without (3-D) septa. The advantage of 3-D acquisition in terms of increased sensitivity is counterbalanced by increases in scatter and random photons that do not contribute useful information and increases in detector deadtime that reduce effective sensitivity. The advantage of either mode remains controversial and is related to the scanner design, the properties of the detector material and the particular application. Three-dimensional acquisition on high-count-rate performance BGO-based systems have proven comparable to 2-D acquisition for myocardial perfusion studies with Rb-825 and quantitative studies have further validated 3-D applications with F-18 fluorodeoxyglucose6 and oxygen-15 (O-15) water.7 Note that detector deadtime issues may mean that administered doses need to be reduced for scanners without septa and protocols should be carefully optimized based upon the count rate characteristics of the particular tomograph. Scanners that have no septa and operate only in 3-D mode can support a larger patient port without increasing the diameter of the detector ring. Under this design, the space previously occupied by the septa, is now unused and available to enlarge the patient port. Increasing the diameter of the detector ring reduces sensitivity and increases the number of detectors required. Whereas older, whole-body PET scanners had a patient port of around 60 cm or less, modern scanners typically have a 70 cm diameter aperture. Although this represents a significant improvement, the limited size of the patient aperture still poses problems when imaging obese patients. Germanium-68 (Ge-68) transmission sources8 have been traditionally used for PET attenuation correction as it is a long lived (271 days) isotope that decays to the positron emitter Gallium-68 (Ga-68). Measuring the attenuation experienced by the annihilation photons within the body using an external source with the same photon energy gave rise to highly accurate corrections. However, the activity of the Ge-68 source was limited by detector deadtime issues and transmission scan times of up to 10 minutes are sometimes required. Cesium-137 (Cs-137) is a single-photon emitter (662 keV) that has been effectively used as an alternative to Ge-68 with acceptable accuracy. Shielding on the detector side of the source allows greater activities of Cs-137 to be used, resulting in shorter transmission scan times of around 1.5 minutes for cardiac studies. Combined PET/CT Scanners Combined PET/CT scanners9 provide a tool for obtaining complimentary anatomical and functional information in a single imaging session. CT angiography provides information on the presence and extent of luminal narrowing of coronary arteries, whereas PET provides information on the functional effects of such lesions. The CT is much better suited to determine whether a stenosis is present. The PET, on the other hand, is more suited to determining whether such a stenosis actually requires interventional treatment. The combination of these two modalities is particularly relevant in patients who have an intermediate finding on either PET or CT. The advantage of the combined scanner is that the corresponding images are spatially aligned and both data sets can be acquired at a single imaging session. Subsidiary benefits include the improved accuracy with which the heart can be positioned within the PET field of view. In addition, appropriate corrections for the energy difference between CT x-rays and annihilation photons have allowed CT images to be used for PET attenuation correction.10 This has allowed the replacement of Ge-68 or Cs-137 transmission scans with faster CT scans, reducing the overall duration of the scanning procedure. One potential problem of using fast CT scans for attenuation correction, however, is the motion of the organs during respiration. The CT scanner “freezes” the heart, lungs and liver at one point in the respiratory cycle, while the PET emission data are an average over many respiratory cycles. Methods to correct this problem are currently under investigation. The effective dose from multichannel CT angiography ranges from 7 to 13 mSv,11 whereas the effective dose for rubidium PET is approximately 5.5 mSv (based upon data in 12) if the maximum allowable activity of 60 mCi is administered at both rest and stress. This falls to 2.75 mSv if the activity for each of the rest and stress phases of the study is decreased to 30 mCi. With current advances in 3D PET instrumentation, diagnostic quality gated rubidium PET images can now be acquired using only 30 mCi at both rest and stress, as illustrated in Figure 4. Gated Myocardial Perfusion PET The ability to acquire cardiac PET images in conjunction with electrocardiogram (ECG) gating (Figure 2) is another important development that has not always been available, particularly on 3-D scanners. Some systems support ECG gating via an acquisition mode referred to as “list-mode.” In such a mode, the positions of all coincidence pairs are recorded along with timing information and input from an ECG machine. These data can be retrospectively processed to produce ECG-gated images, ungated images and, if necessary, dynamic images that represent the activity distribution as a function of time. The flexibility of this mode of acquisition is particularly convenient for quantitative analysis. PET Quantification of Myocardial Blood Flow Conventional SPECT nuclear medicine approaches identify flow-limiting coronary artery stenoses by delineating the relative spatial distribution of myocardial blood flow at rest and again during stress. While this approach has been clinically useful, assessment of only the relative spatial distribution of myocardial blood flow has several serious limitations. First, it does not permit quantitative comparisons between a region at rest and the same region at stress. Second, it may fail to identify patients with: 1) balanced reduction in coronary artery blood flow, 2) diffuse, nonocclusive luminal coronary artery narrowing, or 3) an occlusive lesion in the region with the highest radiotracer uptake by SPECT. A significant advantage of PET over SPECT is in the noninvasive measurement of regional myocardial blood flow in units of milliliters of blood per minute per gram of myocardium, or at least in some absolute unit that can be compared across patients. Although the distribution of a radiotracer may be homogeneous throughout the left ventricular myocardium, absolute myocardial blood flow may be abnormal. Because the majority of acute coronary events originate in coronary arteries without distinct angiographic stenosis, identification of pre-clinical atherosclerosis could prove clinically important. Quantitative measurements of absolute myocardial blood flow with PET might thus identify patients who are at risk for future acute coronary events and thereby provide a strong rationale for aggressive medical and/or therapeutic interventions. PET Perfusion Tracers Flow agents can be divided into two types: freely diffusible tracers and soluble, microsphere-like substances that are trapped in the myocardium. Usually, only freely diffusible tracers provide the ability to carry out repeated studies in rapid sequence. The rapid physiologic washout clears the tracer from the myocardium, permitting a subsequent study to be performed. O-15 labeled water is an example of such a freely diffusible tracer. However, by their very nature, the images of the distribution of such tracers usually do not give clinically meaningful images. Rather, mathematical modeling13 must be performed (Figure 3) at each pixel in order to determine parametric images of flow. Such computations often produce noisy images, especially in the case of O-15 water. An advantage of freely diffusible tracers, however, is that, unlike the microsphere-like tracers, they do not depend on a metabolic-trapping mechanism (a mechanism that might change as a function of a changing metabolic environment or drugs). So-called soluble, microsphere-like tracers are usually easier to scan because the radioactivity is fixed in place for a reasonable length of time and therefore optimal counting statistics can be obtained. However, the persistence of binding usually prevents these tracers from being used in sequential studies. In addition, the amount of binding may depend on a biochemical interaction, and this interaction could change as a function of disease and confound the flow measurement. Nitrogen-13 (N-13) ammonia falls into this category and can be used to produce both high-quality clinical images and quantitative measures of myocardial perfusion.14 Although N-13 ammonia produces excellent perfusion images, it has a 10-minute half-life and, like O-15 water, requires an on-site cyclotron, which currently limits its widespread application. Rb-82 also falls more closely into this second category of flow tracers. However, its short half-life (75 seconds) means that any trapped Rb-82 quickly disappears from the myocardium by physical decay, thus reducing radiation exposure to the patient and permitting repeated studies to be performed. In addition, it has been shown (in animal studies) that its extraction fraction does not change significantly over a wide range of metabolic conditions and is not altered by many drugs, including those that affect sodium-potassium pump function (Rb retention is of course rapidly affected by cell membrane disruption). Despite its short half-life, Rb-82 is easily obtained, as it is generator-produced. The relatively long-lived Sr-82/Rb-82 generator (typically 4-week “shelf life”), puts Rb-82 into the same class as F-18. Namely, it can be used clinically without the need for an onsite cyclotron. In Search of the Ideal Clinical Perfusion Tracer: Rubidium-82? Rb-82 is a cation, the intracellular uptake of which across the sarcolemmal membrane reflects active cation transport. In experimental studies, myocardial uptake of rubidium reflects absolute blood flows up to 2 to 3 ml/gm/minute. However, net uptake of rubidium plateaus at the hyperemic flows often achieved with pharmacologic stress. Nonetheless, qualitative assessment of relative rubidium perfusion defects have correlated well with those obtained from microspheres. Clinically, rubidium PET has both high sensitivity and specificity for detecting CAD15 (Figure 4). As a result, clinical assessment of myocardial perfusion with rubidium PET has received approval by the U.S. Food and Drug Administration and Center for Medicaid and Medicare Services. Although possessing many favorable physiologic and physical properties, Rb-82 is a difficult tracer to image. It emits an unusually high-energy positron, which can travel a considerable distance in tissue before annihilating, thereby reducing image quality. Equally important, its short half-life is both a benefit and a drawback. The dosimetry of rubidium and the feasibility for multiple sequential studies are both helped by its short half-life. However, the short half-life means that little time is available to create an image. This problem is further exacerbated because 1–2 minutes must elapse before imaging is begun in order to allow for clearance of the tracer from the blood. This usually necessitates injecting large doses of rubidium in order to get sufficient counts in the short imaging time available. Although the first few minutes after the infusion of Rb-82 are not usually included in clinical acquisition protocols, it is precisely this period that is of interest if myocardial perfusion is to be quantified. Dynamic imaging of the heart during this time allows for analysis of the Rb-82 concentration in both arterial blood and myocardial tissue as a function of time (Figure 5). The kinetic behavior of Rb-82 in tissue can be described by a two-compartment model16 (Figure 6), which can be fitted to the patient time-activity curves. The parameters of the model, which include flow, can be estimated using nonlinear regression to obtain values that minimize the least-squares difference between the model and the patient data. The large number of free parameters and the high noise levels frequently encountered in Rb-82 images mean that simultaneous estimation of all parameters cannot be performed reliably.17 The variability of flow estimates can be reduced by fixing certain parameters to physiologically realistic values, but the fact that the extraction fraction of Rb-82 is flow-dependent remains a challenge for accurate quantification. Semi-quantitative indices of flow such as dividing the mean tissue uptake over a certain period by the integral of the blood concentration, may prove more practical for routine use. Many of the error sources approximately cancel out when flow reserve is calculated and first-order corrections can be applied for the variable extraction fraction of Rb-82.18 Conclusion While contrast coronary angiography (invasive or noninvasive) provides information on the presence and extent of anatomical luminal narrowing of epicardial coronary arteries, stress myocardial perfusion SPECT or PET provides information on the downstream functional consequences of such anatomic lesions. In addition, nuclear perfusion techniques may provide insight into the regional effects of microvascular endothelial dysfunction and impairment of regional coronary blood flow reserve. As such, nuclear approaches provide complimentary functional and physiologic information to the angiographic data, which is not necessarily redundant or duplicate. With the advent of hybrid PET/CT systems, such complimentary information can be realized immediately at the same imaging session. The combination of these two modalities is particularly relevant in patients who have an intermediate finding on either PET or CT. The implication of this hybrid PET/CT imaging approach is more appropriate utilization of resources and enhanced efficiency of healthcare delivery.
References
1. Dilsizian V. SPECT and PET Techniques. In: Dilsizian V, Narula J (eds). Atlas of Nuclear Cardiology. Braunwald (series edition). Philadelphia: Current Medicine, Inc. 2003, pp 19–46. 2. Links JM, Becker LC, Rigo P, et al. Combined corrections for attenuation, depth-dependent blur and motion in cardiac SPECT: A multicenter trial. J Nucl Cardiol 2000;7:414–425. 3. Hendel RC. Attenuation correction: Eternal dilemma or real improvement? Q J Nucl Med Mol Imaging 2005;49:30–42. 4. Wackers FJ. Should SPET attenuation correction be more widely employed in routine clinical practice? Against. Eur J Nucl Med Mol Imaging 2002;29:409–412. 5. Knesaurek K, Machac J, Krynyckyi BR, et al. Comparison of 2-dimensional and 3-dimensional 82Rb myocardial perfusion PET imaging. J Nucl Med 2003;44:1350–1356. 6. Lubberink M, Boellaard R, van der Weerdt AP, et al. Quantitative comparison of analytic and iterative reconstruction methods in 2- and 3-dimensional dynamic cardiac 18F-FDG PET. J Nucl Med 2004;45:2008–2015. 7. Schafers KP, Spinks TJ, Camici PG, et al. Absolute quantification of myocardial blood flow with H215O and 3-dimensional PET: An experimental validation. J Nucl Med 2002;43:1031–1040. 8. Bailey DL. Transmission scanning in emission tomography. Eur J Nucl Med 1998;25:774–787. 9. Townsend DW, Carney JPJ, Yap JT, et al. PET/CT today and tomorrow. J Nucl Med 2004;45:4S–14S. 10. Koepfli P, Hany TF, Wyss CA, et al. CT attenuation correction for myocardial perfusion quantification using a PET/CT hybrid scanner. J Nucl Med 2004;45:537–542. 11. Hunold P, Vogt FM, Schmermund A, et al. Radiation exposure during cardiac CT: Effective doses at multi-detector row CT and electron-beam CT. Radiology 2003;226:145–152. 12. Cardiogen-82 package insert, Bracco Diagnostics, Inc., Princeton, New Jersey, 1999. 13. Iida H, Kanno I, Takahashi A, et al. Measurement of absolute myocardial blood flow with H215O and dynamic positron emission tomography: Strategy for quantification in relation to the partial volume effect. Circulation 1988;78:104–115. 14. Kitsiou AN, Bacharach SL, Bartlett ML, et al. 13N-Ammonia myocardial blood flow and uptake: Relation to functional outcome of asynergic regions after revascularization. J Am Coll Cardiol 1999;33:678–686. 15. Go RT, Marwick TH, MacIntyre WJ, et al. A prospective comparison of rubidium-82 PET and thallium-201 SPECT in myocardial perfusion imaging utilizing a single dipyridamole stress in the diagnosis of coronary artery disease. J Nucl Med 1990;31:1899–1905. 16. Herrero P, Markham J, Shelton ME, Bergman SR. Implementation and evaluation of a two-compartment model for quantification of myocardial perfusion with rebidium-82 and positron emission tomography. Circ Res 1992;70:496–507. 17. Coxson PG, Huesman RH, Borland L. Consequences of using a simplified kinetic model for dynamic PET data. J Nucl Med 1997;38:660–667. 18. Yoshida K, Mullani N, Gould KL. Coronary flow and flow reserve by PET simplified for clinical applications using rubidium-82 or nitrogen-13-ammonia. J Nucl Med 1996;37:1701–1712.