Myocardial perfusion imaging with PET offers several advantages over SPECT for the detection and characterization of coronary artery disease. With PET, individual tissue-density measurements are routinely used to correct the emission images for photon attenuation and scatter by interposed tissue. And because PET scanners have higher spatial resolutions and count sensitivities than SPECT scanners, PET images can discriminate more readily between areas with normal and abnormal perfusion. As a result, PET imaging has a better diagnostic accuracy for coronary artery disease detection, with reported sensitivities approximately 4%–5% higher and specificities 3%–5% higher than for SPECT imaging (1,2). Moreover, left ventricular function can be assessed during vasodilator stress with PET imaging, permitting measurement of ventricular contractile reserve and identification of ischemia-related deterioration in regional function (3,4). It is also feasible to assess rest and hyperemic myocardial perfusion (in milliliters of blood flow/min/gram of tissue) and to derive myocardial perfusion reserve measurements from dynamic PET
See page 1581
perfusion images using commercially available software (5).This assists in the detection of “balanced coronary artery disease” and in the identification of microvascular disease (6). On a practical level, patients are exposed to significantly less radiation with PET imaging than with SPECT imaging (typically less than 5 mSv for a rest and stress PET study (7)). Finally, a rest–vasodilator stress PET myocardial perfusion study can be completed in a quarter of the time required for a SPECT perfusion study with a 99mTc-labeled tracer.
WHY PET IS NOT USED MORE OFTEN FOR CLINICAL IMAGING
Given the advantages of PET, why is it not used more frequently for clinical myocardial perfusion imaging? One reason is cost. PET/CT scanners are more expensive than SPECT or SPECT/CT scanners, both to purchase and to maintain. Currently, clinical PET myocardial perfusion imaging is performed using either 82Rb or 13N-NH3. 82Rb is eluted from a bedside generator, and there are periodic ongoing costs associated with generator replacement when the generator reaches the end of its service life (usually between 2 and 8 wk). Generator life is determined by its manufacturing characteristics, not its number of uses. As a result, costs per 82Rb imaging procedure are reduced by maximizing the number of imaging studies performed during the useful life of the generator. 13N-NH3, the other tracer used for clinical PET perfusion imaging, is cyclotron-produced and has only a 10-min half-life. If this tracer is to be used for clinical perfusion imaging, the cyclotron has to be near the imaging center. In addition, close coordination between personnel at the imaging center and personnel at the cyclotron is required to ensure that the 13N-NH3 is available at the time it is needed for the stress injection.
Aside from the economic considerations, both 82Rb and 13N-NH3 have other limitations. 82Rb is impractical for exercise stress perfusion imaging because of its short 75-s half-life. Gastrointestinal background activity may be high, even in fasting patients. Images obtained with 13N-NH3 typically exhibit prominent background activity in the liver and may also show prominent pulmonary uptake in many cases. Moreover, some patients without coronary artery disease may exhibit relatively low tracer uptake in the inferolateral region of the ventricle on 13N-NH3 images, possibly because of genetic differences in tracer retention (8,9).
ADVANTAGES OF AN 18F-LABELED PERFUSION TRACER
Because of the limitations of 82Rb and 13N-NH3, a perfusion tracer labeled with 18F for PET imaging is attractive clinically. The half-life of 18F is almost 110 min, which is long enough to permit the transport of unit doses of 18F-labeled perfusion tracers from a regional cyclotron to a PET imaging center. Therefore, either exercise or vasodilator stress myocardial perfusion imaging could be performed at centers that are presently performing only 18F-FDG PET/CT imaging for oncology patients. The extra cost of adding myocardial perfusion imaging to the case mix for such a center would likely be modest if unit doses of an 18F-labeled perfusion tracer were available at a reasonable price. If 18F-labeled perfusion tracers were available as unit doses, more PET imaging centers could perform cardiac perfusion studies. These studies could be performed with or without 18F-FDG for the assessment of myocardial viability (e.g., as part of a 2-d imaging protocol), thereby increasing the numbers of patients with access to PET myocardial imaging studies.
Because the positron emitted by 18F is relatively low-energy, travel distances in tissue before annihilation are significantly shorter than with 82Rb and somewhat shorter than with 13N-NH3. The result is myocardial images that are more sharply defined than those obtained with the other perfusion tracers. Moreover, as initial experience with 18F-flurpiridaz suggests, 18F-labeled tracers may offer better target-to-background ratios than the present tracers, yielding higher-quality images (10).
18F-LABELED FLUOROALKYLPHOSPHONIUM SALTS FOR PERFUSION IMAGING
In a preclinical study appearing in this issue of The Journal of Nuclear Medicine, Kim and colleagues assess the suitability of 3 18F-labeled fluoroalkylphosphonium salts for PET myocardial perfusion imaging (11). Similar to 99mTc-sestamibi and 99mTc-tetrofosmin, these moieties depend on high mitochondrial membrane potentials for retention in cardiac myocytes. In the current study, (5-18F-fluoropentyl)triphenylphosphonium cation (18F-FPTP), (6-18F-fluorohexyl)triphenylphosphonium cation (18F-FHTP), and (2-(2-18F-fluoroethoxy)triphenylphosphonium cation (or 18F-FETP) were compared with 13N-NH3 in Sprague–Dawley rat hearts.
In studies on isolated Langendorff perfused hearts, the authors found higher first-pass extraction fractions for all 3 18F-labeled tracers than for 13N-NH3 at flow velocities exceeding 4.0 mL/min. A higher first-pass extraction fraction indicates that net myocardial uptake of the tracer will more closely parallel tissue perfusion at high flow rates. On perfusion images, therefore, slight differences in hyperemic flow rate should be more readily detectable than on 13N-NH3 images, suggesting that PET imaging with the new perfusion tracers will prove more sensitive for detecting moderate coronary stenoses. The authors also performed dynamic PET imaging on normal rats and rats with infarctions using a small-animal PET/CT tomograph. Areas of infarction were well defined on the 18F-labeled perfusion images. Moreover, at 10 min after tracer injection, myocardium-to-liver ratios were 3–5 times higher for the 18F-labeled images than the 13N-NH3 images, and myocardium-to-lung ratios were approximately 2–3 times higher. Target-to-background ratios were therefore considerably better with the newer tracers, resulting in better image quality. These preclinical studies thus indicate that the 18F-labeled phosphonium cations are promising PET myocardial perfusion tracers.
CHALLENGES AHEAD
Several major hurdles must of course be overcome to transfer the findings of a promising preclinical study into daily practice. Safety and biodistribution studies are a prerequisite to use in humans, and a well-designed hierarchy of clinical trials using appropriately constructed imaging protocols is necessary to confirm efficacy and secure Food and Drug Administration approval for clinical use. 18F-flurpiridaz, another tracer of myocardial perfusion, is in stage III clinical trials and may be the first 18F-labeled PET perfusion tracer to be approved by the Food and Drug Administration for clinical practice. Nevertheless, the new 18F-labeled fluoroalkylphosphonium tracers may also one day prove useful for human studies and could provide an additional option for clinical PET perfusion imaging. Ultimately, it is hoped that cardiac care will benefit from access of greater numbers of patients to PET myocardial perfusion imaging.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
Footnotes
Published online Jul. 30, 2015.
- © 2015 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication June 26, 2015.
- Accepted for publication July 10, 2015.