A growing consensus champions PET as the most effective imaging technique for myocardial perfusion imaging and points to some very clear advantages compared with SPECT for the detection of coronary artery disease (CAD). Such improvements include sharper resolution and the ability to glean meaningful myocardial blood flow reserve and coronary flow reserve measurements. However, whether PET is truly moving in on traditional CT and SPECT territory in clinical practice remains to be seen.
Adoption of PET in recent years has been steady, but slow, and the precision of PET myocardial perfusion imaging (MPI) is contingent on the capability of respective radiotracers to provide a clearer picture of disease.
PET has been considered a first-line and reimbursable option for the diagnosis of CAD since the mid-1990s, but the initial investment and subsequent operating cost is still daunting for small institutions that rely heavily if not solely on SPECT for nuclear stress testing. This is complicated by the need for onsite cyclotrons or contract use of Rubidium-82 (Rb-82) generators, but the promise of a new and potent PET tracer could level the playing field.
Moving a mountain
An estimated 7.2 million people die globally from coronary heart disease each year, according to the World Health Organization. Coronary heart disease is projected to rise from 47 million disability-adjusted life years globally in 1990 to 82 million in 2020. Invasive coronary angiography is by far the most common assessment for CAD. The downstream imaging modality from here is CT, which, unlike invasive coronary angiography, can help determine not only the location of arterial plaque but also its morphology and composition.
The vast majority of physiologic MPI for the detection of CAD is still conducted using SPECT (J Nucl Med 2013;54(2):221-228), but there are very real diagnostic advantages to using PET, despite the challenges the technology faces in its possible path to expansion.
“PET has higher diagnostic accuracy—sensitivity and specificity—than SPECT for the assessment of myocardial perfusion, function and viability,” says Zhonghua Sun, PhD, associate professor of medical imaging in the department of imaging and applied physics at Curtin University of Technology in Perth, Australia. “Myocardial perfusion PET allows assessment of left ventricular function and wall motion, which is of paramount importance for prediction of major cardiac events in coronary artery disease.”
PET has a few features that extend beyond the capabilities of conventional SPECT, including less radiotracer scatter and a greater signal to noise ratio, dual coincidence and higher counts, the oversight of attenuation correction, which is not always or even often used in SPECT, and perhaps most notably, significantly increased spatial and temporal resolution, says Marcelo F. Di Carli, MD, chief of nuclear medicine and molecular imaging at Brigham and Women’s Hospital in Boston.
“The spatial resolution of SPECT after processing is probably in the neighborhood of 12 mm,” says Di Carli. “There are newer SPECT scanners [with cadmium-zinc-telluride detectors] that have superb spatial resolution—about 8 mm. It is 5 and 6 mm with PET, but the problem is that it really depends on the tracer that is used. If Rb-82 is used, the spatial resolution is nowhere near 6 mm.”
The issue with Rb-82 is a diminished number of counts and a higher energy of released positrons. If the energy is too high, the positrons travel through matter longer, resulting in reduced spatial resolution, notes Di Carli.
Even better image quality is available with high-definition PET reconstruction software and other technology enhancements, including ‘‘motion-frozen’’ cardiac images that combine respiratory and ECG gating, corrective software for patient motion and high-definition PET that compensate for distorted data due to circular geometry of PET scanners (J Nucl Card 2012;19[Sup 1]:S38-45).
|Myocardial Perfusion Imaging vs. Coronary CT Angiography: When to Use Which?|
|Potential testing algorithm based on use of stress cardiac SPECT myocardial perfusion imaging.|
|Source: J Nucl Med 2011;52:1079-1086|
Real clinical comparison
The SPARC study, a national registry that includes 40 clinical sites across the U.S., was established for comparative-effectiveness research investigations of SPECT, PET and coronary CT angiography. Researchers found that true clinical performance between SPECT and PET was more similar than previously believed (J Nucl Card 2009;16935-948).
“PET really performed similarly, even though we initially thought that it would do better than SPECT,” says Di Carli, an author of the study. “We didn’t have the power to perform a subgroup analysis, but overall these techniques were equally effective. We are now exploring the issues of costs.”
It is difficult to compare MPI using SPECT or PET with structural CT angiography for their respective diagnostic accuracy, because they provide very different data. While CT conveys essential information about plaque deposition, it does not necessarily provide an accurate picture of ischemia or vascular viability. Physicians should choose between physiologic and structural modalities on a case-by-case basis depending on the unique needs of the patient (Appl Rad 2009;16:935-948).
A study comparing SPECT, PET and PET/CT showed that stand alone PET had the highest mean values of the trifecta of sensitivity (91 percent), specificity (89 percent) and accuracy (89 percent) for the diagnosis of coronary artery disease (Biomed Imaging Interv J 2011;7:E9).
Shortly thereafter another study was published that showed hybrid PET/CT had greater predictive value for coronary artery disease with 81 percent for CT, 86 percent for PET, and 100 percent for PET/CT (J Nucl Med 2011:52:1079-1086).
Challenges & demands
With trials like COURAGE providing a skeptical view of routine revascularization, the number of patients with chronic coronary artery or multivessel disease treated successfully with aggressive drug treatment and appropriate lifestyle changes may increase in years to come. Thus, imaging could play a more important role in preventing unnecessary revascularizations, which could theoretically offset the cost of advanced imaging systems like PET.
A follow-up study of nuclear stress tests conducted on COURAGE trial patients showed that quantitative noninvasive assessment of perfusion improved diagnostic understanding and treatment planning. Those patients who underwent nuclear imaging procedures—SPECT in this case—and who received either medical therapy and revascularization or medical therapy alone and who enjoyed significant improvement of their ischemia were less likely to suffer death or a non-fatal myocardial infarction (Appl Rad 2009;16:935-948).
Currently in the U.S., PET studies require pre-approval for reimbursement, which is not expected to become more generous any time soon.
“Nuclear medicine is under the gun and the more barriers that you put down, the more contraction you will see,” says Di Carli. “Physicians don’t have time to stay on the phone for two hours. Unfortunately, they are under pressure to improve their productivity. In many instances, we are contracting not necessarily because we are providing better care. And we may be placing very high barriers for people to get access to some tests that are quite appropriate. While you are improving appropriateness, you also are reducing access for patients who need these studies.”
PET also could provide diagnostic benefits for those patients even at lower risk for CAD. The capability to ascertain absolute myocardial blood flow and coronary flow reserve measurements provides a more exacting and objective view of multi-vessel and microvascular disease, extent of ischemia and its impact on overall cardiovascular function (J Nucl Med 2009;50:1062-1071).
“The use of absolute quantification of myocardial blood flow enables the detection of early endothelial and vascular changes affecting flow before overt disease develops as well as detecting multi-vessel or left main stem coronary artery disease in situations where there is balanced flow reduction to all coronary arteries,” says Sun. “Thus, cardiac PET is valuable in risk stratification of patients with suspected CAD, in particular for those patients with low and intermediate risk groups.”
But with concerns for costs increasingly in the subtext of decision-making, practical use of PET could be restricted to more exceptional cases.
Tailoring PET to Patients
PET becomes a more salient diagnostic choice for specific patient populations that are apt to get the most benefit from the technology.
“Although cardiac PET has grown in the last five to 10 years, it still represents a rather small segment of all cardiac imaging that one performs, not only in nuclear techniques but also other modalities,” says Di Carli. “The validity of statements about improved technical characteristics of PET over SPECT are well accepted, but if the question is whether it is better than SPECT in all patient groups, the blanket statement is that it is not. It is at least as good—it could be better—but in which patient populations?”
Obese patients might specifically benefit from the use of PET, says Di Carli, because of the difficulties with SPECT in imaging these patients with high body mass indexes. Also PET could find a niche with women, particularly those with intermediate clinical risk.
Women with low clinical risk presenting with chest pain are more often than not sent for a regular ECG stress test without imaging because a significant amount of data show that imaging this patient population does not yield significant diagnostic benefit or predictive value, says Di Carli. However, moderate and high risk patients, female or otherwise, may have more to gain.
“Patients with intermediate risk, particularly women with diabetes or multiple risk factors need imaging, because we need an approach that is more sensitive,” says Di Carli. “PET has a distinct advantage, because it has improved spatial resolutions and can potentially detect smaller areas of perfusion abnormalities and diagnose the presence of coronary artery disease.”
High-risk patients including diabetics, those with chronic renal impairment and patients with multiple risk factors also may be specifically earmarked for cardiac PET imaging.
A recent comparative analysis of currently approved PET tracers showed that using 13N-ammonia for PET MPI was not only more sensitive but also more diagnostically accurate than Rb-82. However, the primary limitation of 13N-ammonia is the need of an onsite cyclotron, whereas Rb-82 can be produced by a generator, making it the obvious choice for logistical reasons alone, but this exacts a price—not only in terms of a fixed annual cost for generators as well as image quality (Biomed Imaging Interv J 2011;7:E9).
The introduction of a new PET radiotracer working its way through regulatory approval would offer a commercially produced single-unit dose with extraction characteristics and levels of resolution comparable to 13N-ammonia. Considering 13N-ammonia has a very short half-life of only 10 minutes, F-18 Flurpiridaz stands out for its ease of use and capability to take PET resolutions beyond those of SPECT for the detection of coronary artery disease (J Nucl Card 2012;19[Sup 1]:S38-45).
““Flurpiridaz F-18 is a nearly ideal myocardial imaging tracer. Its characteristics are superior to those of Rubidium in many respects,” says Daniel S. Berman, MD, director of the department of nuclear imaging at Cedar-Sinai Heart Institute, Los Angeles. In the future, the area where PET will be potentially expanded is with this new agent currently in clinical trials.”
F-18 Flurpiridaz is currently in the third phase of clinical trials and is the only investigational PET tracer that appears to be headed swiftly to the U.S. market, according to Di Carli.
Room for Growth
With an arsenal of technical features and a new radiotracer on the way, PET has the potential to seize a larger slice of the cardiac imaging pie. Even if overall nuclear imaging remains flat or contracts, PET could soon begin supplanting some of SPECT’s caseload, even though it certainly will not happen overnight.
|Can techs order tests more appropriately than docs?|
In the case of ischemia or left ventricle infarction, both stress and rest imaging are often needed to obtain a proper diagnosis of coronary artery disease. Discerning whether just a stress test or both should be performed is typically the physician's decision, but techs may be just as qualified to make that call. Now there is proof.
A Swedish study, assessing three technologists and three physicians at Malmö University Hospital, examined initial non-gated SPECT stress test studies to discern the necessity of a follow-up rest test for 532 patients (J Nucl Med Technol 2008;36:181-185).
Myocardial perfusion imaging (MPI) was performed in 421 cases. In 107 others, MPI was conducted to assess known coronary artery disease, and four other MPI studies were performed for other reasons. Of these, 79 patients had previously suffered myocardial infarction (MI).
An experienced physician reviewed all studies to provide a standard of evaluation for ischemia and MI and the need for a rest study. A "rest-study-required-group" was established. Upon review, technologists were able to accurately identify rest-study-group cases more than 99 percent of the time (171 out of 172). The physicians identified 169. In those cases for which a rest study was deemed unnecessary, physicians were correct 32 percent of the time and technologists in 21 percent of cases.
The risk of technologists withholding a rest study from a patient who would later be diagnosed with ischemia or MI was one out of 75 patients. For physicians, it was three out of 115. The researchers concluded that educated technologists are capable of deciding independently whether a patient should undergo a rest test.
"We believe this could improve clinical workflow," says the study's lead author Lena Johansson, PhD, RN, of the department of clinical physiology at Sahlgrenska University Hospital in Gothenborg, Sweden. "The technologists do not have to wait for a decision to be made by the physicians, and the physicians will have more time to correctly interpret the images and write the final report to the referring clinician."
Physicians still had the final say, and patients were informed that they might be asked to return if their doctor wished that they undergo a rest test.
"I would be delighted if technologists were inspired to assess more nuclear medicine images," says Johansson. "It is, however, important to be properly educated before implementing a new routine."