PET: Not yet an Essential Component of Every Modern-Day Clinical Nuclear Laboratory

Note: This is the Con article of a two-part "Pro-Con" set. Go to the Pro article.

In evaluating whether positron emission tomography (PET)-based myocardial perfusion imaging (MPI) should be the major and essential imaging modality in modern nuclear cardiology laboratories, it is useful to approach the issue from two perspectives. First, what the performance of PET MPI is compared with the more widely used and available single-photon emission computed tomography (SPECT) MPI. Second, what practical issues are involved in bringing PET MPI into a laboratory not currently performing that type of test.

From a test performance perspective, it would be challenging to argue with the concept that PET is the superior technology in almost all ways for MPI compared with SPECT:

  • The perfusion tracer emissions are of much higher energy.
  • The nature of coincidence detection in the acquisition of the emitted positrons results in better resolution and localization.
  • There is inherent attenuation correction.
  • The image quality is superior in general to SPECT and is especially so in challenging patient subsets such as obesity.
  • In analyses comparing the techniques and in meta-analyses, the sensitivity and specificity of PET to detect or rule-out angiographically obstructive coronary disease are superior to SPECT, though modestly so in absolute magnitude.1
  • Radiation exposure with PET is lower than SPECT MPI.
  • Test completion time is faster with PET.
  • Current hardware and software technology allows the potential for quantification of stress perfusion and perfusion reserve, enabling the possibility of better detection of more modest stenoses and overcoming the occasional conundrum of missing multivessel coronary artery disease with "balanced ischemia" seen with SPECT.2

So why is it that more than 25 years after the approval of the PET perfusion tracer rubidium-82 there are relatively few laboratories performing PET MPI compared with those performing SPECT? The answer to that question can be answered by considering the second point regarding the practical and cost issues of a PET MPI program.

The current iteration of PET MPI and the perfusion tracers available mandates the use of pharmacologic stress only. This means that for any patient who is capable of exercising, and especially for those patients referred for evaluation of exertional symptoms, a large amount of clinically relevant information related to exercise is lost. The prognostic power of functional capacity will be missed, as will any ability to connect symptoms with exertion and, even more importantly, connect exertional symptoms with the presence or absence of ischemia. Thus, at the moment, only patients referred for pharmacologic stress would be eligible for PET MPI without substantial data loss.

Then there are the cost issues. In order to do rubidium-82 PET MPI, a generator must be purchased monthly, resulting in a fixed cost to a laboratory of approximately $35,000 per month or more, or over $400,000 per year. It has been suggested that to break even financially, an average of 3-4 outpatients per day needs to be tested every weekday.3 Although many larger practices and community hospitals or academic centers could support that volume from among those referred for pharmacologic stress MPI, many others could not, and risk is shifted to the provider.

There is also the issue of a PET camera. At larger community hospitals or academic centers where that technology is already in place for oncology studies, camera time sufficient to perform at least 3-4 MPI studies is not likely to be available. Purchasing a new camera specifically for PET MPI can be daunting, though cost may be mitigated by combining it with computed tomography (CT) imaging in a PET/CT platform so that a broader usage can be anticipated. In any case, if one has a working SPECT camera, starting a PET program and maintaining adequate volume so as to avoid losing money requires shifting patients from SPECT to PET, resulting in an idle SPECT camera. One might justify this on the basis of the potential to increase throughput, but most laboratories no longer have long waiting times to book a case because referral volumes have generally decreased in recent years based on the proliferation and application of Appropriate Use Criteria and concerns regarding radiation exposure. Thus, in many laboratories, acquiring a PET camera for faster MPI imaging times will not necessarily result in incremental volume. Moreover, in a hospital-based setting, shifting inpatients from SPECT to the higher cost PET technology will undermine any margin within the diagnostic-related grouping payment system for inpatients.

Finally, there is the complex issue of reimbursement. If Ben Franklin were around today, to his oft-quoted statement " this world nothing can be said to be certain, except death and taxes..."4 he would have added the certainty of ongoing reduction in reimbursement. The costs of the camera and the high, fixed, continual costs of providing the perfusion agent required for PET MPI create an ongoing risk for many laboratories.

So who should be doing this? Nuclear cardiology laboratories in large practices or community hospitals or academic centers with a high volume of MPI can make this work, as a good number do with well-trained personnel doing high-quality work. But the volume requirements, the relative lack of well-trained PET perfusion imagers, and the daunting (at the moment) economics are all likely to continue to constrain the much wider availability of this technology. The potential availability of the F-18 tracer flurpiridaz, which can be delivered in unit doses, may address some of these issues.5 Even so, it is not likely to approach the modest current cost of SPECT tracers. It is also important to note that SPECT technology is not sitting still. Recent years have seen new camera technologies allowing high-quality SPECT studies with lower tracer doses at faster imaging times with lower radiation exposure.6

In cardiology practice, there are many examples of a technology that is clearly superior on a technical or performance basis but does not completely or even predominantly displace the lesser technology. For example, cardiac magnetic resonance imaging is superior to echocardiography for many applications and provides much more precise information about structure and function among other aspects, but it is not "essential" in every modern day outpatient laboratory; few cardiology practices have this technology on site, and practitioners can find it if they need it. SPECT MPI is clearly superior to stress electrocardiography testing, yet Appropriate Use Criteria recommend exercise electrocardiography as first line testing for symptomatic low-risk patients who can exercise.7

In much the same way, PET MPI is not essential in every modern laboratory that does high-quality SPECT work. Its marginally better performance is accompanied by many challenging practical issues, particularly at start-up. It is a "nice to have" but not a "need to have." There are certainly emerging applications such as imaging valve infection and tracking disease activity in sarcoidosis,8 but these can best be done in referral centers.

In a laboratory where a SPECT system was approaching end-of-life, would I jump into PET as an essential new direction? I would certainly seriously consider it given the better performance, but I would carefully analyze whether the referral volume could support its use and consider the current local reimbursement environment with some (admittedly highly unstable!) projections of where reimbursement may be going.


  1. Mc Ardle BA, Dowsley TF, deKemp RA, Wells GA, Beanlands RS. Does rubidium-82 PET have superior accuracy to SPECT perfusion imaging for the diagnosis of obstructive coronary disease?: a systematic review and meta-analysis. J Am Coll Cardiol 2012;60:1828-37.
  2. Di Carli MF, Lipton MJ (eds). Cardiac Pet and PET/CT Imaging. New York: Springer Science+Business Media LLC; 2007.
  3. Godt E. PET: Making the Numbers Work (Cardiovascular Business website). 2012. Available at: Accessed 06/24/2016.
  4. Death and taxes (idiom) (Wikipedia website). 2016. Available at: Accessed 06/24/2016.
  5. Maddahi J, Packard RR. Cardiac PET perfusion tracers: current status and future directions. Semin Nucl Med 2014;44:333-43.
  6. Garcia EV. Physical attributes, limitations, and future potential for PET and SPECT. J Nucl Cardiol 2012;19(Suppl I);S19-29.
  7. Wolk MJ, Bailey SR, Doherty JU, et al. ACCF/AHA/ASE/ASNC/HFSA/HRS/ SCAI/SCCT/SCMR/STS 2013 multimodality appropriate use criteria for the detection and risk assessment of stable ischemic heart disease: a report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, and Society of Thoracic Surgeons. J Am Coll Cardiol 2014;63:380-406.
  8. Di Carli MF, Geva T, Davidoff R. The Future of Cardiovascular Imaging. Circulation 2016;133:2640-61.

Clinical Topics: Heart Failure and Cardiomyopathies, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Atherosclerotic Disease (CAD/PAD), Interventions and Coronary Artery Disease, Interventions and Imaging, Angiography, Computed Tomography, Echocardiography/Ultrasound, Magnetic Resonance Imaging, Nuclear Imaging

Keywords: Constriction, Pathologic, Coronary Angiography, Coronary Artery Disease, Echocardiography, Electrocardiography, Electrons, Magnetic Resonance Imaging, Myocardial Perfusion Imaging, Physical Exertion, Positron-Emission Tomography, Rubidium, Sarcoidosis, Tomography, Emission-Computed, Single-Photon, Tomography, X-Ray Computed

< Back to Listings