Patient-centric performance and interpretation of positron emission tomography /computed tomography myocardial perfusion imaging: a clinical consensus statement of the European Association of Cardiovascular Imaging of the European Society of Cardiology

Abstract

Positron emission tomography (PET) is the most advanced myocardial perfusion (MPI) technique for the non-invasive assessment of coronary artery disease and its many manifestations, including ischaemia, hibernation, and scar. This comprehensive overview aims to empower clinicians, technicians, and patients with clear, structured knowledge on performing and interpreting PET MPI. This document will describe stress protocols, patient preparation, tracer pharmacodynamics and nuclear properties, camera capabilities, post-acquisition processing, and a comprehensive and clear reporting system for both perfusion and viability imaging.

Keywords: cardiac stress testing; positron emission tomography myocardial perfusion; viability imaging.

Figure 1

Methods of stress testing. Stress testing can be performed pharmacologically (Q1, Q2, Q4) or with exercise (Q3). Pharmacological stress testing is more common with PET compared with single photon emission CT, the alternative technology to perform nuclear MPI. Patients undergoing vasodilator stress tests may experience side effects such as flushing, headache, shortness of breath, and chest pain. Similarly, dobutamine may cause palpitations, chest pain, headache, flushing, and shortness of breath. Reproduced with permission from Abadie et al.

Figure 2

Examples of medications, foods, and drinks that may interfere with image quality. Some providers may continue beta-blockers or vasodilators during exercise testing to assess the degree of ischaemia while on treatment. Reproduced with permission from Abadie et al.

Figure 3

Example protocol with ⁸²Rb-chloride.

Figure 4

Example protocol with ¹³N-ammonia (¹³N-NH₃). Protocols with shorter wait times can also be performed, using subtraction techniques to adjust for the residual activity.^,

Figure 5

Example protocol with ¹⁸F-flurpiridaz. Stress-first imaging is preferred, as normal stress perfusion obviates the need for rest imaging. This may reduce radiation exposure, which is particularly important for ¹⁸F-flurpiridaz given its higher radiation dose compared with ¹³N-ammonia and ⁸²Rb-chloride. Rest-first protocols are also feasible/acceptable. Protocols with shorter wait times can be performed, using subtraction techniques to adjust for the residual activity, although these techniques have not specifically been validated with ¹⁸F-flurpiridaz.^,

Figure 6

Example of protocol with ¹⁵O-water.

Figure 7

Comparison of myocardial extraction between different commercially available PET and SPECT tracers.

Figure 8

Iterative reconstruction and filtering. Cardiac PET horizontal long axis images reconstructed using OSEM algorithm. (A) 1 iteration, 5 subsets; (B) 3 iterations, 5 subsets; (C) 7 iterations, 5 subsets; (D) 10 iterations, 5 subsets; (E) 3 iterations, 5 subsets with an 8 mm Gaussian smoothing filter.

Figure 9

Elements of a cardiac PET myocardial perfusion study. (A) Gated images at stress (top row) and rest (bottom row) with quantitative analysis of left ventricular ejection fraction and volumes. (B) Static perfusion images at stress (top row) and rest (bottom row) displayed in short axis, horizontal long axis, and vertical long axis planes. Perfusion defects are scored using a semi-quantitative scale on a bulls-eye plot, with summed stress, summed rest, and summed difference scores. (C) Coronary artery calcium score. (D) MBF analysis using tracer kinetic modelling, reporting rest, stress, and reserve MBF. ⁸²Rb-chloride is the tracer used in this study.

Figure 10

Standard views of the left ventricle. Traditionally, stress images are placed above rest images. SA, short axis that runs from the apex (left) to the base (right); HLA, horizontal long axis that runs from the inferior (left) to anterior wall (right), VLA, vertical long axis that runs from the septum (left) to the lateral wall (right).

Figure 11

Standard 17-segment model and polar map for reporting semi-quantitative perfusion. Each segment corresponds to a particular vascular distribution. Reproduced with permission from Abadie et al.

Figure 12

Myocardial dyssynchrony can be accessed via gated PET images. (1A) The patient has a narrow QRS on electrocardiogram with a narrow range of R-R intervals (1B). Panel 1C demonstrates that all the myocardial segments are reaching peak displacement at the same time. Conversely, patient B has a left bundle branch block and premature ventricular contraction (2A) with a wide range of R-R intervals (2B). Panel 2C shows a delayed and dyssynchronous peak displacement of the myocardial segments.

Figure 13

Example of quantitative MBF. Top left: retention images are used to ensure the myocardial border is properly registered and aligns with tracer uptake. Top right: figure demonstrates degree of patient motion during the exam which may interfere with accuracy of blood flow data. Bottom left: time-activity curves showing MBF at stress and rest. Note the early blood pool peak (green) that quickly drops to lower levels of activity. Conversely, the myocardial flow persists throughout the curve (other colours). Bottom right: resultant stress, rest, and reserve flow by vascular distribution, with adjacent global value. ⁸²Rb-chloride is the tracer used in this study.

Figure 14

Normal myocardial flow reserve is typically defined as ≥2.0. In the absence of obstructive epicardial disease, an abnormal myocardial flow reserve (<2.0) can indicate microvascular dysfunction. Two main categories of microvascular dysfunction have been described, classical/structural with low stress flow, and functional/endogen with high rest flow.

Figure 15

The PET/CT fusion images at stress (A) show a decrease in tracer uptake in the spleen compared with the rest images (B). This sign, termed splenic switch-off, can be used to ensure the patient had an adequate response to the vasodilator stress agent, which in this case was regadenoson. ⁸²Rb-chloride is the tracer used in this study.

Figure 16

(A) The patient had a poorly functioning intravenous catheter and extravasation of the radiotracer. As a result, the curves showing the uptake of the tracer at rest (yellow and blue) are delayed compared with stress (red and green). (B) Correct time-activity curve where the rest and stress tracer activity peaks at the same time.

Figure 17

Demonstrative examples of the most common artefacts encountered at PET MPI. (A) Typical appearance of a motion artefact (butterfly appearance) evident at both stress and rest images of a restless patient with claustrophobia undergoing PET MPI. (B) Typical fixed lateral wall defect in a patient undergoing rest/stress ¹³N-NH₃ PET MPI. Note that the defect is more evident at stress due to higher tracer dose applied. (C) Mis-registration of the stress PET images with the CT transmission images resulting in a reversible perfusion defect at the inferolateral basal wall. (D) Typical apical thinning defect at a patient undergoing PET MPI with normal wall motion and thickening. (E) Increased lung uptake at a patient undergoing rest/stress ¹³N-NH₃ PET MPI directly after cigarette smoking. Note the more prominent extracardiac uptake during rest. Asterisks depict the artefact at the slices, and arrow heads depict the corresponding artefact appearance at the polar plots.

Figure 18

The spectrum of myocardial ischaemic dysfunction.

Figure 19

Examples of viability imaging protocols for PET with ¹⁸F-FDG and SPECT with ⁹⁹m-technetium (^99mTc). For FDG-PET (top), patients consume a low-fat meal the evening prior. Rest imaging is performed with traditional perfusion tracers. If there are fixed perfusion defects, the patient can proceed with viability imaging. Patients undergo glucose loading followed by injection of IV insulin to target goal serum glucose. FDG is then injected and viability imaging performed. Hibernating myocardium will have uptake of FDG. For SPECT (bottom), no dietary prep is necessary. After rest imaging demonstrates fixed perfusion defects, patients can undergo viability imaging. Nitroglycerine is administered prior to the second injection of technetium to augment MBF and tracer uptake. SL NTG, sublingual nitroglycerine.

Figure 20

(A) Large fixed perfusion defect in the inferior and inferolateral segments on resting ⁸²Rb-chloride imaging. (B) Viability imaging shows FDG uptake in the inferior and inferolateral wall demonstrating hibernating myocardium.

Figure 21

(A) Large fixed perfusion defect in a left anterior descending distribution on resting ⁸²Rb-chloride imaging. (B) Viability imaging shows no FDG uptake in these segments, demonstrating scar.

Figure 22

Stress/rest/viability imaging with ⁸²Rb. Stress (A) and rest (B) imaging shows a small, fixed defect in the inferior segments along with ischaemia in the lateral wall. (C) FDG images show scar in the inferior segments. FDG uptake in the lateral wall, corresponding to the ischaemic territory, is an example of ischaemic memory.

Figure 23

CT transmission-emission misalignment. Misalignment of CT transmission and ¹³N-NH₃ emission scan inferiorly and inferolateral causes an apparent perfusion defect (A) that disappears with proper alignment (B).

Figure 24

Incidental findings on CT performed for attenuation correction. (A) Coronary artery calcification (dotted line) and bilateral pleural effusions (solid line). (B) Breast mass (dotted line) and severe mitral annular calcification (solid line).

Figure 25

Incidentally found, aneurysmal dilation of the ascending aorta.

Figure 26

Pillars of a comprehensive PET-MPI report.