Cardiac Positron Emission Tomography: Myocardial Perfusion, Myocardial Blood Flow, Coronary Flow Reserve Imaging

Cardiac Positron Emission Tomography: Overview of Myocardial Perfusion, Myocardial Blood Flow and Coronary Flow Reserve Imaging Darrell D. Burckhardt, Ph.D. SIEMENS www.siemens.com/mi 3 4 Table of Contents Introduction 1 Unique Imaging Systems for a Unique Imaging Workflow 1 Quality Assurance for Cardiac PET/CT 3 List-mode Acquisition Protocol 3 Dynamic LM PET Image Formation 4 Myocardial Perfusion, Myocardial Blood Flow, and Coronary Flow Reserve 4 Processing of dynamic PET data to obtain MBF and CFR 6 Analysis and Interpretation of Clinical Cases 8 Conclusion 10 Disclaimer 10 About the authors 11 Reference 11 1 Cardiac Positron Emission Tomography: Overview of Myocardial Perfusion, Myocardial Blood Flow and Coronary Flow Reserve Imaging Introduction Unique Imaging Systems for a Unique Imaging Workflow The clinical use of 82Rb and 13NH3 cardiac positron emission Proper equipment selection, along with thorough quality computed tomography (PET) for semi-quantitative myocardial assurance is essential to acquire quantitative clinical data perfusion imaging (MPI) has increased significantly over the for optimizing diagnostic accuracy8,19,22. Owing to its widely past decade1,2,3,4,5,6,7,11, due primarily to the increased diag- accepted role in oncology, clinical PET is almost entirely avail- nostic accuracy and prognostic value of PET MPI over stress able in a hybrid multi-modality design with computed tomog- echocardiography and single photon emission computed tomog- raphy (CT). Siemens Biograph™ TruePoint™ and mCT (Figure raphy (SPECT) in clinical assessment of patients with known or 1), respectively, are leading edge 3-dimensional (3D) lutetium suspected obstructive coronary artery disease (CAD)3,4. In efforts oxyorthosilicate (LSO) high resolution and time-of-flight PET/ to provide objective physiological means of managing patients CT devices capable of simultaneous static, dynamic, and gated with CAD, quantitative measures reflecting myocardial blood cardiovascular imaging through list mode (LM) acquisition9. flow (MBF) and coronary flow reserve (CFR) using dynamic 82Rb Cardiovascular PET/CT clinical applications can provide assess- and 13NH3 PET have likewise received recent attention4,5,29,30. ment of myocardial perfusion, viability, blood flow, coronary Patients with chronic stable CAD, as well as those affected by flow reserve, neuronal and receptor function, and inflamed and diffuse and/or balanced multi-vessel disease, often achieve vulnerable plaque as well as state of the art CT angiography better outcomes when revascularization is deferred and coro- (CTA), perfusion and calcium scoring (CAC)2,10,11. Integrating nary blood flow is improved through intensive pharmacological the anatomic, hemodynamic, and metabolic characteristics of therapy and lifestyle modification. An automated quantitative the interacting effects of coronary artery stenosis and diffuse analysis of ischemia and stenosis severity utilizing MBF and CFR atherosclerosis with myocardial function and morphology repre- integrated with routine visual or semi-quantitative MPI can be sents a major advance toward comprehensive diagnostic and used objectively to determine whether or not a stenosis requires prognostic workup of CAD12. Although, standard cardiac gated revascularization thus providing better patient outcomes at stress MPI is the mainstay of cardiovascular imaging13, an alter- lower costs. native imaging methodology provided by Siemens dynamic PET/ CT cardiac workflow (Figure 2) is potentially “game changing.” Several leading academic institutions have demonstrated, in This technique, described in the following sections, allows humans, the feasibility of MBF and CFR achieved with dynamic clinical decisions regarding coronary interventions to be based 13NH3 and 82Rb PET4,5,6,16. Siemens, in its commitment to contin- on additional objective, physiological measures of functional uously develop patient focused, cutting-edge PET technologies, significance of coronary artery stenosis without any complexity, has taken the next step. This paper describes the first* ever, additional radiopharmaceutical administration, or imaging time commercially available, comprehensive clinical workflow for the added to the standard PET/CT MPI examination. non-invasive, quantitative physiological assessment of CAD and other cardiac disease characterized by angiographically normal coronaries6. Figure 1. Siemens Biograph TruePoint (a) and mCT (b) PET/CT devices. *Siemens is the only vendor to offer an integrated myocardial blood flow solution (i.e. PET/CT scanner plus FDA-approved software). Of the major PET/CT vendors to date (GE, Philips, Siemens), only Siemens has released an FDA-approved myocardial blood flow solution. 2 Figure 2. Dynamic PET/CT List-mode PET PET Image Analysis and cardiac workflow. syngo Acquisition Formation Dynamic PET Interpretation CT scout scan List-mode Replay Data Loading Quantitative analysis CT AC scan o Static Localization Qualitative review Radionuclide injection . Dynamic Reorientation Rest PET LM scan Gated Interpretation 333 Stress agent injection Quality Assurance Segmentation Reporting Motion correction Delay . PET/CT alignment TAC generation Radionu clide injection o Image Quality MBF kinetic modeling Stress PET LM scan Reconstruction CFR computation Optional CT AC scan Perfusion image Optional CAC and/or CTA Polar Plots Quality Assurance for Cardiac PET/CT Quality assurance is imperative to providing optimal clinical standard 82Rb and 13NH3 MPI acquisition start delays of ~90 s results. It is demanding and requires routine maintenance and and 3 min post-radionuclide administration for blood pool clear- calibration of the PET/CT and ancillary devices (e.g., infusion ance, respectively, dynamic LM acquisition starts with radionu- pumps, dose calibrators, well counters, gating devices, etc). clide administration to obtain data upon radionuclide arrival in Clinical PET/CT data must also be inspected routinely for artifacts the right and left ventricle (LV). The LM acquisition continues typically caused by spatio-temporal misalignment resulting from for 6-8 min and 10-20 min for 82Rb and 13NH3, respectively. Rest a variety of factors2,14,15,16,17. Assuring that early blood pool and stress arterial radiopharmaceutical administration consist of radioactivity has cleared from the heart is also necessary to approximately 1110–1480 MBq (30-50 mCi) 82Rb and 370-740 avoid adversely affecting PET perfusion image quality and subse- MBq (10-20 mCi) 13NH3 (Table 1). Stress imaging is performed quent semi-quantitative MPI polar plots analysis2,11,16,19,29. A pharmacologically after rest imaging, allowing for radionuclide complete description of quality assurance procedures is beyond decay, and the administration of adenosine, dypridamole, or this paper’s scope and the reader is encouraged to review the dobutamine. Heart rate, blood pressure and ECG are typically American Society of Nuclear Cardiology (ASNC) guideline on the recorded at baseline rest and throughout pharmacological stress subject19. imaging. The deliberate generic nature of the protocol described above is meant only to serve as a basis for following sections. Detailed recommendations for cardiac PET imaging procedures, List-mode Acquisition Protocol including pharmacologic stress agent and radionuclide adminis- Patients are placed head first and supine on the PET/CT patient tration, are available in the referenced peer reviewed literature handling system. An initial low-dose CT scout scan is obtained 19,20,21,22,23 . for proper patient placement in subsequent low-dose CT attenu- ation correction (CT AC) and PET scans. As stated above, the During LM acquisition the PET system records incoming timing, PET LM imaging protocol does not introduce any additional event location, and other (phase) information in order of their complexity over standard static or gated MPI acquisitions. Unlike occurrence (Figure 3). Dynamic LM Replay Data Stream > Time Event Other Frame 1 Frame 2 Frame 3 Figure 3. List-mode data and replay example. The duration of each frame is the same (2 time tags) while the number of events and other tags in the frames are different. 3 Dynamic LM PET Image Formation Myocardial Perfusion, Myocardial Blood Flow, and Coronary List-mode (LM) imaging is an ideal approach, in that, multiple Flow Reserve image reconstructions (i.e., cardiac and/or respiratory gated, Understanding the basic properties of PET MPI agents is neces- static, dynamic, and summed frames) can be obtained from a sary in order to choose the proper agent for a particular clinical single radionuclide administration and LM acquisition. With setting and to fully understand the results of myocardial perfu- dynamic LM replay, all data stream events having the same loca- sion and blood flow imaging (19). Flow imaging agents fall tion over a specified time interval or frame are added together into two basic categories: 1) freely diffusible and 2) soluble and arranged according to projection position and angle to substances “trapped” in the myocardium. Usually, only freely form sinograms. These sinograms are corrected for a variety of diffusible tracers allow for repeated studies due to their rapid physical, geometrical, electronic, and anatomical factors prior physiological washout from the myocardium. H215O is an to image reconstruction. When properly calibrated, the recon- example of a freely diffusible agent (24 ). Images of a freely diffus- structed dynamic image sequence represents the continuous ible agent often suffer from poor counting statistics resulting in time course of radionuclide concentration (Bq/ml) distributed in noisy images of the radionuclide distribution. Soluble agents, the blood pool and myocardial tissue regions-of-interest (ROIs), however, remain fixed in the myocardium for a sufficient length sampled over specific intervals in time as depicted in Figure 4. of time, usually by a metabolic-trapping mechanism, and produce images of reasonable image quality. 82Rb and 13NH3 Activity fall into this second category. 82Rb has a short half life (75 s) and therefore quickly disappears from the myocardium allowing rapid repeated imaging studies. Blood Input Function, Cp(t) The amount of perfusion agent taken up in a specific area of the myocardium is determined by the tracer’s physical properties, Events delivery (regional coronary blood flow), and myocyte extrac- tion and retention. Both myocyte cell membrane integrity and energy utilization are necessary for intracellular tracer extrac- Tissue uptake, CT(t) tion and retention. Thus, retained or “trapped” tracer radioac- tivity not only indicates the pattern of blood supply but also indicates regional myocyte viability25. Accordingly, myocardial regions supplied by “normal” coronary blood flow will possess a “normal” pattern of tracer uptake. Myocardial regions supplied by stenosed coronary arteries will have lower coronary blood flow and reduced uptake relative to normal regional uptake (26). Frames Time Under resting conditions, however, the reduction in CBF is not always obvious in relative regional uptake even in patients with moderate to severe stenosis (<90%) or with systemic, balanced Figure 4. Time-activity curves. Accumulated events per unit volume CAD. However, when myocardial metabolic demand for nutrients over time at an image ROI location are proportional to the radionuclide and oxygen is increased during exercise or stress conditions, the concentration at the corresponding anatomical location. relative difference between normal and stenotic vascular blood flow is considerable and produces a distinct relative reduction in tracer uptake (i.e., perfusion defect) between normal and ischemic myocardium. This difference in tracer uptake or non- In the preceding example, images are reconstructed for the homogeneity between normal and stenotic vascular territories specific purpose of determining the target tissues radionuclide during stress is the underlying principle establishing myocardial pharmacodynamics (i.e., blood pool, myocardium time-activity perfusion imaging27. Typically, perfusion images, normalized curves). Generally, cardiac and respiratory gated images can to peak myocardial uptake, are compared to recognized normal be formed using LM replay and other physiological information population perfusion pattern averages. Normalized perfusion inserted into the LM data stream obtained from ancillary equip- values falling outside their “normal population” regional pattern ment (e.g., ECG, pressure sensors). While methods for obtaining confidence intervals are quantified as defect extent and severity. image derived physiological gating information are reported in For the same region, defects present at stress and absent at rest the literature9 they are not currently commercially available and indicate reversible ischemia while matching rest-stress defects are therefore beyond the scope of this paper. suggestive of myocardial scarring or infarction. Detection of CAD with MPI, therefore, relies on regional radiotracer uptake differ- ences resulting from stress-induced coronary flow heterogeneity. 4 82Rb 13NH3 Administered radioactivity range GBq (mCi) 1.11-1.85 (30-35) 0.37-0.74 (10-20) patient < 91 kg (200 lbs) 1.11 (30) patient < 113 kg (250 lbs) 1.48 (40) 0.37 (10) patient > 113 kg (250 lbs) 1.85 (50) 0.74(20) Radioactivity infusion/injection maximum rate (ml/min) 65 12 maximum volume (ml) 50 6 Scan duration (min) 6-8 10-15 Low-dose CT attenuation correction protocol free-breathing free-breathing Tube current (effective mAs) 11 11 Tube voltage (kVp) 120 120 CTDI vol (mGy) 0.74 0.74 Slice thickness (mm) 3.0 3.0 Scan time (sec) 4 4 Rest-Stress scan separation (min) 10 50 Stress agent rate and period of administration (mg/kg/min min) Adenosine Dypridamole 0.14 6 0.16 6 Dobutamine (see guidelines) 0.14 4 0.14 4 Radionuclide administration delay post-stress agent administration start (min) Adenosine 1.5 3 Dypridamole 3-5 3-5 Dobutamine (≥ 85% age predicted maximum heart rate) MPI delay post-radionuclide administration (s) 90 - 120 90 -180 Dynamic frame strategy 12x5s 12x10s 6x10s 6x30s 4x20s 2x60s 4x40s 1x180s Reconstruction method OSEM 2D-OSEM iterations 3 4 subsets 8 8 Zoom (+X direction offset mm) 2.2/2.6 + 30 2.2/2.6 + 30 matrix size (pixels) 168x168/128x128 168x168/128x128 Gaussian filter FWHM (mm) 5 5 Table 1. Typical parameters associated with 82Rb and 13NH3 dynamic LM studies.19,20,21,22,23 5 Figure 5. The one- and two-compartment tracer K kinetic models for 82Rb (a) and 13NH3 (b), repec- tively. CP(t) is the blood input function and CT(t) Cp(t) CT(t) the myocardial tissue compartment. The 13NH3 K2 model contains extra-vascular or interstitial ”free” and intracellular or “trapped” spaces. K1 represents myocardial extraction, k2 the rate of transport back to the blood, and k3 the rate of a) 13NH3 metabolic conversion into 13N-glutamine. CT(t) ............... Extra K3 Cp(t) Metabolically .......................... Vascular bound K2 ................................... (b) The assessment of perfusion tracer uptake patterns, however physiologically realistic values, in order to improve the reliability useful in the identification of normal or ischemic myocardium of the parametric model estimates. under stress or exercise conditions, has its limitations. Relative myocardial uptake may appear to be homogeneous (“normal”) Coronary flow reserve (CFR), defined as the amount of additional when myocardial blood flow is abnormal as in patients with coronary blood flow that can be supplied to the heart above balanced coronary blood flow reduction, diffuse or non-occlu- baseline, is calculated as the ratio of near-maximal vasodilation sive epicardial luminal narrowing of coronary arteries, or an to basal (stress to rest) MBF. CFR is a measure of both the larger occlusive coronary stenosis in the region with the highest tracer epicardial coronary vessels and the microcirculation29. There is uptake(assumed to have normal blood flow and used as a refer- a rich amount of literature and excitement on the PET computa- ence region for normalization). Acute coronary events, majorly, tion of myocardial blood flow and coronary reserve. Impaired do not originate in coronary arteries with distinct stenosis. flow reserve, in some situations such as, early stage athero- The reduction or impairment of blood flow in acute coro- sclerosis, is thought to relate to dysfunction of the endothelial nary syndrome (ACS) result from plaque rupture, subsequent control of the microvasculature. Coronary flow reserve, thus, is thrombus formation, and vascular occlusion28. Measurements important not only in evaluating individual coronary stenosis of myocardial blood flow provided by dynamic PET can provide but also the evaluation of a number of disease processes which a more objective assessment of the physiological severity and affect the heart. functional consequences of coronary stenosis in the former example and the in the latter example, helps identify patients Coronary flow reserve can be measured invasively at the time with pre-clinical arthrosclerosis. of coronary catheterization. PET, however, has the advantage of making this measurement non-invasively. Siemens seamlessly The individual images of a dynamic PET sequence do not neces- integrates the compartment kinetic analysis of PET data, without sarily yield clinically relevant information as they are often noisy complexity, into the clinical workflow using state-of-the-art PET/ due to their short duration and image pixels of the myocardium CT scanners, acquisition modes, reconstruction methods, and obtained early in the dynamic sequence contain blood pool kinetic analysis for quantitative MBF and CFR quantification and radioactivity (spill-over). However, tracer kinetic data (TACs) review. derived from the entire ensemble of early to late dynamic frames can produce a quantitatively accurate estimate of the Processing of dynamic PET data to obtain MBF and CFR desired physiological parameter (i.e., regional myocardial blood A number of image processing steps are required to obtain the flow) by a processing procedure called mathematical “tracer blood input function and tissue uptake time activities curves for kinetic” modeling. The pharmaco-kinetic behavior of 82Rb and the compartmental tracer kinetic analysis in the determination 13NH3 described by one-30 and two-tissue compartment math- of MBF and CFR. These steps, described below in the context of ematical models31, respectively (Figure 5), are fitted to the their syngo implementation, start with the selection of stress discretely sampled patient time-activity curves derived from the and rest dynamic PET (transaxial) image sequences from the dynamic PET series. The model parameters are estimated using syngo patient browser followed by loading the data into the MI non-linear regression by minimizing the differences between Circulation application and then launching syngo Dynamic PET. the model results and patient time-activity data. The model Automatically, the weighted sum of the later frames in each parameters, which include blood flow, are often constrained to dynamic sequence is used to locate delineate, and re-orient the 6 SA ant HLA apex VLA 161910 Fant sep lat sep lat base apex = E Frame 21/21 Frame 21/21 Frame 21/21 00:05:30 00:05:30 00:05:30 inf 30s base 30s inf 30s SA ant HLA apex VLA ant 178920 = sep lat sep lat base apex = Frame 21/21 Frame 21/21 Frame 21/21 00:05:30 00:05:30 00:05:30 inf 30s base 30s inf 30s Process Cancel Figure 6. Automatic heart localization, alignment, and display in standard cardiac orientation. LV myocardium along its long-axis for all frames in the dynamic Maximum activity points are obtained along each of the 540 sequence. Stress and rest sequences are stereotactically aligned radial profiles of the summed late frame image. Due to the to each other in standard cardiac orientation. The syngo Dynamic noise in areas of lower myocardial tracer uptake, the location PET application also allows for manual adjustments of apex, base of maximum uptake along the radial profile is not always within and long axis location and angle as needed (Figure 6). the LV myocardium. In this case, special smoothing and inter- polation techniques are used to derive the mid-myocardial iso- The syngo Dynamic PET application uses the conventional surface. The mid-myocardial iso-surface is shifted to form endo- cylindrical-spherical model32 for dynamic myocardial sampling cardial and epicardial boundaries outlining the myocardium as described below and illustrated in Figure 7. A cylindrical (Figure 8). For quality assurance, a persistent overlay of this sampling pattern of 36 radial profiles (every 10º) perpendicular outline in the three orthogonal views while stepping through to the long-axis is used in each of 9 slices equally spaced over the dynamic frame sequence, allows for the assessment of the basal two-thirds of the long-axis. A spherical sampling at 18º cardiac movement. If necessary, cardiac motion compensation is intervals in the apical zone is employed, to generate 6 spherical applied automatically to later frames. slices. A 10º radial angular sampling angle is also used in the spherical or apical one-third of the myocardium region. 7 Long axis Cylindrical slice Base Mid-line iso-surface Long axis Cylindrical Sampling 100 ....... Sampling rays Intensity Endocardial Maxima = 10º boundary ................... Spherical Sampling Epicardial boundary Q= 18º Apex Sampling ray intensity profile Figure 7. Left ventricular myocardial sampling strategy. Myocardial time-activity curve data points are obtained at The screen capture of the syngo Dynamic PET application each time frame along the radial line segments defined by the (Figure 9) displays an abnormal patient examination. The data epicardial and endocardial boundaries as the averaged value of demonstrate the impaired pharmacologic-stimulated myocardial the neighborhood around the maximum radionuclide concen- blood flow, resulting in abnormal coronary flow reserve in the tration. Averaging is performed to reduce the effects of noise, LAD and LCX territories and remarkably abnormal CFR in the blood partial volume and spill-over effects. The partial volume RCA. The MBF (QMP) for LAD (1.06 ml/g/min) and LCX (1.08 and spill-over effects are also estimated and accounted for ml/g/min) territories at stress are over twice as high as that of during kinetic modeling. A total of 540 time-activity curves are the RCA (0.47 ml/g/min) territory. The rest MBF (QMP) values generated from the myocardium in this fashion. for the three vascular territories are similar (LAD 0.87, LCX 0.93, RCA 0.7 ml/g/min) with a slightly reduced value for the RCA. The The blood input function used for all kinetic analyses is obtained CFR (MPR) values are LAD 1.21, LCX 1.13, and RCA 0.65. The from the dynamic sequence by averaging the activity in a cylin- quantitative MBF (QMP) and CFR (MPR) polar plots and values drical region-of-interest placed in the middle of the left ventricle augment as well as correlate well with the concurrent qualita- in the most basal slice for each time frame. This blood pool ROI tive MPI information, thus demonstrating the potential for a is displayed as a green cylindrical overlaid on the three orthog- more comprehensive clinical examination and interpretation onal views (Figure 8). of the consequences of atherosclerosis on the myocardium and coronary vasculature. Analysis and Interpretation of Clinical Cases The results of tracer kinetic analysis include polar plots for stress The screen capture of the syngo Dynamic PET application MBF, rest MBF and CFR. For each of the polar plots, an overlay (Figure 10) displays an abnormal patient examination. Despite demarking the three main vascular territories from which a a rather homogenous perfusion distribution at rest, there is a parametric chart of the vascular territory statistics (mean and large hypo-perfused fixed apical defect and reduced anterior standard deviation) are derived along with a marked out section and lateral wall uptake at stress. The MBF (QMP) for LAD (1.48 close to the basal septum is shown Figure 9. For rest and stress, ml/g/min) and LCX (1.52 ml/g/min) territories at stress are lower (summed late frame) semi-quantitative MPI polar plots, three than RCA (2.06 ml/g/min) territory. The rest MBF (QMP) values orthogonal view displays of dynamic frames, graphed plots for the three vascular territories are similar (LAD 1.20, LCX 1.02, of the fitted and actual input function and tissue time-activity RCA 1.25 (ml/g/min)) with slightly reduced value for the LCX. curves are also provided. Selecting a segment on any one the The CFR (MPR) values are LAD 1.24, LCX 1.52, and RCA 1.65. polar plots, highlights the same segment in each polar plot and The MBF (QMP) and CFR (MPR) polar plots and quantitative updates the lower left MPI, MBF (QMP) CFR (MPR) values, and values correlate well with the MPI images and clinical diagnosis. the plotted TACs, accordingly. 8 HLA apex HLA apex HLA apex HLA apex HLA apex sep lat sep at sep at sep lat sep lat 41/81 41/81 41/81 41/81 41/81 Frame 1/21 Frame 2/21 Frame 3/21 Frame 4/21 Frame 5/21 00:00:00 00:00:10 00:00:20 00:00:30 00:00:40 5222 Bq/ml base 10s 73989 Bq/ml base 10s 296911 Bq/ml base 10s 257572 Bq/ml base 10s 243240 Bq/ml base 10s HLA apex HLA apex HLA apex HLA apex HLA apex sep lat sep at sep lat sep lat sep lat 41/81 41/81 41/81 41/81 41/81 Frame 6/21 Frame 7/21 Frame 8/21 Frame 10/21 Frame 12/21 00:00:50 00:01:00 00:01:10 00:01:30 00:01:50 172264 Bq/ml base 10s 130210 Bq/ml base 10s 91822 Bq/ml base 10s 81866 Bq/ml base 10s 49744 Bq/ml base 10s HLA apex HLA apex HLA apex HLA apex HLA apex sep lat sep lat sep lat sep lat sep at 41/81 41/81 41/81 41/81 41/81 Frame 14/21 Frame 16/21 Frame 18/21 Frame 20/21 Frame 21/21 00:02:20 00:03:00 00:04:00 00:05:00 00:05:30 51624 Bq/ml base 20s 57732 Bq/ml base 30s 36096 Bq/ml base 30s 37919 Bq/ml base 30s 42548 Bq/ml base 30s Figure 8. Segmented myocardial and blood pool regions in dynamic PET frames from which the blood input function and tissue time-activity curves are derived. The radioactivity bolus can be seen entering the right ventricle, the left ventricle, washing out of the blood pool and taken up in the myocardium (upper left then across and down to lower right). Stress : Dynamic stress NH3 progressif AC_CT#1 (AC) : 6/11/08 16:12:37 : Radionucide Total Dose 1005.0 MBq SA ant HLA apex VLA ant 161910 Uptake 205372 QMP 2.7 LAD LCX LAD LCX sep lat sep lat base apex O 138/256 36/81 149/256 Frame 21/21 Frame 21/21 Frame 21/21 00:05:30 00:05:30 00:05:30 = RCA RCA 120239 Bq/ml inf 30s 120239 Bq/ml base 30s 120239 Bq/ml inf 30s 9 118746 Bq/ml 00:05:00 6OS 0 0.86 ml/g/min Outlier % = 1.8 0.0 Rest : Dynamic Rest NH3 progressif AC_CT#2 (AC) : 6/11/08 17:03:41 : Radionuclide Total Dose 1008.0 MBq SA ant HLA apex VLA ant 178920 Uptake QMP LAD LCX LAD LCX sep at I sep lat | base apex 137/256 36/81 150/256 Frame 21/21 Frame 21/21 Frame 21/21 00:05:30 00:05:30 00:05:30 RCA RCA 173776 Bq/ml inf 30s 173776 Bq/ml base 30s 173776 Bq/ml 30s 166892 Bq/ml 10:05:00 0 0.96 ml/g/min Outlier % = 0,0 0.0 QMP (ml/g/min) 3.50E5- Reserve Reserve 3.00E5 LAD Stress Rest Ammonia 2.50E5 LCX mean std dev. mean std dev. mean std dev. 2.00E5 - blood stress LAD 1.06 0.34 0.87 0.11 1.21 0.32 1.50E5 uptake LCx 1.08 0.39 0.12 1.13 0.32 1.00E5 rest 0.93 Intensity (Bq/ml) blood input RCA 0.47 0.70 0.11 0.65 0 19 6.00E4 rest uptake Global 0.93 0.41 0.84 0.14 1.06 0.37 0.00E0- 100 200 300 400 RCA Time (s) 0.90 Oullier % = 1.8 0.0 stress X2 = 1.88E9 rest X = 1.05E9 Figure 9. syngo Dynamic PET display of MBF (QMP) and MPR(CFR) of a 13NH3 clinical case showing reversible larger reversible inferio-apical defect. 9 WE SUIT new meup NH3_THAL01 : Anonymous : Unknown age : Unknown DoB : O : Unknown BMI : Stress : : 4/9/08 14:23:25 : Radionuclide Total Dose 1044.0 MBq SA ant HLA apex VLA ant 202472 Uptake 208972 QMP 3.0 LAD LAD = LCX LCX sep lat sep lat |base -- - - - pex | = 129/256 41/81 129/256 Frame 21/21 Frame 21/21 Frame 21/21 00:05:30 00:05:30 00:05:30 RCA RCA 46187 Bq/ml inf 30s 46187 Bq/ml base 30s 46187 Bq/ml inf 30s 00:05:00 60s 0 Outlier % = 0.2 0.0 Rest : : 4/9/08 15:37:54 : Radionuclide Total Dose 1061.0 MBq- SA ant HLA apex VLA 212663 Uptake ant 174183 QMP 1.5 LAD LCX LAD LCX sep lat sep - - lat base ---- apex = 129/256 41/81 129/256 Frame 21/21 Frame 21/21 Frame 21/21 00:05:30 00:05:30 00:05:30 RCA RCA 41968 Bq/ml inf 30s 41968 Bq/ml base 30s 41968 Bq/ml inf 30s 00:05:00 Outlier % = 0.8 0.0 60s 3.00E5- QMP (ml/g/min) Reserve Reserve 3.0 2.50E5- LAD Stress Rest LCX 2.00E5 Ammonia mear std dev. mean std dev. mean std dev. 1.50E5 stress LAD 1.48 0.30 1.20 0.12 1.24 0.24 blood input 1.00E5 rest Intensity (Bq/ml) LCX 1.52 0.35 1.02 0.17 1.52 0.35 blood 5.00E4- input RCA 2.06 0.30 1.25 0.09 1.65 0.21 Figure 8. Segmented myocardial and blood pool regions in dynamic PET frames from which the blood input function and tissue time-activity curves are Globa 1.63 0.40 1.16 0.16 1.41 0.32 0.00E0- 100 200 derived. The radioactivity bolus can be seen entering the right ventricle, the left ventricle, washing out of the blood pool and taken up in the myocardium 300 400 RCA Time (s) Oullier % = 1.0 0.0 (upper left then across and down to lower right). Figure 10. syngo Dynamic PET display of MBF (QMP) and MPR(CFR) of a 13NH3 clinical case showing an apical defect in the stress study. Conclusion Disclaimer The Biograph PET/CT and the syngo Dynamic PET application The ultimate judgment about the propriety of any specific proce- provides for non-invasive assessment of myocardial blood dure or course of action must be made by the physician when flow and coronary flow reserve with 82Rb* and 13NH3 using considering the circumstances presented. one- and two- tissue compartment tracer kinetic models. The automated technique, provided by Siemens, allows clinical All that should be expected is that the practitioner will follow a decisions regarding coronary interventions to be based on reasonable course of action based on current knowledge, avail- additional objective, physiological measures of myocardial func- able resources, and the needs of the patient to deliver effective tion without any complexity, additional radiopharmaceutical and safe medical care. The sole purpose of this white paper is to administration, or imaging time added to the standard PET/CT assist practitioners in achieving this objective. MPI examination. * syngo Dynamic PET is FDA-cleared for ammonia and pending validation for rubidium. 10 Contributing Authors References Darrell D. Burckhardt, Ph.D. received his graduate degree from 1 Kneasurek K, Machac J, Krynyckyi, BR, Almeida OD, The Johns Hopkins University School of Hygiene and Public Comparison of 2-dimensoional and 3-dimensional Health, Department of Environmental Health Sciences, Division 82Rb myocardial perfusion PET imaging. J Nucl Med. of Radiation Health Sciences. Since 1999, Dr. Burckhardt has had 2003; 44:1350-1356. a number of responsibilities at Siemens. 2 Di Carli MF, Dorball S, Meserve J, El Fakhri G, Sitek A, Dr. Jérôme Declerck received a Ph.D. in Engineering Sciences Moore SC. Clinical myocardial perfusion PET/CT. from the Institut National de Recherche en Informatique et J Nucl Med 2007; 48:783-793. Automatique (INRIA) and University of Nice (France). He was a visiting scientist at the Johns Hopkins University (Baltimore, 3 Sampson UK, Dorbala S, Limaye A, Kwong R, Di Carli USA) in 1998-1999 and has been directing the Science and MF. Diagnostic accuracy of rubidium-82 myocardial perfu- Technology Team of Siemens Molecular Imaging (formerly sion imaging with hybrid positron emission tomography/ Mirada Solutions) since 1999. computed tomography in the detection of coronary artery disease. Am Coll Cardiol, 2007; 49:1052-1058. Dr. Xiao-Bo Pan received a Ph.D. in flow imaging instrumenta- tion and image analysis from Heriot-Watt University (UK) and a 4 Bateman TM, Heller GV, McGhie AI, Friedman JD, Case BA in Information and Control Engineering from Xian JiaoTong JA, Bryngelson JR, Hertenstein GK, Moutray KL, Reid K, University (China). She was a Research Associate at the Depart- Cullom SJ. Diagnostic accuracy of rest/stress ECG gated ment of Engineering, University of Oxford, from 1999 and has Rb-82 myocardial perfusion PET: comparison with ECG-gated worked in the Science and Technology Team of Siemens Molec- Tc-sestamibi SPECT. J Nucl Cardiol 2006; 13:24-33. ular Imaging (formerly Mirada Solutions) since 2002. 5 El Fakhri G, Kardan a, Sitek A, Dorbala S, Abi-Hatem N, Lahoud Y, Fischman a, Coughlan M, Yasuda T, Di Carli MF. Reproducibility and accuracy of quantitative myocardial Acronyms and notations blood flow assessment with 82Rb PET: Comparison with 13N-ammonia PET. J Nucl Med 2009; 50:1062-1071. BIF blood input function 6 Camici PG, Gropler RJ, Jones T, Abbate AL, Maseri A, CAD coronary artery disease Melin JA, Merlet P, Parodi O, Schelbert HR, Schwaiger M, Wijns W. The impact of myocardial blood flow quantitation CT computed tomography with PET on the understanding of cardiac diseases. Euro- pean Heart Journal 1996; 17:25-34. CTA CT angiography 7 Yoshinaga K, Chow BJW, Williams K, Chen L, deKemp CFR coronary flow reserve RA, Garrard L, Zeto ALT, Aung M, Davies R, Ruddy TD, Beanlands RSB. What is the prognostic value of myocardial FWHM full width at half maximum perfusion imaging using rubidium-82 positron emission tomography? J Am Coll Cardiol 2006; 48:1029-39. MBF myocardial blood flow 8 Nichols KJ, Bacharach S, Bergmann S, Chen J, Cullom SJ, MPR myocardial perfusion reserve Dorbala S, Ficaro EP, Galt JR, Green Conaway DL, Heller GV, Hyun MC, Links J, Machac J. Instrumentation quality 13NH3 Nitrogen - 13 radiolabeled ammonia assurance and performance. J Nucl Cardiol 2007; 14:e61-78. PET positron emission tomography 9 Büther F, Dawood M, Stegger L, Wübbeling F, Schäfers M, Schober O, Schäfers KP. List Mode – Driven Cardiac and QMP quantitative myocardial perfusion Respiratory Gating in PET. J Nucl Med 2009; 50:674–681. 82Rb Rubidium - 82 10 Wykrzykowska J, Lehman S, Williams G, Parker JA, Palmer MR, Varkey S, Kolodny G, Laham R. Imaging of SPECT single photon emission computed tomography inflamed and vulnerable plaque in coronary arteries with 18F-FDG PET/CT in patients with suppression of myocardial TAC time-activity curve uptake using a low-carbohydrate, high-fat preparation. J Nucl Med 2009; 50:563-568. 11 Di Carli MF, Hachamovitch R. New Technology for Noninvasive Evaluation of Coronary Artery Disease. Circulation 2007;115;1464-1480. 11 12 Gould KL. Physiological severity of coronary artery disease. 22 Bokhari S, Ficaro EP, McCallister BD. Adenosine stress Am J Physiol Heart Circ Physiol 2006 291:2583-2585. protocols for myocardial perfusion imaging. J Nucl Cardiol 2007; 14:415-6. 13 Bengal FM. Clinical Cardiovascular Molecular Imaging. J Nucl Med 2009; 50:837–840. 24 Iida H, Kanno I, takahashi A, Miura S, Murakami M, Takahashi K, Ono Y, Shishido F, Inuganmi A, Tomura N. 14 Nye JA, Esteves F, Votaw J. Minimizing artifacts resulting Measurement of absolute myocardial bolld flow with H215O from respiratory and cardiac motion by optimization of and dynamic positron-emission tomography. Strategy for the transmission san in cardiac PET/CT. Med. Phys. 2007; quantification in relation to the partial volume effect. Circu- 34:1901-1906. lation 1988; 78:104-115. 15 Martenz-Moller A, Souvatzloglou M, Navab N, Schwaiger 25 Bhola R, Quaife RA. (2001). Nuclear cardiology, magnetic M, Nekolla S. Artifacts from misaligned CT in cardiac resonance imaging and computed tomography. In: Adair OV perfusion PET/CT studies: frequency, effects and potential (Ed.), Cardiology secrets. (2nd ed. pp. 55-65). Philadelphia, solutions. J Nucl Med 2007; 48:188-193. PA: Elsevier Health Sciences. 16 Loghin C, Sdringola S, Gould KL. Common Artifacts in PET 26 Di Carli MF, Czernin J, Hoh CK, Gerbaudo VH, Brunken RC, myocardial perfusion images due to attenuation-emission Huang , Phelps, and ME, Schelbert HR. Relation among mis-registration: Clinical significance, causes, and solutions. stenosis severity, myocardial blood flow, and flow reserve J Nucl Med 2004; 45:1029–1039. in patients with coronary artery disease. Circulation 1995; 91:1944-51. 17 Schuster DM, halkar RK, Esteves FP, Garcia, EV, Cook CD, Syed MA, Bowman FD, Votaw JR. Investigation of emis- 27 Jadvar H, Parker JA. (2005). Chapter 3: Cardiology. In: sion-transmission misalignment artifacts on rubidium-82 Parker JA (Ed) Clinical PET and PET/CT. (pp.69-83) New York, cardiac PET with adenosine pharmacologic stress. Mol NY: Springer. Imaging Biol 2008; 10:201-208. 28 Vesely MR, Dilsizian V. Nuclear cardiac stress testing in the 18 Lodge MA, Braess H, Mahmoud F, Suh J, Englar N, era of molecular medicine. J Nucl Med 2008; 49:399-413. Geyser-Stoops S, Jenkins J, Bacharach SL, Dilsizian V. Developments in nuclear cardiology: Transition from 29 Kaufmann PA, Camici PG. Myocardial blood flow measure- single photon emission computed tomography to positron ment by PET: Technical aspects and clinical applications. emission tomography/computed tomography. J Invasive J Nucl Med 2005; 46:75-88. Cardiol 2005; 17:491-496. 30 Lortie M, Beanlands RS, Yoshinaga K, Klein R, Dasilva 19 Dilsizian V, Bacharach SL, Beanlands RS, Bergmann SR, JN, DeKemp RA. Quantification of myocardial blood flow Delbeke D, Gropler RJ, Knuuti J, Schelbert HR, Travin MI. with 82Rb dynamic PET imaging. Eur J Nucl Med Mol Imaging PET myocardial perfusion and metabolism clinical imaging. 2007; 34:1765-1774. J Nucl Cardiol. 2009; Jun 4 (Online First). 31 Hutchins GD, Schwaiger M, Rosenspire KC, Krivokapich, 20 Strauss HW Miller DD, Wttry MD, Cerqueira MD, Garcia Schelbert H, Kuhl DE. Non-invasive quantification of EV, Iskandrian AS, Schelbert HR, Wackers FJ, Balon HR, regional blood flow in human heart using N-13 ammonia Lang O, Machac J. Procedure Guideline for Myocardial and dynamic positron emission tomographic imaging. Perfusion Imaging 3.3. J Nucl Med Technol 2008; J Am Coll Cardiol 1990; 15:1032-1042. 36:155-161. 32 Nekolla SG, Miethaner C, Nguyen N, Zegler SI, 21 Henzlova MJ, Cerqueira MD, Hansen CL, Taillefer R, Schwaiger M. Reproducibility of a polar map generation Yao S. Imaging guidelines for nuclear cardiology proce- and assessment of defect severity and extent assessment dures: Stress protocols and tracers. J Nucl Cardiol 2009; in myocardial perfusion imaging using positron emission 16. doi: 10.1007/ s12350-009-9062-4. tomography. Eur J Nucl Med 1998; 25:1313-1321. 22 Hesse B, Tägil K, Cuocolo A, Anagnostopoulos C, Bardiés M, Bax J, Bengel F, Busemann Sokole E, Davies G, Dondi M, Edenbrandt L, Franken P, Kjaer A, Knuuti J, Lassmann M, Ljungberg M, Marcassa C, Marie PY, McKiddie F, O’Connor M, Prvulovich E, Underwood R, van Eck-Smit B; EANM/ESC Group. EANM/ESC procedural guidelines for myocardial perfusion imaging in nuclear cardiology. Eur J Nucl Med Mol Imaging. 2005;32:855–897. 12 Some of the radiopharmaceuticals mentioned in this document are not available in all regula- tory jurisdictions and/or may not be approved for use for the indications discussed in all countries where they are marketed. Siemens makes no claims of safety or effectiveness and does not endorse any off-label clinical use of radiophar- maceuticals discussed in this paper. Trademarks and service marks used in this ma- terial are property of Siemens Medical Solutions USA or Siemens AG. Siemens Medical Solutions USA, Inc. Global Business Unit Address © 2009 Siemens Medical Solutions USA, Inc. All rights reserved. Siemens Medical Solutions USA, Inc. Molecular Imaging All photographs © 2009 Siemens Medical 2501 N. Barrington Road Solutions, USA, Inc. All rights reserved. Hoffman Estates, IL 60192-2061 USA Note: Original images always lose a certain Telephone: +1847 304 7700 amount of detail when reproduced. www.siemens.com/mi Global Siemens Headquarters Global Siemens Headquarters Address of legal manufacturer Healthcare Headquarters Siemens Medical Solutions USA Siemens AG Molecular Imaging Wittelsbacherplatz 2 Siemens AG 2501 N. Barrington Road 80333 Munich Healthcare Sector Hoffman Estates, IL 60192-2061 Germany Henkestrasse 127 USA 91052 Erlangen Germany Telephone: +49 9131 84-0 www.siemens.com/healthcare www.siemens.com/mi Order No. A91MI-10212-1T-7600 | Printed in USA | PA 0909 | © 09.2009, Siemens AG