Assessment of myocardial viability in dysfunctional myocardium by resting myocardial blood flow determined with oxygen 15 water PET

Assessment of myocardial viability in dysfunctional myocardium by resting myocardial blood flow determined with oxygen 15 water PET Bernd Nowak, MD,a Wolfgang M. Schaefer, MD PhD,a Karl-Christian Koch, MD,b Hans-Juergen Kaiser, PhD,a Stephan Block, MD,a Christian Knackstedt, MD,b Michael Zimny, MD,a Juergen vom Dahl, MD,b and Udalrich Buell, MDa Background. There is controversy about the role of decreased resting blood flow as the pathophysiologic correlate of hibernating myocardium. The aim of this study was an absolute quantification of volumetric myocardial blood flow (MBFvol) in dysfunctional myocardium with different viability conditions as defined by fluorine 18 deoxyglucose (FDG) positron emission tomography (PET) while taking into consideration the functional recovery after revascularization. The impact of MBFvol in the diagnosis of functional recovery was also investigated. Methods and Results. Forty-two patients with severe coronary artery disease and dysfunctional myocardium underwent resting oxygen 15 water PET, as well as FDG PET and technetium 99m tetrofosmin single photon emission computed tomography, all attenuation-corrected. Relative FDG and Tc-99m tetrofosmin uptake (normalized to the segment with 100% Tc-99m tetrofosmin uptake), as well as MBFvol (myocardial blood flow multiplied by the water-perfusable tissue fraction to account for the flow to the entire segment volume), were determined in 18 myocardial segments per patient. Viability in dysfunctional segments (estimated by ventriculography) with reduced Tc-99m tetrofosmin uptake of 70% or lower was classified as viable (FDG >70%, mismatch) or nonviable (FDG <70%, match). Fifteen patients underwent revascularization and were followed up. Mismatch segments with improved function were classified as hibernating myocardium. Mean MBFvol in viable myocardium was slightly reduced (0.60 ⴞ 0.02 mL 䡠 min–1 䡠 mL–1) compared with that in normokinetic myocardium (0.64 ⴞ 0.01 mL 䡠 min–1 䡠 mL–1) (P ⴝ .036) and was significantly higher than in nonviable myocardium (0.36 ⴞ 0.01 mL 䡠 min–1 䡠 mL–1) (P < .001). Receiver operating characteristic analysis confirmed an FDG uptake greater than 70% as the optimal threshold to predict functional recovery (diagnostic accuracy [ACC], 76%). MBFvol in hibernating myocardium (0.62 ⴞ 0.04 mL 䡠 min–1 䡠 mL–1) was not significantly reduced compared with that in normokinetic myocardium (0.66 ⴞ 0.02 mL 䡠 min–1 䡠 mL–1) and was significantly higher than in persistently dysfunctional myocardium (0.51 ⴞ 0.04 mL 䡠 min–1 䡠 mL–1) (P < .05). The ACC of MBFvol greater than 0.40 mL 䡠 min–1 䡠 mL–1 as the threshold to predict functional recovery was 61% but did not improve the accuracy of FDG PET by itself. Conclusions. In patients with severe coronary artery disease and dysfunctional myocardium, MBFvol as determined with O-15 water differs significantly between viable and nonviable myocardium as determined by FDG PET and is not significantly reduced in hibernating compared with normokinetic myocardium. Therefore chronically reduced resting blood flow appears unlikely to be the pathophysiologic correlate of the functional state of hibernation. However, MBFvol does not improve the ACC of FDG PET by itself. (J Nucl Cardiol 2003;10:34-45.) Key Words: Oxygen 15 water • positron emission tomography • myocardial blood flow • myocardial viability • myocardial hibernation

From the Departments of Nuclear Medicinea and Internal Medicine I (Cardiology),b University Hospital, Aachen University of Technology, Aachen, Germany Received for publication Feb 7, 2002; final revision accepted July 10, 2002. Reprint requests: Bernd Nowak, MD, Department of Nuclear Medicine, University Hospital, Aachen University of Technology, Pauwelsstrasse 30, 52074 Aachen, Germany; [email protected]. 34

Because of new therapeutic strategies that reduce the mortality rate associated with acute coronary syndromes due to coronary artery disease (CAD), more patients Copyright © 2003 by the American Society of Nuclear Cardiology. 1071-3581/2003/$30.00 ⫹ 0 doi:10.1067/mnc.2003.22

Journal of Nuclear Cardiology Volume 10, Number 1;34-45

Nowak et al Blood flow in hibernating myocardium

35

Figure 1. Study design. H2-15O, O-15 water.

enter the chronic phase of this disease.1 Thus CAD patients with impaired left ventricular function and regional wall motion abnormalities need to be stratified for revascularization procedures more often and more frequently. It has been shown that myocardial regions with chronic contractile dysfunction but preserved viability (hibernating myocardium) may improve in function after coronary revascularization, in contrast to persistently dysfunctional nonviable myocardium resulting from infarction. This was first reported by Diamond et al2 in 1978. Rahimtoola3,4 popularized this concept of hibernating myocardium as a state of chronically reduced contractility in response to a reduction in resting blood supply that can be partially or completely restored to normal either by improving blood flow or by reducing oxygen demand. A number of nuclear imaging studies have validated this mismatch pattern of reduced perfusion but preserved viability in dysfunctional myocardium as an important predictor of recovery of wall motion abnormalities after surgical revascularization.5-8 Positron emission tomography (PET) with the radiolabeled glucose analog 2-[fluorine-18]-fluoro-2-deoxyD-glucose (FDG) is considered one of the most accurate nuclear cardiology methods for detecting regional myocardial viability and is widely used, whereas myocardial perfusion can be assessed by PET or single photon emission computed tomography (SPECT) with the use of several flow markers such as nitrogen 13 ammonia, technetium 99m methoxy-isobutyl-isonitrile (Tc-99m MIBI), Tc-99m tetrofosmin, or thallium 201 chloride. Cellular uptake values of these flow markers are proportional to myocardial blood flow (MBF) over a wide range but also depend on metabolic retention or integrity of cell membrane.9,10 Thus they not only show MBF alone but also give evidence of myocardial viability. Therefore reduced uptake of these flow markers in hibernating myocardium might not be caused solely by reduced blood flow. Hence, decreased resting blood flow as the pathophysiologic explanation for hibernating myocardium has been called into question.1

The aim of our study was an absolute quantification of volumetric myocardial blood flow (MBFvol) with oxygen 15 water PET in dysfunctional myocardium with different viability conditions as determined by FDG PET while taking into consideration the functional recovery after revascularization. We also investigated the impact of resting MBF on the diagnosis of reversibly dysfunctional myocardium. METHODS Patients Forty-two consecutively admitted patients (35 men and 7 women; age, 63 ⫾ 11 years [mean ⫾ SD]; range, 40-78 years; body weight, 77 ⫾ 12 kg) with severe CAD and regional wall motion abnormalities who were scheduled for FDG PET to assess myocardial viability were included in the study. Thirtythree patients had a history of at least one previous myocardial infarction (22 anterior/apical wall, 10 posterior wall, 4 posterolateral wall, 2 anteroseptal wall, and 2 anterolateral/lateral wall). Ten patients had single-vessel disease, twelve had 2-vessel disease, and twenty had 3-vessel disease. The left anterior descending artery was affected in 40 patients, the left circumflex artery in 25, and the right coronary artery in 29. The mean left ventricular ejection fraction of all 42 patients was reduced to 38% ⫾ 13%. Diabetes mellitus was previously diagnosed in 9 patients. The mean fasting glucose level of all patients was 6.9 ⫾ 2.1 mmol/L (range, 4.6-17.7 mmol/L).

Patient Preparation All studies were done on the same day with the patients in the supine position (Figure 1). All patients had fasted previously and were taking their regular cardiac medication. A light meal was served between injection of Tc-99m tetrofosmin and SPECT acquisition to reduce biliary activity. In order to reduce myocardial fatty acid metabolism, all patients received 250 mg acipimox approximately 2 hours before administration of FDG. About 1 hour before injection of FDG, nondiabetic patients received a 50-g oral glucose loading dose. Insulin was admin-

36

Nowak et al Blood flow in hibernating myocardium

istered intravenously to patients with diabetes (2-8 IU) 5 to 10 minutes before administration of FDG.

FDG PET Static FDG PET scans (ECAT EXACT 922/47; SiemensCTI, Knoxville, Tenn) were done 60 minutes after intravenous administration of 223 ⫾ 57 MBq FDG. The acquisition time was 30 minutes for emission (2-dimensional mode) and 15 minutes for transmission (germanium 68/gallium 68 rod sources). Attenuation-corrected images were reconstructed by filtered backprojection (Hann; cutoff, 0.4 Nyquist). The matrix size was 128 ⫻ 128 pixels, with a reconstruction zoom of 2.154.

O-15 Water PET An RDS 111 cyclotron (CTI, Knoxville, Tenn) was used to irradiate nitrogen 15 with 11-MeV protons for production of O-15, which reacted to O-15 water in a water module (CTI). Two dynamic resting O-15 water PET studies were done in every patient and started immediately with the beginning of intravenous infusion of 700 to 1000 MBq O-15 water for 30 seconds with the use of an ECAT EXACT 922/47 PET scanner (Siemens-CTI). Each dynamic study (2-dimensional mode) consisted of 26 frames (10 ⫻ 6 seconds and 16 ⫻ 15 seconds) to a total scan time of 5 minutes. A 15-minute period between both dynamic studies to allow the O-15 radioactivity to decay was used for a transmission scan (Ge-68/Ga-68). Attenuationcorrected images were reconstructed by filtered backprojection (Hann; cutoff, 0.4 Nyquist). The matrix size was 128 ⫻ 128 pixels, with a reconstruction zoom of 2.154.

SPECT Myocardial perfusion SPECT imaging was done 60 minutes after injection of 425 ⫾ 49 MBq Tc-99m tetrofosmin with use of a dual-head gamma camera (Solus; ADAC Laboratories, Milpitas, Calif) equipped with a low-energy, all-purpose collimator. Acquisition parameters, attenuation, and scatter-corrected reconstruction in a 128 ⫻ 128 matrix have been described in detail elsewhere.11 Emission was done with a 360° rotation (180° per head) in 64 steps (128 projections) of 20 seconds each in three independent energy windows: 140 ⫾ 14 keV for emission, 120 ⫾ 6 keV for scatter detection, and 90 ⫾ 11 keV for backscatter detection. Data sets of windows 1 and 2 were processed to obtain a scatter-corrected data set, which was then reconstructed with use of a Butterworth filter (cutoff, 0.7 Nyquist; order, 5; 128 ⫻ 128 matrix) and processed with the data set of window 3 (filtered backprojection; Ramp, 1) to obtain a final segmented attenuation- and scatter-corrected transaxial data set.

Left Ventricular Angiography Initial left ventricular angiography and ventriculography were obtained in the 30° right anterior oblique view and 60° left

Journal of Nuclear Cardiology January/February 2003

anterior oblique view at an interval of 20 ⫾ 20 days (mean ⫾ SD) from the nuclear imaging. Two observers assessed the ventriculography results without knowledge of clinical, angiographic, or nuclear imaging findings. The left ventricular ejection fraction was calculated by the area-length method from the 30° right anterior oblique view, and regional wall motion was expressed as normokinetic, hypokinetic, akinetic, or dyskinetic. A wall motion score (WMS) was assigned to dysfunctional segments as follows: 1, hypokinetic; 2, akinetic; and 3, dyskinetic.

Nuclear Imaging Data Analysis FDG and O-15 water PET data were acquired primarily with the use of ECAT 7.1 software (Siemens-CTI). Transaxial SPECT slices were first converted to ECAT 6.4 file format (including adaption of the pixel size to the pixel size of the PET data sets) and then imported into the ECAT 7.1 software. Next, all PET and SPECT images were reoriented in left ventricular short-axis slices of 1.2 cm in thickness. For reorientation of the dynamic O-15 water data, the reorientation axes of the corresponding FDG PET data were applied to the O-15 water data and controlled for accuracy against early O-15 water study frames for assessment of the O-15 water–filled left ventricle volume. Each of three FDG short-axis slices (apical, midventricular, and basal) was divided into 6 regions of interest (segments) measuring 7 mm in radial diameter (anteroseptal, septal, posteroseptal, posterolateral, lateral, and anterolateral) for a total of 18 segments per data set (Figure 2). These segments were also transferred to the O-15 water PET and Tc-99m tetrofosmin SPECT images. Transfer of these segments to the O-15 water PET data was controlled on dynamic frames showing right and left ventricular filling of O-15 water. Motion artifacts were corrected manually on the short-axis planes. The mean movement of all segments between FDG and O-15 water data was 2.4 ⫾ 0.2 mm (mean ⫾ SD) in the anterior-posterior direction and 1.9 ⫾ 0.1 mm in the medial-lateral direction. Distribution of myocardial Tc-99m tetrofosmin or FDG uptake was assessed by calculating the count densities (counts per pixel) for every segment. Relative segmental Tc-99m tetrofosmin or FDG uptake was then expressed as the percentage of the uptake in the segment with maximum uptake. According to the validated perfusion/metabolism mismatch pattern,12 the diagnosis of myocardial viability by FDG PET was indicated in dysfunctional segments with reduced relative Tc-99m tetrofosmin uptake of 70% or lower and was done as follows: Relative FDG uptake was normalized to the segment with the maximum Tc-99m tetrofosmin uptake (normalized FDG uptake), and myocardial viability was classified as viable (mismatch [normalized FDG uptake ⬎70%]) or nonviable (match [normalized FDG uptake ⱕ70%]). Dysfunctional segments with perfusion/metabolism mismatch (viable) and proven recovery in function after successful revascularization were classified as hibernating myocardium. In addition, the threshold of normalized FDG uptake greater than 70% to define viability was subsequently confirmed by receiver operating characteristic (ROC) analysis in revascularized dysfunctional segments.

Journal of Nuclear Cardiology Volume 10, Number 1;34-45

Nowak et al Blood flow in hibernating myocardium

37

Figure 2. Definition of 18 myocardial segments in 3 short-axis slices (apical, midventricular, and basal). LV, Left ventricular input region of interest.

MBF and water-perfusable tissue fraction (PTF) were calculated with the use of a modified single-tissue compartment kinetic model13,14 that was developed by Iida et al,15,16 who introduced the concept of PTF. PTF reflects the ratio of the mass of water-perfusable tissue within a given segment and the volume of that segment (in grams per milliliter).15 The calculation of PTF has been described in detail previously.15,16 The ratio of influx to efflux rate constants in the model describing the behavior of water in the myocardium is directly proportional to PTF, with the constant of proportionality being the tissue/blood pool partition coefficient of water. Therefore PTF can be determined by this kinetic model analysis assuming a constant tissue/blood pool partition coefficient of 0.91 g/mL, based on in vitro water content data.16 Segmental MBF reflects the myocardial flow rate per mass of PTF only, whereas flow rate per total volume of the myocardial segment can be calculated by multiplying segmental MBF by PTF. This results in non–partial volume– corrected blood flow values reflecting the mean blood flows to the entire segment volumes—therefore called volumetric myocardial blood flow (MBFvol)16—which are directly comparable to volumetric data like FDG or Tc-99m tetrofosmin uptake values. Therefore this parameter was the focus of attention in our comparative investigations. MBFvol, MBF, and PTF were determined for every myocardial segment with a previously validated fitting routine17 with the use of a left ventricular input function.18 Values for every segment were calculated as the mean from values of both dynamic O-15 water PET studies. Mean MBFvol, MBF, and PTF were calculated for normokinetic remote myocardium as well as for all dysfunctional segments classified as viable, nonviable, hibernating, or unimproved in function after revascularization.

Follow-up Twenty patients underwent successful revascularization 4.9 ⫾ 4.8 weeks (mean ⫾ SD) after the nuclear imaging

studies. Fifteen of these patients could be recruited for follow-up examination to evaluate the recovery of regional wall motion abnormalities. Follow-up assessment by two observers was done by ventriculography in 7 patients 6.4 ⫾ 0.7 months (mean ⫾ SD) after percutaneous transluminal angioplasty (n ⫽ 6) or coronary bypass graft surgery (n ⫽ 1). Transthoracic echocardiography with the use of an HP Sonos 5500 (Hewlett Packard, Andover, Mass) with harmonic capabilities (s4 transducer) was carried out in 8 patients 17.1 ⫾ 4.5 months (mean ⫾ SD) after percutaneous transluminal angioplasty (n ⫽ 2) or coronary bypass graft surgery (n ⫽ 6) to estimate recovery of regional wall motion. Parasternal long- and short-axis views and apical 2- and 4-chamber views were recorded on videotape, analyzed, and visually matched with the preoperative ventriculograms. Two observers unaware of preoperative angiographic, functional, or nuclear imaging data and postoperative clinical findings analyzed the data. Dysfunctional segments were classified as having improved in function if the corresponding WMS was reduced for at least one point after revascularization and as unimproved if it remained unchanged.

Statistics All statistical analyses except ROC analyses were done with SPSS 10 (SPSS Inc, Chicago, Ill). Data are shown as mean ⫾ SEM unless stated otherwise. Differences of mean values were tested for significance by a nonparametric rank-sum test (Mann-Whitney U test). P ⬍ .05 was considered significant. ROC analysis was done with Rockit 9.1 software (provided by Charles E. Metz et al, Department of Radiology, University of Chicago, Chicago, Ill19), which estimates areas below the ROC curves and corresponding asymmetric 95% confidence intervals, indicating the likelihood of correct test decisions. Sensitivity, specificity, positive and negative predictive values (PPV and NPV, respectively), and diagnostic accuracy (ACC) were calculated for the most appropriate threshold value of the normalized relative FDG uptake.

38

Nowak et al Blood flow in hibernating myocardium

Journal of Nuclear Cardiology January/February 2003

Table 1. Regional distribution of mean MBFvol, MBF, and PTF values in normokinetic segments of all 42 patients, with segments of 3 short-axis slices grouped by location

Segments (Values for 3 short-axis slices)

n

MBFvol (mL 䡠 min–1 䡠 mL–1)

MBF (mL 䡠 min–1 䡠 g–1)

PTF (g/mL)

Anteroseptal (1⫹7⫹13) Septal (2⫹8⫹14) Posteroseptal (3⫹9⫹15) Posterolateral (4⫹10⫹16) Lateral (5⫹11⫹17) Anterolateral (6⫹12⫹18)

41 38 60 58 88 53

0.58 ⫾ 0.03 0.62 ⫾ 0.03 0.61 ⫾ 0.03 0.69 ⫾ 0.03* 0.68 ⫾ 0.02‡ 0.61 ⫾ 0.02

0.84 ⫾ 0.04 0.85 ⫾ 0.03 0.84 ⫾ 0.04 0.98 ⫾ 0.04† 1.07 ⫾ 0.04§ 0.92 ⫾ 0.03

0.70 ⫾ 0.01 0.74 ⫾ 0.02 0.74 ⫾ 0.02 0.76 ⫾ 0.02 0.65 ⫾ 0.01㛳 0.67 ⫾ 0.01¶

*P ⬍ .05 vs anteroseptal and posteroseptal. † P ⬍ .01 vs posteroseptal, and P ⬍ .05 vs anteroseptal and septal. ‡ P ⬍ .05 vs posteroseptal. § P ⬍ .001 vs anteroseptal, septal, and posteroseptal, and P ⬍ .05 vs anterolateral. 㛳 P ⬍ .001 vs septal, posteroseptal, and posterolateral, and P ⬍ .01 vs anteroseptal. ¶ P ⬍ .01 vs septal, posteroseptal, and posterolateral.

RESULTS Regional Variation of MBF in Normokinetic Myocardium Regional distributions of mean MBFvol, MBF, and PTF in normokinetic segments of all patients are summarized in Table 1. MBFvol in lateral or posterolateral segments was slightly but significantly higher than MBFvol in anteroseptal or posteroseptal segments. Left ventricular distribution of MBFvol was more homogenous than distribution of MBF. MBF in Viable or Nonviable Myocardium in All Patients The results and distribution of normokinetic remote myocardium and dysfunctional myocardial segments with mismatch (viable) or match (nonviable) pattern in all 42 patients are summarized in Table 2. Mean MBFvol in normokinetic remote myocardial segments was 0.64 ⫾ 0.01 mL 䡠 min–1 䡠 mL–1. Mean MBFvol in segments classified as viable myocardium was slightly reduced, at 0.60 ⫾ 0.02 mL 䡠 min–1 䡠 mL–1 (P ⫽ .036). When compared with MBFvol of remote myocardium and viable myocardium, MBFvol of nonviable myocardium was significantly reduced (P ⬍ .001), at 0.36 ⫾ 0.01 mL 䡠 min–1 䡠 mL–1 (– 44% vs remote myocardium and – 40% vs viable myocardium). Relative Tc-99m tetrofosmin uptake was significantly lower (P ⬍ .001) in viable (58% ⫾ 1.0%) and nonviable myocardium (46% ⫾ 1.2%) than in remote myocardium (85% ⫾ 0.6%), with slightly but significantly higher values for viable versus nonviable myocardium (P ⬍ .001).

ACC of FDG PET for Predicting Functional Recovery of Dysfunctional Segments (ROC Analysis) As improved contractile performance is considered to be the gold standard for viability of dysfunctional myocardium, we performed analyses in revascularized patients. Fifteen patients underwent successful revascularization and were available for follow-up examination. These patients had 72 dysfunctional segments with match or mismatch pattern, as determined by nuclear imaging techniques. Of these 72 segments, 40 improved in function whereas 32 remained unchanged. Figure 3 and Table 3 show the results of FDG PET in predicting functional recovery. Thresholds for normalized FDG uptake of greater than 60% and greater than 70% yielded comparable ACC (78% and 76%), but 60% as the threshold was accompanied by an unacceptably low specificity of 63%; therefore a normalized FDG uptake of greater than 70% was picked as the best value to predict functional recovery with well-balanced sensitivity (80%) and specificity (72%). MBF in Viable or Nonviable Myocardium in Revascularized Patients The results and distribution of normokinetic remote myocardium and dysfunctional myocardial segments with mismatch (viable) or match (nonviable) patterns in 15 revascularized patients are summarized in Table 4. Mean MBFvol in normokinetic segments was 0.66 ⫾ 0.02 mL 䡠 min–1 䡠 mL–1. Mean MBFvol in segments classified as viable myocardium was not significantly reduced, at 0.61 ⫾ 0.03 mL 䡠 min–1 䡠 mL–1. When compared with MBFvol of remote myocardium and

Journal of Nuclear Cardiology Volume 10, Number 1;34-45

Nowak et al Blood flow in hibernating myocardium

39

Table 2. Summary of results from normokinetic remote myocardial segments and dysfunctional myocardial segments of 42 patients classified by FDG PET as mismatch (viable) or match (nonviable)

All patients (n ⴝ 42)

Remote

Segments Normokinesia Hypokinesia Akinesia Dyskinesia WMS Normalized FDG uptake (%) Tetrofosmin uptake (%) MBFvol (mL 䡠 min–1 䡠 mL–1) MBF (mL 䡠 mL–1 䡠 g–1) PTF (g/mL)

338 338 — — — — 96 ⫾ 1.1 85 ⫾ 0.6 0.64 ⫾ 0.01 0.94 ⫾ 0.02 0.70 ⫾ 0.01

Viable 103 — 49 50 4 1.56 ⫾ 0.06 100 ⫾ 4.4† 58 ⫾ 1.0‡ 0.60 ⫾ 0.02§ 0.87 ⫾ 0.03㛳 0.72 ⫾ 0.02

Nonviable 143 — 29 109 5 1.83 ⫾ 0.04* 50 ⫾ 1.3*‡ 46 ⫾ 1.2*‡ 0.36 ⫾ 0.01*‡ 0.64 ⫾ 0.02*‡ 0.65 ⫾ 0.02*‡

Viable, Dysfunctional segments with tetrofosmin/FDG mismatch; Nonviable, dysfunctional segments with tetrofosmin/FDG match, Normalized FDG uptake, relative FDG uptake normalized to segment with maximum tetrofosmin uptake. *P ⬍ .001 vs viable. † P ⬍ .01 vs remote. ‡ P ⬍ .001 vs remote. § P ⫽ .036 vs remote. 㛳 P ⫽ .025 vs remote.

Figure 3. FDG PET–fitted ROC curve for predicting improved contractile performance of dysfunctional myocardial segments after revascularization. A normalized relative FDG uptake greater than 70% was estimated as the threshold with the highest ACC and the best-balanced sensitivity-specificity proportion for predicting functional improvement. Data from 15 revascularized patients who were recruited for follow-up are shown.

viable myocardium, MBFvol of nonviable myocardium was significantly reduced, at 0.46 ⫾ 0.04 mL 䡠 min–1 䡠 mL–1 (–30% vs remote myocardium [P ⬍ .001] and

–25% vs viable myocardium [P ⬍ .01]). The mean WMS did not differ significantly in the viable and nonviable groups; however, recovery was significantly better in the viable group (⌬WMS ⫽ –1.20 ⫾ 0.14) than in the nonviable group (⌬WMS ⫽ – 0.29 ⫾ 0.10) (P ⬍ .001). Again, relative Tc-99m tetrofosmin uptake was significantly lower (P ⬍ .001) in viable (62% ⫾ 1.1%) and nonviable myocardium (53% ⫾ 2.0%) compared with remote myocardium (85% ⫾ 0.8%), with slightly but significantly higher values for viable versus nonviable myocardium (P ⬍ .001). Hibernating segments versus normokinetic and unimproved segments are shown in Table 5. Mean MBFvol in hibernating myocardium was not significantly reduced, at 0.62 ⫾ 0.04 mL 䡠 min–1 䡠 mL–1, whereas MBFvol in persistently dysfunctional myocardium was significantly reduced, at 0.51 ⫾ 0.04 mL 䡠 min–1 䡠 mL–1 (–23% vs remote myocardium [P ⬍ .001] and –18% vs viable myocardium [P ⬍ .05]). In contrast, mean relative Tc-99m tetrofosmin uptake values did not differ significantly between hibernating (61% ⫾ 1.4%) and persistently dysfunctional myocardium (56% ⫾ 2.2%) (P ⫽ not significant [NS]). As shown in Table 6, the time interval between PET and revascularization did not differ significantly between the hibernation group and the unimproved group and thus did not contribute to differences in functional recovery. The optimum threshold for MBFvol alone to predict functional recovery was greater than 0.4 mL 䡠 min–1 䡠 mL–1. With the use of this threshold, sensitivity, speci-

40

Nowak et al Blood flow in hibernating myocardium

Journal of Nuclear Cardiology January/February 2003

Table 3. Results of ROC analysis according to Figure 3

AUC (95% CI) Normalized FDG uptake

Threshold Sensitivity Specificity

0.79 (0.65-0.89)

⬎50%

38/40 (95%)

7/32 (22%)

⬎60% ⬎70% ⬎80%

36/40 (90%) 32/40 (80%) 22/40 (55%)

20/32 (63%) 23/32 (72%) 27/32 (84%)

PPV 38/63 (60%)

NPV

ACC

7/9 (78%) 45/72 (63%)

36/48 (75%) 20/24 (83%) 56/72 (78%) 32/41 (78%) 23/31 (74%) 55/72 (76%) 22/27 (81%) 27/45 (60%) 49/72 (68%)

AUC, Area under curve; CI, confidence interval; Normalized FDG uptake, relative FDG uptake (%) normalized to segment with maximum tetrofosmin uptake.

Table 4. Summary of results of normokinetic remote myocardial segments and dysfunctional myocardial segments of 15 revascularized patients classified by FDG PET as mismatch (viable) or match (nonviable)

Revascularized patients (n ⴝ 15)

Remote

Segments Normokinesia Hypokinesia Akinesia Dyskinesia WMS ⌬WMS Normalized FDG uptake (%) Tetrofosmin uptake (%) MBFvol (mL 䡠 min–1 䡠 g–1) MBF (mL 䡠 min–1 䡠 g–1) PTF (g/mL)

141 141 — — — — — 93 ⫾ 1.0 85 ⫾ 0.8 0.66 ⫾ 0.02 0.95 ⫾ 0.03 0.71 ⫾ 0.01

Viable 41 — 14 23 4 1.76 ⫾ 0.10 –1.20 ⫾ 0.14 86 ⫾ 1.0† 62 ⫾ 1.1† 0.61 ⫾ 0.03 § 0.85 ⫾ 0.05 0.77 ⫾ 0.04

Nonviable 31 — 11 15 5 1.81 ⫾ 0.13 –0.29 ⫾ 0.10* 52 ⫾ 2.3*† 53 ⫾ 2.0*† 0.46 ⫾ 0.04†‡ 0.74 ⫾ 0.06㛳 0.70 ⫾ 0.06‡¶

Viable, Dysfunctional segments with tetrofosmin/FDG mismatch; Nonviable, dysfunctional segments with tetrofosmin/FDG match; ⌬WMS, difference in WMS after revascularization; Normalized FDG uptake, relative FDG uptake normalized to segment with maximum tetrofosmin uptake. *P ⬍ .001 vs viable. † P ⬍ .001 vs remote. ‡ P ⬍ .01 vs viable. § P ⫽ .046 vs remote. 㛳 P ⬍ .01 vs remote. ¶ P ⬍ .05 vs remote.

ficity, PPV, NPV, and ACC of MBFvol to predict recovery were 80%, 38%, 62%, 60%, and 61%, respectively. The addition of MBFvol did not improve the ACC of FDG PET by itself. DISCUSSION MBF in Viable and Hibernating Myocardium The concept of hibernating myocardium popularized by Rahimtoola3,4 included chronically reduced resting MBF as the pathophysiologic correlate of left ventricular dysfunction. This hypothesis was based mainly on flow measurements using Tl-201 SPECT imaging, which does not permit absolute quantification of MBF. Tl-201

SPECT, however, as well as Tc-99m MIBI and Tc-99m tetrofosmin SPECT, tends to overestimate the severity of perfusion defects due to partial volume effects resulting from low spatial resolution (hibernating myocardium does not thicken during systole).1 Furthermore, it is well known that myocardial retention of all perfusion markers used for identifying hibernating myocardium as perfusion/metabolism mismatch5-7—Tl-201, Tc-99m flow markers, and N-13 ammonia— depend not only on myocardial perfusion but also on cell integrity,9,10 which has been shown to be impaired in hibernating myocardium.20 This might explain the comparable reduction in relative Tc-99m tetrofosmin uptake values in hibernating and persistently dysfunctional myocardium despite signifi-

Journal of Nuclear Cardiology Volume 10, Number 1;34-45

Nowak et al Blood flow in hibernating myocardium

41

Table 5. Summary of results from normokinetic remote myocardial segments and dysfunctional myocardial segments of 15 revascularized patients classified as hibernating myocardium or persistently dysfunctional (unimproved) myocardium

Revascularized patients (n ⴝ 15)

Remote

Hibernating

Unimproved

Segments Normokinesia Hypokinesia Akinesia Dyskinesia WMS ⌬WMS Normalized FDG uptake (%) Tetrofosmin uptake (%) MBFvol (mL 䡠 min–1 䡠 mL–1) MBF (mL 䡠 min–1 䡠 g–1) PTF (g/mL)

141 141 — — — — — 93 ⫾ 1.0 85 ⫾ 0.8 0.66 ⫾ 0.02 0.95 ⫾ 0.03 0.71 ⫾ 0.01

32 — 8 21 3 1.84 ⫾ 0.10 –1.53 ⫾ 0.12 84 ⫾ 1.5* 61 ⫾ 1.4* 0.62 ⫾ 0.04 0.88 ⫾ 0.05 0.76 ⫾ 0.04

32 — 14 12 6 1.75 ⫾ 0.13 — 62 ⫾ 4.3*† 56 ⫾ 2.2* 0.51 ⫾ 0.04*‡ 0.76 ⫾ 0.07‡§ 0.75 ⫾ 0.05

Hibernating, Dysfunctional segments with tetrofosmin/FDG mismatch and improved function after revascularization; Unimproved, dysfunctional segments with reduced tetrofosmin uptake without improvement in function after revascularization; ⌬WMS, difference in WMS after revascularization; Normalized FDG uptake, relative FDG uptake normalized to segment with maximum tetrofosmin uptake. *P ⬍ .001 vs remote. † P ⬍ .001 vs hibernating. ‡ P ⬍ .05 vs hibernating. § P ⬍ .01 vs remote.

Table 6. Time interval between nuclear imaging studies and revascularization procedure

Revascularized patients (n ⴝ 15) Dysfunctional segments Time from PET to revascularization (wk)

Hibernating

Unimproved

32 4.69 ⫾ 0.83

32 3.53 ⫾ 0.37

The time interval did not differ significantly between hibernating and persistently dysfunctional (unimproved) myocardial segments and, therefore, did not contribute to the differences observed in functional recovery. Hibernating, Dysfunctional segments with tetrofosmin/FDG mismatch and improved function after revascularization; Unimproved, dysfunctional segments with reduced tetrofosmin uptake without improvement in function after revascularization.

cant differences in blood flow values observed in our study. PET with O-15 water used as a freely diffusible flow tracer overcomes these physical and biochemical limitations and allows absolute quantification of MBF and MBFvol.15,16,18,21 Indeed, a correlation in dysfunctional myocardium has been described between resting MBF and contractile reserve during infusion of dobutamine as an indicator of preserved viability.22 Almost normal ranges of volumetric blood flow in hibernating myocardium compared with remote myocardial regions in patients with left ventricular dysfunction have been reported.23,24 In 30 patients with previous myocardial infarction, Marinho et al23 re-

ported MBFvol values of 0.71 ⫾ 0.23 mL 䡠 min–1 䡠 mL–1 (mean ⫾ SD) in normal myocardial segments versus 0.64 ⫾ 0.27 mL 䡠 min–1 䡠 mL–1 (P ⫽ NS) in dysfunctional segments that improved in function after revascularization. In 16 patients with chronic dysfunction of the anterior wall studied by Gerber et al,24 MBFvol was 0.56 ⫾ 0.05 mL 䡠 min–1 䡠 mL–1 (mean ⫾ SEM) in normokinetic remote myocardial segments versus 0.47 ⫾ 0.05 mL 䡠 min–1 䡠 mL–1 in reversibly dysfunctional segments (P ⫽ NS). Summarized on the basis of 46 patients, these two studies indicate a slight 12% reduction of MBFvol in hibernating versus remote myocardium (0.58 mL 䡠 min–1 䡠 mL–1 vs 0.66 mL 䡠 min–1 䡠 mL–1), which is still within

42

Nowak et al Blood flow in hibernating myocardium

the range of flow measured in normal subjects and cannot account for the extend of dysfunction in patients with severe chronic CAD. These results are confirmed by our study. If resting MBF is not significantly reduced in hibernating myocardium, the concept of myocardial stunning introduced by Braunwald and Kloner25 as reversible contractile dysfunction after acute myocardial ischemia followed by reperfusion might be responsible for chronic dysfunctional myocardium. Indeed, in a pig model, Shen and Vatner26 observed repetitive episodes of ischemia followed by cumulative stunning that led to chronic reduction in ventricular function. Pagano et al27 found normal resting blood flow but a significantly reduced coronary flow reserve in hibernating myocardium after infusion of dipyridamole in 22 patients. Therefore repetitive stunning as the trigger of chronic left ventricular dysfunction is considered an important component of the definition of hibernating myocardium.1

Impact of Resting MBF in Prediction of Functional Recovery MBFvol was the parameter estimated from dynamic water studies that differentiated hibernating from persistently dysfunctional myocardium with the highest selectivity and was considered to be the likeliest parameter to contribute to detection of myocardial viability and to prediction of contractile recovery of dysfunctional myocardium after revascularization. However, despite the observation that MBFvol was not significantly reduced in hibernating and reversibly dysfunctional myocardium, the ACC of MBFvol for predicting functional recovery was rather low and did not improve the accuracy of FDG PET by itself. Seen in this context, it is worth mentioning that the accuracy of FDG PET in predicting functional recovery in our study matches the results reported in a meta-analysis by Di Carli28 (sensitivity, specificity, PPV, NPV, and ACC of 84%, 70%, 76%, 82%, and 79%, respectively). Limited data are available on the assessment of myocardial viability by resting blood flow alone. Most investigators used dynamic N-13 ammonia PET. Gewirtz et al29 examined 26 patients with chronic myocardial infarction and showed that the presence of viable myocardium was very unlikely if blood flow was less than 0.25 mL 䡠 min 䡠 g–1. They also noted a moderate correlation between relative FDG uptake and relative regional blood flow (r ⫽ 0.63, P ⬍ .001). Using a threshold MBF of 0.64 mL 䡠 min 䡠 g–1, Grandin et al30 correctly classified viable and nonviable myocardium

Journal of Nuclear Cardiology January/February 2003

with a sensitivity of 70% (12/17 cases) and a specificity of 88% (7/8 cases). Finally, Gerber et al31 observed basal MBF within normal limits in 20 of 24 patients (83%) with reversible anterior wall dysfunction. We are aware of at least one study attempting to define myocardial viability by resting blood flow with the use of carbon 11 acetate as a flow marker.32 When 0.5 mL 䡠 min 䡠 g–1 was defined as the cutoff value, MBF turned out to be an independent predictor of reversible dysfunction in 30 patients with chronic myocardial infarction with positive and negative predictive accuracies of 79% and 90%, respectively. To our knowledge, there are only two studies that have assessed myocardial viability with the use of MBF as determined by O-15 water PET.23,33 Both used the partial volume– corrected MBF value, which reflects blood flow in the water-PTF only and thus tends to overestimate blood flow to segments consisting of a mixture of perfusable and viable tissue and fibrotic tissue.24 Marinho et al23 used a very low threshold of 0.42 mL 䡠 min 䡠 g–1, which predicted functional recovery in 30 patients with a sensitivity, specificity, PPV, and NPV of 90%, 10%, 55%, and 45%, respectively. Conversano et al33 used a threshold of 0.5 mL 䡠 min 䡠 g–1 and predicted functional recovery in 17 patients with a sensitivity of 63% (20/32 segments). However, Conversano et al also noted a markedly variable blood flow in reversibly dysfunctional segments, which generally impairs viability classification by resting blood flow as determined by O-15 water PET. Furthermore, blood flow deficits of intermediate severity are difficult to interpret, as they may represent extensive subendocardial necrosis with normal myocardium (unlikely to show improved function after revascularization) or else may reflect the coexistence of ischemic but viable areas with either normal or scar tissue, which are likely to improve in function after revascularization.28 Different approaches by which to assess myocardial viability with parameters estimated from O-15 water PET and carbon O-15 PET use the PTF and the perfusable tissue index (PTI), calculated as the ratio of PTF and anatomic tissue fraction (which equals total extravascular tissue, obtained from transmission and C–O-15 data). Our PTF values of remote myocardium match well with those reported in the literature.16,24 However, although PTF has been proposed as a measure of tissue viability,16 the PTF values in our study were only slightly lower in nonviable compared with viable dysfunctional or remote myocardium. This is in contrast to reports from Yamamoto et al34 and Gerber et al,24 who describe distinctly higher PTF values in viable compared with nonviable myocardium. Nonviable myocardium in these studies con-

Journal of Nuclear Cardiology Volume 10, Number 1;34-45

sisted of complete necrosis, as the enrolled patients had transmural Q-wave infarction. However, significantly diminished PTF in transmural infarcted myocardium compared with remote myocardium has been shown by Iida et al35 in canine studies of old myocardial infarction, whereas PTF in nontransmural infarction was similar to that in remote myocardium. Nontransmural infarcted myocardium consisting of subendocardial necrosis and normal myocardium may contribute substantially to the unimproved segments in our study and therefore explain the relatively high PTF in these segments. This hypothesis is further supported by the relatively high FDG uptake of 62% ⫾ 4.3% in this group. Moreover, PTF not only reflects the amount of water-perfusable tissue but also depends on heterogeneity of myocardial perfusion.36 The influence of heterogeneous blood flow on our PTF values cannot be ruled out, as a considerable number of the patients in our study had 3-vessel disease. Another approach to delineate viability is the use of PTI. A PTI threshold value of 0.7 differentiated viable from nonviable myocardium with high accuracy34,37 but required an additional blood pool scan with C–O-15. Study Limitations Follow-up evaluations of revascularized patients differed with regard to the time and method of examination. Seven patients were followed up with the use of ventriculography 6.4 ⫾ 0.7 months (mean ⫾ SD) after revascularization, and eight underwent transthoracic echocardiography 17.1 ⫾ 4.5 months after revascularization. Recovery of regional and global left ventricular function after successful revascularization of hibernating myocardium has been shown to be progressive, following a monoexponential time course.38 However, currently available data from two studies only proved further functional improvement from 2 to 6 months and from 3 to 14 months after revascularization.38,39 Analysis of these two publications suggests that rather negligible changes in contractile function of hibernating myocardium between 6 and 14 months after revascularization can be expected. It is therefore unlikely that the different times of follow-up in our study significantly affected our findings. Assessment of regional wall motion abnormalities was done by two observers who read and matched ventriculography and echocardiography films without knowledge of accompanying clinical, angiographic, or nuclear imaging findings. Thus a most accurate comparison of both methods was ensured. Only short-axis slices could be evaluated with our overlay algorithm. Therefore the apical segment was not

Nowak et al Blood flow in hibernating myocardium

43

included in our investigation. The aim of our study was to define viable and nonviable myocardium as determined by FDG PET and to confirm hibernating myocardium by follow-up investigations after revascularization. Mean MBFvol in these groups and the diagnostic impact of MBFvol ought to be established. This requires consistent definitions of viability by using established nuclear imaging procedures (ie, normalization of relative uptake values to segments with maximum uptake). Because none of our patients had maximum apical FDG or Tc-99m tetrofosmin uptake in the clinical routine evaluation, our normalization algorithm for relative quantification and normalization of FDG and Tc-99m tetrofosmin uptake was not affected by the absence of the apical segment.

Conclusions The results of this study indicate that in patients with severe CAD and left ventricular dysfunction, resting MBFvol as determined with O-15 water PET differs significantly between viable and nonviable myocardium as defined by FDG PET and is not reduced in hibernating compared with normokinetic myocardium. Therefore other mechanisms such as repeated episodes of stunning might instead be the underlying cause of the functional state of hibernating myocardium. On an individual basis, resting MBFvol allows assessment of myocardial viability with low accuracy and does not improve ACC of FDG PET by itself in predicting functional recovery after revascularization.

Acknowledgment We thank Anne Bol and Ann Coppens, Positron Emission Tomography Laboratory, University of Louvain Medical School, Louvain-la-Neuve, Belgium, for their kind assistance with O-15 water PET studies and for providing the software (Mediman 4.0 and Models 3.1) for calculating MBF. Thanks are also due to A. Rodo´n for language and style editing. The authors have indicated they have no financial conflicts of interest.

References 1. Camici PG, Rimoldi O. Blood flow in myocardial hibernation. Curr Opin Cardiol 1998;13:409-14. 2. Diamond GA, Forrester JS, deLuz PL, Wyatt HL, Swan HJ. Post-extrasystolic potentiation of ischemic myocardium by atrial stimulation. Am Heart J 1978;95:204-9.

44

Nowak et al Blood flow in hibernating myocardium

3. Rahimtoola SH. A perspective on the three large multicenter randomized clinical trials of coronary bypass surgery for chronic stable angina. Circulation 1985;72:V123-35. 4. Rahimtoola SH. The hibernating myocardium. Am Heart J 1989; 117:211-21. 5. Tillisch J, Brunken R, Marshall R, Schwaiger M, Mandelkern M, Phelps M, et al. Reversibility of cardiac wall-motion abnormalities predicted by positron tomography. N Engl J Med 1986; 314:884-8. 6. Tamaki N, Yonekura Y, Yamashita K, Saji H, Magata Y, Senda M, et al. Positron emission tomography using fluorine-18 deoxyglucose in evaluation of coronary artery bypass grafting. Am J Cardiol 1989;64:860-5. 7. Lucignani G, Paolini G, Landoni C, Zuccari M, Paganelli G, Galli L, et al. Presurgical identification of hibernating myocardium by combined use of technetium-99m hexakis 2-methoxyisobutylisonitrile single photon emission tomography and fluorine-18 fluoro-2-deoxy-D-glucose positron emission tomography in patients with coronary artery disease. Eur J Nucl Med 1992;19: 874-81. 8. vom Dahl J, Altehoefer C, Sheehan FH, Buechin P, Uebis R, Messmer BJ, et al. Recovery of regional left ventricular dysfunction after coronary revascularization. Impact of myocardial viability assessed by nuclear imaging and vessel patency at follow-up angiography. J Am Coll Cardiol 1996;28:948-58. 9. Rauch B, Helus F, Grunze M, Braunwell E, Mall G, Hasselbach W, et al. Kinetics of 13N-ammonia uptake in myocardial single cells indicating potential limitations in its applicability as a marker of myocardial blood flow. Circulation 1985;71:387-93. 10. Arbab AS, Koizumi K, Toyama K, Arai T, Araki T. Technetium99m-tetrofosmin, technetium-99m-MIBI and thallium-201 uptake in rat myocardial cells. J Nucl Med 1998;39:266-71. 11. Kupferschlager J, Muller B, Schulz G, Kaiser HJ, Wagenknecht G, Bares R, et al. A method for combined scatter and attenuation correction without transmission measurement for myocardial SPECT with 99m-Tc binding [in German]. Nuklearmedizin 1997; 36:56-64. 12. vom Dahl J, Altehoefer C, Sheehan FH, Buechin P, Schulz G, Schwarz ER, et al. Effect of myocardial viability assessed by technetium-99m-sestamibi SPECT and fluorine-18-FDG PET on clinical outcome in coronary artery disease. J Nucl Med 1997;38: 742-8. 13. Kety S. Theory of blood-tissue exchange and its application to measurement of blood flow. Methods Med Res 1960;8:223-7. 14. Kety S. Measurement of local blood flow by the exchange of an inert, diffusible substance. Methods Med Res 1960;8:228-36. 15. Iida H, Kanno I, Takahashi A, Miura S, Murakami M, Takahashi K, et al. Measurement of absolute myocardial blood flow with H215O and dynamic positron-emission tomography. Strategy for quantification in relation to the partial-volume effect. Circulation 1988;78:104-15. 16. Iida H, Rhodes CG, de Silva R, Yamamoto Y, Araujo LI, Maseri A, et al. Myocardial tissue fraction— correction for partial volume effects and measure of tissue viability. J Nucl Med 1991;32:216975. 17. Bol A, Melin JA, Vanoverschelde JL, Baudhuin T, Vogelaers D, De Pauw M, et al. Direct comparison of [13N]ammonia and [15O]water estimates of perfusion with quantification of regional myocardial blood flow by microspheres. Circulation 1993;87:51225. 18. Iida H, Rhodes CG, de Silva R, Araujo LI, Bloomfield PM, Lammertsma AA, et al. Use of the left ventricular time-activity curve as a noninvasive input function in dynamic oxygen-15-water positron emission tomography. J Nucl Med 1992;33:1669-77.

Journal of Nuclear Cardiology January/February 2003

19. Metz CE, Herman BA, Roe CA. Statistical comparison of two ROC-curve estimates obtained from partially-paired datasets. Med Decis Making 1998;18:110-21. 20. Schwarz ER, Schaper J, vom Dahl J, Altehoefer C, Grohmann B, Schoendube F, et al. Myocyte degeneration and cell death in hibernating human myocardium. J Am Coll Cardiol 1996;27:1577-85. 21. Iida H, Takahashi A, Tamura Y, Ono Y, Lammertsma AA. Myocardial blood flow: comparison of oxygen-15-water bolus injection, slow infusion and oxygen-15-carbon dioxide slow inhalation. J Nucl Med 1995;36:78-85. 22. Panza JA, Dilsizian V, Curiel RV, Unger EF, Laurienzo JM, Kitsiou AN. Myocardial blood flow at rest and contractile reserve in patients with chronic coronary artery disease and left ventricular dysfunction. J Nucl Cardiol 1999;6:487-94. 23. Marinho NV, Keogh BE, Costa DC, Lammertsma AA, Ell PJ, Camici PG. Pathophysiology of chronic left ventricular dysfunction. New insights from the measurement of absolute myocardial blood flow and glucose utilization. Circulation 1996;93:737-44. 24. Gerber BL, Melin JA, Bol A, Labar D, Cogneau M, Michel C, et al. Nitrogen-13-ammonia and oxygen-15-water estimates of absolute myocardial perfusion in left ventricular ischemic dysfunction. J Nucl Med 1998;39:1655-62. 25. Braunwald E, Kloner RA. The stunned myocardium: prolonged, postischemic ventricular dysfunction. Circulation 1982;66:1146-49. 26. Shen YT, Vatner SF. Mechanism of impaired myocardial function during progressive coronary stenosis in conscious pigs. Hibernation versus stunning? Circ Res 1995;76:479-88. 27. Pagano D, Townend JN, Parums DV, Bonser RS, Camici PG. Hibernating myocardium: morphological correlates of inotropic stimulation and glucose uptake. Heart 2000;83:456-61. 28. Di Carli MF. Predicting improved function after myocardial revascularization. Curr Opin Cardiol 1998;13:415-24. 29. Gewirtz H, Fischman AJ, Abraham S, Gilson M, Strauss HW, Alpert NM. Positron emission tomographic measurements of absolute regional myocardial blood flow permits identification of nonviable myocardium in patients with chronic myocardial infarction. J Am Coll Cardiol 1994;23:851-9. 30. Grandin C, Wijns W, Melin JA, Bol A, Robert AR, Heyndrickx GR, et al. Delineation of myocardial viability with PET. J Nucl Med 1995;36:1543-52. 31. Gerber BL, Vanoverschelde JL, Bol A, Michel C, Labar D, Wijns W, et al. Myocardial blood flow, glucose uptake, and recruitment of inotropic reserve in chronic left ventricular ischemic dysfunction. Implications for the pathophysiology of chronic myocardial hibernation. Circulation 1996;94:651-59. 32. Wolpers HG, Burchert W, van den Hoff J, Weinhardt R, Meyer GJ, Lichtlen PR. Assessment of myocardial viability by use of 11Cacetate and positron emission tomography. Threshold criteria of reversible dysfunction. Circulation 1997;95:1417-24. 33. Conversano A, Walsh JF, Geltman EM, Perez JE, Bergmann SR, Gropler RJ. Delineation of myocardial stunning and hibernation by positron emission tomography in advanced coronary artery disease. Am Heart J 1996;131:440-50. 34. Yamamoto Y, de Silva R, Rhodes CG, Araujo LI, Iida H, Rechavia E, et al. A new strategy for the assessment of viable myocardium and regional myocardial blood flow using 15O-water and dynamic positron emission tomography. Circulation 1992;86:167-78. 35. Iida H, Tamura Y, Kitamura K, Bloomfield PM, Eberl S, Ono Y. Histochemical correlates of 15O-water-perfusable tissue fraction in experimental canine studies of old myocardial infarction. J Nucl Med 2000;41:1737-45. 36. Herrero P, Staudenherz A, Walsh JF, Gropler RJ, Bergmann SR. Heterogeneity of myocardial perfusion provides the physiological basis of perfusable tissue index. J Nucl Med 1995;36:320-7.

Journal of Nuclear Cardiology Volume 10, Number 1;34-45

37. de Silva R, Yamamoto Y, Rhodes CG, Iida H, Nihoyannopoulos P, Davies GJ, et al. Preoperative prediction of the outcome of coronary revascularization using positron emission tomography. Circulation 1992;86:1738-42. 38. Vanoverschelde JL, Depre C, Gerber BL, Borgers M, Wijns W, Robert A, et al. Time course of functional recovery after coronary

Nowak et al Blood flow in hibernating myocardium

45

artery bypass graft surgery in patients with chronic left ventricular ischemic dysfunction. Am J Cardiol 2000;85:1432-9. 39. Bax JJ, Visser FC, Poldermans D, Elhendy A, Cornel JH, Boersma E, et al. Time course of functional recovery of stunned and hibernating segments after surgical revascularization. Circulation 2001;104:I314-l318.