Human Pathologic Correlation with PET in Ischemic and Nonischemic

Jonathan J. Berry, John M. Hoffman, Charles Steenbergen, Jay A. Baker, Carey Floyd, Peter Van Trigt, Michael W. Hanson and R. Edward Coleman

Division ofCardiology, Department ofMedicine, Section ofNuclear , Department ofRadiology, Department of , Department ofSurgery, Duke University Medical Center, Durham, North Carolina

34), echocardiographic (35—37)and magnetic resonance To assess the correlation between myocardial perfusion, me imaging (38—40)methods, all of which are fraught with tabolism and histologic findings in patients with cardiomyop difficulties which have precluded widespread clinical use. athy, we evaluated myocardial perfusion and metabolism Positron emission tomography (PET), utilizing a variety using positron emission tomography (PET) with 13NH3(am of metabolic and perfusion tracers, is an imaging modality monia) and 18FDG(fluoro-2-deoxy-glucose) in nine patients which shows promise in accurately addressing the impor prior to undergoing orthotopic cardiac transplantation. Four tant question of tissue viability. Marshall et al. (41 ) used patients had ischemic cardiomyopathy (ISCM) and five had ‘8F-2-fluorodeoxyglucose (FDG) and ‘3N-ammonia in nonischemiccardiomyopathy (NISCM).Normalizedcircumfer patients with to distinguish is ential profile analyses of representative mid-ventricular per chemic-viable from infarcted-nonviable tissue by compar fusion and metabolism PET images were performed for each patient. A corresponding mid-ventricular transaxial slice was ing regional count density with a normal population data obtained from the formalin fixed explanted heart and proc base and defined PET infarction as a matched defect in essed for routine histology. Hematoxylin-eosin stained and perfusion and metabolism. This concept has been vali Masson trichrome stained sections were evaluated and the dated in animal models (42—44)and has been used pro percentage of the slice occupied by infarct was determined spectively to accurately predict postrevascularization re planimetricallyat 10-degree intervals in a circumferential man turn of function in reversibly injured, viable tissue and the ner. A significant correlation was found between circumfer lack of functional benefit to infarcted segments (45,46). ential normalized PET count density profile of perfusion and Pathologic validation of the PET criteria used for the metabolism in both the ischemic and nonischemic groups identification of infarcted myocardium was performed in (ISCM range r = 0.65—0.75;NISCM range, r = 0.70—0.87). this study using ‘3NH3for myocardial perfusion and Furthermore,there was a correlationin the ISCM group ‘8p1y@for viability imaging in patients with cardiomyop between the extent of matched perfusion/metabolism defects and transmural infarct involvement (r = 0.66—0.88).PET per athy who subsequently had orthotopic cardiac transplan fusion and metabolic data closely correlate with pathologic tation. infarction in human hearts of ischemic cardiomyopathy patients. METHODS J NucIMed 1993;34:39—47 Patients on the cardiac transplantation list at Duke University Medical Center underwent PET imaging with ‘3NH3and ‘8FDG prior to orthotopic cardiac transplantation in order to determine the extent of myocardial infarction. The study population con sisted of nine patients [four with ischemic cardiomyopathy he ability to distinguish nonviable or infarcted myo (15CM)and five with nonischemiccardiomyopathy(NISCM)] cardial tissue from either normal or viable ischemic myo who had perfusion and metabolic images obtained prior to car cardial tissue continues to be a clinical challenge. Accurate diac transplantation. 15CM and NISCM are defined pathologi and reliable quantification of irreversibly damaged myo cally as the presence or absence respectively of histologic myo cardium has both therapeutic and prognostic importance cardial infarction. The average time between PET imaging and (1—6). transplantation was 39 days (range = 7 to 87 days). There were Numerous methods of sizing and localizing the extent no clinical events suggestive of myocardial infarction in any patient between the date of PET scanning and transplantation. ofirreversible myocardial injury exist, including enzymatic Clinical and demographic characteristics are listed in Table I. (6—14),electrocardiographic ( 15—21), scintigraphic (22— ImagingProtocol Patients were positioned in the gantry of the ECAT III PET ReceivedJun. 4, 1992; revision accepted Aug. 17, 1992. tomograph (CT!, Knoxville, TN) and a transmission rectilinear Forreprintsorcorrespondencecontact:JohnM. Hoffman,MD, Department of , Division of , Duke University Medical Center, scan was performed utilizing a 68Ga external ring source. This Box3949, Durham,NC 27710. scan allowed localization of the cardiac silhouette for correct

PathologicCorrelationwith PETinCardiomyopathy•Berryet al. 39 TABLE I ClinicalCharacteristics PatientAgeMlno. CABG EF M/FM (yr)(V/N) (V/N) (%) # Vess.Dis. EKG G/F MI/LVH/RBBBF2M51YV142Ant.MlG3M46YV103Ant.1F48VV1 83Ant.

MI/RBBBF4M52VN103LVHG5M41NN41LVHG6F62NN230LBBBF7M34NN130Ant.MlG8F33NN140Ant./Sept.

MI/LVHF9F56NN141LBBBG

Ant. = anterior;CABG= coronaryarterybypassgrafting;EF= ejectionfraction;F = fasted;G = glucoseloaded;Inf.= inferior;LBBB= left ; LVH = left ; Ml = myocardial infarction; RBBB = right bundle branch block; Sept. = septal; Vess.Dis.= 75%stenosisin anyepicardialcoronaryartery. positioningto obtainthe tomographictransmissionand emission ventricular transaxial slice corresponding to a PET imaging level. scans. The collimators were set at 65 mm giving a full width at This slice was then divided into eight or nine cross sections which half maximum (FWHM) resolution of 16 mm. Three transmis were processed for routine histology. Two paraffin embedded sion planes were acquired for 10 mm (two direct planes, one sections were prepared from each tissue block; one was stained cross-plane) for attenuation correction. Ten millicuries (370 with hematoxylin and eosin, the other with a Masson trichrome MBq) of ‘3NH3were administered intravenously and data were stain, and the circumferential transmural extent of the infarcted acquired for 10 mm after 3 mm of equilibration. Immediately tissue was determined (Figs. lB—C).Beginning at the same start followingthe perfusionstudy acquisition, 10mCi (370 MBq)of ing point of 0 degrees established in the PET ROI analysis, the ISp-@c@ were injected intravenously. After a 30 mm period for percent of noninfarcted tissue was determined by planimetry in accumulation, emission data were acquired for 10 mm. Patients each 10-degree sector (Fig. lB). Histologic myocardial infarction were studied in the fasting state (l2 hr fasting) (n = 4) or after was defined as dense, confluent areas of with loss of receiving 50 g of oral glucose (Glucola) 30 mm prior to the normal myocardial architecture. injectionof 18pj@(n = 5). Data Analysis PETImageAnalysis Pair-wise correlation, using Spearman rank order correlation Perfusion and metabolic PET images were reconstructed, at coefficients(for nonparametricdata), wasused to correlatenor tenuation corrected,and transferredonto a SUN III workstation malized l8f1@j@and ‘3NH3ROl count density with percent non where a circumferential region of interest (ROl) count density infarcted tissue at similar radial locations (±5degrees) around profile was constructed using an interactive program. The initial the circumference of the myocardium. ROI (containing 21 pixels) was assignedto the intersectionof the RESULTS right ventricular free wall, the interventricular septum and the left ventricular anterior wall at the mid ventricular level (Fig. There were four patients in the 15CM group and five lA). Contiguous ROIs of the same size were then placed in a patients in the NISCM group. The average age of the clockwise fashion around the circumference of the perfusion and ISCM group was 49 yr (range 46—52)and for the NISCM metabolic PET images. By assigning the first ROl as 0 degrees, group 45 yr (range 33—62).All patients in the 15CM group the radial angle was automatically computed by the program for had a history of one or more documented myocardial the line joining the geometric center of the left ventricular cavity infarctions and three of four patients had EKG criteria (Q with the center of the RO!. This method of analysis allowed waves) of myocardial damage. The mean EF was 13% for paired data of mean count density of all pixels within a given ROl and radialangle to be generated.The circumferentialcount both the 15CM and NISCM groups (Table 1). No history density profile for each ROl was then normalized for the ROI ofprior myocardial infarctions were present in the patients with the highest mean count density. In addition, the PET images with NISCM, although two patients had anterior precor were interpreted by two experienced observers without benefit of dial Q-waves. All patients with 15CM had significant cor clinical information. The interpretation of either 15CM or onary disease pathologically (75% stenosis in at NISCM was based on the uniformity oftracer distribution in the least one epicardial coronary artery). None of the five myocardium. patients with NISCM had angiographic evidence of signif Pathology icant (>75%) , but two patients had The hearts were removed at the time of transplantation and pathologic evidence of significant one-vessel coronary ar placed in phosphate-buffered formalin for approximately 48 hr tery disease. Neither of these patients had a totally oc to allow fixation. A from the PET laboratory with the cluded vessel. study pathologist oriented the heart in its anatomic position. Pathologic examination of all hearts with NISCM re While viewing the transaxial PET images, they selected a mid vealed no gross or histologic evidence of myocardial in

40 The Journalof NuclearMedicine•Vol. 34 •No. 1 •January1993 PET/PathologyCorrelation @] I PET FPathology

50° 60° S 21% Normal Myocardlum

240° 360° —Normalmyocardlum Degrees @ Myocardlal InfarctIon = Fat A B C

FIGURE 1. AOlanalysisof PETimage(A)andpathology(B).NotetheassignmentoftheinitialAOl(labeled0°)at theintersection of the right ventricular free wall, the interventricular septum, and the left ventricular anterior wall. (C) Quantitative circumferential histopathologicalanalysis of a midventricular,transaxial slice of myocardiumand the extent of transmural involvementby the infarctionprocessis planimetricallyquantitatedat 10-degreeintervalsin a clockwise,circumferentialmannerto correspondto the PET AOl.

farction. There was evidence ofinterstitial and perivascular count density profiles and pathology (Table 2). Figure 3 fibrosis in most cases. In contrast to this, all hearts in the shows the perfusion and metabolism PET images of Pa ISCM group showed extensive areas of myocardial infarc tient 1 with an arrow pointing to the initial RO! at 0 tion with varying degrees of transmural involvement. All degrees. Figure 4A shows the schematic representation of infarcts were old (>6 mo) histologically and characterized the gross pathology and Figure 4B shows the corresponding by extensive areas of confluent fibrosis with thinning of PET pathology correlation for Patient 1. This individual the affected area. had an extensive nontransmural anterior infarct with a In the NISCM group, the percent of normal tissue total left anterior descending and left circumflex coronary (circumferential pathologic analysis) was uniformly 100% artery occlusion and anterior Q-waves on EKG. A close in all hearts; therefore, a correlation could not be per correlation is seen between the extent of transmural path formed between PET and pathologic data. However, the ologic involvement with the PET count density profile. correlation between the perfusion and metabolism circum Figures 5A and B show the pathology schematic and ferential profile analysis was significant (r 0.70: p < the PET pathology correlation of a patient with a lateral 0.001) for all hearts, indicating a very close coupling and infero-posterior myocardial infarction. There are two between perfusion and metabolism. Figure 2 shows an small areas which are transmurally involved with fibrosis example of the count density profile for Patient 5 but probably appear no different on the count profiles (NISCM). The perfusion and metabolism ROl profiles are because of partial volume effects. closely coupled in all segments. Of particular interest is The location of Q-waves on EKG correlated well with the decline in count density in the infero-lateral, inferior the pathologic location of myocardial infarction in all and infero-septal walls (between 180—270degrees). patients with ISCM. In the patients with NISCM however, All patients in the ISCM group had perfusion and two patients had significant Q-waves in the anterior pre metabolic images which could be evaluated, except for cordial leads indicative of an anterior myocardial infarc Patient 4 who had an uninterpretable FDG image because tion. These patients had no evidence for significant coro of persistent blood-pool activity. Patients 1, 2 and 3 had nary artery disease by coronary angiography. Furthermore, excellent correlation between PET perfusion/metabolism these two patients had no evidence of myocardial damage

PathologicCorrelation with PET in Cardiomyopathy °Berry et al. 41 PET/Pathology Correlation I NISCM AnterIor Lateral Inferior Septal 100 100

@80 @c——Pathology-@Perfusion.—MetabolismII 80z

)I@ 0 @6O 60@ 0 @4O 40— C 0 FIGURE2. PETAOlpathologycorral ative graphicalrepresentationof a patient withnonischemiccardiomyopathy.DOtted line acrossthe top represents100% non @2:______IIII ______2:@: infarcted tissue by pathology. Circumfer ential perfusion and metabolism PET 0 40 80 120 160200 240 280 320 360 count profileshows the tight couplingbe Radial Angle (degrees) tween perfusionand metabolism.Decline in counts in the inferior wall is discussed Patient 5 indetailintext. pathologically and had normal PET count density profiles segments which display contractile dysfunction that may in the anterior wall. represent either irreversibly injured tissue or acutely stunned and/or chronically hibernating (49—52) but re DISCUSSION versibly dysfunctional tissue. Myocardial viability is a diagnostic problem which has Scintigraphic methods using 20Tl, 99mTc@pyrophosphate been studied by a multitude of noninvasive modalities. (22—34)and “In-antimyosin antibodies (53—57)have Critical decisions regarding patient management depend been used to address the question of viability but each on differentiating viable, ischemic, and possibly jeopard method has limitations. Further innovations to address ized myocardium from nonviable, irreversibly damaged the question of myocardial viability have been described tissue. Risk stratification and prognostic information (1— that combine scintigraphic imaging modalities such as dual 6) areprovidedby assessingthe amount ofinfarcted tissue isotope imaging with @mTc@pyrophosphateand 201Tl(58— and its impact on regional and global LV function. Dc 61) or ‘‘‘In-antimyosin antibody and 201Tl (53). These signing effective means ofreducing myocardial infarct size approaches have shown promising results. A fixed 201Tl (47—48)and preserving jeopardized myocardium requires perfusion defect is thought to represent scar, however, a method of distinguishing viable from nonviable tissue. between 20% (62) and 75% (63,64) of persistent 201Tl This distinction becomes even more crucial when formu defects improve after . Serial or delayed lating a revascularization strategy involving myocardial imaging (65—67)or 201T1reinjection (68—70)may improve the sensitivity and specificity of detecting viable myocar dium. Brunken et al. ( 71) and Tamaki et al. ( 72) showed TABLE2 I8p-j@ metabolic activity by PET in segments with per PET: Path Correlation sistent 201Tl perfusion defects, implying that persistent PatientFDG:NH3no. PathNH3: PathFDG: defects may represent severe in addition to in IS/NIS r=r=r= farction. Thus, persistent defects on 201Tl imaging may 1IS0.820.720.752IS0.880.600.693IS0.730.840.654IS—0.72—5NIS——0.746NIS——0.867NIS—0.708NIS—0.879NIS—0.70overestimate the amount of nonviable tissue. Myocardial infarct size, assessed pathologically in man, correlates fairly well with scintigraphic methods utilizing 201Tl (22) and 99mTcpyrophosphate (23). Other methods have been used for detecting the extent of myocardial damage, such as electrocardiographic QRS scoring (15-21), intervention ventriculography ( 73), echocardiographic assessment of wall thickening (37), back scatter analysis ( 74,75), and magnetic resonance imaging (38—40). FDG= fluorodeoxyglucose;IS = ischemic;NIS = nonischemic; PET is a metabolic imaging modality which shows r = Spearmanrankordercorrelationcoefficient. promise in differentiating normal from ischemic and in

42 The Journal of Nuclear Medicine•Vol.34 •No. 1 •January1993 @...

@[email protected] C@@ISI@) c@)

FIGURE 3. PET perfusion(left)and metabolism(right)imagesof Patient1 (ISCM).The arrowspointto the initial0-degreeAOl assignmentat the intersectionof the right ventricularfree wall, the interventncularseptum,and the left ventricularanteriorwall.

farcted myocardium. In a canine model oftransient (3 hr) utilized PET with ‘8FDGand ‘3NH3to differentiate ische ischemia followed by reperfusion, Schwaiger et al. (42) mia from infarction in man. They compared the circum showed that myocardial perfusion using ‘3NH3and meta ferential profile analysis of patients with a history of acute bolic activity using ‘‘C-palmitate (CPA) correlated well myocardial infarction to that of a normal data base. In with the degree of functional recovery at 4 wk and the farction was defined as a matched perfusion and metabolic extent of necrosis found at autopsy. Fluorine-l 8-deoxyglu defect that fell below two standard deviations over a 60- cose uptake correlated less well with functional recovery. degree sector. Histologic confirmation of this definition In patients with acute myocardial infarction, studied with has not been shown in humans. Indeed, in a canine model PET utilizing ‘3NH3and ‘8FDGwithin 72 hr after onset involving a 2-hr left anterior descending coronary artery of symptoms, segments with a matched reduction in per occlusion followed by reperfusion, Melin et al. ( 77) showed fusion and metabolism showed no improvement in wall that ‘8FDGuptake was associated with viable tissue by motion at 12 wk, whereas a significant improvement in TTC staining, while depressed 8fl@@uptake indicated wall motion occurred in 50% of the segments showing irreversible injury. Weiss et al. ( 78) quantified myocardial persistent metabolic activity ( 76). Marshall et al. (41) infarct size in a chronic canine left anterior descending

@ A Pathology B PET/PathologyCorrelation

Anterior Lateral InferIor Septal 100 100

@ @80 80 >@ 0 U) C 60fr 0 @40 40@ C 0. 0 C.) 20 20@

0 0 0 40 80 120 160 200 240 280 320 360 —Normalmyocardlum RadialAngle (degrees) @ Myocardlal Infarction @ Fat Patient 1

FIGURE 4. (A) Pathologyof Patient1 showingextensiveantero-septal,anterior,and antero-lateralnontransmuralmyocardial infarction.(B) PET pathology correlativecircumferentialAOl analysisfor Patient 1 with ischemiccardiomyopathy.Note the close correlationof PETand pathologycorrespondingto areasof nontransmuralmyocardialinfarction.

PathologicCorrelationwith PETinCardiomyopathy•Berryet aI. 43 A I Pathoiog@

20 30 40 50' so. 10 70' 80' 350' 340' 90'

110' i— 120'

140' 150' 150' —Normalmyocardlum 170' —MyocardlalInfarction = Fat B IPET/PathologyCorrelation AnterIor Lateral InferIor Septal 100 100

80

U) C 60 ::@@ 0 0 4..C 40 0 -I C-) 20 20@ FIGURE5. (A)Schematicpathological representationof Patient3 withextensive anterolateral, lateral and infero-posterior 0 0 myocardial infarction. (B) PET-pathology 0 40 80 120 160 200 240 280 320 360 circumferentialcorrelativeAOl analysisof Patient 3. Shadedareas representtrans Patient #3 Radial Angle (degrees) muralextensionof fibrosis. coronary artery occlusion model by PET using “C-pal lation between PET and pathology in all patients. mitate and found a high correlation between PET esti In the group with NISCM, there was excellent correla mated infarct size and both histologic and CPK depletion tion between perfusion and metabolism. In all five hearts, estimates. CPA has also been used with PET in patients the inferior wall showed diminished perfusion and metab with acute myocardial infarction to compare PET derived olism. The reason for this apparent decrease is not clear. myocardial infarct size with CPK release (79). PET, using When compared to circumferential profile analysis of per ‘‘C-palmitic acid, has been used to distinguish transmural fusion and metabolism in normal volunteers (unpublished from non-transmural myocardial infarction (80,81 ) and data), a similar matched decrease in count density was ischemic from nonischemic cardiomyopathy (82,83). observed in the inferior wall. A possible explanation for Ter-Pogossian et al. (79) likewise found a high correla this finding could be partial volume effect in this segment tion between PET estimated infarct size and enzymatic because of its tangential orientation with the tomographic (CK-MB)estimatesin patientswith acute myocardial data acquired with our PET tomograph. Newer PET sys infarction. tems allowing oblique reconstruction (along the short-axis In this study, nine patients (four ISCM, five NISCM) of the heart) may be helpful in determining the true with comparably severe left ventricular dysfunction had significance of these findings. Marshall et al. (41 ) showed histologic correlation of PET perfusion and metabolic a similar decrease in perfusion and metabolism using data. In the group with ISCM, there was excellent corre ‘3NH3and ‘8FDGin the inferior segments oftheir normal

44 The Journal of Nuclear Medicine•Vol.34 •No. 1 •January 1993 patients. Visual comparison of the perfusion and meta REFERENCES bolic PET images of the ISCM group with the NISCM 1. PageDL. Caulfield JB, KastoriA, Dc SanctisRW. SandersCA. Myocardial group reveals a homogeneous pattern in the latter and a changesassociatedwith cardiogenic shock. N Engi J Med 197l;285:133— heterogeneous, segmentally abnormal pattern in the for 137. mer. Similar findings have been observed in two previous 2. Alonzo DR. Scheidt S. Post M, Killip T. Pathophysiology of cardiogenic shock: quantification of myocardial necrosis, clinical, pathologic and dcc reports (82,83) using ‘‘C-palmitate and PET and in one trocardiographic correlations. Circulation 1973;48:588—596. report using ‘3NH3and ‘8FDGwith PET (84). Our findings 3. Geltman EM, EhsaniAA, Campbell MK, SchechtmanK, RobertsR, Sobel on the visual interpretations agree with these previous BE. The influence oflocation and extent of myocardial infarction on long term ventricular dsyrhythmia and mortality. Circulation 1979:60:805—814. reports, and the patients in our study had pathologic 4. Perez-Oonzolez J. Botvinick EH, Dunn R, ci al. The late prognostic value confirmation. of acute scintigraphic measurement of myocardial infarct size. Circulation Because of the limited resolution (FWHM = 16mm) of 1982;66:960—971. 5. Holman BL, Goldhaber SZ, Kirsch KM. et al. Measurement ofinfarct size the ‘8FDGand ‘3NH3studies with our tomograph, deter using single photon emission computed tomography and technetium-99m mination of the transmural extent of infarction is difficult pyrophosphate: a description of method and comparison with patient to assess. Three tomographic PET slices were obtained, prognosis. Am J Cardiol 1982:50:503-5 I 1. 6. Sobel BE, Bresnahan GF, Shell WE, Yoder RD. Estimation of infarct size but oblique reorientation with the long-axis of the heart is in man and its relation to prognosis. Circulation 1972:46:640—648. not possible with only three imaging planes obtained at a 7. Shell WE. Kjekshus JK, Sobel BE. Quantitative assessment of the extent ofmyocardial infarction in the conscious dog by means ofanalysis of serial single bed position. Reorientation would permit better changes in serum creatinine phosphokinase activity. J Clin Invest evaluation of the inferior wall. The images were obtained 1971:50:2614—2625. in an ungated mode and thus regional asynergy could 8. Roberts R. Henry PD. Sobel BE. An improved basis for enzymatic esti mation of infarct size. Circulation 1975:52:743—753. contribute to apparent differences in regional perfusion 9. RogersWi, McDaniel HG, Smith LR, Mantel JA, Russell RO. Rackley and metabolism by PET (85,86). The raw PET data were CE. Correlation of angiographic estimates of myocardial infarct size and not adjusted for wall thickness. Areas ofchronic infarction accumulated release ofcreatine kinase MB isoenzyme in man. Circulation 1977;56: 199—205. exhibit wall thinning which may artifactually diminish 10. Morrison J. Coromilas J. Munsey D. et al. Correlation of radionuclide PET count density because of partial volume effects (87). estimates of myocardial infarction size and release of creatine kinase-MB Patients in this study were imaged in either the fasted in man. Circulation 1980;62:277—287. I I. DevriesSR,JaffeAS,Geltman EM. SobelBE,AbendscheinDR. Enzymatic or glucose loaded state. Metabolic preparation of patients estimation of the extent of irreversible myocardial injury early after reper (fasting versus glucose loading) undergoing PET imaging fusion. Am Heartf 1989:117:31—36. with ‘8FDGhas been shown by Gropler et al. (88) to 12. Roe CR, Cobb FR. Starmer CF. The relationship between enzymatic and histologic estimates of the extent of myocardial infarction in conscious critically affect the regional distribution of ‘8FDGmyocar dogswithpermanentcoronaryocclusion.Circulation1977:55:438—449. dial uptake. These differences could affect both visual 13. Grande P. Hansen BF, Christiansen C. Naestoft J. Estimation of acute interpretation of the PET images (89) as well as the myocardial infarct size in man by serum CK-MB measurements. Circula tion 1982:65:756—764. quantitative PET/pathology correlation. However, they 14. Hackel DB, Reimer KA. Ideker RE. et al. Comparison of enzymatic and demonstrated no effect of glucose loading on regional anatomic estimates of myocardial infarct size in man. Circulation myocardial perfusion. The metabolic state of the patients 1984:70:824—835. 15. Savage RM, Wagner GS, Ideker RE, Podolsky SA, Hackel DB. Correlation in this study did not seem to have an effect on the results of postmortem anatomic findings with electrocardiographic changes in because the metabolism and perfusion were congruent in patients with myocardial infarction. Circulation I977:55:279—285. every patient. In our series only a single study, however, 16. Eisen Hi, Barzilai B, iaffe AS, Geltman EM. Relationship ofQRS scoring system to enzymatic and pathologic infarct size: the role ofinfarct location. was compromised by an uninterpretable FDG study. Am HeartJ 1988:115:993—1001. The ‘8FDGand ‘3NH3studies in patients with 15CM 17. SeinoY. Staniloff HM, Shell WE, Mickle D, Shah PK, Vyden JK. Evalu correlated well with pathology. Areas ofmatched perfusion ation of a QRS scoring system in acute myocardial infarction. Relation to infarct size, early stage left ventricular , and exercise and metabolic defects on PET corresponded in location to performance. Am J Cardiol 1983:52:37—42. areas of pathologically confirmed infarction. In patients 18. Roubin G5, Shen WF, Kelly DT, Hams Pi. The QRS scoring system for with NISCM, there were no areas ofabnormalities on PET estimating myocardial infarct size: clinical. angiographic and prognostic correlations. J Am Co/I Cardiol 1983:2:38—44. and no pathologic evidence of infarction. Since a good 19. ldeker RE. Wagner G5. Ruth WK, et al. Evaluation of a QRS scoring correlation was found between a matched perfusion and system for estimating myocardial infarct size: 11.Correlation with quanti tative anatomic findings for anterior infarcts. Am J Cardiol 1982:49:1604— metabolic defect and irreversible tissue damage patholog 1614. ically, the matched defects by PET reflect infarcted myo 20. Ward RM, White RD. Ideker RE, et al. Evaluation of a QRS scoring cardium. system for estimating myocardial infarct size: IV. Correlation with quan titative anatomic findings for posterolateral infarcts. Am J Cardiol 1984:53:706—714. 21. Roark SF, ldeker RE, Wagner GS, et al. Evaluation of a QRS scoring ACKNOWLEDGMENTS system for estimating myocardial infarct size: III. Correlation with quanti tative anatomic findings for inferior infarcts. .4m J Cardiol I983:5 I: 382— The authors would like to thank Sandra Bowling and Suzanne 389. Ealy-Romey for their expert secretarial support and John Need 22. Wackers Fi, Becker AE, Samson G, et al. Localization and size of acute transmural myocardial infarction estimated from thallium-201 scintiscans: PhD, Craig Harris MS. Sharon Hamblen, David Coates and Tom a clinicopathological study. Circulation 1977:56:72—78. Hawk for technical support. This research was partially supported 23. Poliner LR, Buja M. Parkey RW, Bonte Fi, Willerson iT. Clinicopatho. by NIH SCOR grant P50-HL 17760. logic findings in 52 patients studies by technetium-99m stannous pyro

PathologicCorrelation with PET in Cardiomyopathy •Berry et al. 45 phosphate myocardial scintigraphy. Circulation l979;59:257—267. cardiac wall-motion abnormalities predicted by positron tomography. N 24.Mahmarianii, PrattCM,Borges-NetoS.CashionWR,RobertsR,Verani EnglJMed l986;314:884—888. MS. Quantificationof infarct size by @°‘Tlsingle-photonemissioncorn 46. TarnakiN, YonekuraY, YamashitaK, etal. Positronemissiontomography puted tomography during acute myocardial infarction in humans: corn using fluorine-I 8 deoxyglucose in evaluation of coronary artery bypass parison with enzymatic methods. Circulation 1988;78:831—839. grafting. Am J Cardiol 1989:64:860—865. 25.Tarnaki5,MurakamiT,KadotaK, KambaraH,etal.Effectsofcoronary 47. Buja LM, Willerson JT. Infarct size. Can it be measured or modified in artery reperfusion on relation between creatine kinase-MB and infarct size humans? Prog in Cardiovasc Dis 1987:29:271—289. estimated by myocardial emission tomography with thalliurn-20l in man. 48. Rude RE, Muller JE, Braunwald E. Efforts to limit the size of myocardial JAm CollCardiol1983;2:1031—1038. infarcts.Ann InternalMed1981:95:736—761. 26. Tamaki 5, Nakajirna H, Murakirni T, et al. Estimation of infarct size by 49. Braunwald E, Rutherford JD. Reversible ischemic left ventricular dysfunc myocardialemission computed tomography with thalliurn-201and its tion: evidence for the “hibernatingmyocardium―. J Am Coil Cardiol relation to creatine kinase-MB release after myocardial infarction in man. 1986;8:1467—1470. Circulation1982;66:994.-lOOl. 50. Rahimtoola SH. The . Am Heart J 1989:1 17:211— 27. Henning H, Schelbert HR. Righeni A, Ashburn WL, O'Rourke RA. Dual 220. myocardial imaging with technetium-99m pyrophosphate and thallium 5 I . Kloner RA, Przyklenk K, Patel B. Altered myocardial states: the stunned 201 for detecting, localizing and sizing acute myocardial infarction. Am J and hibernating myocardiurn. Am J Med 1989: 86(suppl.IA): 14—22. Cardiol1977;40:147—155. 52. Braunwald E, Kloner RA. The stunned myocardium: prolonged, post 28. Prigent F, Maddahi J, Garcia EV, Satoh Y, Van Train K, Berman DS. ischemic ventricular dysfunction. Circulation 1982:66: 1146—1149. Quantitation of myocardial infarct size by thalliurn-201 single-photon 53. Johnson LL, L.errick KS, Coromilas J, et al. Measurement of infarct size emission computed tomography: experimental validation in the dog. Cir and percentageof myocardiuminfarctedin a dog preparationwith single, culation1986;74:852—861. photon emission computed tomography, thallium-20l, and indium-I 11 29.BulkleyBH,SilvermanK,WeisfeldtML,BurowR,PondM, BeckerLC. monoclonal antirnyosin Fab. Circulation 1987:76:I81—190. Pathologicbasisofthalliurn-20l scintigraphicdefectsin patientswithfatal 54. Antunes ML, Seldin DW, Wall RM, Johnson LL. Measurement of acute rnyocardial injury. Circulation 1979;60:785—792. Q-wave myocardial infarct size with single photon emission computed 30.BulkleyBH,HutchinsGM,BaileyI, StraussHW,PinB.Thalliurn-20l tomography imaging of indium-l 11 antimyosin. Am J Cardiol imaging and gated cardiac blood pool scans in patients with ischernic and 1989;63:777—783. idiopathic congestive cardiornyopathy. A clinical and pathologic study. 55. van Vlies B, Bass i, Visser CA, et al. Predictive value of indium-I I I Circulation1977;55:753—760. antimyosin uptake for improvement of left ventricular wall motion after 31. Keyes JW, Leonard PF, Brady SL, SvetkoffDL, Rogers WL, Lucchesi BR. thrornbolysis in acute myocardial infarction. Am J Cardiol 1989:64:167— Myocardial infarct quantification in the dog by single-photon emission 171. computed tomography. Circulation 1978:58:227—232. 56. Tamaki N, Yamada T, Matsumori A, et al. tndiurn-l 11 antimyosin 32. Jansen DE, Corbeti JR. Wolfe CL, et al. Quantification of rnyocardial antibody imaging for detecting different stages of myocardial infarction: infarction: a comparison of single-photon emission computed tomography comparison with technietium-99m pyrophosphate imaging. J Nucl Med with pyrophosphate to serial plasma MB-creatinine kinase measurements. l990;3l:l36—142. Circulation l985;72:327—333. 57. Jam D, Crawley iC, Lahiri A, Raftery EB. Indium-l I 1antimyosin images 33. Wolfe CL, LewisSE, CorbettJR. ParkeyRW, Buja LM, WillersonJT. compared with triphenyl tetrozolium chloride staining in a patient six days Measurement of myocardial infarction fraction using single photon ernis after myocardial infarction. J NucI Med 1990:31:231—233. sion computed tomography. JAm CoilCardiol 1985;6:145—151. 58. Hashimoto T, Kambara H, Fudo T, et al. Significance of technetiurn-99rn/ 34.LewisSE,DeVousMD,CorbettJR.etal.Measurementofinfarctsizein Thallium-201 overlap on simultaneous dual emission tomography in acute acute canine rnyocardial infarction by single-photon emission computed myocardial infarction. Am J Cardiol 1988:61: 1181—1186. tomography with technetium-99rn pyrophosphate. Am J Cardiol 59. Sochor H, Schwaiger M, Schelbert HR. et al. Relationship between @°‘Tl, 1984;54: 193—199. @“Tc(Sn)pyrophosphateand F-I8 2-deoxyglucoseuptakein ischemically 35. WeissJL, Bulkley BH, Hutchins GM, Mason Si. Two-dimensional echo injured dog myocardium. Am Heart J 1987:1 14:1066—1077. cardiographic recognition of myocardial injury in man: a comparison with 60. Schofer J, Speilmann RP, Bromel T, Bleifeld W, Mathey 0G. Thallium postmortem studies. Circulation l981;63:40l—408. 201/technetium-99m pyrophosphate overlap in patients with acute myo 36.WyattHL,Meerbaum5, HengMK, RitJ,GueretP.CordayE.Experi cardial infarction after thrombolysis: prediction of depressed wall motion mental evaluation of the extent of rnyocardial dyssynergy and infarct size despite thallium uptake. Am Heart J 1986:1 12:291—295. by two-dimensional echocardiography. Circulation 198l;63:607—6l4. 61. Schofer J, Mathey DO, Montz R, Bleifeld W, Stritzke P. Use of dual 37. Lieberman AN, WeissJL, Jugduu BI, et al. Two-dimensional echocardiog intracoronary scintigraphy with thallium-20l and technetium-99m pyro raphy and infarct size: relationship of regional wall motion and thickening phosphate to predict improvement in left ventricular wall motion irnme to the extent of rnyocardial infarction in the dog. Circulation 1981;63:739— diately after intracoronary thrombolysis in acute myocardial infarction. J 746. Am Coil Cardiol 1983:2:737—744. 38.CaputoGR,SechlernU, TscholakoffD, HigginsCB.Measurementof 62. Gibson RS, Watson DD, Taylor Gi, et al. Prospective assessment of myocardialinfarctsizeat earlyand late time intervalsusingMR imaging: regional myocardial perfusion before and after coronary revascularization an experimental study in dogs. Am JRadiol 1987:149:237—243. by quantitative thallium-201 scintigraphy. J Am Coil (‘ardiol 39. BouchardA, ReevesRC, CranneryG, BishopSP,PohostGM. Assessment 1983:1:804—815. of myocardial infarct size by means of T2-weighted H Nuclear magnetic 63. Liu P. Kiess MC, Okada RD. et al. The persistent defect on exercise resonanceimaging.Am HeartJ 1989;l17:281—289. thallium imaging and its fate after myocardial revascularization: does it 40.JohnsJA,LeavittMB,NewellJB,etal.Quantitationofacutemyocardial represent scar or ischemia?Am Heart J 1985:110:996—1001. infarct size by nuclear magnetic resonance imaging. J Am Coil Cardiol 64. BergerBC, WatsonDD, BurwellLR, et al. Redistributionof thallium at l990;15: 143—149. rest in patients with stable and unstable and the effect of coronary 41. Marshall RC, Tillish JH, Phelps ME, et al. Identification and differentiation artery bypass surgery. Circulation 1979:60: 1114—1125. of resting myocardial ischemia and infarction in man with positron corn 65. Kiat H, Berman DS, Maddahi J, et al. Late reversibility of tomographic puted tomography, l8F-labeled fluorodeoxyglucose and N-l3 ammonia. myocardial thallium-201 defects: an accurate marker of myocardial viabil Circulation l983;67:766—778. ity. JAm CoilCardiol 1988;l2:1456—1463. 42. Schwaiger M, Schelbert HR. Ellison D, et al. Sustained regional abnor 66. OkadaRD. BoucherCA. Differentiationof viableand nonviablemyocar matitiesin cardiacmetabolismafter transientischemiain the chronicdog dium after acute reperfusion using serial thallium-20l imaging. Am Heart model. JAm CoilCardiol 1985;6:336—347. J 1987:113:241—249. 43. Schwaiger M, Neese RA, Araujo L, et al. Sustained non oxidative glucose 67. Pohost GM, Zir LM, Moore RH, McKusick KA, Guiney TE, BelIer GA. utilization and depletion of glycogen in reperfused canine rnyocardium. J Differentiation oftransiently ischemic from infarcted myocardium by serial Am CoilCardioll989;13:745—754. imaging after a single dose ofthallium-201. Circulation 1977:55:294—302. 44. Schelbert HR, Henze E, Phelps ME, Kuhl DE. Assessment of regional 68. BonowRO, DilsizianV.CuocoloA, BacharachSL.Identificationof viable myocardial ischemia by positron emission computed tomography. Am myocardium in patients with chronic coronary artery disease and left Heart I 1982;103:588—597. ventricular dysfunction: comparison of thallium scintigraphy with reinjec 45. Tillisch J, Brunken R, Marshall R, Schwaiger M, et al. Reversibility of tion and PET imaging with ‘8F-fluorodeoxyglucose. Circulation

46 The Journal of Nuclear Medicine •Vol. 34 •No. 1 •January 1993 1991:83:26—37. 79. Ter-Pogossian MM, Klein MS. Markham J, Roberts R, Sobel BE. Regional 69. Dilsizian V. Smeltzer WR, Dextras R, Bonow RO. Does twenty-four hour assessment of myocardial metabolic integrity in vivo by positron emission delayed imaging following thallium reinjection enhance detection of viable tomography with ‘‘C-labeledpalmitate. Circulation 1980:61:242—255. myocardium? [Abstract]. J Am Coil Cardiol 1990:15(suppl A):248A. 80. Geltman EM, Biello D, Welch Mi, Ter-Pogossian MM, Roberts R, Sobel 70. Dilsizian V, Rocco TP, Freeman N. SmeltzerWR, Bonow RO. Thallium BE. Characterization of nontransmural myocardial infarction by positron reinjection after stress redistribution imaging improves detection of is emission tomography. Circulation 1982:65:747—755. chemic myocardium: a qualitative and quantitative SPECT study [Ab 8 1. Hashimoto T, Kambara H, Fudo T, et al. Non-Q wave versus Q wave stract]. J hi, CoIl Cardiol I990: 15(suppl A): 147A. myocardial infarction: regional myocardial metabolism and blood flow 71. Brunken R, Schwaiger M, Grover-McKay M. Phelps ME, Tillish J, Schel assessed by positron emission tomography. JAm CoilCardiol 1988:12:88— bert HR. Positron emission tomography detects tissue metabolic activity 93. in myocardial segments with persistent thallium perfusion defects. J Am 82. Eisenberg JD, Sobel BE, Geltman EM. Differentiation of ischemic from (oil Cardiol 1987:10:557—567. non-ischemic cardiomyopathy with positron emission tomography. Am J 72. Tamaki N, Yonekura Y, Yamashita K, Ct al. Relation of left ventricular Cardiol1987:59:1410—1414. perfusion and wall motion with metabolic activity in persistent defects on 83. Geltman EM, Smith IL, Beecher D, Ludbrook PA, Ter-Pogossian MM, thallium-201 tomography in healed myocardial infarction. Am J Cardiol Sobel BE. Altered regional myocardial metabolism in congestive cardio 1988:62:202—208. myopathy detected by positron tomography. Am J Med 1983:74:773—785. 73.BankaVS.BodenheimerMM,ShahR,HelfantRH.Interventionventric 84. Mody P/, Brunken RC, Stevenson LW, Nienaber CA, Phelps ME. Schel ulography: comparative value of nitroglycerin, post-extrasystolic potentia bert HR. Can positron tomography distinguish dilated from ischemic tion and nitroglycerin plus post-extrasystolic potentiation. Circulation cardiomyopathy? [Abstract]. Circulation 1988:78(suppl Il):92. 1976:53:632—637. 85. Yamashita K, Tamaki N, Yonekura Y, et al. Quantitative analysis of 74. FitzgeraldPi, McDaniel MD, Rolett EL, StruhbehnJW, JamesDH. Two regional wall motion by gated positron emission tomography: validation dimensional ultrasonic tissue characterization: backscatter power. endocar and comparison with left ventriculography. J Noel Med 1989:30:1775— dial wall motion, and their phase relationship for normal, ischemic, and 1786. infarcted myocardium. Circulation 1987:76:850—859. 86. Parodi 0, Schelbert HR. Schwaiger M, Hansen H. Selin C, Hoffman EJ. 75. Milunski MR. Mohr GA. Perez iE, et al. Ultrasonic tissue characterization Cardiac emission computed tomography: underestimation of regional with integrated backscatter: acute myocardial ischemia, reperfusion, and tracer concentrations due to wall motion abnormalities. J Comput Assist stunned myocardium in patients. Circulation 1989:80:491—503. Tomogr 1984:8:1063—1092. 76. SchwaigerM, Brunken R. Grover-McKay M, et al. Regional myocardial 87. Hoffman EJ, Huang S-C, Phelps ME. Quantitation in positron emission metabolism in patients with acute myocardial infarction assessed by posi computed tomography: I. Effects of object size. J Comput Assist Tomogr tron emission tomography. J Am (‘oilCardiol 1986:8:800—808. 1979:3:299—308. 77. Melin IA. Wijns W, Keyeux A, et al. Assessment ofthallium-201 redistri 88. Gropler Ri, Siegel BA. Lee KJ, et al. Nonuniformity in myocardial bution versus glucose uptake as predictors of viability after coronary artery accumulation of fluorine-l 8-fluorodeoxyglucose in normal fasted humans. occlusion and reperfusion. Circulation 1988:77:927—934. J NuelMed 1990:31:1749—1756. 78. Weiss ES, Ahmed SA, Welch MJ. Williamson JR. Ter-PogossianMM, 89. Berry ii, Baker JA, Pieper KS, Hanson MW, Hoffman JM, Coleman RE. Sobel BE. Quantification of infarction in cross sections of canine myocar The effect of metabolic milieu on cardiac PET imaging using tluorine-18- dium in vivo with positron emission transaxial tomography and “C- deoxyglucose and nitrogen-l 3-ammonia in normal volunteers. J Nucl Med palmitate. Circulation 1977:55: 66—73. 1991:32:1518—1525.

PathologicCorrelation with PET in Cardiomyopathy ‘Berry et al. 47