concordance discordance stress-redistribution-reinj...

941

Concordance and Discordance BetweenStress-Redistribution-Reinjection andRest-Redistribution Thallium Imaging

for Assessing Viable MyocardiumComparison With Metabolic Activity by

Positron Emission Tomography

Vasken Dilsizian, MD; Pasquale Perrone-Filardi, MD; James A. Arrighi, MD;Stephen L. Bacharach, PhD; Arshed A. Quyyumi, MD;Nanette M.T. Freedman, PhD; Robert 0. Bonow, MD

Background. Stress thallium scintigraphy provides important diagnostic and prognostic information inpatients with coronary artery disease by demonstrating regional myocardial ischemia. However, if theclinical question being addressed is whether a region is viable and not whether there is inducible ischemia,then it may be more reasonable to perform rest-redistribution imaging rather than stress-redistributionimaging followed by either reinjection or late redistribution. Therefore, we determined whether stress-redistribution-reiijection and rest-redistribution imaging provide the same information regardingmyocardial viability.Methods and Results. Both stress-redistribution-reinijection and rest-redistribution thallium single photon

emission computed tomographic imaging was performed in 41 patients with chronic stable coronary arterydisease, with quantitative analysis of regional thallium activity. Thallium reinjection was performedimmediately after the 3- to 4-hour redistribution Images were completed. Ofthe 155 myocardial regions withperfusion defects on the stress images, 91 (59%) were irreversible on conventional 3- to 4-hour redistributionimages. When the outcomes of these irreversible regions were assessed after reinjection and compared withrest-redistribution images, there was concordance ofdata regarding myocardial viability (normal/reversibleor irreversible) in 72 of the 91 (79%) irreversible defects. Twenty of the 41 patients also underwent positronemission tomography at rest with [18Flfluorodeoxyglucose and [11Owater. In these patients, stress-redistribution-reinjection and rest-redistribution imaging provided concordant information regardingmyocardial viability in 427 (72%) of594 myocardial regions and discordance in 167 regions. However, whenirreversible thallium defects were further analyzed according to the severity of the thallium defect in thesediscordant regions, 149 of 167 (89%1) demonstrated only mild-to-moderate reduction in thallium activity(51% to 85% of normal activity), and positron emission tomography verified 98% of these regions to bemetabolically active and viable. Thus, when the severity of thallium activity was considered withinirreversible thallium defects, the concordance between stress-redistribution-reinjection and rest-redistribu-tion imaging regarding myocardial viability increased to 94%.

Conclusions. These data indicate that one oftwo imaging modalities, either stress-redistribution-rei'jectionor rest-redistribution imaging, may be used for identifying viable myocardium. However, if there are nocontraindications to stress testing, stress-redistribution-rewinection imaging provides a more comprehensiveassessment of the extent and severity of coronary artery disease by demonstrating regional myocardialischemia without jeopardizing information on myocardial viability. (Cirultion. 1993;88:941-952.)KEY WORDs * coronary artery disease * myocardium * ischemia * scintigraphy * tomography

J n many patients with chronic coronary artery dis- after revascularization suggests that many asynergicease, impaired left ventricular function at rest is a myocardial regions represent hibernating myocardiumpotentially reversible process that may improve or secondary to chronic hypoperfusion.5,6 However, iden-

normalize after myocardial revascularization.1-4 Recov- tification of patients with such potentially reversible leftery of regional or global left ventricular function at rest ventricular dysfunction has been problematic, since

regional dysfunction arising from ischemic myocardiummay be clinically indistinguishable from that arisingReceived January 13, 1992; revision accepted April 30, 1993. from infarcted myocardium.

From the Cardiology Branch, National Heart, Lung, and Blood Exercie my scinhInstitute, and the Department of Nuclear Medicine, National Exerclse thallium scmntography has been used fre-Institutes of Health, Bethesda, Md. quently for the assessment of myocardial viability by a

Correspondence to Vasken Dilsizian, MD, National Institutes number of methods, including early (3-4 hours) or lateof Health, Building 10, Room 7B-15, Bethesda, MD 20892. (24 hours) redistribution imaging.7 Recently, thallium

by guest on June 21, 2018http://circ.ahajournals.org/

Dow

nloaded from

http://circ.ahajournals.org/

942 Circulation Vol 88, No 3 September 1993

reinjection after a 3- to 4-hour redistribution8.9 or afterlate redistribution'0 has improved the accuracy in iden-tifying viable myocardium over that achieved by redis-tribution imaging alone. Rest-redistribution thalliumimaging has also been used successfully in the past todistinguish viable from nonviable myocardium.11-15 Ifthe clinical question being addressed is whether a regionis viable and not whether there is inducible ischemia,then it may be more reasonable to perform thalliumscintigraphy at rest rather than during stress. Such anapproach, however, will not provide the additionaldiagnostic and prognostic information gained by evi-dence of exercise-induced ischemia. Therefore, the aimof this study was to determine whether stress-redistri-bution-reinjection imaging provides the same informa-tion as rest-redistribution imaging for identifying viablemyocardium.

MethodsPatient SelectionWe studied 41 patients with chronic stable coronary

artery disease. The patients ranged in age from 38 to 76years (mean, 59 years); there were 38 men and 3women. All patients underwent a history and physicalexamination, chest x-ray, ECG, thallium single photonemission computed tomography (SPECT), and coro-nary arteriography. Coronary artery disease was de-fined as >50% reduction in luminal diameter of at leastone major epicardial coronary artery as determined bycoronary angiography. All cardiac medications werewithdrawn before exercise studies in 76% of patients. Inthe remaining 24% of patients, cardiac medicationsincluded either one or a combination of nitrates, cal-cium channel blockers, and 3-blockers; one patient wason digoxin and another on amiodarone. We studied onlypatients with chronic stable coronary artery disease; nopatient with recent acute myocardial infarction or un-stable angina was included in the study. Thirty-twopatients had either a clinical history of prior myocardialinfarction or Q waves on the ECG. Fifteen patients hadundergone previous coronary artery bypass surgery.Patients were included in the study only if they demon-strated an irreversible thallium defect on postexercise 3-to 4-hour redistribution images. Patients demonstrating"fixed" thallium defects after conventional redistribu-tion images who were eligible for this study then gaveinformed consent for a second thallium study performedat rest.

Thallium SPECT ImagingAll patients underwent exercise thallium SPECT as

previously described.8 After an overnight fast, patientsexercised on a treadmill, and 2 mCi of thallium wasinjected at peak exercise. SPECT thallium images wereobtained with a wide-field-of-view rotating gamma cam-era equipped with a low-energy, medium-resolution,high-sensitivity, parallel-hole collimator (Apex 415,APC-3, Elscint Co, Boston, Mass) centered on the68-keV photo peak with a 20% window. The camerawas rotated over a 1800 arc in an elliptical orbit aboutthe patient's thorax at 60 increments for 30 secondseach. Redistribution images were acquired 3 to 4 hoursafter exercise. Immediately after redistribution, a 1-mCiadditional thallium dose was administered at rest, andreinjection images were acquired 10 to 15 minutesthereafter.

The rest-redistribution study involved 2 mCi of '111administered at rest and was performed 1 to 2 weeks afterthe stress-redistribution-reinjection thallium study.SPECT thallium data were acquired 10 to 20 minutesafter the rest-administered dose and 3 to 4 hours later.The reconstructed images of both stress-redistribution-reinjection and rest-redistribution images were readblindly, grading each segment as either normal, reversible,or irreversible. Short-axis tomograms from the five sets ofthallium images (stress, redistribution, reinjection, rest,and redistribution) were also analyzed objectively by useof a semiautomatic quantitative circumferential profile aspreviously described.8"16"17

Quantitative Thallium AnalysisBriefly, for each patient, an operator-defined region

of interest was drawn around the left ventricular activityof each short-axis slice on the stress images and thecorresponding tomograms of the redistribution, reinjec-tion, rest, and redistribution images. The myocardialactivity was subdivided into 64 sectors, each emanatingfrom the center of the tomograms. All 64 sectors were ofequal arc and constructed beginning at the 3 o'clockposition (midlateral wall) and proceeding counterclock-wise. The sectors from two consecutive, three-pixel-thick, short-axis tomograms representing the midpor-tion of the heart were then grouped and averaged intofour myocardial regions: anterior, septal, inferior, andlateral. The sectors from the apical short-axis tomogramwere analyzed to assess apical perfusion defects.

Assessment of regional thallium activity on stressimages was carried out with reference to mean thalliumactivity for groups of normal male and female subjects.The normal subjects, who have been previously report-ed,8 consisted of 26 male and 24 female volunteers whowere screened for coronary artery disease by completephysical examination, ECG, two-dimensional echocar-diogram, and exercise treadmill testing. Subjects whohad a strong family history of coronary artery diseasewere excluded from the normal data base. The malesubjects ranged in age from 21 to 58 years (mean age, 39years), and the female subjects ranged in age from 24 to57 years (mean age, 38 years).For each of the 41 patients in the present study, the

myocardial region with the maximum mean counts perpixel on the stress study was normalized to the value forthe corresponding region for normal subjects of thesame sex, and this was used as a normal referenceregion for that patient. The corresponding regions inthe redistribution, reinjection, rest, and redistributionthallium studies were identified and used as the refer-ence region for those studies. The thallium activity in allother myocardial regions was then expressed as a per-cent of the activity measured in that reference regionfor each of the stress, redistribution, reinjection, rest,and redistribution image series. A myocardial regionwas considered abnormal in a patient with coronaryartery disease if the thallium uptake on the stress imagewas greater than 2 SD below the mean observed in thesame region for normal volunteers of the same sex. Onthe basis of previous reproducibility measurements inour laboratory, a region with reduced activity on thestress study was considered reversibly ischemic if theincrease of normalized thallium uptake on the redistri-bution or reinjection image exceeded the reproducibil-ity limit for that region.'7 Alternatively, a region with


Dow

nloaded from


Dilsizian et al Stress vs Rest Thallium Imaging 943

reduced activity on the stress study was consideredirreversibly abnormal if the normalized thallium activityin that region on subsequent images did not increasemore than the reproducibility limit for that region.These irreversible thallium defects on redistributionimaging were then subgrouped on the basis of theseverity of reduction in thallium activity17,18: mild tomoderate (51% to 85% of peak activity) and severe(s50% of peak) thallium defects. For analysis of indi-vidual patients, myocardial regions were grouped foreach patient as viable or nonviable on the basis of theseverity of reduction of thallium activity and the pres-ence or absence of thallium reversibility.

Positron Emission TomographyTwenty of the 41 patients underwent positron emission

tomography (PET) studies to assess regional myocardialperfusion with 15O-labeled water ([150]H20) and exoge-nous glucose utilization with [18F]fluorodeoxyglucose(FDG) as previously described.18 Imaging was per-formed with a whole-body PET camera producing 21contiguous tomograms spaced 5.1 mm apart with a slicethickness of 13 mm and an in-plane resolution of 6.5mm. Images were obtained perpendicular to the longaxis of the body to create a series of transaxial tomo-grams. All patients were pretreated with 50 g of oralglucose 1 hour before the study after an overnight fast.After a 20-minute transmission scan to correct forattenuation, two separate bolus injections of 12 to 15mCi of [150]H2O were administered intravenously 12minutes apart. Like most PET scanners, our scanner isunable to accurately handle the bolus phase of aninjection of more than 15 to 20 mCi when the heart is inthe field of view. However, since an injection of 15 to 20mCi of [150]H20 does not give sufficient counts toobtain an accurate value of flow, two [`50]H20 studieswere performed, back to back. Each study was analyzedseparately, and the flow values were averaged together.This resulted in average flow values with SDs compara-ble to what would have been obtained from a single 30-to 40-mCi injection. The two separate bolus injectionsof [150]H2O were followed by the administration of 5mCi of FDG 15 minutes later. Dynamic PET data wereacquired continuously for 5 minutes after each [150]H20injection and for 60 to 75 minutes after the FDGinjection. The data acquired at 30 minutes after FDGinjection, corresponding to the final 30 to 45 minutes ofdata acquisition, were reconstructed to create tomo-graphic images of regional myocardial FDG uptake.

Regional myocardial FDG uptake. For each patient, amean of six transaxial tomograms from the five sets ofthallium images (stress, redistribution, reinjection, rest,and redistribution) and the corresponding transaxialtomograms of myocardial FDG uptake from the PETstudy were visually aligned for direct comparison.18 Tocompare relative regional FDG uptake and thalliumactivity objectively, five myocardial regions of interestrepresenting the posterolateral, anterolateral, an-teroapical, anteroseptal, and posteroseptal myocardiumwere drawn on each FDG tomogram and on each of thefive corresponding thallium images. FDG and thalliumactivities were then computed within each region.

In each patient, the myocardial region with the max-imum counts on the exercise thallium study was used asthe normal reference region for that patient. The cor-responding regions in the redistribution, reinjection,

rest, and redistribution thallium studies were identifiedand used as the reference regions for those studies. Thethallium activity in all other myocardial regions wasthen expressed as a percentage of the activity measuredin the reference region for each of the exercise, redis-tribution, reinjection, rest, and redistribution imageseries. For each exercise study, thallium activity in anymyocardial region measuring <85% of the normalreference region was considered reduced and defined asa thallium perfusion defect. Perfusion defects on exer-cise were defined as irreversible if relative thalliumactivity was unchanged or increased <10% on thesubsequent redistribution study. The myocardial regionon the FDG series that corresponded to the normalreference region on the thallium stress image series wasused as the normal reference region for relative FDGuptake. FDG uptake in all other myocardial regions wasexpressed as a percent of the activity in this referenceregion.

Regional myocardial blood flow. Absolute regionalmyocardial blood flow was computed from the dynamic[150JH20 data as previously described.18 The FDGimage series was reviewed for each patient to identifythe appropriate tomographic levels in which the leftventricular cavity was well defined. On average, foursuch tomographic levels were identified per patient.Left ventricular cavity regions of interest were manuallyconstructed on these FDG tomograms and were thenapplied directly to the corresponding tomographic[150]H20 data to derive a composite ventricular bloodpool time-activity curve of the tracer. This ventriculartime-activity curve was then used as the [`50]H20arterial input function. Previous studies have demon-strated that the blood pool time-activity curve com-puted in this manner accurately represents instanta-neous arterial concentrations of [150]H20.19'20 Using the['50]H20 arterial input function, the myocardial['50]H2O time-activity curve, and an assumed partitioncoefficient for [`50]H20 of 0.92, absolute regional myo-cardial blood flow was computed by a modification ofthe methods of Iida et a121 and Herrero et a122 thatautomatically accounts for partial volume and spill-overeffects.

Regional FDG uptake relative to blood flow. RegionalFDG uptake was then interpreted in relation to regionalmyocardial blood flow assessed by [`50]H20. Fourgroups of myocardial regions were identified: (1) normal(.80% of activity in the normal reference region asso-ciated with normal blood flow), (2) mismatch (reducedmyocardial blood flow with FDG:blood flow ratio.110% of that of the normal reference region), (3)moderately reduced FDG uptake (50% to 79% ofnormal reference FDG activity with reduced or normalblood flow and FDG:blood flow ratio <110% of that ofthe normal reference region), and (4) severely reducedFDG uptake (<50% of normal reference FDG activitywith reduced blood flow and FDG:blood flow ratio<110% of that of the reference region). As previouslydescribed,23 a region could be defined as showing mis-match if FDG activity was normal, increased, or lessthan normal, as long as FDG activity was disproportion-ately increased relative to the reduced regional bloodflow. The cutoff for the FDG:blood flow ratio of 110%was derived from previously published data obtained innormal volunteers in whom FDG:blood flow ratio valuesranged from 1.02 to 1.12 mg/g per minute,24 which is


Dow

nloaded from



similar to the 1.20 cutoff value used by Vanoverscheldeet al.25 We have previously shown that when regionswith reduced FDG uptake on PET images are sub-grouped on the basis of the severity of reduction inFDG activity, regions with moderately reduced FDGuptake have greater end-diastolic wall thickness andregional systolic wall thickening by gated magneticresonance imaging than regions with severely reducedFDG uptake.26 Thus, regions with moderately reducedFDG uptake were interpreted as representing a mixtureof viable and fibrotic myocardium, and they weregrouped together with normal and mismatch FDG-blood flow patterns as regions with viable myocardi-um.18 Regions with severely reduced FDG activity wereconsidered to represent myocardial fibrosis. For analysisof individual patients, regions were grouped for eachpatient as normal, mismatch, moderately reduced FDG,or severely reduced FDG activity.

Characterization of Patients WhoUnderwent PET StudiesWhen the characteristics of the 20 patients who under-

went PET studies were compared with those of the entiregroup of 41 patients, there was no difference with respectto extent of coronary artery stenosis. Among the 20patients, S patients (25%) had single-vessel disease, 6(30%) had marked narrowing of two vessels, and 9 (45%)of three vessels. Similarly, among the 41 patients, 12patients (29%) had single-vessel disease, 11 (27%) hadmarked narrowing of two vessels, and 18 (44%) of threevessels. The exercise duration on treadmill testing(6.4+3.0 vs 6.6±2.9 minute, P=NS), rate-pressure prod-uct achieved during exercise (21±6x103 vs 22±6x103,P=NS), and the percentages of patients with anginalsymptoms and medical therapy were the same in bothgroups. There were also no differences between the twogroups with respect to the left ventricular ejection fractionor wall motion abnormality at rest. Left ventricular ejec-tion fraction in the 41 patients ranged from 6% to 64%(mean, 32%) and was below the normal range in 31patients. In the subgroup of 20 patients who underwentPET studies, left ventricular ejection fraction ranged from6% to 50% (mean, 26%) and was below the normal rangein 18 patients.

Radionuclide AngiographyGated blood pool cardiac scintigraphy was performed

to assess left ventricular ejection fraction and regionalwall motion at rest by use of red blood cells labeled invivo with 20 to 25 mCi of 99mTc, as previously de-scribed.27 Left ventricular ejection fraction was derivedby computer analysis of the scintigraphic data, andregional wall motion was assessed qualitatively by twoexperienced observers from the images displayed incineangiographic format. The images were graded on athree-point scale as follows: grade 0, akinetic or dyski-netic; grade 1, hypokinetic; and grade 2, normal. Aregion was considered to have improved wall motionwhen the assigned abnormal regional grade increasedor normalized after revascularization. The lower limitof normal for resting ejection fraction by our tech-nique is 45%.

Coronary ArteriographyCardiac catheterization was performed by the percu-

taneous femoral technique. Coronary artery stenosis

and graft patency were assessed by experienced cardi-ologists without knowledge of exercise thallium results.In patients with bypass grafts, a vessel was consideredpatent if there was no significant narrowing within thegraft or in the native coronary artery distal to the graftanastomosis. For direct comparison of the severity ofstress and rest thallium defects with the results ofcoronary angiography, the five myocardial regions perpatient were grouped into the three major coronaryvascular territories, with the anterior and septal regionsrepresenting the left anterior descending artery terri-tory, the lateral region representing the left circumflexartery territory, and the inferior region representing theposterior descending artery territory. Assignment of theapex to a vascular territory was variable and based onthe results of coronary angiography and the presence ofadjacent perfusion defects.

Revascularization StudiesSeven patients underwent myocardial revasculariza-

tion (two percutaneous transluminal coronary angio-plasty and five coronary artery bypass surgery). Thesepatients were restudied within a mean of 10 monthsafter revascularization by coronary arteriography, thal-lium scintigraphy, and radionuclide angiography.

Statistical AnalysisData are presented as mean±SD. Differences be-

tween mild-to-moderate and severe regional thalliumuptake in concordant and discordant regions were ana-lyzed by the two-tailed unpaired t test. Group compar-isons between stress-redistribution-reinjection and rest-redistribution images were performed by x2 analysis.Differences between the entire group and the subgroupof patients who underwent PET studies with respect toexercise duration, rate-pressure product, symptoms ofangina, medical therapy, left ventricular function, wallmotion abnormalities, and the extent of coronary arterystenosis were analyzed by either two-tailed unpaired ttest or x2 analysis. Changes in left ventricular ejectionfraction and regional thallium activity from before toafter revascularization were analyzed by the two-tailedpaired t test. We did not perform statistical analysisbased on the results obtained from an analysis ofindividual patients. Since many myocardial regionswithin each patient had different categories of thalliumuptake and FDG:blood flow relations, each patientcould not be considered an independent unit ofobservation.

ResultsAnalysis of Regional Thallium ActivityAmong the 41 patients studied, a total of 205 myo-

cardial regions (5 regions per patient) were analyzed.The exercise thallium SPECT study identified 50 regionsto be normal and 155 regions to be abnormal during stress.Of the 155 abnormal regions, 91 defects (59%) wereirreversible on conventional 3- to 4-hour redistributionimages. After the reinjection of thallium at rest, 32 of the91 regions with irreversible defects (35%) had improvedor normalized thallium uptake (suggestive of viable myo-cardium), and 59 remained irreversible (Fig 1). Similarly,rest-redistribution imaging identified 39 of 91 regions(43%) with irreversible thallium defects on conventionalstress-redistribution images to be viable (P=NS) on the


Dow

nloaded from



z91Irreversible

Defect35% /\

32 LIF59Normalized or Irreversible

Improved

9.5jIt

Normal orReversible

Irreible

Rest-Redistribution

56%.

81%

Rest

Redistribution

,44%12%.

1

88%'

'19% 78%

Activity compared to stress-redistribution defectIncreased

Unchanged or Decreased

FIG 1. Flow diagram displaying the prevalence of reversibleand irreversible thallium perfusion defects by stress-redistri-bution-reinjection and rest-redistribution studies in the 41patients studied.

basis of either normal uptake initially at rest or an initialdefect that showed redistribution at 3 to 4 hours. Whenthe rest-redistribution results were compared with stress-redistribution-reinjection results, there was concordanceof data regarding detection of myocardial viability (nor-mal/reversible vs irreversible) in 72 (79%) of the 91irreversible defects, with 26 (28%) identified as viable and46 (51%) identified as scar. In the 19 remaining regions(21%) with discordance between the stress-redistribution-reinjection and rest-redistribution studies, viable myocar-dium was suggested in 6 by stress-redistribution-reinjec-tion alone and in 13 by rest-redistribution studies alone.

Initial rest thallium images (without exercise) de-tected only 18 (56%) of the 32 regions determined to beviable by reinjection (Fig 1), although they also identi-fied 7 additional regions with apparent viability amongthe 59 irreversible regions by stress-redistribution-rein-jection imaging. Redistribution images acquired 3 to 4hours after the initial rest thallium study identified anadditional 8 regions among the 32 reversible regions(identified by reinjection) and 6 regions among the 59irreversible regions to be viable.To investigate the apparent discordance between

stress-redistribution-reinjection and rest-redistributionimaging regarding myocardial viability and also to verifythe concordance of the two imaging protocols, regionalmetabolic data from the subset of 20 patients who alsounderwent PET studies with FDG and [150]H20 wereanalyzed.

Analysis of Regional Thallium Activity in PatientsWho Underwent PET Studies

Like the findings obtained in all 41 patients, whenstress-redistribution-reinjection and rest-redistributionimages were classified as normal/reversible (viable) or

n = 594

28047%

121

20%

46 147

I 8% 25%

Normal or IrreversibleReversible

Rest-Redistribution

FIG 2. Chart showing concordance and discordance be-tween stress-redistribution-reinjection and rest-redistributionimages in the 20 patients who underwent positron emissiontomographic studies. Five myocardial regions of interest weredrawn on the transaxial tomograms from the five sets ofthallium images, and thallium activities were then computedwithin each region. Thallium defects were classified as nor-

mal/reversible or irreversible.

irreversible (nonviable) in the 20 patients with PETstudies, the two imaging methods provided concordantinformation in 72% of myocardial regions (Fig 2).Among the 28% discordant regions, 20% were identi-fied to be reversible by stress-redistribution-reinjectionimaging and irreversible by rest-redistribution imaging,whereas 8% were identified to be normal or reversibleby rest-redistribution imaging and irreversible by stress-redistribution-reinjection imaging.

This 28% discordance between the two thalliumimaging methods resulted when all irreversible thalliumdefects were grouped together, without considering theseverity of the reduction in thallium activity within thedefect. We have previously shown that when irreversiblethallium defects on redistribution images are sub-grouped on the basis of the severity of reduction inthallium activity, mild-to-moderate (51% to 85% ofnormal activity) irreversible defects were shown by PETto be metabolically active and viable, whereas regionswith severe irreversible thallium defects were shown byPET to be metabolically inactive, and hence scarred.17,18Thus, we reanalyzed the concordant and discordantmyocardial regions after subgrouping the regions ac-cording to the severity of thallium activity within irre-versible defects.

Myocardial Viability Assessed According to Severityof Thallium DefectOf the 314 regions with irreversible thallium defects

by either stress-redistribution-reinjection or rest-redis-tribution imaging, 147 regions were determined to beirreversible by both imaging techniques, whereas discor-dant information regarding reversibility was obtained inthe other 167 regions.

Regions with concordant 20'T1 results. Among the 147regions in which both stress-redistribution-reinjectionand rest-redistribution imaging techniques were concor-dant in demonstrating irreversible thallium defects, 64had mild-to-moderate reduction in thallium activity(66+10%), and 83 had severely reduced thallium activ-ity (27+12%). PET identified 91% of the regions withmild-to-moderate irreversible thallium defects to beviable. In contrast, only 17% of regions with severe

Stress-Redistribution-Reinjection

Stress-redistribution

Reinjection

m~~~~~~~~~~~~~~~~~~~~~~~~~


Dow

nloaded from


Stress-Redistribution-Reinjection Rest-Redistribution

S RD RI R RD

FIG 3. Tomograms demonstrating concordance between positron emission tomographic (PET), stress-redistribution-reinjection,and rest-redistribution imaging in this patient example. Two consecutive transaxial tomograms are displayed for['8F]fluorodeoxyglucose (FDG) and myocardial bloodflow (MBF) by PET, with corresponding thallium tomograms ofstress (S),redistribution (RD), reinjection (RI), and rest (R)-redistribution. On the PET study, myocardial blood flow is reduced in theanteroapical, anteroseptal, and posteroseptal regions. FDG uptake in the corresponding regions demonstrates a mismatch in theposteroseptal region (shown by the arrowhead) and a match between FDG uptake and blood flow in the anteroapical andanteroseptal regions. Corresponding single photon emission computed tomographic thallium images reveal extensive perfusionabnormalities involving the apical and septal regions during stress, which persist on redistribution images. However, as in thefindings on PET, thallium reinjection images show improved thallium uptake in theposteroseptal region (shown by the arrowhead),whereas the apical region remains fixed. On rest-redistribution images, the apical region has severely reduced thallium activity,which remains fixed, whereas the posteroseptal region, which was abnormal on the initial rest study, shows significant improvementon 3- to 4-hour redistribution study, suggestive of viable myocardium.

irreversible thallium defects had preserved metabolicactivity. A representative example of a patient demon-strating concordance between PET, stress-redistribu-tion-reinjection, and rest-redistribution images is shownin Fig 3.

Regions with discordant 20'T1 results. Among the 167regions in which the viability information (normal/reversible vs irreversible) was discordant betweenstress-redistribution-reinjection and rest-redistributionstudies, the majority (149 regions, or 89%) had onlymild-to-moderate reduction in thallium activity. Therelative thallium activity within the mild-to-moderatediscordant regions was 69±9%, similar to that observedin mild-to-moderate concordant regions (66±10%,P=NS). PET verified 146 (98%) of these regions to bemetabolically active and viable.Of the 18 discordant regions with severe reduction in

thallium activity, 15 were judged to be viable by stress-redistribution-reinjection studies, and 3 were viable byrest-redistribution studies (Fig 4). All 18 regions withsevere defects were supplied by totally occluded orcritically stenosed coronary arteries. In 14 (78%) ofthese 18 discordant regions, stress-redistribution-rein-jection imaging provided concordant information withPET (12 viable and 2 scarred), whereas rest-redistribu-tion imaging underestimated viable myocardium. These14 regions represent 15% (14/92) of all (concordant anddiscordant) severe irreversible thallium defects and only2% (14/594) of all myocardial regions studied. Althoughthese regions were present in 7 of the 20 patientsstudied, they represent only 1 to 4 regions (from a meanof 30 regions) per patient. It is noteworthy that themean thallium activity in these severe defects was40±7%. Although this meets the criteria we havedefined for a severe defect (<50% activity), this level ofactivity was significantly greater than that observed inthe severe defects that were concordant (27±12%,P<.001).

Thallium Activity in Relation to PatternsofPET ViabilityThe results of PET patterns of viability in relation to

stress-redistribution-reinjection and rest-redistributionthallium data, analyzed according to myocardial regionsand patients, are shown in Tables 1 and 2, respectively.A total of 594 myocardial regions were evaluated in the20 patients, of which 209 (35%) had normal FDGuptake and regional blood flow, 188 (32%) had anischemic pattern with increased FDG uptake relative toreduced regional blood flow, 138 (23%) had moderatelyreduced FDG uptake, and 59 (10%) had severely re-duced FDG uptake. The 59 regions with severely re-duced FDG uptake were considered to represent myo-cardial fibrosis by PET.

Regional analysis. On the basis of the stress-redistri-bution-reinjection thallium studies (Table 1), 50 of the59 myocardial regions (85%) with severely reducedFDG uptake by PET had severe irreversible thalliumdefects. In contrast, only 2% of the normal myocardialregions by PET, 4% of the mismatch, and 13% of themoderately reduced FDG regions had severe irrevers-ible thallium defects. Similarly, on the basis of therest-redistribution thallium studies (Table 2), 49 of the59 myocardial regions (83%) with severely reducedFDG uptake by PET had severe irreversible thalliumdefects, and only 2% of the normal and 5% of themismatch myocardial regions by PET had severe irre-versible thallium defects. However, in regions withmoderately reduced FDG uptake, 26% of the regionshad severe irreversible defects on the rest-redistributionthallium studies, compared with 13% by stress-redistri-bution-reinjection studies.

Patient analysis. In this analysis, regions for eachpatient were grouped as having viable or nonviablemyocardium on the basis of relative thallium activity.Similarly, for the PET data, regions were grouped as


PET

FDG MBF


Dow

nloaded from



16DiscordantRegions

Mild-Moderate Severe(51%-85% of Peak) (C50% of Peak)

Vial Icar

2%

PET

Viable 1

ARest-Redistribution Stress-Redistrlbution-Relnjection Rest-Redistribution

FIG 4. Flow diagram displaying the results of positron emission tomography (PET) in regions with discordant viabilityinformation between stress-redistribution-reinjection and rest-redistribution thallium studies. Myocardial regions are subgroupedaccording to the severity of thallium activity within discordant defects.

normal, mismatch, moderately reduced FDG activity, orseverely reduced FDG activity. The results of thepatient analysis were similar to those of the regionalanalysis (Tables 1 and 2). Normal, mismatch, andmoderately reduced FDG PET patterns were observedin all 20 patients, whereas a severely reduced FDGpattern was observed in 10 patients. Among these 10patients with myocardial fibrosis by PET, 8 (80%) wereidentified to have nonviable myocardium both by thestress-redistribution-reinjection and rest-redistributionthallium studies. In contrast, only 1 of the 20 patientswith normal and 2 of the 20 patients with mismatchpattern by PET had severe irreversible thallium defectsby the stress-redistribution-reinjection imaging. Simi-larly, only 2 of the 20 patients with normal and 4 of the

20 patients with mismatch pattern by PET had severeirreversible thallium defects by the rest-redistributionimaging. However, among the 20 patients in whomregions with moderately reduced FDG uptake were

observed, 8 (40%) of the patients were identified tohave nonviable myocardium by the rest-redistributionthallium studies, compared with only 3 patients (15%)by stress-redistribution-reinjection studies.

Effects of RevascularizationAmong the seven patients who underwent myocardial

revascularization, six (86%) had an increase in restingleft ventricular ejection fraction, from 33±12% beforeto 39±13% after revascularization (P=.02). Six patientshad subnormal ejection fractions at rest (c45% for our

TABLE 1. Relation of Stress-Redistribution-Reinjection Thallium Results to Patterns of PET ViabilityPET

Stress- FDG: blood flow Moderately Severelyredistribution- Normal mismatch reduced FDG reduced FDGreinjectionthallium SPECT No. % No. S No. % No. S

Regional analysisn 209 188 138 59Normal 98 47 38 20 26 19 1 2Reversible 65 31 115 61 48 35 7 12Irreversible

Mild-to-moderate 42 20 28 15 46 33 1 2Severe 4 2 7 4 18 13 50 85

Patient analysisN 20 20 20 10Viable 19 95 18 90 17 85 2 20Nonviable 1 5 2 10 3 15 8 80

SPECT, single photon emission computed tomography; FDG, [18F]fluorodeoxyglucose; n, number of myocardial regions; N, number ofpatients in whom the corresponding positron emission tomography (PET) category was observed.

bcar 1


Dow

nloaded from



TABLE 2. Relation of Rest-Redistribution Thallium Results to Patterns of PET ViabilityPET

Rest- FDG: blood flow Moderately Severelyredistribution Normal mismatch reduced FDG reduced FDGthallium SPECT No. % No. % No. S No. %

Regional analysisn 209 188 138 59Normal 108 52 56 30 19 14 0 0Reversible 48 23 50 26 26 19 8 14Irreversible

Mild-to-moderate 49 23 73 39 57 41 2 3Severe 4 2 9 5 36 26 49 83

Patient analysisN 20 20 20 10Viable 18 90 16 80 12 60 2 20Nonviable 2 10 4 20 8 40 8 80

SPECT, single photon emission computed tomography; FDG, ['8F]fluorodeoxyglucose; n, number of myocardial regions; N, number ofpatients in whom the corresponding positron emission tomography (PET) category was observed.

laboratory) before revascularization, of whom five man-ifested an increase in ejection fraction from 30±8%before to 36±11% after revascularization (P<.04).A total of 14 regions were identified with abnormal

regional wall motion before revascularization, 7 withirreversible thallium defects despite reinjection and 7with improved thallium uptake after either 3- to 4-hourredistribution or the reinjection study. Three of the 7reversible defects (43%) were reversible only afterreinjection. All 7 regions with irreversible thalliumdefects on redistribution-reinjection imaging before re-vascularization (mean, 45±21%) again demonstratedirreversible thallium defects after revascularization(45±20%, P=NS) that were associated with persistentabnormalities in regional wall motion. In contrast, 6 of7 regions (86%) with improved thallium uptake afterredistribution or reinjection studies before revascular-ization (61±6%) had improved thallium uptake afterrevascularization (71±8%, P=.04) that were associatedwith improved regional wall motion at rest on the basisof radionuclide angiography. The remaining one regionwith persistent wall motion abnormality after revascu-larization had a reversible thallium defect that wasunchanged compared with the prerevascularizationstudy. Similarly, when rest-redistribution thallium up-take was assessed among the 14 regions with abnormalregional wall motion before revascularization, 7 hadirreversible thallium defects and 7 were either normalor reversible. Six of 7 regions (86%) with irreversiblethallium defects on redistribution imaging before revas-cularization (44±13%) again demonstrated irreversiblethallium defects after revascularization (48±25%,P=NS) that were associated with persistent abnormal-ities in regional wall motion. In contrast, 5 of 7 regions(71%) with normal or reversible thallium defects beforerevascularization (74±4%) had normal thallium uptakeafter revascularization (71±8%, P=NS) that was asso-ciated with improved regional wall motion at rest. Anexample of a patient with three-vessel coronary arterydisease in whom global and regional left ventricularfunction improved after revascularization is shown inFig 5.

Regional Thallium Uptake Relativeto Coronary AnatomyWhen the severity of stress and rest thallium defects

was compared with the severity of coronary arterystenosis, as anticipated, the stress studies identifiedmore myocardial regions subtended by stenotic coro-nary arteries than rest studies alone. Of the 205 myo-cardial regions analyzed, 155 regions (76%) were iden-tified to be abnormal during stress compared with 109regions (53%) with the rest-redistribution protocol.When the relative thallium activity after the stress andrest injections were assessed in each coronary arteryvascular territory, there was a significant associationbetween the severity of thallium uptake during stressand percent coronary artery stenosis. Relative to peakactivity, thallium activity in territories with 100% coro-nary occlusion was 42±14% (mean±SD) comparedwith 54±20% in territories supplied by 75% to 99%coronary stenosis (P<.01) and 61%±19% in territoriessupplied by 50% to 74% coronary stenosis (P<.003). Incontrast, there was no association between severity ofthallium uptake at rest and percent coronary arterystenosis. Thallium activity was 54±19% in territorieswith 100% coronary occlusion, 65±21% in territoriessupplied by 75% to 99% coronary stenosis, and64±20% in territories supplied by 50% to 74% coronarystenosis (P=NS).

DiscussionStress-redistribution 201T1 scintigraphy provides im-

portant diagnostic and prognostic information in pa-tients with coronary artery disease. However, conven-tional stress-redistribution imaging does not alwaysdistinguish between regions of ischemic or scarredmyocardium. Thallium reinjection immediately afterstress-redistribution imaging identifies ischemic but vi-able myocardium in 31% to 49% of myocardial regionsthat appear irreversible on conventional 3- to 4-hourredistribution images.8,9,28,29 However, elimination ofthe 3- to 4-hour redistribution images and reliance onreinjection images alone may incorrectly assign 8% to10% of thallium defects identified on stress images to beirreversible and scarred.8'30 These regions represent


Dow

nloaded from



PRE-OPStress-Redistribution-Reinjection Rest-Redistribution

S RD RI R RD

POST-CABGStress-Reinjection

S

ED

RI

ES

LVEF =40% LVEF = 54%FIG 5. 201Tl scintigraphy and radionuclide angiography studiesfrom a patient with significant three-vessel coronary artery diseasebefore and after coronary artery bypass surgery (CABG). Top, two consecutive short-axis tomograms ofpreoperative (pre-op)stress (S), redistribution (RD), reinjection (RI), and rest (R)-redistribution images with correspondingpostoperative (post-CABG)stress and reinjection tomograms. The preoperative thallium stress-redistribution images (on the left) reveal reversible anterior andminimally reversible inferior defects (shown by the arrowhead) that normalize after reinjection of thallium at rest. Rest-redistribution images identified reversible anterior and minimally reversible inferior defects. After revascularization (on the right),stress-reinjection images reveal normal thallium uptake in the anterior and inferior regions. Global and regional left ventricularfunction at rest, before and after operation, is shown in the bottom panel. Left ventricular ejection fraction (LVEF) was calculatedfrom end-diastolic (ED) and end-systolic (ES) frames in the left anterior oblique view. Before operation, the left ventricularcontraction was diffusely hypokinetic with a calculated ejection fraction of 40%. After revascularization, left ventricularcontraction significantly improved with an increase in ejection fraction to 54%.

19% to 25% of myocardial regions that would bereversible on 3- to 4-hour redistribution images.30 Thisresults from a disproportionately smaller increment inregional thallium activity after reinjection in some isch-emic regions compared with the uptake in normalregions, a phenomenon we have called "differentialuptake."8

Because of the logistical problems in performingthree sets of thallium images in all patients, as well asthe problem in interpreting the results if only one set ofimages is obtained after the stress images (either redis-tribution without reinjection or reinjection without re-distribution), it would be reasonable to perform thal-lium imaging with resting injections of the isotope if theclinical question to be addressed is one of myocardialviability, not exercise-induced ischemia. In 1979,Gewirtz et all' reported that thallium defects may occuron resting images in patients with severe coronaryartery disease in the absence of an acute ischemicprocess or previous myocardial infarction. They alsorecognized that many of these resting perfusion defectsredistribute over the next 2 to 4 hours, especially inregions with normal or hypokinetic wall motion. Sincethis initial report, three other studies evaluated theefficacy of rest-redistribution imaging in predicting theoutcome of myocardial regions after revasculariza-tion.12-14 Berger and coworkers12 investigated the out-come of rest-redistribution thallium abnormalities in 22patients undergoing coronary artery bypass surgery.Among the 48 regions with reversible defects before

surgery, 37 (77%) had completely normal thalliumuptake after surgery and 10 had persistent reversibledefects. However, 12 of 18 regions with irreversiblethallium defects before surgery (67%) also showednormal thallium uptake after surgery. These data sug-gest that although rest-redistribution thallium imagingidentifies viable myocardium in the majority of revers-ible regions, it may underestimate viable myocardium inirreversible regions. The second study, by Iskandrianand colleagues,13 examined the value of rest-redistribu-tion thallium imaging in predicting improvement in leftventricular ejection fraction after coronary artery by-pass surgery. In 14 patients with left ventricular dys-function at rest and normal or reversible thalliumabnormalities, 12 (86%) showed improvement in leftventricular function after surgery. In contrast, amongthe 9 patients with left ventricular dysfunction andirreversible perfusion defects on rest-redistribution im-aging, only 2 (22%) showed improvement in left ven-tricular function after surgery (P<.01). The third study,by Mori et al,14 showed that although 79% of asynergicregions with thallium redistribution had improved wallmotion after revascularization, 38% of asynergic regionswithout thallium redistribution before surgery alsoshowed improved wall motion after surgery. Thesethree studies were consistent in demonstrating that themajority of regions (77% to 86%) with reversible thal-lium defects before surgery have normal thallium up-take and/or improved left ventricular function aftersurgery. However, these studies were also consistent in


Dow

nloaded from



showing that 22% to 67% of regions with irreversibledefects also improve after revascularization. Hence, theavailable data suggest that although rest-redistributionthallium imaging identifies viable myocardium in themajority of reversible regions, it may underestimateviable myocardium in up to two thirds of irreversibleregions.However, it should be emphasized that none of the

three studies used quantitative thallium scintigraphicmethods for assessment of the severity of the irrevers-ible thallium defects. Thallium defects in each of thesestudies were classified as being reversible, partiallyreversible, or irreversible. More recently, improved re-sults were obtained with quantitative analysis in whichthe severity of reduction in thallium activity was con-sidered within irreversible rest-redistribution thalliumdefects.15 When myocardial viability was defined asthallium activity >50% of activity in normal regions,57% of severely asynergic regions that were viable bythallium demonstrated improved wall motion after sur-gery, compared with only 23% of severely asynergicregions that were considered to be nonviable bythallium.'5The results of the present study indicate that in

patients with chronic stable coronary artery disease,stress-redistribution-reinjection and rest-redistributionimaging provide concordant information in 72% ofirreversible defects when regions are classified simply aseither normal/reversible or irreversible thallium defects.This suggests a relatively high rate of discordance (28%)between the two methods of thallium imaging. How-ever, these results can be improved considerably byanalyzing irreversible defects further according to theseverity of the thallium defect. Of the discordant re-gions, 89% demonstrated only mild-to-moderate reduc-tion in thallium activity. We have previously demon-strated that regions with mild-to-moderate irreversiblethallium defects represent predominantly viable myo-cardium compared with PET17,18.26and have substantialuptake of thallium (using the differential uptake ratio)after reinjection17 and preserved regional systolic wallthickening and end-diastolic wall thickness by magneticresonance imaging.26 These findings are also supportedby data reported by Gibson and colleagues31 in patientsbefore and after revascularization using planar thalliumscintigraphy. Thus, the finding of only mild-to-moderatereduction in thallium activity within irreversible regionsconnotes predominantly viable myocardium and ap-pears to be independent of whether one is performing astress study or a rest-redistribution study. In our series,this quantitative approach increased the concordancebetween stress-redistribution-reinjection and rest-redis-tribution imaging regarding myocardial viability to 94%.That reversible or normal thallium uptake on stress-

redistribution-reinjection and rest-redistribution studiesrepresents viable myocardium is substantiated by im-proved regional and global function in the small sub-group of patients undergoing revascularization. Al-though the findings in these patients do support ourconclusion, we acknowledge that the number of patientsand myocardial regions submitted to revascularization issmall. However, it is also important to point out that thethallium reinjection protocol used in this study has beenshown to predict the functional recovery of regions withapparently irreversible defects on conventional stress-redistribution imaging828,32 with an accuracy similar to

that reported when FDG imaging with PET wasused.33-37 Thus, there are a number of prerevasculariza-tion and postrevascularization studies to date that sup-port the findings of FDG uptake on PET and thalliumuptake after reinjection as reliable and accurate mark-ers of viable myocardium.

In regions with severely reduced thallium activity,rest-redistribution imaging appears to underestimateviable myocardium in 15% of irreversible thallium de-fects compared with stress-redistribution-reinjectionimaging and PET. These severe discordant defectsoccurred in 2% of all myocardial regions studied, rep-resenting only 1 to 4 regions from a mean of 30 regionsper patient. Hence, the clinical relevance of theseregional differences for an individual patient is likely tobe negligible. Moreover, among these severe defects,the mean thallium activity in the concordant regions(27±12%) was significantly lower than the activitywithin the discordant regions (40±7%). It is likely thatwithin the range of thallium activity from 40% to 50% ofnormal activity, there is some overlap of informationregarding myocardial viability.Another possible explanation for the underestimation

of viable myocardium by rest-redistribution imaging insome discordant regions may relate to a rather severereduction in regional blood flow even under restingconditions. Such myocardial regions, supplied by totallyoccluded or critically stenosed coronary arteries, mayexhibit severely reduced thallium uptake at rest andremain irreversible on the redistribution study. How-ever, exercise stress may produce even greater regionalmyocardial blood flow heterogeneity. When followed byaugmentation of the serum thallium levels with reinjec-tion after 3 to 4 hours, the magnitude of improvement inthallium uptake from stress to reinjection will differen-tiate ischemic and viable myocardium from scarredmyocardium. Hence, it is not surprising that a smallnumber of regions with severely reduced and irrevers-ible thallium defect on rest-redistribution images maybe identified as ischemic rather than fibrotic myocar-dium on stress-redistribution-reinjection imaging.

In most cases, the identification of inducible myocar-dial ischemia is a much more important clinical variablein terms of patient management and risk assessmentthan the knowledge of myocardial viability. The influ-ence of exercise stress in differentiating an asynergicregion with admixture of scarred and viable myocar-dium from a region with underperfused but viable(hibernating) myocardium is outlined in Fig 6. A hypo-contractile segment may represent either an admixtureof scarred and viable myocardium that is nonischemic ora completely viable but hibernating myocardium. Thesetwo hypocontractile segments may be differentiated bydemonstrating exercise-induced ischemia (reversiblethallium defect) in the case of hibernating myocardiumand absence of ischemia (irreversible thallium defect) inthe case of the admixture of fibrotic and viable myocar-dium. Accurate distinction between these two hypocon-tractile segments has important clinical implications,since impaired regional function is a potentially revers-ible process in the case of hibernating myocardium butnot in regions with admixture of viable and scarredmyocardium.

Preliminary data regarding the frequency of late (18-to 72-hour) redistribution imaging after the injection ofthallium at rest suggest that 13% to 30% of patients


Dow

nloaded from



Coronary Anatomy

Pathology

Thallium-201Stress-Reinjection

Post-OpRegional Wall Motion

Unperfused butViable Myocardium(Hibernating)

FIG 6. Schematic ofhow exercise stress mayproduce a greater regional myocardial bloodflow heterogeneity compared with rest-redis-tribution imaging and thereby differentiate anasynergic region with an admixture ofscarredand viable myocardium from a region withunderperfused but viable (hibernating) myo-cardium. A preoperative asynergic myocar-dial region that exhibits reduced thalliumuptake at rest and remains irreversible on theredistribution study may represent a regionwith patent (40% stenosis) coronary arteryafter thrombolytic therapy (on the left) or aregion with critically narrowed (90% stenosis)coronary artery without myocardial infarc-tion (on the right). In the case of the patientwith 40% coronary artery stenosis and priormyocardial infarction, the dysfunctionalmyocardium (assessed several months afterthe acute infarction) represents an admixtureof scarred and viable myocardium that willnot recover after revascularization. In con-trast, in the patient with 90% coronary arterystenosis without prior myocardial infarction,the dysfunctional myocardium perfused bythis artery represents underperfused but via-ble myocardium that will recover completelyafter revascularization. These two situationscan be differentiated by a stress-redistribu-tion-reinjection study but not by a rest-redis-tribution study.

studied for myocardial infarction, unstable angina, orheart failure had additional redistribution on late imag-ing.38'39 However, the frequency of late redistribution inour study population with chronic stable coronary ar-tery disease is unknown.

Thallium reinjection image is a composite of restingperfusion superimposed on a stress-redistribution im-age. Although initial rest thallium images (withoutexercise) detected only 56% of the regions determinedto be viable by stress-redistribution-reinjection, theyidentified 7 additional viable regions among the 59regions determined to be irreversible (and thereforeinterpreted to be scarred) by stress-redistribution-rein-jection imaging. Since reinjection of thallium afterstress-redistribution imaging does not provide the sameinformation as a rest thallium study (without exercise),the terms reinjection and rest-injected thallium shouldnot be used interchangeably. Redistribution imagesacquired 3 to 4 hours after the initial rest thallium studyidentified an additional 14 viable regions, of which 8were identified to be viable by stress-redistribution-reinjection regions. Therefore, our results are consistentwith previous reports indicating that 3- to 4-hour redis-tribution images after initial rest images are required todifferentiate viable from nonviable myocardium. 1-15

In conclusion, one of two imaging modalities, eitherstress-redistribution-reinjection or rest-redistributionimaging, may be used for identifying viable myocar-

dium. However, if there are no contraindications tostress testing, stress-redistribution-reinjection imagingprovides a more comprehensive assessment of the ex-tent and severity of coronary artery disease by demon-strating regional myocardial ischemia without jeopar-dizing information on myocardial viability.

References1. Rozanski A, Berman D, Gray R, Diamond G, Raymond M, Prause

J, Maddahi J, Swan HJC, Matloff J. Preoperative prediction ofreversible myocardial asynergy by postexercise radionuclide ven-triculography. N Engl J Med. 1982;307:212-216.

2. Brundage BH, Massie BM, Botvinick EH. Improved regional ven-tricular function after successful surgical revascularization. J AmCoil Cardiol. 1984;3:902-908.

3. Dilsizian V, Bonow RO, Cannon RO, Tracy CM, Vitale DF,McIntosh CL, Clark RE, Bacharach SL, Green MV. The effect ofcoronary artery bypass grafting on left ventricular systolic functionat rest: evidence for preoperative subclinical myocardial ischemia.Am J Cardiol. 1988;61:1248-1254.

4. Bonow RO, Dilsizian V. Thallium-201 for assessing myocardialviability. Semin Nucl Med. 1991;21:230-241.

5. Rahimtoola SH. Coronary bypass surgery for chronic angi-na- 1981: a perspective. Circulation. 1982;65:225-241.

6. Braunwald E, Rutherford JD. Reversible ischemic left ventriculardysfunction: evidence for the "hibernating myocardium." J AmCoil Cardiol. 1986;8:1467-1470.

7. Dilsizian V, Bonow RO. Current diagnostic techniques of assessingmyocardial viability in hibernating and stunned myocardium. Cir-culation. 1993;87:1-20.

8. Dilsizian V, Rocco TP, Freedman NM, Leon MB, Bonow RO.Enhanced detection of ischemic but viable myocardium by the


Dow

nloaded from



reinjection of thallium after stress-redistribution imaging. N Engl JMed. 1990;323:141-146.

9. Rocco TP, Dilsizian V, McKusick KA, Fischman AJ, Boucher CA,Strauss HW. Comparison of thallium redistribution with rest "rein-jection" imaging for the detection of viable myocardium. Am JCardiol. 1990;66:158-163.

10. Kayden DS, Sigal S, Soufer R, Mattera J, Zaret BL, Wackers FJ.Thallium-201 for assessment of myocardial viability: quantitativecomparison of 24-hour redistribution imaging with imaging afterreinjection at rest. JAm Coll CardioL 1991;18:1480-1486.

11. Gewirtz H, Beller GA, Strauss HW, Dinsmore RE, Zir LM, Mc-Kusick KA, Pohost GM. Transient defects of resting thallium scansin patients with coronary artery disease. Circulation. 1979;59:707-713.

12. Berger BC, Watson DD, Burwell LR, Crosby IK, Wellons HA,Teates CD, Beller GA. Redistribution of thallium at rest inpatients with stable and unstable angina and the effect of coronaryartery bypass surgery. Circulation. 1979;60:1114-1125.

13. Iskandrian AS, Hakki A, Kane SA, Goel IP, Mundth ED, Hakki A,Segal BL, Amenta A. Rest and redistribution thallium-201 myo-cardial scintigraphy to predict improvement in left ventricularfunction after coronary artery bypass grafting. Am J Cardiol. 1983;51:1312-1316.

14. Mori T, Minamiji K, Kurogane H, Ogawa K, Yoshida Y. Rest-injected thallium-201 imaging for assessing viability of severeasynergic regions. J Nucl Med. 1991;32:1718-1724.

15. Ragosta M, Beller GA, Watson DD, Kaul S, Gimple LW. Quan-titative plenar rest-redistribution aHTl imaging in detection of myo-cardial viability and prediction of improvement in left ventricularfunction after coronary bypass surgery in patients with severelydepressed left ventricular function. Circulation. 1993;87:1630-1641.

16. Dilsizian V, Smeltzer WR, Freedman NMT, Dextras R, BonowRO. Thallium reinjection after stress-redistribution imaging: does24-hour delayed imaging after reinjection enhance detection ofviable myocardium? Circulation. 1991;83:1247-1255.

17. Dilsizian V, Freedman NMT, Bacharach SL, Perrone-Filardi P,Bonow RO. Regional thallium uptake in irreversible defects: mag-nitude of change in thallium activity after reinjection distinguishesviable from nonviable myocardium. Circulation. 1992;85:627-634.

18. Bonow RO, Dilsizian V, Cuocolo A, Bacharach SL. Identificationof viable myocardium in patients with coronary artery disease andleft ventricular dysfunction: comparison of thallium scintigraphywith reinjection and PET imaging with 18F-fluorodeoxyglucose.Circulation. 1991;83:26-37.

19. Bergmann SR, Herrero P, Markham J, Weinheimer CJ, WalshMN. Noninvasive quantitation of myocardial blood flow in humansubjects with oxygen-15-labeled water and positron emissiontomography. JAm Coll Cardiol. 1989;14:639-652.

20. Bacharach SL, Cuocolo A, Bonow RO, Carson RE, Unger EF,Sheffield C, Finn R, Stein S, Herscovitch P, Green MV. Arterialblood concentration curves by cardiac PET without arterialsampling or image reconstruction. Computers in Cardiology. Wash-ington, DC: IEEE Computer Society Press; 1989:219.

21. Iida H, Kanno I, Tamahashi A, Miura S, Murakami M, Takahishi K,Ono Y, Shishido F, Inugami A, Tomura N, Higano S, Fujita H, SasakiH, Nakamichi H, Mizusawa S, Kondo Y, Uemura K. Measurement ofabsolute myocardial blood flow with H2150 and dynamic positron-emission tomography: strategy for quantification in relation to thepartial-volume effect. Circulation. 1988;78:104-115.

22. Herrero P, Markham J, Bergmann SR. Quantitation of myocardialblood flow with H2150 and positron emission tomography:assessment and error analysis of a mathematical model. J ComputAssist Tomogr. 1989;13:862-873.

23. Marshall R, Tillisch JH, Phelps ME, Huang SC, Carson R, HenzeE, Schelbert HR. Identification and differentiation of resting myo-cardial ischemia and infarction in men with positron computedtomography, F-18 labeled fluorodeoxyglucose, and N-13 ammonia.Circulation. 1983;67:766-778.

24. De Landsheere C, Raets D, Peirard L, Degueldre C, Legrand V,Lemaire C, Guillaume M, Lamotte D, Kulbertus HE, Rigo P.

Regional myocardial perfusion and glucose uptake: clinical expe-rience in 92 cases studied with positron tomography. In: SchmidtHAE, Chambron J, eds. Nuclear Medicine: Quantitative Analysis inImaging and Function. Stuttgart: Schattauer Verlag; 1990:245-247.

25. Vanoverschelde JU, Merlin JA, Bol A, Vanbutsele R, Cogneau M,Labar D, Robert A, Michel C, Wijns W. Regional oxidativemetabolism in patients after recovery from reperfused anteriormyocardial infarction: relation to regional blood flow and glucoseuptake. Circulation. 1992;85:9-21.

26. Perrone-Filardi P, Bacharach SL, Dilsizian V, Maurea S, FrankJA, Bonow RO. Regional left ventricular wall thickening: relationto regional uptake of 18fluorodeoxyglucose and 1Tl in patientswith chronic coronary artery disease and left ventricular dys-function. Circulation. 1992;86:1125-1137.

27. Borer JS, Bacharach SL, Green MV, Kent KM, Epstein SE,Johnston GS. Real-time radionuclide cineangiography in the non-invasive evaluation of global and regional left ventricular functionat rest and during exercise in patients with coronary artery disease.N Engl I Med. 1977;296:839-844.

28. Ohtani H, Tamaki N, Yonekura Y, Mohiuddin IH, Hirata K, BanT, Konishi J. Value of thallium-201 reinjection after delayedSPECT imaging for predicting reversible ischemia after coronaryartery bypass grafting. Am J Cardiol. 1990;66:394-399.

29. Tamaki N, Ohtani H, Yonekura Y, Nohara R, Kambara H, KawaiC, Hirata K, Ban T, Konishi J. Significance of fill-in afterthallium-201 reinjection following delayed imaging: comparisonwith regional wall motion and angiographic findings. J Nucl Med.1990;31:1617-1623.

30. Dilsizian V, Bonow RO. Differential uptake and apparent 'Tlwashout after thallium reinjection: options regarding early redis-tribution imaging before reinjection or late redistribution imagingafter reinjection. Circulation. 1992;85:1032-1038.

31. Gibson RS, Watson DD, Taylor GJ, Crosby IK, Wellons HL, HoltND, Beller GA. Prospective assessment of regional myocardialperfusion before and after coronary revascularization surgery byquantitative thallium-201 scintigraphy. JAm Coll Cardiol. 1983;1:804-815.

32. Nienaber CA, de la Roche J, Carnarius H, Montz R. Impact of201thallium reinjection imaging to identify myocardial viability aftervasodilation-redistribution SPECT. J Am Coll Cardiol. 1993;21:283A. Abstract.

33. Tillisch JH, Brunken R, Marshall R, Schwaiger M, Mandelkorn M,Phelps M, Schelbert H. Reversibility of cardiac wall-motion abnor-malities predicted by positron tomography. N Engl J Med. 1986;314:884-888.

34. Tamaki N, Yonekura Y, Yamashita K, Saji H, Magata Y, Senda M,Konishi Y, Hirata K, Ban T, Konishi J. Positron emissiontomography using fluorine-18 deoxyglucose in evaluation ofcoronary artery bypass grafting. Am J Cardiol. 1989;64:860-865.

35. Nienaber CA, Brunken RC, Sherman CT, Yeatman LA, GambhirSS, Krivokapich J, Demer LL, Ratib 0, Child JS, Phelps ME,Schelbert HR. Metabolic and functional recovery of ischemichuman myocardium after coronary angioplasty. JAm Coll Cardiol.1991;18:966-978.

36. Marwick TH, MacIntyre WJ, Lafont A, Nemec JJ, Salcedo EE.Metabolic responses of hibernating and infarcted myocardium torevascularization: a follow-up study of regional perfusion, function,and metabolism. Circulation. 1992;85:1347-1353.

37. Gropler RJ, Geltman EM, Sampathkumaran K, Perez JE,Moerlein SM, Sobel BE, Bergmann SR, Siegel BA. Functionalrecovery after coronary revascularization for chronic coronaryartery disease is dependent on maintenance of oxidativemetabolism. JAm Coll Cardiol. 1992;20:569-577.

38. Takemoto KA, Kiat H, Friedman J, Silber H, Biasio Y, Palmas W,Berman DS. The frequency of late redistribution in restthallium-201 scintigraphic studies. Circulation. 1991;84:II-88.Abstract.

39. McCallister BD, Clements IP, Hauser MF, Gibbons RJ. The valueof 24-hour images after rest thallium injection. JAm Coll Cardiol.1992;19:128A. Abstract.


Dow

nloaded from


BonowV Dilsizian, P Perrone-Filardi, J A Arrighi, S L Bacharach, A A Quyyumi, N M Freedman and R O

positron emission tomography.thallium imaging for assessing viable myocardium. Comparison with metabolic activity by

Concordance and discordance between stress-redistribution-reinjection and rest-redistribution

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1993 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.88.3.941

1993;88:941-952Circulation.

http://circ.ahajournals.org/content/88/3/941World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Circulation Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:


Dow

nloaded from

http://circ.ahajournals.org/content/88/3/941

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


concordance discordance stress-redistribution-reinj...

Documents