- Email: [email protected]
Correlation of Resting First-Pass Left Ventricular Ejection Fraction and Resting Myocardial Infarct Size
Correlation of Resting First-Pass Left Ventricular Ejection Fraction and Resting Myocardial Infarct Size Panithaya Chareonthaitawee, MD, Timothy F. Christian, MD, Todd D. Miller, David O. Hodge, MS, and Raymond J. Gibbons, MD
MD,
This study determined the correlation between the extent of the resting perfusion defect by technetium99m sestamibi tomographic imaging and the firstpass left ventricular (LV) ejection fraction (EF). A total of 1,955 patients underwent technetium-99m sestamibi tomographic imaging with measurement of firstpass resting LVEF. Twenty-five percent of patients had a prior history of myocardial infarction. First-pass LVEF was measured using a peripheral intravenous injection and a multicrystal gamma camera with standard software. Resting tomographic perfusion defect size (infarct size) was quantitated using previously published methods. Mean LVEF for the study group was 0.60 6 0.11. Mean LV infarct size was 5 6 11%. For the 1,265 patients (65% of the study group) with no measurable perfusion defect, the preva-
lence of a normal (>0.50) LVEF was 96% (1,212 of 1,265 patients). For patients with a measurable defect (n 5 690, 35%), the inverse linear correlation with LVEF was highly significant (r 5 20.60, p <0.0001) but with wide confidence limits (SEE 5 10 LVEF points), thereby limiting the predictive value in individual patients. Thus, in the absence of known cardiomyopathy, valvular heart disease, or left bundle branch block, patients without a quantifiable resting perfusion defect are highly likely to have a normal resting LVEF and may not require determination of LV function. For patients with resting perfusion defects, LVEF cannot be predicted with confidence and should therefore be measured. Q1998 by Excerpta Medica, Inc. (Am J Cardiol 1998;81:1281–1285)
eft ventricular (LV) ejection fraction (EF) is a major determinant of prognosis in patients with L coronary artery disease. Assessment of LVEF is
METHODS
1,2
therefore an important component in the evaluation of the cardiac patient. Measurement of LVEF is now possible in combination with perfusion owing to the higher count statistics obtained with technetium (Tc)99m– based perfusion agents. These perfusion tracers have previously been demonstrated to provide a reliable and accurate means of assessing LVEF.3–5 However, depending on how this information is obtained, there may be added processing time and cost. Previous studies have demonstrated that in certain patient subsets, LVEF may be reliably predicted from clinical and electrocardiographic variables.6 – 8 In addition, prior reports from this institution have shown a correlation between infarct size by Tc-99m sestamibi tomographic imaging and LVEF by first-pass radionuclide ventriculography and by electron beam computed tomography in the first year after an acute myocardial infarction.9,10 This study determines the correlation between the extent of the resting tomographic perfusion defect and the first-pass LVEF, and assesses the ability of the resting perfusion defect to predict LVEF.
From the Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Manuscript received September 16, 1997; revised manuscript received and accepted February 5, 1998. Address for reprints: Timothy F. Christian, MD, Mayo Clinic, 200 First Street SW, East 16-B, Rochester, Minnesota 55905. ©1998 by Excerpta Medica, Inc. All rights reserved.
Study group: The study group consisted of a series of patients referred to the nuclear cardiology laboratory at the Mayo Clinic for Tc-99m sestamibi perfusion imaging between March 1994 and April 1996. Before the stress portion of each study, the responsible monitoring personnel (physician or nurse) completed a database questionnaire, which was subsequently entered into a computer by a research nurse. The study group was identified from this computer database and included all patients with (1) a 2-day Tc-99m sestamibi rest-stress protocol, and (2) technically adequate Tc-99m sestamibi tomographic perfusion imaging and first-pass resting LVEF. Patients were excluded if they had known cardiomyopathy (56 patients), valvular heart disease (123 patients), left bundle branch block (99 patients), a permanent pacemaker (40 patients), or an arrhythmia. A total of 1,955 patients were included in the study group. Tc-99m sestamibi infarct (defect) size: Tc-99m sestamibi perfusion studies were performed as a 2-day (rest-stress) protocol at the Mayo Clinic Nuclear Cardiology Laboratory. The resting image was recorded on day 1. Approximately 30 to 45 minutes after injection of 20 to 30 mCi of Tc-99m sestamibi, tomographic imaging was performed with 30 images over a 180° arc, 45° right anterior oblique to 45° left posterior oblique postion using a step and shoot mode.11 Perfusion defect size was quantitated from the sum of 5 short-axis slices and determined from the number of pixels with ,60% of maximal counts over the 5 slices (Figure 1).11 Infarct size was defined as the extent of the perfusion defect and expressed as a percentage of the LV wall.11,12 This method of quantifying infarct 0002-9149/98/$19.00 PII S0002-9149(98)00156-8
1281
and dynamic acquisition was performed with a 32 3 32 matrix and a frame rate of 30 to 40 frames/s. LVEF was calculated from $4 cardiac cycles using a count-based method from generated time activity curves and commercially available software (Scinticor).17 Studies were considered technically suboptimal if there was evidence of arrhythmia during the acquisition, a poor bolus, poor separation of right and left ventricular phases, or poor counting statistics. Each study was reviewed by at least 1 trained observer for technical quality. A first-pass LVEF $0.50 is considered normal using this approach in our laboratory. FIGURE 1. Tc-99m sestamibi tomographic imaging of an inferior defect. Short-axis count profile starts at the anterior wall and progresses clockwise at 6& intervals. Perfusion defect size was quantitated from the sum of the 5 short-axis slices and determined from the number of pixels with <60% of maximal counts.
TABLE I Clinical Characteristics of Study Group (n 5 1,955) No. (%) Women/men Prior myocardial infarction Diabetes mellitus Insulin-requiring Systemic hypertension* Hyperlipidemia† Smoking history Normal resting electrocardiogram Prior coronary artery bypass grafting Medication use b blocker Calcium antagonist Nitrates
675 (35%)/1,280 (65%) 496 (25%) 451 (23%) 188 (10%) 1,132 (58%) 1,169 (60%) 1,126 (58%) 700 (36%) 372 (19%) 634 (32%) 708 (36%) 407 (21%)
*Systemic hypertension defined by history or by a blood pressure at the outpatient clinic of .140/90 mm Hg. † Hyperlipidemia defined by history or by either a cholesterol or triglycerides measurement at .90th percentile.
size has been validated by both animal and clinical trials,9,13,14 and the 60% threshold has previously been shown to provide the optimal separation between viable and nonviable myocardium.15,16 First-pass radionuclide angiography: First-pass LVEF was measured after the sestamibi injection for the same resting study. A large bore (18-gauge, 1.3mm) intravenous catheter was placed in a medial antecubital vein. The patient was positioned in the anterior view of a multicrystal gamma camera (Scinticor SIM-400, Milwaukee, Wisconsin) with a medium-sensitivity collimator and a 50% energy window centered over the 140-keV photo peak. A 0.5-mCi test dose of Tc-99m sestamibi was administered to define the left ventricular region of interest. Tc-99m sestamibi (20 to 30 mCi) was then injected as a bolus dose, 1282 THE AMERICAN JOURNAL OF CARDIOLOGYT
VOL. 81
Clinical and historical variables: Prior to stress imaging,
clinical and historical variables were prospectively recorded for all patients.
Other estimations of left ventricular ejection fraction:
A systematic search of the database was performed retrospectively to identify LVEF measurements by other modalities (contrast left ventriculogram by biplane technique and echocardiogram by Simpson’s rule) within 3 months of the first-pass LVEF of interest in the study population. Statistical analysis: Data are presented as mean 6 SD. Simple linear regression analysis was used to correlate infarct size by Tc-99m sestamibi tomographic imaging with the first-pass LVEF. Chi-square and 2-tailed probability tests were performed to compare the clinical characteristics of patients with measurable resting perfusion defects with and without a history of prior infarction. Spearman rank correlation coefficients were calculated to quantify the strength of the correlation between first-pass LVEF and LVEF measurements by contrast left ventriculography and echocardiography.
RESULTS Pertinent clinical characteristics of the study group are listed in Table I. The mean age of the study group was 63 6 11 years. Most patients were men (65%). Twenty-five percent had prior myocardial infarction. Thirty-six percent had a normal resting electrocardiogram. Table II summarizes the comparison between the clinical characteristics of patients with a resting perfusion defect (n 5 690) versus those of patients without a resting perfusion defect (n 5 1,265). Patients with a measurable perfusion defect were older and were more likely to be men, to be diabetic, to take b blockers, to have a history of infarction, and to have a history of cigarette smoking. Patients without a meaJUNE 1, 1998
In patients with no measurable perfusion defect by Tc-99m sestamibi tomographic imaging (n 5 With Defect Without Defect p 1,265, 65%), the prevalence of a nor(n 5 690) (n 5 1,265) Value mal resting LVEF was 96% (Figure Age (median, yr) 66 64 0.0003 3). Conversely, in patients with a Women/men 99 (14%)/591 (86%) 576 (45%)/689 (54%) 0.001 measurable perfusion defect, the Prior myocardial infarction 319 (46%) 177 (14%) 0.001 prevalence of a normal LVEF was Diabetes mellitus 179 (26%) 272 (22%) 0.035 only 65%. Insulin-requiring 81 (12%) 107 (8%) 0.013 Systemic hypertension* 404 (59%) 728 (58%) 0.605 For the 690 patients with a meaHyperlipidemia† 386 (57%) 780 (62%) 0.026 surable defect, the inverse linear corSmoking history 448 (65%) 678 (54%) 0.001 relation between defect extent and Normal resting 133 (19%) 567 (45%) 0.001 first-pass LVEF (Figure 4) was electrocardiogram highly significant (r 5 20.60, p Medication use b blocker 253 (37%) 381 (30%) 0.003 ,0.0001); however, the standard erCalcium antagonist 259 (37%) 449 (35%) 0.369 ror was 10 LVEF points, thereby Nitrates 213 (31%) 194 (15%) 0.001 limiting the predictive value in indi*Systemic hypertension defined by history or by a blood pressure at the outpatient clinic of .140/90 vidual patients. mm Hg. A larger proportion of patients † Hyperlipidemia defined by history or by either a cholesterol or triglycerides measurement at .90th (35%) had a measurable perfusion percentile. defect than those who had a history Data are expressed as number (%) except for age (median, in years). of prior myocardial infarction (25%). We therefore compared the clinical characteristics of patients with a measurable perfusion defect but no history of myocardial infarction (n 5 369) to patients with a measurable defect and a history of myocardial infarction (n 5 319). The results are summarized in Table III. There were no significant differences in the 2 groups with respect to gender or diabetes. Patients with a measurable resting defect and no history of myocardial infarction were significantly older, tended to be hypertensive, and were more likely to have an inferior defect. There were 144 and 431 patients with LVEF measurements by contrast left ventriculography and echocardiography, respectively, within 3 months of the firstpass study. The correlation between firstFIGURE 2. Histogram showing the distribution of infarct size determined by pass and contrast left ventriculographic Tc-99m sestamibi for the study group of 1,955 patients. The distribution was highly skewed toward no resting defect (65% of study group) and smaller LVEF (r 5 0.61, SEE 5 0.07) was similar (<10% of the left ventricle) defects (17.5% of study group). to the correlation between first-pass and echocardiographic LVEF (r 5 0.66, SEE 5 0.04). TABLE II Comparison of Clinical Characteristics of Patients With or Without a Resting Defect
surable perfusion defect were more likely to have a normal resting electrocardiogram. Tc-99m sestamibi perfusion imaging and first-pass ejection fraction: The distribution of infarct size deter-
mined by Tc-99m sestamibi tomographic imaging for the study group is shown in Figure 2. The mean infarct size for the group of 1,955 patients was 5 6 11% of the left ventricle (median 5 0), with a range of 0% (n 5 1,265) to 76% of the left ventricle. The distribution of infarct size was highly skewed toward no resting defects (65% of study group) and smaller (#10% of the left ventricle) defects (17.5% of study group). The mean Tc-99m sestamibi first-pass LVEF was 0.60 6 0.11 (median 5 0.62), and the range 0.15 to 0.89.
DISCUSSION LVEF assessment using the first-pass technique has been adopted as part of Tc-99m sestamibi imaging in many nuclear cardiology laboratories. This method of assessing LV function has previously been shown to be accurate and reproducible.3–5 First-pass LVEF can be acquired either with a multicrystal gamma camera or with electronically enhanced single-crystal detectors and ultra high sensitivity collimators.4 Commercial software to perform the gated analysis of LVEF from tomographic images is readily available, several of which are almost fully automatic.18 In addition to global LVEF, assessment of regional wall motion may also provide valuable information.19,20 However, for laboratories using ungated analysis, the
CORONARY ARTERY DISEASE/EJECTION FRACTION AND INFARCT SIZE
1283
third of the study population), this study shows a highly significant inverse linear correlation between defect size and LVEF, but with wide confidence limits. Therefore, in a patient with a resting perfusion defect, LVEF cannot be predicted with confidence from the extent of the defect and should be measured. A normal resting electrocardiogram and certain clinical variables may be used to predict normal LV function.6 – 8,27 However, in this study and a previous report27 from this laboratory, the prevalence of a normal resting electrocardiogram in patients referred for Tc-99m sestamibi perfusion imaging was 36% and 33%, respecFIGURE 3. Bar graph showing the percentage of patients with normal LV functively. As a result, in one third of patients tion (defined as LVEF >50%) in the presence or absence of a measurable perfusion defect. For the 1,265 patients (65%) with no measurable defect, the referred for perfusion imaging, a normal prevalence of normal resting LV function was 96%. In patients with a measurLVEF cannot be predicted using this apable defect (n 5 690; 35%), the prevalence of normal resting LV function was proach. In addition, the strongest clinical only 65%. predictor of a normal LVEF (no history of prior myocardial infarction) was associated with only an 87% prevalence of normal LV function.27 Thus, data from this study allow further identification of patients at low risk for LV dysfunction. When the first-pass LVEF study is inadequate on day 1 of a 2-day Tc-99m sestamibi protocol, clinicians may use the absence of a measurable tomographic perfusion defect to forego repeat gated analysis for LV function on the following day. In patients undergoing ungated studies, the absence of a measurable perfusion defect also allows the clinician to predict a normal LVEF without the additional time or cost of the measurement (the first-pass study to obtain LVEF adds 2.52 Medicare relative value units). In patients FIGURE 4. Scatterplot showing the inverse correlation of resting LV function undergoing same-day rest-stress imaging, and infarct size: r 5 20.60, p <0.0001, SEE 10 LVEF points. the absence of a measurable defect on the small but additional patient and laboratory acquisition resting image precludes the need for gating during stress and processing time for routine LVEF assesment by imaging. Christian et al28 previously demonstrated that thalthis first-pass technique may not be justified. Additionally, although recent reports suggest that Tc-99m lium-201 imaging provides significantly larger estisestamibi may be a marker of myocardial viability,15 mates of infarct size than Tc-99m sestamibi imaging. particularly if used in conjunction with nitrate infu- Hence, in the absence of a measurable defect by sion,21–23 other studies have shown that Tc-99m ses- thallium-201, clinicians can also infer the presence of tamibi does not provide as complete viability infor- normal LV function and avoid other diagnostic studies mation when compared with thallium-201.24 –26 Con- to assess LVEF. Last, the finding of an abnormal EF in sequently, a shift to Tc-99m– based perfusion agents the absence of a measurable tomographic perfusion to obtain LVEF measurements may result in decreased defect would be unexpected and the quality of the study should be carefully scrutinized. detection of myocardial viability. Limitations to this study include the exclusion of In the present study, performed at a tertiary referral center, nearly two thirds of patients had no measurable patients with a known cardiomyopathy, valvular disresting perfusion defect. The prevalence of a normal ease, left bundle branch block, arrhythmias, and carLVEF in this group was 96%. Therefore, clinicians diac pacemaker. In these cases, LV function may be can be .95% confident (confidence interval 93% to influenced by factors unrelated to myocardial perfu96%) that a patient without a measurable perfusion sion. A second limitation is the less likely use of defect (two thirds of the referral population) will have Tc-99m sestamibi when viability is a key issue. Thalnormal global resting LV function, and measurement lium-201 is used preferentially in these instances. A third limitation is the potential attenuation of counts of LVEF is most likely not justified. In patients with a measurable perfusion defect (one from overlying soft tissue (diaphragm and breast), 1284 THE AMERICAN JOURNAL OF CARDIOLOGYT
VOL. 81
JUNE 1, 1998
8. Rihal CS, Davis KB, Kennedy JW, Gersh BJ. The utility of clinical, electrocardiographic, and roentgenographic variables in the prediction of left ventricular function. Am J Cardiol 1995;75:220 –223. With History Without History p 9. Christian TF, Behrenbeck T, Gersh BJ, Gibbons RJ. Relation of left ventricular volume and function over (n 5 319) (n 5 369) Value one year after acute myocardial infarction to infarct size Age (median, yr) 67 64 0.009 determined by technetium-99m sestamibi. Am J Cardiol 1991;68:21–26. Weight (median, kg) 94 89 0.01 10. Chareonthaitawee P, Christian TF, Hirose K, GibWomen/men 44 (14%)/275 (86%) 54 (15%)/315 (85%) 0.753 bons RJ, Rumberger JA. Relation of initial infarct size Diabetes mellitus 90 (28%) 88 (24%) 0.184 to extent of left ventricular remodeling in the year after Systemic hypertension* 174 (55%) 229 (62%) 0.057 acute myocardial infarction. J Am Coll Cardiol 1995; Hyperlipidemia† 188 (59%) 200 (54%) 0.209 25:567–573. Smoking history 211 (67%) 236 (64%) 0.440 11. Christian TF, Gibbons RJ, Gersh BJ. Effect of infarct Inferior defect location 183 (73%) 145 (83%) 0.030 location in myocardial salvage assessed by technetium99m-isonitrile. J Am Coll Cardiol 1991;17:1303–1308. *Systemic hypertension defined by history or by a blood pressure at the outpatient clinic of .140/90 12. O’Connor MK, Hammell TC, Gibbons RJ. In vitro mm Hg. validation of a simple tomographic technique for esti† Hyperlipidemia defined by history of either a cholesterol or triglycerides measurement at .90th mation of percent myocardium “at risk” following administration of Tc-99m isonitrile. Eur J Nucl Med 1990; percentile. 17:69 –76. Data are expressed as number (%) except for age (median, in years), and weight (median, in kg). 13. Christian TF, Behrenbeck T, Pellikka PA, Huber KC, Chesebro JH, Gibbons RJ. Mismatch of left ventricular function and infarct size demonstrated by technetium-99m isonitrile imaging after reperfusion therapy which can occur with all cardiac isotopes and produce for acute myocardial infarction: identification of myocardial stunning and hyperkinesia. J Am Coll Cardiol 1990;16:1632–1638. a resting defect in the absence of myocardial infarc- 14. Behrenbeck T, Pellikka PA, Huber KC, Breshnahan JF, Gersh BJ, Gibbons tion. However, a previous analysis29 by our laboratory RJ. Primary PTCA in myocardial infarction: assessment of myocardial salvage examining 100 consecutive patients (mean body with Tc-99m sestamibi. J Am Coll Cardiol 1991;17:365–372. 15. Udelson JE, Coleman PS, Metherall J, Pandian NG, Gomez AR, Griffith JL, weight 89 6 19 kg) with normal resting electrocar- Shea NL, Oates E, Konstam MA. Predicting recovery of severe regional dysdiograms and no clinical history of myocardial infarc- function: comparison of resting scintigraphy with 201-Tl and 99mTc-sestamibi. 1994;89:2552–2561. tion resulted in 8 patients with trivial defects measur- Circulation 16. Medrano R, Lowry RW, Young JB, Weilbaecher DG, Michael LH, Afridi I, ing between 1% and 3% of the left ventricle and 11 He ZX, Mahmarian JJ, Verani MS. Assessment of myocardial viability with patients with inferior defects measuring between 4% 99mTc sestamibi in patients undergoing cardiac transplantation: a scintigraphic/ study. Circulation 1996;94:1010 –1017. and 21% of the left ventricle. Seven of the latter 11 pathologic 17. O’Connor MK. The Mayo Clinic Manual of Nuclear Medicine. New York: patients underwent further cardiac evaluation and Churchill Livingstone, 1996:197–203. demonstrated either a significant stenosis in the artery 18. Germano G, Kiat H, Kavanagh PB, Moriel M, Mazzanti M, Su HT, VanTrain Berman DS. Automatic quantification of ejection fraction from gated myosupplying the defect and/or a wall motion abnormality KF, cardial perfusion SPECT. J Nucl Med 1995;36:2138 –2147. on echocardiography in the same territory. Thus, tis- 19. DePuey EG, Rozanski A. Using gated technetium-99m-sestamibi SPECT to charfixed myocardial defects as infarct or artifact. J Nucl Med 1995;36:952–955. sue attenuation appeared to play a minor role in this acterize 20. Port SC. The first pass ventricular function study: why now? J Nucl Med series of patients, many of whom were obese.29 1994;35:1302–1303. Although the assessment of global LVEF is impor- 21. Maurea S, Cuolo A, Soricelli A, Castelli L, Nappi A, Squame F, Imbriaco M, B, Salvatore M. Enhanced detection of viable myocardium by technetant for prognosis, other information facilitated by Trimarco tium-99m-MIBI imaging after nitrate administration in chronic coronary artery gated imaging including regional wall motion analysis disease. J Nucl Med 1995;36:1945–1952. to assess viability and attenuation artifacts may also be 22. Sciagra` R, Bisi G, Santoro GM, Zerauschek F, Sestini S, Pedenovi P, R, Fazzini PF. Comparison of baseline-nitrate technetium-99msestavital.19,20 This study does not compromise these goals Pappagallo mibi with rest-redistribution thallium-201 tomography in detecting viable hiberbut attempts to provide the clinician with more infor- nating myocardium and predicting postrevascularization recovery. J Am Coll Cardiol 1997;30:384 –391. mation to allow selective testing of patients, and thus, 23. Bisi G, Sciagra` R, Giovanni GM, Fazzini PF. Rest technetium-99m sestamibi potentially to reduce the cost, time, and effort to both tomography in combination with short-term administration of nitrates: feasibility and reliability for prediction of postrevascularization outcome of asynergic terpatients and laboratories. ritories. J Am Coll Cardiol 1994;24:1282–1289. 24. Dilsizian V, Arrighi JA, Diodati JG, Quyyumi AA, Alavi K, Bacharach SL, Marin-Neto JA, Katsiyiannis PT, Bonow RO. Myocardial viability in patients 1. White HD, Norris RM, Brown MA, Brandt PWT, Whitlock RML, Wild CJ. with chronic coronary artery disease: comparison of 99mTc-sestamibi with thalLeft ventricular end-systolic volume as a major determinant of survival after lium reinjection and [18f]fluorodeoxyglucose. Circulation 1994;89:578 –587. 25. Marzullo P, Parodi O, Reisenhofer B, Sambuceti G, Picano E, Distante A, recovery from myocardial infarction. Circulation 1987;76:44 –51. 2. The Multicenter Postinfarct Research Group. Risk stratification and survival Gimelli A, L’Abbante A. Value of rest thallium-201/technetium-99m sestamibi after myocardial infarction. N Engl J Med 1983;309:331–336. scans and dobutamine echocardiography for detecting myocardial viability. Am J 3. Chua T, Kiat H, Germano G, Maurer G, VanTrain K, Friedman J, Berman D. Cardiol 1993;71:166 –172. 26. Altehoefer C, vom Dahl J, Biedermann M, Uebis R, Beilin I, Sheehan F, Gated technetium-99m sestamibi for simultaneous assessment of stress myocarHanrath P, Buell U. Significance of defect severity in technetium-99m-MIBI dial perfusion, post exercise regional ventricular function and myocardial viabilSPECT at rest to assess myocardial viability: comparison with fluorine-18-FDG ity. Correlation with echocardiography and rest thallium-201 scintigraphy. J Am PET. J Nucl Med 1994:35;569 –574. Coll Cardiol 1994;23:1107–1114. 4. Nichols K, DePuey EG, Gooneratne N, Salensky H, Friedman M, Cochoff S. 27. Christian TF, Miller TD, Chareonthaitawee P, Hodge DO, O’Connor MK, First pass ventricular ejection fraction using a single-crystal nuclear camera. Gibbons RJ. The prevalence of normal resting left ventricular function in patients J Nucl Med 1994;35:1292–1300. with normal rest electrocardiograms. Am J Cardiol 1997;79:1295–1298. 5. DePuey EG, Nichols K, Dobrinsky C. Left ventricular ejection fraction assessed 28. Christian TF, O’Connor MK, Hopfenspirger MR, Gibbons RJ. Comparison of from gated technetium-99m sestamibi SPECT. J Nucl Med 1993;34:1871–1876. reinjection thallium 201 and resting technetium 99m sestamibi tomographic 6. O’Keefe JH, Zinsmeister AR, Gibbons RJ. Value of electrocardiographic images for the quantification of infarct size after acute myocardial infarction. findings in predicting rest left ventricular function in patients with chest pain and J Nucl Card 1994;1:17–28. 29. Miller TD, Christian TF, Hopfenspirger MR, Hodge DO, Gersh BJ, Gibbons suspected coronary artery disease. Am J Med 1989;86:658 – 662. 7. Connolly DC, Elveback LR, Oxman HA. Coronary heart disease in residents RJ. Infarct size after acute myocardial infarction measured by quantitative tomoof Rochester, Minnesota, IV. Prognostic value of the resting electrocardiogram at graphic 99mTc sestamibi imaging predicts subsequent mortality. Circulation 1995; 92:334 –341. the time of initial diagnosis of angina pectoris. Mayo Clin Proc 1984;59:247–250.
TABLE III Comparison of Clinical Characteristics of Patients With Resting Defects With or Without History of Prior Myocardial Infarction
CORONARY ARTERY DISEASE/EJECTION FRACTION AND INFARCT SIZE
1285
MD,
This study determined the correlation between the extent of the resting perfusion defect by technetium99m sestamibi tomographic imaging and the firstpass left ventricular (LV) ejection fraction (EF). A total of 1,955 patients underwent technetium-99m sestamibi tomographic imaging with measurement of firstpass resting LVEF. Twenty-five percent of patients had a prior history of myocardial infarction. First-pass LVEF was measured using a peripheral intravenous injection and a multicrystal gamma camera with standard software. Resting tomographic perfusion defect size (infarct size) was quantitated using previously published methods. Mean LVEF for the study group was 0.60 6 0.11. Mean LV infarct size was 5 6 11%. For the 1,265 patients (65% of the study group) with no measurable perfusion defect, the preva-
lence of a normal (>0.50) LVEF was 96% (1,212 of 1,265 patients). For patients with a measurable defect (n 5 690, 35%), the inverse linear correlation with LVEF was highly significant (r 5 20.60, p <0.0001) but with wide confidence limits (SEE 5 10 LVEF points), thereby limiting the predictive value in individual patients. Thus, in the absence of known cardiomyopathy, valvular heart disease, or left bundle branch block, patients without a quantifiable resting perfusion defect are highly likely to have a normal resting LVEF and may not require determination of LV function. For patients with resting perfusion defects, LVEF cannot be predicted with confidence and should therefore be measured. Q1998 by Excerpta Medica, Inc. (Am J Cardiol 1998;81:1281–1285)
eft ventricular (LV) ejection fraction (EF) is a major determinant of prognosis in patients with L coronary artery disease. Assessment of LVEF is
METHODS
1,2
therefore an important component in the evaluation of the cardiac patient. Measurement of LVEF is now possible in combination with perfusion owing to the higher count statistics obtained with technetium (Tc)99m– based perfusion agents. These perfusion tracers have previously been demonstrated to provide a reliable and accurate means of assessing LVEF.3–5 However, depending on how this information is obtained, there may be added processing time and cost. Previous studies have demonstrated that in certain patient subsets, LVEF may be reliably predicted from clinical and electrocardiographic variables.6 – 8 In addition, prior reports from this institution have shown a correlation between infarct size by Tc-99m sestamibi tomographic imaging and LVEF by first-pass radionuclide ventriculography and by electron beam computed tomography in the first year after an acute myocardial infarction.9,10 This study determines the correlation between the extent of the resting tomographic perfusion defect and the first-pass LVEF, and assesses the ability of the resting perfusion defect to predict LVEF.
From the Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Manuscript received September 16, 1997; revised manuscript received and accepted February 5, 1998. Address for reprints: Timothy F. Christian, MD, Mayo Clinic, 200 First Street SW, East 16-B, Rochester, Minnesota 55905. ©1998 by Excerpta Medica, Inc. All rights reserved.
Study group: The study group consisted of a series of patients referred to the nuclear cardiology laboratory at the Mayo Clinic for Tc-99m sestamibi perfusion imaging between March 1994 and April 1996. Before the stress portion of each study, the responsible monitoring personnel (physician or nurse) completed a database questionnaire, which was subsequently entered into a computer by a research nurse. The study group was identified from this computer database and included all patients with (1) a 2-day Tc-99m sestamibi rest-stress protocol, and (2) technically adequate Tc-99m sestamibi tomographic perfusion imaging and first-pass resting LVEF. Patients were excluded if they had known cardiomyopathy (56 patients), valvular heart disease (123 patients), left bundle branch block (99 patients), a permanent pacemaker (40 patients), or an arrhythmia. A total of 1,955 patients were included in the study group. Tc-99m sestamibi infarct (defect) size: Tc-99m sestamibi perfusion studies were performed as a 2-day (rest-stress) protocol at the Mayo Clinic Nuclear Cardiology Laboratory. The resting image was recorded on day 1. Approximately 30 to 45 minutes after injection of 20 to 30 mCi of Tc-99m sestamibi, tomographic imaging was performed with 30 images over a 180° arc, 45° right anterior oblique to 45° left posterior oblique postion using a step and shoot mode.11 Perfusion defect size was quantitated from the sum of 5 short-axis slices and determined from the number of pixels with ,60% of maximal counts over the 5 slices (Figure 1).11 Infarct size was defined as the extent of the perfusion defect and expressed as a percentage of the LV wall.11,12 This method of quantifying infarct 0002-9149/98/$19.00 PII S0002-9149(98)00156-8
1281
and dynamic acquisition was performed with a 32 3 32 matrix and a frame rate of 30 to 40 frames/s. LVEF was calculated from $4 cardiac cycles using a count-based method from generated time activity curves and commercially available software (Scinticor).17 Studies were considered technically suboptimal if there was evidence of arrhythmia during the acquisition, a poor bolus, poor separation of right and left ventricular phases, or poor counting statistics. Each study was reviewed by at least 1 trained observer for technical quality. A first-pass LVEF $0.50 is considered normal using this approach in our laboratory. FIGURE 1. Tc-99m sestamibi tomographic imaging of an inferior defect. Short-axis count profile starts at the anterior wall and progresses clockwise at 6& intervals. Perfusion defect size was quantitated from the sum of the 5 short-axis slices and determined from the number of pixels with <60% of maximal counts.
TABLE I Clinical Characteristics of Study Group (n 5 1,955) No. (%) Women/men Prior myocardial infarction Diabetes mellitus Insulin-requiring Systemic hypertension* Hyperlipidemia† Smoking history Normal resting electrocardiogram Prior coronary artery bypass grafting Medication use b blocker Calcium antagonist Nitrates
675 (35%)/1,280 (65%) 496 (25%) 451 (23%) 188 (10%) 1,132 (58%) 1,169 (60%) 1,126 (58%) 700 (36%) 372 (19%) 634 (32%) 708 (36%) 407 (21%)
*Systemic hypertension defined by history or by a blood pressure at the outpatient clinic of .140/90 mm Hg. † Hyperlipidemia defined by history or by either a cholesterol or triglycerides measurement at .90th percentile.
size has been validated by both animal and clinical trials,9,13,14 and the 60% threshold has previously been shown to provide the optimal separation between viable and nonviable myocardium.15,16 First-pass radionuclide angiography: First-pass LVEF was measured after the sestamibi injection for the same resting study. A large bore (18-gauge, 1.3mm) intravenous catheter was placed in a medial antecubital vein. The patient was positioned in the anterior view of a multicrystal gamma camera (Scinticor SIM-400, Milwaukee, Wisconsin) with a medium-sensitivity collimator and a 50% energy window centered over the 140-keV photo peak. A 0.5-mCi test dose of Tc-99m sestamibi was administered to define the left ventricular region of interest. Tc-99m sestamibi (20 to 30 mCi) was then injected as a bolus dose, 1282 THE AMERICAN JOURNAL OF CARDIOLOGYT
VOL. 81
Clinical and historical variables: Prior to stress imaging,
clinical and historical variables were prospectively recorded for all patients.
Other estimations of left ventricular ejection fraction:
A systematic search of the database was performed retrospectively to identify LVEF measurements by other modalities (contrast left ventriculogram by biplane technique and echocardiogram by Simpson’s rule) within 3 months of the first-pass LVEF of interest in the study population. Statistical analysis: Data are presented as mean 6 SD. Simple linear regression analysis was used to correlate infarct size by Tc-99m sestamibi tomographic imaging with the first-pass LVEF. Chi-square and 2-tailed probability tests were performed to compare the clinical characteristics of patients with measurable resting perfusion defects with and without a history of prior infarction. Spearman rank correlation coefficients were calculated to quantify the strength of the correlation between first-pass LVEF and LVEF measurements by contrast left ventriculography and echocardiography.
RESULTS Pertinent clinical characteristics of the study group are listed in Table I. The mean age of the study group was 63 6 11 years. Most patients were men (65%). Twenty-five percent had prior myocardial infarction. Thirty-six percent had a normal resting electrocardiogram. Table II summarizes the comparison between the clinical characteristics of patients with a resting perfusion defect (n 5 690) versus those of patients without a resting perfusion defect (n 5 1,265). Patients with a measurable perfusion defect were older and were more likely to be men, to be diabetic, to take b blockers, to have a history of infarction, and to have a history of cigarette smoking. Patients without a meaJUNE 1, 1998
In patients with no measurable perfusion defect by Tc-99m sestamibi tomographic imaging (n 5 With Defect Without Defect p 1,265, 65%), the prevalence of a nor(n 5 690) (n 5 1,265) Value mal resting LVEF was 96% (Figure Age (median, yr) 66 64 0.0003 3). Conversely, in patients with a Women/men 99 (14%)/591 (86%) 576 (45%)/689 (54%) 0.001 measurable perfusion defect, the Prior myocardial infarction 319 (46%) 177 (14%) 0.001 prevalence of a normal LVEF was Diabetes mellitus 179 (26%) 272 (22%) 0.035 only 65%. Insulin-requiring 81 (12%) 107 (8%) 0.013 Systemic hypertension* 404 (59%) 728 (58%) 0.605 For the 690 patients with a meaHyperlipidemia† 386 (57%) 780 (62%) 0.026 surable defect, the inverse linear corSmoking history 448 (65%) 678 (54%) 0.001 relation between defect extent and Normal resting 133 (19%) 567 (45%) 0.001 first-pass LVEF (Figure 4) was electrocardiogram highly significant (r 5 20.60, p Medication use b blocker 253 (37%) 381 (30%) 0.003 ,0.0001); however, the standard erCalcium antagonist 259 (37%) 449 (35%) 0.369 ror was 10 LVEF points, thereby Nitrates 213 (31%) 194 (15%) 0.001 limiting the predictive value in indi*Systemic hypertension defined by history or by a blood pressure at the outpatient clinic of .140/90 vidual patients. mm Hg. A larger proportion of patients † Hyperlipidemia defined by history or by either a cholesterol or triglycerides measurement at .90th (35%) had a measurable perfusion percentile. defect than those who had a history Data are expressed as number (%) except for age (median, in years). of prior myocardial infarction (25%). We therefore compared the clinical characteristics of patients with a measurable perfusion defect but no history of myocardial infarction (n 5 369) to patients with a measurable defect and a history of myocardial infarction (n 5 319). The results are summarized in Table III. There were no significant differences in the 2 groups with respect to gender or diabetes. Patients with a measurable resting defect and no history of myocardial infarction were significantly older, tended to be hypertensive, and were more likely to have an inferior defect. There were 144 and 431 patients with LVEF measurements by contrast left ventriculography and echocardiography, respectively, within 3 months of the firstpass study. The correlation between firstFIGURE 2. Histogram showing the distribution of infarct size determined by pass and contrast left ventriculographic Tc-99m sestamibi for the study group of 1,955 patients. The distribution was highly skewed toward no resting defect (65% of study group) and smaller LVEF (r 5 0.61, SEE 5 0.07) was similar (<10% of the left ventricle) defects (17.5% of study group). to the correlation between first-pass and echocardiographic LVEF (r 5 0.66, SEE 5 0.04). TABLE II Comparison of Clinical Characteristics of Patients With or Without a Resting Defect
surable perfusion defect were more likely to have a normal resting electrocardiogram. Tc-99m sestamibi perfusion imaging and first-pass ejection fraction: The distribution of infarct size deter-
mined by Tc-99m sestamibi tomographic imaging for the study group is shown in Figure 2. The mean infarct size for the group of 1,955 patients was 5 6 11% of the left ventricle (median 5 0), with a range of 0% (n 5 1,265) to 76% of the left ventricle. The distribution of infarct size was highly skewed toward no resting defects (65% of study group) and smaller (#10% of the left ventricle) defects (17.5% of study group). The mean Tc-99m sestamibi first-pass LVEF was 0.60 6 0.11 (median 5 0.62), and the range 0.15 to 0.89.
DISCUSSION LVEF assessment using the first-pass technique has been adopted as part of Tc-99m sestamibi imaging in many nuclear cardiology laboratories. This method of assessing LV function has previously been shown to be accurate and reproducible.3–5 First-pass LVEF can be acquired either with a multicrystal gamma camera or with electronically enhanced single-crystal detectors and ultra high sensitivity collimators.4 Commercial software to perform the gated analysis of LVEF from tomographic images is readily available, several of which are almost fully automatic.18 In addition to global LVEF, assessment of regional wall motion may also provide valuable information.19,20 However, for laboratories using ungated analysis, the
CORONARY ARTERY DISEASE/EJECTION FRACTION AND INFARCT SIZE
1283
third of the study population), this study shows a highly significant inverse linear correlation between defect size and LVEF, but with wide confidence limits. Therefore, in a patient with a resting perfusion defect, LVEF cannot be predicted with confidence from the extent of the defect and should be measured. A normal resting electrocardiogram and certain clinical variables may be used to predict normal LV function.6 – 8,27 However, in this study and a previous report27 from this laboratory, the prevalence of a normal resting electrocardiogram in patients referred for Tc-99m sestamibi perfusion imaging was 36% and 33%, respecFIGURE 3. Bar graph showing the percentage of patients with normal LV functively. As a result, in one third of patients tion (defined as LVEF >50%) in the presence or absence of a measurable perfusion defect. For the 1,265 patients (65%) with no measurable defect, the referred for perfusion imaging, a normal prevalence of normal resting LV function was 96%. In patients with a measurLVEF cannot be predicted using this apable defect (n 5 690; 35%), the prevalence of normal resting LV function was proach. In addition, the strongest clinical only 65%. predictor of a normal LVEF (no history of prior myocardial infarction) was associated with only an 87% prevalence of normal LV function.27 Thus, data from this study allow further identification of patients at low risk for LV dysfunction. When the first-pass LVEF study is inadequate on day 1 of a 2-day Tc-99m sestamibi protocol, clinicians may use the absence of a measurable tomographic perfusion defect to forego repeat gated analysis for LV function on the following day. In patients undergoing ungated studies, the absence of a measurable perfusion defect also allows the clinician to predict a normal LVEF without the additional time or cost of the measurement (the first-pass study to obtain LVEF adds 2.52 Medicare relative value units). In patients FIGURE 4. Scatterplot showing the inverse correlation of resting LV function undergoing same-day rest-stress imaging, and infarct size: r 5 20.60, p <0.0001, SEE 10 LVEF points. the absence of a measurable defect on the small but additional patient and laboratory acquisition resting image precludes the need for gating during stress and processing time for routine LVEF assesment by imaging. Christian et al28 previously demonstrated that thalthis first-pass technique may not be justified. Additionally, although recent reports suggest that Tc-99m lium-201 imaging provides significantly larger estisestamibi may be a marker of myocardial viability,15 mates of infarct size than Tc-99m sestamibi imaging. particularly if used in conjunction with nitrate infu- Hence, in the absence of a measurable defect by sion,21–23 other studies have shown that Tc-99m ses- thallium-201, clinicians can also infer the presence of tamibi does not provide as complete viability infor- normal LV function and avoid other diagnostic studies mation when compared with thallium-201.24 –26 Con- to assess LVEF. Last, the finding of an abnormal EF in sequently, a shift to Tc-99m– based perfusion agents the absence of a measurable tomographic perfusion to obtain LVEF measurements may result in decreased defect would be unexpected and the quality of the study should be carefully scrutinized. detection of myocardial viability. Limitations to this study include the exclusion of In the present study, performed at a tertiary referral center, nearly two thirds of patients had no measurable patients with a known cardiomyopathy, valvular disresting perfusion defect. The prevalence of a normal ease, left bundle branch block, arrhythmias, and carLVEF in this group was 96%. Therefore, clinicians diac pacemaker. In these cases, LV function may be can be .95% confident (confidence interval 93% to influenced by factors unrelated to myocardial perfu96%) that a patient without a measurable perfusion sion. A second limitation is the less likely use of defect (two thirds of the referral population) will have Tc-99m sestamibi when viability is a key issue. Thalnormal global resting LV function, and measurement lium-201 is used preferentially in these instances. A third limitation is the potential attenuation of counts of LVEF is most likely not justified. In patients with a measurable perfusion defect (one from overlying soft tissue (diaphragm and breast), 1284 THE AMERICAN JOURNAL OF CARDIOLOGYT
VOL. 81
JUNE 1, 1998
8. Rihal CS, Davis KB, Kennedy JW, Gersh BJ. The utility of clinical, electrocardiographic, and roentgenographic variables in the prediction of left ventricular function. Am J Cardiol 1995;75:220 –223. With History Without History p 9. Christian TF, Behrenbeck T, Gersh BJ, Gibbons RJ. Relation of left ventricular volume and function over (n 5 319) (n 5 369) Value one year after acute myocardial infarction to infarct size Age (median, yr) 67 64 0.009 determined by technetium-99m sestamibi. Am J Cardiol 1991;68:21–26. Weight (median, kg) 94 89 0.01 10. Chareonthaitawee P, Christian TF, Hirose K, GibWomen/men 44 (14%)/275 (86%) 54 (15%)/315 (85%) 0.753 bons RJ, Rumberger JA. Relation of initial infarct size Diabetes mellitus 90 (28%) 88 (24%) 0.184 to extent of left ventricular remodeling in the year after Systemic hypertension* 174 (55%) 229 (62%) 0.057 acute myocardial infarction. J Am Coll Cardiol 1995; Hyperlipidemia† 188 (59%) 200 (54%) 0.209 25:567–573. Smoking history 211 (67%) 236 (64%) 0.440 11. Christian TF, Gibbons RJ, Gersh BJ. Effect of infarct Inferior defect location 183 (73%) 145 (83%) 0.030 location in myocardial salvage assessed by technetium99m-isonitrile. J Am Coll Cardiol 1991;17:1303–1308. *Systemic hypertension defined by history or by a blood pressure at the outpatient clinic of .140/90 12. O’Connor MK, Hammell TC, Gibbons RJ. In vitro mm Hg. validation of a simple tomographic technique for esti† Hyperlipidemia defined by history of either a cholesterol or triglycerides measurement at .90th mation of percent myocardium “at risk” following administration of Tc-99m isonitrile. Eur J Nucl Med 1990; percentile. 17:69 –76. Data are expressed as number (%) except for age (median, in years), and weight (median, in kg). 13. Christian TF, Behrenbeck T, Pellikka PA, Huber KC, Chesebro JH, Gibbons RJ. Mismatch of left ventricular function and infarct size demonstrated by technetium-99m isonitrile imaging after reperfusion therapy which can occur with all cardiac isotopes and produce for acute myocardial infarction: identification of myocardial stunning and hyperkinesia. J Am Coll Cardiol 1990;16:1632–1638. a resting defect in the absence of myocardial infarc- 14. Behrenbeck T, Pellikka PA, Huber KC, Breshnahan JF, Gersh BJ, Gibbons tion. However, a previous analysis29 by our laboratory RJ. Primary PTCA in myocardial infarction: assessment of myocardial salvage examining 100 consecutive patients (mean body with Tc-99m sestamibi. J Am Coll Cardiol 1991;17:365–372. 15. Udelson JE, Coleman PS, Metherall J, Pandian NG, Gomez AR, Griffith JL, weight 89 6 19 kg) with normal resting electrocar- Shea NL, Oates E, Konstam MA. Predicting recovery of severe regional dysdiograms and no clinical history of myocardial infarc- function: comparison of resting scintigraphy with 201-Tl and 99mTc-sestamibi. 1994;89:2552–2561. tion resulted in 8 patients with trivial defects measur- Circulation 16. Medrano R, Lowry RW, Young JB, Weilbaecher DG, Michael LH, Afridi I, ing between 1% and 3% of the left ventricle and 11 He ZX, Mahmarian JJ, Verani MS. Assessment of myocardial viability with patients with inferior defects measuring between 4% 99mTc sestamibi in patients undergoing cardiac transplantation: a scintigraphic/ study. Circulation 1996;94:1010 –1017. and 21% of the left ventricle. Seven of the latter 11 pathologic 17. O’Connor MK. The Mayo Clinic Manual of Nuclear Medicine. New York: patients underwent further cardiac evaluation and Churchill Livingstone, 1996:197–203. demonstrated either a significant stenosis in the artery 18. Germano G, Kiat H, Kavanagh PB, Moriel M, Mazzanti M, Su HT, VanTrain Berman DS. Automatic quantification of ejection fraction from gated myosupplying the defect and/or a wall motion abnormality KF, cardial perfusion SPECT. J Nucl Med 1995;36:2138 –2147. on echocardiography in the same territory. Thus, tis- 19. DePuey EG, Rozanski A. Using gated technetium-99m-sestamibi SPECT to charfixed myocardial defects as infarct or artifact. J Nucl Med 1995;36:952–955. sue attenuation appeared to play a minor role in this acterize 20. Port SC. The first pass ventricular function study: why now? J Nucl Med series of patients, many of whom were obese.29 1994;35:1302–1303. Although the assessment of global LVEF is impor- 21. Maurea S, Cuolo A, Soricelli A, Castelli L, Nappi A, Squame F, Imbriaco M, B, Salvatore M. Enhanced detection of viable myocardium by technetant for prognosis, other information facilitated by Trimarco tium-99m-MIBI imaging after nitrate administration in chronic coronary artery gated imaging including regional wall motion analysis disease. J Nucl Med 1995;36:1945–1952. to assess viability and attenuation artifacts may also be 22. Sciagra` R, Bisi G, Santoro GM, Zerauschek F, Sestini S, Pedenovi P, R, Fazzini PF. Comparison of baseline-nitrate technetium-99msestavital.19,20 This study does not compromise these goals Pappagallo mibi with rest-redistribution thallium-201 tomography in detecting viable hiberbut attempts to provide the clinician with more infor- nating myocardium and predicting postrevascularization recovery. J Am Coll Cardiol 1997;30:384 –391. mation to allow selective testing of patients, and thus, 23. Bisi G, Sciagra` R, Giovanni GM, Fazzini PF. Rest technetium-99m sestamibi potentially to reduce the cost, time, and effort to both tomography in combination with short-term administration of nitrates: feasibility and reliability for prediction of postrevascularization outcome of asynergic terpatients and laboratories. ritories. J Am Coll Cardiol 1994;24:1282–1289. 24. Dilsizian V, Arrighi JA, Diodati JG, Quyyumi AA, Alavi K, Bacharach SL, Marin-Neto JA, Katsiyiannis PT, Bonow RO. Myocardial viability in patients 1. White HD, Norris RM, Brown MA, Brandt PWT, Whitlock RML, Wild CJ. with chronic coronary artery disease: comparison of 99mTc-sestamibi with thalLeft ventricular end-systolic volume as a major determinant of survival after lium reinjection and [18f]fluorodeoxyglucose. Circulation 1994;89:578 –587. 25. Marzullo P, Parodi O, Reisenhofer B, Sambuceti G, Picano E, Distante A, recovery from myocardial infarction. Circulation 1987;76:44 –51. 2. The Multicenter Postinfarct Research Group. Risk stratification and survival Gimelli A, L’Abbante A. Value of rest thallium-201/technetium-99m sestamibi after myocardial infarction. N Engl J Med 1983;309:331–336. scans and dobutamine echocardiography for detecting myocardial viability. Am J 3. Chua T, Kiat H, Germano G, Maurer G, VanTrain K, Friedman J, Berman D. Cardiol 1993;71:166 –172. 26. Altehoefer C, vom Dahl J, Biedermann M, Uebis R, Beilin I, Sheehan F, Gated technetium-99m sestamibi for simultaneous assessment of stress myocarHanrath P, Buell U. Significance of defect severity in technetium-99m-MIBI dial perfusion, post exercise regional ventricular function and myocardial viabilSPECT at rest to assess myocardial viability: comparison with fluorine-18-FDG ity. Correlation with echocardiography and rest thallium-201 scintigraphy. J Am PET. J Nucl Med 1994:35;569 –574. Coll Cardiol 1994;23:1107–1114. 4. Nichols K, DePuey EG, Gooneratne N, Salensky H, Friedman M, Cochoff S. 27. Christian TF, Miller TD, Chareonthaitawee P, Hodge DO, O’Connor MK, First pass ventricular ejection fraction using a single-crystal nuclear camera. Gibbons RJ. The prevalence of normal resting left ventricular function in patients J Nucl Med 1994;35:1292–1300. with normal rest electrocardiograms. Am J Cardiol 1997;79:1295–1298. 5. DePuey EG, Nichols K, Dobrinsky C. Left ventricular ejection fraction assessed 28. Christian TF, O’Connor MK, Hopfenspirger MR, Gibbons RJ. Comparison of from gated technetium-99m sestamibi SPECT. J Nucl Med 1993;34:1871–1876. reinjection thallium 201 and resting technetium 99m sestamibi tomographic 6. O’Keefe JH, Zinsmeister AR, Gibbons RJ. Value of electrocardiographic images for the quantification of infarct size after acute myocardial infarction. findings in predicting rest left ventricular function in patients with chest pain and J Nucl Card 1994;1:17–28. 29. Miller TD, Christian TF, Hopfenspirger MR, Hodge DO, Gersh BJ, Gibbons suspected coronary artery disease. Am J Med 1989;86:658 – 662. 7. Connolly DC, Elveback LR, Oxman HA. Coronary heart disease in residents RJ. Infarct size after acute myocardial infarction measured by quantitative tomoof Rochester, Minnesota, IV. Prognostic value of the resting electrocardiogram at graphic 99mTc sestamibi imaging predicts subsequent mortality. Circulation 1995; 92:334 –341. the time of initial diagnosis of angina pectoris. Mayo Clin Proc 1984;59:247–250.
TABLE III Comparison of Clinical Characteristics of Patients With Resting Defects With or Without History of Prior Myocardial Infarction
CORONARY ARTERY DISEASE/EJECTION FRACTION AND INFARCT SIZE
1285