(Radiology. 2000;215:529-533.)
© RSNA, 2000
Poststress Measurements of Left Ventricular Function with Gated Perfusion SPECT: Comparison with Resting Measurements by Using a Same-Day Perfusion-Function Protocol1
Salvador Borges-Neto, MD,
Aamir Javaid, MD,
Linda K. Shaw, BS,
David F. Kong, MD,
Michael W. Hanson, MD,
Robert A. Pagnanelli, CNMT,
Gregory Ravizzini, MD and
R. E. Coleman, MD
1 From the Departments of Radiology—Nuclear Medicine (S.B.N., M.W.H., R.A.P., G.R., R.E.C.) and Medicine—Cardiology (S.B.N., A.J., L.K.S., D.F.K., M.W.H.), Duke University Medical Center and Health Systems, Duke North Rm 1410, PO Box 3949, Durham, NC 27710. Received May 3, 1999; revision requested June 16; final revision received October 18; accepted October 20. Address correspondence to S.B.N. (e-mail: borge001@mc.duke.edu.)
 |
Abstract
|
|---|
PURPOSE: To investigate the relationship between the development of ischemia during stress testing and the changes in left ventricular ejection fraction (LVEF) measurements obtained after stress and at rest with a same-day perfusion-function imaging protocol.
MATERIALS AND METHODS: One hundred twenty-six patients underwent a same-day rest-stress (61%) or stress-rest (39%) protocol and gated single photon emission computed tomography (SPECT). Perfusion analysis was performed with a 12-segment model. Defects were scored (0 = no defect, 1 = mild defect, 2 = moderate defect, and 3 = severe defect); differences between the summed stress and resting scores of greater than three indicated substantial ischemia.
RESULTS: Resting and poststress LVEFs correlated significantly (r = 0.97, P < .001); however, patients with and patients without ischemia had significant differences in poststress versus resting LVEFs (-4.0 vs 1.0, respectively; P < .01). In patients with ischemia versus patients without ischemia, subgroup analysis stress-rest (-2.5 vs 1.0, P = .047) and rest-stress (-4.0 vs 1.0, P = .006) protocols yielded similar results.
CONCLUSION: In patients with clinically important stress-induced perfusion abnormalities, the LVEF after stress was significantly lower than the LVEF at rest with same-day rest-stress and stress-rest imaging protocols. In the clinical setting, poststress LVEFs may be lower than true resting measurements, particularly in patients with moderate to severe stress-induced ischemia.
Index terms: Heart, ejection fraction • Heart, function • Myocardium, ischemia, 511.1939 • Myocardium, radionuclide studies, 511.1939, 511.12171 • Myocardium, SPECT, 511.1939, 511.12162 • Receiver operating characteristic (ROC) curve • Single photon emission computed tomography (SPECT), 511.12162
|
Introduction
|
|---|
The combined assessment of myocardial perfusion and left ventricular function has been a major advance in the field of nuclear cardiology. Measurements of left ventricular function and myocardial perfusion imaging can be performed with a single injection of a technetium 99m–labeled radiopharmaceutical at rest and at peak stress by using first-pass radionuclide angiography followed by single photon emission computed tomography (SPECT) perfusion imaging (1–3). This technique has been validated, and the results demonstrate the potential effect on diagnostic and prognostic assessment of patients who are suspected of having or known to have coronary artery disease (4–7).
Gated images of the perfused myocardium also enable analysis of wall motion, wall thickening, and ejection fraction measurements with a combined assessment of myocardial perfusion and ventricular function during a single injection of the radiopharmaceutical (8–10). The advantage of this technique is that it relies on a single acquisition for both perfusion and functional imaging. Cardiac function, however, is assessed at rest or during a resting state after stress. Because of the higher amount of radioactivity administered at peak stress with a single-day imaging protocol by using a rest-stress 99mTc radiopharmaceutical study or a dual-isotope resting thallium-stress 99mTc perfusion tracer protocol, poststress tomograms are most commonly gated, with the assumption that they reflect resting function.
In a recent study by Johnson et al (11), who used a 2-day protocol for rest and stress, abnormal left ventricular function was documented on poststress gated tomograms. This finding probably reflects postischemic stunning in patients who develop severe perfusion abnormalities during stress and suggests that poststress scanning does not always represent basal left ventricular function in patients with severe stress-induced ischemia.
The 2-day rest-stress protocol is not convenient for patients; therefore, most laboratories perform the low-dose and high-dose same-day protocol (1). At present, data are not available concerning the effect of ischemia on left ventricular ejection fractions (LVEFs) obtained with gated SPECT and a same-day perfusion-function protocol. The purpose of this investigation was to assess the relationship between the development of ischemia during a stress test and the changes between LVEF measurements obtained after stress and at true resting with gated SPECT of the perfused myocardium by using a same-day rest-stress or stress-rest perfusion function protocol.
|
MATERIALS AND METHODS
|
|---|
Study Population
Between June 1997 and April 1998, 126 consecutive patients successfully underwent a same-day rest-stress or stress-rest sequence for the assessment of both myocardial perfusion and left ventricular function with a gated SPECT perfusion imaging technique. Patients were referred by their clinician to the nuclear cardiology laboratory, and all patients underwent the imaging procedure as part of their clinical evaluation. The imaging study was performed after our routine clinical studies protocol. Before the studies were performed, a signed clinical informed consent form was obtained from all patients in accordance with standard clinical practice. Results of the studies were reviewed according to our institutional confidentiality policy. All patients had a wide range of LVEFs (11%–88%) and were suspected of having or known to have coronary artery disease.
Radionuclide Study Protocol
A same-day rest-stress or stress-rest perfusion-function study was performed by using 10 mCi (370 MBq) of 99mTc perfusion tracer (99mTc sestamibi; Du Pont Pharmaceutical, North Billerica, Mass; intravenous injection) for the initial portion of the examination and 30 mCi (1,110 MBq) for the second portion of the examination (1). For the resting studies, images were obtained 60 minutes after injection. For the stress portion of the examination, images were obtained 30 minutes after injection for patients undergoing an exercise stress test and 45 minutes after injection for those undergoing a pharmacologic stress test. The mean time from stress to imaging was 38 minutes. All resting and poststress perfusion SPECT studies were gated.
Most patients (74%) underwent an exercise stress test with a treadmill. Twenty-three patients underwent a pharmacologic stress test: 18% received dobutamine, 4% received dipyridamole, and 4% received adenosine. Most patients (61%) underwent a rest-stress protocol; 39% underwent a stress-rest protocol.
SPECT Imaging
SPECT images were obtained over a 180° circular orbit by using a dual-head detector gamma camera (Cardial; Elscint, Haifa, Israel) equipped with high-resolution collimators, with the detectors placed at right angles. Images were obtained with 60 projections at 30 seconds (low-dose protocol) and 20 seconds (high-dose protocol). The camera was set for a 140-keV photopeak with a 20% window. Images were obtained with a matrix size of 64 x 64 x 8. During acquisition, images were gated for eight frames per cardiac cycle with 100% beat acceptance. The projection data were prefiltered with a two-dimensional Butterworth filter (frequency of 5.0 with 0.35 cycles per pixel and a pixel size of 0.64 cm). Images were reconstructed with filtered back-projection and no attenuation correction.
After reconstruction of the transverse images, the short-axis images underwent preprocessing for gated LVEF quantification by using previously validated software (9). Briefly, the algorithm uses the gated short-axis images sets. It segments the left ventricle with estimation and further definition of the epicardial surface, endocardial surface, and valve plane for every gating interval. It then calculates the endocardial and epicardial volumes. In the final steps, it isolates the intervals corresponding to end diastole and end systole and derives the LVEF.
The data set was also processed for qualitative interpretation of the perfusion images by using a two-dimensional Butterworth filter or a frequency of 5.0 with 0.35 cycles per pixel and a pixel size of 6.4 cm, with further reconstruction and reorientation in the short, vertical, and horizontal long axes. Resting and poststress tomograms were semiqualitatively analyzed by using a 12-segment model. Defects were scored as follows: 0, no defect; 1, mild defect; 2, moderate defect; and 3, severe defect. The number of abnormal segments and the defect score were determined to calculate a summed resting score and a summed stress score. A difference of three or greater between the summed resting and stress scores was arbitrarily considered to indicate substantial ischemia. The perfusion images were interpreted independently by two experienced nuclear medicine physicians (S.B.N., M.W.H.) who were blinded to the ejection fraction results.
Reproducibility
Fifteen patients underwent two sequential resting acquisitions to document the reproducibility of the test. LVEFs ranged from 16% to 81% for the first measurement and from 14% to 78% for the second measurement. Differences between the two measurements expressed as arbitrary units ranged from 0 to 4 (mean, 2.7).
Statistics and Data Analysis
The baseline characteristics of the population study are given as percentages for categorical variables and as the mean for continuous variables. Regression analysis was performed to determine the Pearson product moment reflecting the degree of random error between the two measurements, and Bland-Altman plots were used to assess systematic errors among measurements as well as their degree of agreement (12,13). The Wilcoxon rank sum test was used to compare LVEFs obtained at rest and after stress in the overall population and to assess differences between patients with and patients without ischemia. A P value of less than .05 was considered to indicate a statistically significant difference for all variables compared.
The logistic regression model was used to assess the predictive value of the differences between LVEFs obtained at rest and after stress for the prediction of the likelihood of ischemia. The area under the receiver operating characteristic curve was examined to assess the ability of the model to enable differentiation of patients with ischemia from those without ischemia. The area under the curve can vary from 0.5 to 1.0, where 1.0 indicates a perfectly predictive model and 0.5 indicates a purely random discriminatory model (14).
|
RESULTS
|
|---|
Demographics and Clinical Descriptors
Of the 126 patients examined, 95 (75%) were men, and 31 (25%) were women (mean age, 59 years; age range, 51–69 years). One hundred six (84%) of the patients were receiving cardiac medications; forty-two (33%) had a history of myocardial infarction. Ninety-eight (78%) patients underwent cardiac catheterization a mean of 385 days after the nuclear imaging study. Twenty (20%) of the patients who underwent cardiac catheterization had one-vessel disease, 21 (21%) had two-vessel disease, and 42 (43%) had three-vessel disease. Percutaneous transluminal coronary angioplasty had been performed in 32 (25%) of the patients a mean of 255 days before the perfusion study; 44 (35%) had previously undergone coronary artery bypass grafting.
Overall LVEFs at Rest and after Stress
A high correlation was shown between resting and poststress LVEFs in all 126 patients (Fig 1). The Bland-Altman plot (Fig 2) demonstrated that in only a few patients was the difference between the resting and poststress measurements above or below 2 SDs from the overall mean difference between those obtained for the entire group. Furthermore, results of the Wilcoxon rank sum test did not reveal any significant differences between the two measurements when all patients were considered (P = .669).
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Plot shows the regression line and high correlation between rest and poststress LVEFs (EF) obtained at gated SPECT (GSPECT) of the perfused myocardium in all 126 patients. Data are percentages. SEE = standard error of the estimate.
|
|
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Bland-Altman plot of rest and poststress LVEFs (EF) shows an overall mean difference of three points (middle line) for all 126 patients. In only a small number of patients was the difference between the rest and poststress LVEF more than 2 SDs (top and bottom lines) from the overall mean difference.
|
|
Similar results had initially led some investigators (15) to conclude that, when no other patient population characteristics or perfusion scanning results are considered, the poststress ejection fraction could potentially be interpreted or reported as a true resting measurement.
LVEFs in Patients with and Patients without Ischemia
The Table demonstrates the LVEF measurements obtained at rest and after stress. A statistically significant difference was observed for the changes between resting and poststress measurements when patients with and patients without ischemia were considered. When patients were further analyzed according to the type of protocol performed (ie, stress-rest or rest-stress), significant differences were also observed. With the stress-rest protocol, the LVEF was -2.5 for patients with ischemia and 1.0 for patients without ischemia (P = .047). With the rest-stress protocol, the LVEF was -4.0 for patients with ischemia and 1.0 for patients without ischemia (P =.006).
Predictive Ability of Change in LVEF
The predictive value of the change in LVEF measurements for estimating the likelihood of ischemia with gated SPECT was assessed with a univariable logistic regression model. A decrease in the poststress LVEF compared with the LVEF at rest was associated with an increased probability of an ischemic perfusion scan pattern with a 
2 value of 21 (P = .001, df = 1). The area under the model receiver operating characteristic curve was 0.711 (on a scale of 0.5 to 1.0), which indicated that the model had moderate ability in the differentiation of patients with stress-induced ischemia from patients without ischemia.
|
DISCUSSION
|
|---|
Johnson et al (11) examined 81 sequential patients by using a 2-day stress-rest sequence protocol with gated perfusion SPECT. In 22 (36%) of 61 patients who demonstrated reversible perfusion defects, they found that the poststress LVEF was more than 5% lower than the LVEF at rest. They also found a significant difference in chordal shortening at rest and after stress in the reversible perfusion defect territories compared with that in the nonischemic perfusion defect territories. The authors did not find any statistically significant difference among groups with regard to age, sex, previous myocardial infarction, or type of stress. Results of univariable analysis showed that the variables that correlated substantially with changes in LVEF were the sum defect reversibility score on the perfusion scan and the presence of left anterior descending coronary artery disease. Results of multivariable analysis showed that only the sum defect reversibility score was significant. The authors concluded that if only poststress gated SPECT images are obtained, left ventricular function would not necessarily represent basal function in the presence of stress-induced ischemia.
A larger amount of radioactivity is injected at stress with the rest-stress 99mTc perfusion tracer protocol or with a dual isotope study, and the poststress tomograms are gated more commonly than are the resting images, with the assumption that the poststress result would reflect resting function. The results of our study, however, did not support this assumption. By using a same-day rest-stress or stress-rest sequence, we documented persistent functional abnormalities after an episode of acute ischemia (Figs 3, 4).
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Representative SPECT scans obtained at the middle of the left ventricle at rest and early after stress demonstrate severe perfusion abnormalities (arrows) in the anteroseptal wall (short-axis [SA] view), apical wall (horizontal long-axis [HLA] view), and anterior wall (vertical long-axis [VLA] view) during stress. Images obtained at rest are almost normal.
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Three-dimensional solid contoured images obtained in the same patient as in Figure 3 represent the left ventricular cavity in end diastole (ED) and end systole (ES). The mesh part of the images defines the epicardial borders. An anterior wall motion abnormality (arrow) is depicted on the poststress images with greater volumes and lower ejection fractions (EF, in percentages) than those of the rest images.
|
|
Findings from initial studies (16,17) have demonstrated that transient development of regional wall motion abnormalities and a global decrease in ejection fraction during stress resolves early in recovery once the stress is discontinued. Findings from most recent studies (18), however, have demonstrated a persistence of regional wall motion abnormalities up to 2 hours after completion of exercise stress testing. There are several possible explanations for the discrepancies between the results of these studies. High levels of circulating cathecholamines present immediately after exercise may reduce ventricular volumes and increase the ejection fraction, particularly in patients with mild or moderate ischemia. Using echocardiographic imaging, investigators (19,20) have demonstrated the persistence of regional wall motion abnormalities up to 30 minutes after completion of treadmill exercise in a high proportion of patients with multivascular disease.
Persistence of functional abnormalities after an episode of ischemia has been well documented in animal models and in humans (21–23). The phenomenon of postischemic stunning has been well described and consists of the presence of abnormal regional function in the absence of necrosis (24,25). Persistence of functional abnormalities are certainly proportional to the degree of ischemia induced and the duration of ischemia.
One of the potential technical explanations for our findings (rather than a true physiologic phenomenon) is the presence of severe perfusion abnormalities and the problems with the edge detection algorithm in finding the correct endocardial border because of the presence of decreased counts. The edge detection algorithm used in our study, originally described by Germano et al (9), identifies perfusion in underperfused areas of the myocardium by extracting count profiles from the nonthreshold image. Despite the predominant perfusion abnormality in the endocardium during ischemia, some degree of myocardial perfusion is always seen in the epicardium and detected in the count profiles. Furthermore, an asymmetric Gaussian curve is fitted to each profile, and the inner and outer SDs of the curve are measured. When perfusion is severely depressed along a profile, these SDs are combined with those of each of its four spatial neighboring profiles. Therefore, it is unlikely that technical factors are the explanation for our findings. This perspective is supported by Johnson et al (11).
Our study findings demonstrate that there is a high correlation between LVEFs obtained at rest and after stress with gated SPECT and same-day rest-stress or stress-rest protocols. However, a decrease in the LVEF obtained after stress compared with the LVEF at rest is associated with a very ischemic perfusion scan pattern. In the clinical setting, LVEFs obtained after stress with gated SPECT and a same-day protocol may not reflect true resting measurements. This finding carries important practical and clinical considerations in the interpretation of basal left ventricular function from poststress measurements in patients undergoing gated myocardial perfusion imaging studies.
|
Acknowledgments
|
|---|
We thank Candie Stewart, Tracy Schneider, and Thomas Hawk for their support in the preparation of the manuscript.
|
Footnotes
|
|---|
Abbreviation: LVEF = left ventricular ejection fraction
Author contributions: Guarantor of integrity of entire study, S.B.N.; study concepts and design, S.B.N.; definition of intellectual content, S.B.N.; literature research, G.R., A.J.; clinical studies, S.B.N., M.W.H.; data acquisition, G.R., R.A.P.; data analysis, A.J., S.B.N.; statistical analysis, L.K.S., D.F.K.; manuscript preparation and editing, S.B.N.; manuscript review, D.F.K., R.E.C.
|
References
|
|---|
-
Borges-Neto S, Coleman RE, Jones RH. Perfusion and function at rest and treadmill exercise using technetium-99m sestamibi: comparison of one- and two-day protocols in normal volunteers. J Nucl Med 1990; 31:1128-1132.[Abstract/Free Full Text]
-
Jones RH, Borges-Neto S, Potts JM. Simultaneous measurement of myocardial perfusion and ventricular function during exercise from a single injection of technetium-99m sestamibi in patients with coronary artery disease. Am J Cardiol 1990; 66:68E-71E.[Medline]
-
Friedman JD, Berman DS, Kiat H, et al. Rest and exercise treadmill first-pass radionuclide ventriculography: validation of left ventricular ejection fraction measurements. J Nucl Cardiol 1994; 1:382-388.[Medline]
-
Borges-Neto S, Coleman RE, Potts JM, Jones RH. Combined treadmill exercise radionuclide angiography and SPECT perfusion studies for assessment of coronary artery disease. Semin Nucl Med 1991; 21:223-229.[Medline]
-
Borges-Neto S, Shaw L, Kesler KL, et al. Prediction of severe coronary artery disease by combined rest and exercise radionuclide angiocardiography and tomographic perfusion imaging with technetium 99m–labeled sestamibi: a comparison with clinical and electrocardiographic data. J Nucl Cardiol 1997; 4:189-194.[Medline]
-
Borges-Neto S, Shaw LK, Javaid A. Prediction of cardiovascular death and non-fatal myocardial infarction in patients with angiographically documented coronary artery disease: a comparison between clinical, stress test, perfusion and functional data. Circulation 1998; 98(suppl):I654.
-
Borges-Neto S, Shaw LK, Coleman RE. Prediction of death and non-fatal myocardial infarction by combined perfusion and radionuclide ventricular function studies during pharmacologic stress testing. J Am Coll Cardiol 1999; 33(suppl):470A.
-
Nichols K, DePuey EG, Rozanski A. Automation of gated tomographic left ventricular ejection fraction. J Nucl Cardiol 1996; 3:475-482.[Medline]
-
Germano G, Kiat H, Kavanagh PB, et al. Automatic quantification of ejection fraction from gated myocardial perfusion SPECT. J Nucl Med 1995; 36:2138-2147.[Abstract/Free Full Text]
-
Chua T, Kiat H, Germano G, et al. Gated technetium-99m sestamibi for simultaneous assessment of stress myocardial perfusion, post-exercise regional ventricular function and myocardial viability: correlation with echocardiography and rest thallium-201 scintigraphy. J Am Coll Cardiol 1994; 23:1107-1114.[Abstract]
-
Johnson LL, Verdesca SA, Aude WY, et al. Postischemic stunning can affect left ventricular ejection fraction and regional wall motion on post-stress sestamibi tomograms. J Am Coll Cardiol 1997; 30:1641-1648.[Abstract]
-
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307-310.[Medline]
-
Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet 1995; 346:1085-1087.[Medline]
-
Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143:29-36.[Abstract/Free Full Text]
-
Di Leo C, Bestetti A, Tagliabue L, et al. 99m Tc-tetrofosmin gated-SPECT LVEF: correlation with echocardiography and contrastographic ventriculography. J Nucl Cardiol 1997; 4(suppl):S56.
-
Rozanski A, Elkayam U, Berman DS, Diamond GA, Prause J, Swan HJC. Improvement of resting myocardial asynergy with cessation of upright bicycle exercise. Circulation 1983; 67:529-535.[Abstract/Free Full Text]
-
Rozanski A, Berman D, Gray R, et al. Preoperative prediction of reversible myocardial asynergy by postexercise radionuclide ventriculography. N Engl J Med 1982; 307:212-216.[Abstract]
-
Ambrosio G, Betocchi S, Pace L, et al. Prolonged impairment of regional contractile function after resolution of exercise-induced angina: evidence of myocardial stunning in patients with coronary artery disease. Circulation 1996; 94:2455-2463.[Abstract/Free Full Text]
-
Robertson S, Geigenbaum H, Armstrong WF, Dillon JC, O'Donnell J, McHenry PW. Exercise echocardiography: a clinically practical addition in the evaluation of coronary artery disease. J Am Coll Cardiol 1983; 2:1085-1091.[Abstract]
-
Kata A, Force T, Folland E, Aebischer N, Sharma S, Parisi A. Echocardiographic assessment of ventricular systolic function. In: Marcus ML, Braunwald E, eds. Marcus cardiac imaging: a comparison to Braunwald's heart disease. Philadelphia, Pa: Saunders, 1996; 297-324.
-
Heyndrickx GR, Millard RW, McRitchie RJ, Maroko PR, Vatner SF. Regional myocardial functional and electrophysiological alterations after brief coronary artery occlusion in dogs. J Clin Invest 1975; 56:978-985.
-
Braunwald E, Kloner RA. The stunned myocardium: prolonged, postischemic ventricular dysfunction. Circulation 1982; 66:1146-1149.[Abstract/Free Full Text]
-
Kloner RA, Allen J, Cox TA, Zheng Y, Ruiz CE. Stunned left ventricular myocardium after exercise treadmill testing in coronary artery disease. Am J Cardiol 1991; 68:329-334.[Medline]
-
Dilsizian V, Bonow RO. Current diagnostic techniques of assessing myocardial viability in patients with hibernating and stunned myocardium. Circulation 1993; 87:1-20.[Free Full Text]
-
Vanoverscheld JL, Wikns W, Depre C, et al. Mechanism of chronic regional postischemic dysfunction in humans. Circulation 1993; 87:1513-1523.[Abstract/Free Full Text]
TOP
Abstract
Introduction
