Angiography. at least two of the following three criteria: 1) A history of

Transcription

1 Right Ventricular Ejection Fraction in Patients with Acute Anterior and Inferior Myocardial Infarction Assessed by Radionuclide Angiography EDWARD TOBINICK, M.D., HEINRICH R. SCHELBERT, M.D., HARTMUT HENNING, M.D., MARTIN LEWINTER, M.D., ANDREW TAYLOR, M.D., WILLIAM L. ASHBURN, M.D., AND JOEL S. KARLINER, M.D. SUMMARY We measured right and left ventricular ejection fraction (EF) from high frequency time-activity curves obtained during the initial passage of an intravenous bolus of 99mTc (Sn) pyrophosphate. In 22 normal controls right ventricular EF averaged ± 0.04 (). In 24 acute anterior or lateral infarction patients right ventricular EF was normal ( ± 0.10), while left ventricular EF was reduced (0.45 ± 0.10, P < vs controls). In 19 acute inferior infarction patients left ventricular EF also was depressed (0.51 ± 0.09, P < vs controls). Among 7 of 19 inferior infarc- tion patients with right ventricular infarction by scintigraphy, right ventricular EF was reduced (0.39 ± 0.05; P < vs normals; P < 0.01 vs inferior infarction patients without right ventricular involvement). In the latter group right ventricular EF averaged 0.51 ± 0.10 (NS vs normals). We conclude 1) a single injection of 19mTc (Sn) pyrophosphate can identify right and left ventricular dysfunction and infarct location in acute myocardial infarction, 2) right ventricular EF is well-preserved except when inferior infarction involves the right ventricle. ALTHOUGH LEFT ventricular performance has been extensively studied after acute myocardial infarction,1 there is little information concerning right ventricular function under these circumstances.2 3 The ejection fraction is a well accepted measure of ventricular function, but right ventricular performance has been difficult to quantitate by conventional means. This is in large part due to the complex geometry of the right ventricle which makes calculation of the right ventricular ejection fraction by the standard angiographic methods extremely difficult. Therefore, we developed a radionuclide technique for determination of right ventricular ejection fraction that is free of assumptions regarding the three-dimensional geometry of the right ventricular chamber. It was the purpose of this investigation to assess the usefulness and limitations of this virtually noninvasive technique in patients with acute myocardial infarction. Methods Patient Population Forty-three patients with an acute myocardial infarction were studied. Thirty-two patients were male and 11 female, with an average age of 63 years (range 34 to 89 years). All patients were admitted to the Ischemic Heart Disease SCOR Unit at the University of California Medical Center, San Diego, with a diagnosis of acute myocardial infarction. The diagnosis of acute myocardial infarction was based on From the Divisions of Nuclear Medicine and Cardiology, University of California, San Diego, School of Medicine and the Veterans Administration Hospital, San Diego, California. Supported by Specialized Center of Research on Ischemic Heart Disease, NIH Research Grant HL awarded by the National Heart, Lung, and Blood Institute. The present address of Drs. Tobinick and Schelbert is UCLA School of Medicine, Los Angeles, California; Dr. Henning's address is University of British Columbia, Vancouver, B.C. Address for reprints: Joel S. Karliner, M.D., University of California San Diego Medical Center, 225 W. Dickinson Street, San Diego, California Received June 30, 1977; revision accepted December 26, at least two of the following three criteria: 1) A history of typical prolonged chest pain; 2) electrocardiographic changes indicative of acute myocardial infarction; and 3) characteristic elevations of serum enzymes (CK, SGOT, LDH). Twenty-two patients, 11 men and 11 women, with an average age of 45 years (range 17 to 73 years), served as a control group to establish normal values for right ventricular ejection fraction (table 1). All 22 patients were referred to the Division of Nuclear Medicine for routine diagnostic bone scans using 15 mci of 99mTc (Sn) pyrophosphate. None of the patients had evidence of cardiopulmonary disease as determined by history, review of prior medical records, physical examination, resting electrocardiogram and chest X-ray. Informed written consent was obtained from all 65 patients participating in this study. Infarct Localization Radionuclide scintigraphy using 99mTc (Sn) pyrophosphate was performed in all 43 patients with acute myocardial infarction 1 to 8 days (mean 2.6 days) after the onset of acute symptoms. The location of the acute myocardial infarction was determined from both the electrocardiogram and the '9mTc (Sn) pyrophosphate images. Imaging was performed using a gamma scintillation camera* equipped with a low energy, high resolution parallel hole collimator two hours after the injection of 15 mci of 99mTc (Sn) pyrophosphate. Images were obtained in the anterior, 450 left anterior oblique and left lateral projections. All images were interpreted independently by two observers without knowledge of the clinical findings. Images were considered positive if there was discrete myocardial uptake of 99mTc (Sn) pyrophosphate in intensity equal to or greater than the activity of the ribs (i.e., 2+ or greater) and if the uptake was seen in at least two of the three projections.4 *Pho-Gamma HP, Searle Radiographics, Inc., or Dyna Mo, Picker Corporation.

2 1 079 RV EF IN MI/Tobinick et al. TABLE 1. Control Group: Patients without Cardiopulmonary Disease Pt/Age/Sex (yr) 1/44/M 2/28/F 3/67/F 4/72/M 5/63/F 6/35/M 7/35/F 8/43/F 9/54/M 10/17/M 11/45/M 12/56/F 13/22/F 14/45/F 15/73/M 16/51/M 17/38/F 18/25/M 19/57/F 20/32/M 21/66/F 22/21/M Mean 45 LVEF RVEF LVEF/RVEF * for difference between LVEF and RVEF *P < Abbreviations: M male, F female; LVEF left ventricular ejection fraction; RVEF right ventricular ejection fraction; - standard deviation. Left Ventricular Ejection Fraction Left ventricular ejection fraction was calculated from the initial passage of `9mTc (Sn) pyrophosphate injected rapidly as a bolus through a large median antecubital vein or through the jugular vein. The volume of the bolus usually was less than 1 ml but did not exceed 2 ml. The first transit of the radioactive bolus through the central circulation was recorded in the 300 right anterior oblique projection using the previously described gamma scintillation cameras and stored in an event-by-event mode (one word list mode) on a small dedicated digital computer (Med II, General Electric Co.). Left ventricular ejection fraction was calculated as described previously from our laboratory.5 In brief, the im- data were replayed, a region-of-interest assigned to the left ventricular blood pool and a high frequency left ventricular time-activity curve was generated. After correcting this curve for background activity, left ventricular ejection fraction was derived from the early downslope of the left ventricular time-activity curve using the root mean square age technique as previously described.5 Right Ventricular Ejection Fraction Right ventricular ejection fraction was obtained from the time-activity curves derived from the right ventricular blood pool in a manner similar to the calculation of the left ventricular ejection fraction. The image data recorded during the initial bolus passage through the central circulation were replayed and the time segment during which the bolus passed through the right ventricle was integrated to a 64 X 64 matrix point digital image (fig. 1). A region-ofinterest was precisely assigned with the light pen to the right ventricular blood pool. Care was taken to exclude the right atrium and the pulmonary artery from this region-ofinterest. The 300 right anterior oblique projection was found to be most suitable because it permitted good separation between the right ventricle, the right atrium, and the pulmonary artery. To correct for activity originating from noncardiac background structures a second semi-anular region-of-interest was then assigned two matrix points away from the anterior, apical and inferior portions of the right ventricular wall (fig. 1). The width of this background region-of-interest was two matrix points. High frequency time-activity curves (25 curve points/sec) then were generated from both the right ventricular and the background regions-of-interest. In order to correct for the difference in size for both regions-of-interest, the background time-activity curve was multiplied by the following ratio: number of matrix points in the right ventricular region-of-interest/number of matrix points in the background region-of-interest. The resulting curve was then subtracted from the right ventricular time-activity curve and a weighted 5 point curve-smoothing routine was applied. This routine was employed for all right ventricular ejection fraction calculations. A typical background corrected FIGURE 1. A) Right heart radionuclide angiogram, obtained after integration of the image data recorded during the transit of the bolus Of 990"Tc (Sn) pyrophosphate through the right heart. S VC superior atrium; RV right vena cava; ventricle; PA RA right pulmonary artery. The two small arrows indicate the position of the tricuspid and pulmonic valves, respectively. B) Assignment ofregions-of-interest with the light pen to the right ventricular cavity (R V-ROI) and the surrounding noncardiac background (BKGR Ol).

3 1080 CIRCULATION VOL 57, No 6, JUNE i,>100 - A) A A A A* ED A ED A A ED A A AA* A A% Aft ES e i ES EF: m ES A A% time (sec) FIGURE 2. Typical right ventricular time activity curve after correction for background activity. Each curve point represents the activity during 40 msec intervals ( 25 curve-points/sec). The peaks on the curve correspond to end diastole (ED), the valleys to end systole (ES). In this patient, ejection fraction was derivedfrom three beats. and smoothed right ventricular time-activity curve is shown in figure 2. Each point on the curve represents the number of counts during a 40 msec period (25 curve points/sec). The curve is characterized by cyclic fluctuations with each peak corresponding to end diastole and each valley to end systole. Right ventricular ejection fraction was then calculated by dividing the difference in counts between end diastole and end systole (a value proportional to stroke volume) by the number of counts at end diastole (a value proportional to end-diastolic volume). For the calculation only beats from the early downslope of the right ventricular time-activity curve were used. Beats with low activity or contaminated by activity arising from the lung were excluded from analysis. In general, three or four cardiac cycles were used for deriving the ejection fraction. In five patients, ejection fraction was calculated from only two beats. In each patient, the values obtained from the beats analyzed were then averaged. To determine the effects of activity originating from other cardiac structures on the right ventricular time-activity curve and to establish the most easily reproducible background region-of-interest, additional regions-of-interest were assigned to the right atrium, the tricuspid valve plane, the pulmonary artery and the area surrounding the entire right ventricle in six patients. The curves obtained from these additional regions-of-interest were normalized with respect to the size of the right ventricular region-of-interest. However, the activity of these normalized curves exceeded that obtained from the right ventricle and indicated that the activity sampled from these regions-of-interest did not represent background counts. Therefore, the background region-of-interest found to be most suitable was the semi-anular ring, described above, that surrounds the free portion of the right ventricular chamber.. Unlike recordings obtained with multicrystal cameras, peak end-diastolic right ventricular count rates averaged only 250 counts/40 msec curve point. These relatively low count rates raise concern about the statistical reliability of ejection fraction calculations obtained from first pass timeactivity curves recorded with the standard single crystal gamma camera. In a previous report from our laboratory the statistical uncertainty of left ventricular ejection fraction calculations was considerably improved by employing the root mean square technique.5 This approach determines the root mean square of the deviations of the raw data curve points from an "average" curve obtained by applying a 14 point weighted curve-smoothing procedure to the raw data curve. However, because of the very rapid upstroke and sharp peak of the right ventricular time-activity curve, the 14 point curve-smoothing procedure did not provide satisfactory average curves and the root mean square approach often resulted in erroneous right ventricular ejection fractions (e.g., values > 1.00). For this reason, and in view of higher right ventricular count rates, right ventricular ejection fraction was calculated from the end-diastolic and end-systolic curve points only. As mentioned earlier, peak right ventricular count rates averaged 250 counts/40 msec curve point. As shown in the appendix, the standard deviation for an ejection fraction of calculated from a typical right ventricular time-activity curve is equal to approximately The standard deviation of the calculation was reduced further to approximately 0.02 by applying a weighted 5 point-smoothing to the right ventricular time-activity curve. The relatively high degree of statistical certainty appears substantiated by the reproducibility of the technique in the patients with repeat studies. While the curve-smoothing procedure improves the statistical reliability of the calculation, it consistently reduces right ventricular ejection fraction. Thus, in 14 patients with heart rates ranging from 56 to 89 beats/min, ejection fractions derived from smoothed time activity curves were 8.9% ± 1.7 (SE) lower than those obtained directly from the raw data curves. This raises the question as to whether right ventricular ejection fraction can be compared to left ventricular ejection fraction. Therefore, left ventricular ejection fractions were determined in the same 14 patients by three different approaches: the root mean square technique; directly from the end-diastolic and endsystolic curve points of the unmodified, raw data timeactivity curve; and lastly, from the smoothed time-activity curve. Compared to ejection fraction values obtained from the raw data curves, left ventricular fraction values derived by the root mean square technique were 14.3% ± 1.9 (SE) lower and ejection fractions derived from the smoothed curves were 11.3% ± 2.0 (SE) lower. Ejection fractions derived from the smoothed curves did not differ significantly from those calculated by the root mean square technique. Because the reduction in the values of left ventricular ejection fraction introduced by the root mean square technique is similar in magnitude to that resulting from smoothing the right ventricular time-activity curve, the ratio between left and right ventricular ejection fraction remains unaffected. Moreover, the consistently lower values for right compared with left ventricular ejection fraction observed in this study agree well with the findings by Berger et al.6 in 20 normal subjects in whom left ventricular ejection fraction averaged 0.70 ± 0.01 (SE) and right ventricular ejection fraction 0.54 ± All data were analyzed using Student's t-test for paired and unpaired data. Results Control Group In the 22 patients without evidence of cardiopulmonary disease right ventricular ejection fraction ranged from 0.44

4 RV EF IN MI/Tobinick et al to 0.60 with a mean of ± 0.04 () (table 1). Left ventricular ejection fraction in these 22 patients ranged from to 0.73 with a mean of 0.60 ± Thus right ventricular ejection fraction was moderately but significantly lower (P < 0.001). The mean ratio of left-to-right ventricular ejection fraction was (range 1.00 to 1.38). None of these 22 patients showed abnormal myocardial activity on the subsequently recorded '9mTc (Sn) pyrophosphate images. Reproducibility of the Method To evaluate the reproducibility of our method, six patients with ischemic heart disease who had normal right ventricular ejection fractions were injected twice with 8 mci of 99mTc DTPA* within a 30 min time interval without changing position. Because of the smaller amount of activity injected the gamma camera was equipped with a high sensitivity low energy parallel hole collimator. Right ventricular ejection fraction averaged 0.55 ± 0.05 for the first measurement; the average value for the second measurement was identical. The variation between the first and second study was small and averaged only 3.4%. It should be emphasized, however, that the close agreement between two subsequent studies in the same patient was heavily dependent on the correction of the right ventricular time-activity curves for activity originating from noncardiac background structures. For the first measurement the effect of background correction on the right ventricular ejection fraction was small, i.e., correcting for background activity resulted in values for ejection fraction that were higher by an average of 4.8% (range 1.9 to 9.3%). However, analysis of sequential studies revealed that ejection fractions were lower by an average of 10.7% (range 2.3 to 22.6%) in the second study when no background correction was employed. To evaluate interobserver variation, studies from nine of the patients were processed independently by two investigators. Calculated right ventricular ejection fraction displayed close agreement with a mean difference of 2.7%. Acute Anterior or Anterolateral Wall Myocardial Infarction In the 24 patients with anterior or lateral myocardial infarctions, left ventricular ejection fraction averaged 0.45 ± 0.10 (range 0.31 to 0.66) and was lower than in the normal control group (P < 0.001, [table 2]). The average right ventricular ejection fraction was ± 0.10 (range 0.32 to 0.80) and did not differ significantly from that found in the control group. The ratio of left-to-right ventricular ejection fraction was 0.83 ± 0.17, a value that was decreased when compared to control values (P < 0.001). Acute Inferior Myocardial Infarction Nineteen patients had an acute inferior myocardial infarction. Left ventricular ejection fraction in these patients averaged 0.51 ± 0.09 and was reduced compared with control values (P < 0.001) (table 3). Right ventricular ejection fraction averaged 0.47 ± 0.10 and was lower than in the normal group (P < 0.03), and in the group with anterior or anterolateral infarction (P < 0.005). This difference was due *Diethylene triamine penta acetate TABLE 2. Patients with Acute Anterior or Lateral Myocardial Infarction Pt/Age/Sex Time* LVEF RVEF LVEF/RVEF (yr) (Days) 1/69/F /52/F /81/M /58/M /68/M /55/M /78/M /65/M /76/M /64/M /70/F /48/M /46/M /56/M /53/F /59/F /75/M /64/M /60/M /58/M /64/M /63/M /85/M /41/M Mean (P <0.001) (NS) (P <0.001) *Time from onset of acute symptoms. For abbreviations see table 1. The P values indicate the difference from the control group (table 1). NS not significant. standard deviation. largely to the patients with right ventricular involvement (see below). The ratio of left-to-right ventricular ejection fraction averaged , a value that was not significantly different from the normal group. Right Ventricular Infarction Utilizing the 99mTc (Sn) pyrophosphate images, patients with inferior wall myocardial infarction were divided into two groups with respect to right ventricular involvement: ten patients exhibited no right ventricular involvement on the 99mTc (Sn) pyrophosphate images (group 1), but in seven patients an additional region of increased ssmtc (Sn) pyrophosphate uptake was noted indicating right ventricular involvement (group 2). As shown in figure 3, this additional region of abnormal 99mTc (Sn) pyrophosphate uptake was usually noted best in the 450 left anterior oblique projection and appeared as an area of activity anterior and to the right of the left ventricular uptake and adjacent to and in line with the inferior wall of the left ventricle.7 One of the 19 patients with acute inferior myocardial infarction by electrocardiography had a negative 99mTc (Sn) pyrophosphate image and in a second patient with an inferior myocardial infarction by electrocardiogram the inferior wall pyrophosphate activity could not be localized to either ventricular chamber. Mean left ventricular ejection fraction was in group 1 and in group 2. These values were depressed compared to the normal group (P < 0.001). Right ventricular ejection fraction averaged 0.51 ± 0.10 (NS vs normals) in group 1 but was only 0.39 ± 0.05 in group 2 (P < vs normals; P < 0.01 vs group 1).

5 1082 CIRCULATION VOL 57, No 6, JUNE 1978 TABLE 3. Group I. Left 1/75/F 2/56/M 3/89/F 4/53/F 5/53/F 6/69/M 7/34/M 8/65/M 9/48/M 10/76/F Mean 62 Patients with Acute Inferior Myocardial Infarction Pt/Age/Sex Time* LVEF RVEF LVEF/RVEF (yr) (Days) Ventricular Infarction Only by Scintigraphy (P <0.001) (NS) (P <0.001) Group 2. Left and Right Ventricular Infarction by Scintigraphy 1/60/M /57/M /74/M /73/M /53/M /80/M /73/M Mean (P <0.001) (P <0.001) (NS) Scintigraphy Negative or Ventricular Localization Not Possible 1/42/M /59/F All Patients (N 19) Mean (P <0.001) (P <0.03) (NS) *Time from onset of acute symptoms. The P values indicate the difference from the normal control patients (table 1). Abbreviations same as table 1. Discussion Technical and Statistical Considerations Left ventricular ejection fraction determined from high frequency time-activity curves recorded during the initial transit of a radioactive bolus through the heart has been shown to correlate closely with the left ventricular ejection fraction determined from single or biplane cineventriculograms using the area-length method." 89 9 First pass radionuclide angiocardiography has an important advantage over other methods for measuring ejection fraction: it does not require approximation of the ventricular chamber by a three-dimensional geometrical model, as is necessary when ejection fraction is derived from single or biplane contrast angiography, ECG gated cardiac blood pool images, or echocardiography. Thus, first pass radionuclide angiography is more accurate than echocardiography in measuring left ventricular ejection fraction in patients with wall motion abnormalities, in whom the geometrical model does not closely approximate the shape of the left ventricular cavity.10 Because first pass radionuclide angiocardiography does not depend on variations in ventricular geometry, it is particularly suitable for measuring right ventricular ejection fraction. Moreover, count rates obtained from the right ventricle are two to three times higher than those obtained from the left ventricle because the radioactive bolus is still rather well confined during its passage through the right ventricle and because of the greater proximity between the right ventricle and the collimator in the right anterior oblique projection. Correction of the right ventricular time-activity curves for background activity affected right ventricular ejection fraction by less than 5% because little activity has reached the lungs when count rates are highest in the right ventricle. Therefore, right ventricular ejection fraction could be determined without background correction. However, for serial measurements of ejection fraction, e.g., to evaluate the effects of interventions on right ventricular performance, residual activity may cause an upward shift of the right ventricular time-activity curve and result in lower ejection fraction values. In this study, almost identical values for ejection fraction measurements performed within less than 30 minutes were obtained only after correcting the right ventricular time-activity curves for background activity. Calculation of right and left ventricular ejection fraction from a recording of the initial bolus passage through the central circulation is based on temporal separation of the right from the left ventricular activity. Due to superimposition of the left and right ventricle in the right anterior oblique projection, delayed clearance of activity from the right ventricle, such as may occur when right ventricular function is depressed, may interfere with accurate measurement of left ventricular fraction. However, even in the seven patients with evidence of right ventricular infarction, right and left ventricular activity were sufficiently separated in time to permit assessment of left ventricular ejection fraction. Right Ventricular Volumes and Ejection Fraction in Normal Subjects There are few studies of right ventricular volumes in normal subjects. Calculation of right ventricular volumes and ejection fraction from biplane cineangiograms in these studies was based on approximating the shape of the right ventricular cavity by various geometric models, e.g., elliptical or rectangular cross-sections or a prism with triangular base."-"3 Mean right ventricular ejection fraction reported in these studies ranged from 0.51 to This wide range may be largely due to differences in determining right ventricular volumes and patient selection. Steele and co-workers'4 initially introduced first pass radionuclide angiocardiography to determine right ventricular function. In 14 patients without coronary artery disease, they observed a mean right ventricular ejection fraction of 0.57, with a range of 0.51 to The lower average value for right ventricular ejection fraction for the 22 patients without evidence of cardiopulmonary disease in our study may be related to differences in data analysis although it is very similar to the value of 0.54 recently reported by Berger et al.6 The same authors reported the ejection fraction to be higher for the left than for the right ventricle (leftto-right ratio of 1.30) which is consistent with our findings (average ratio of left-to-right ventricular ejection fraction of 1.17).

6 1083 RV EF IN MI/Tobinick et alt FIGURE 3. Typical "tmtc (Sn) pyrophosphate images in two patients with acute inferior myocardial infarction. The patient shown in the upper panel displays involvement of the left ventricle only, whereas the patient shown in the lower panel shows involvement of the inferior wall of the left ventricle and the right ventricular free wall, as indicated by the arrows. Right and Left Ventricular Ejection Fraction in Acute Myocardial Infarction There are no previously published studies of right ventricular ejection fraction in acute myocardial infarction. Although isolated right ventricular infarction has been reported to occur in only 2.5 to 4.6% of patients,'5 1" the incidence of concurrent right and left ventricular infarction has been reported to be as high as 43%.17 As expected, the average left ventricular ejection fraction in our patients with acute myocardial infarction was reduced. Right ventricular ejection fraction, however, was relatively well preserved in patients with anterior or lateral infarction, as reflected by both a normal mean right ventricular ejection fraction and a significantly decreased ratio of left-to-right ventricular ejection fraction. The absence of right ventricular functional depression in the presence of anterior or lateral left ventricular functional impairment is supported by the data of Rigo et al.,'8 who reported a mean right to left ventricular area of 0.75 in 13 patients with acute anterior infarction. However, it should be pointed out that three patients with acute anterior or lateral myocardial infarction had a reduced value for both right and left ventricular ejection fraction (table 2). All three patients had clinical evidence for congestive cardiac failure during the course of their acute myocardial infarction and required digitalis and diuretics as part of their management. Two of these patients have subsequently died of an ischemic cardiomyopathy. Right ventricular ejection fraction was not well preserved in our patients with inferior infarction and right ventricular free wall involvement, however. In contrast to the patients with anterior or lateral infarcts, none of whom had involvement of the right ventricular free wall, seven of 19 patients (37%) with inferior infarction had evidence of right ventricular free wall involvement as visualized by 9smTc (Sn) pyrophosphate imaging. Identification of right ventricular infarction by "9mTc (Sn) pyrophosphate has been previously reported by Sharpe et al.,7 who also found a 35% incidence of right ventricular infarction in 14 consecutive patients with acute inferior myocardial infarction. Separation of patients with acute myocardial infarction confined only to the inferior wall of the left ventricle from those with additional involvement of the right ventricle was supported further by measurements of reduced right ventricular performance. Thus, right ventricular ejection fraction was significantly depressed in patients with right ventricular infarction when compared with the normal group. In contrast, right ventricular ejection fraction was normal in patients with inferior infarction not involving the right ventricle. In our patients a reduction in preload due to volume depletion cannot be ruled out entirely as a possible reason for the biventricular reduction in ejection fraction in the patients with both left and right ventricular infarctions. Nevertheless, it is of interest that two of the seven patients with a depressed right ventricular ejection fraction exhibited clinical signs consistent with right ventricular dysfunction; both patients had an elevated jugular venous pressure on physical examination and in one patient direct measurement of the right atrial pressure revealed a mean value of 10 mm Hg. Our observations and those of others7' 17 indicate that right ventricular infarction is more common than previously suspected, but the resulting depression in right ventricular performance only rarely leads to clinical signs of right ventricular failure. Acknowledgments We are grateful to Ms. Sue Swanson for technical assistance, to Mrs. Elizabeth Gilpin for statistical advice, and to Mrs. Marianne Lindley, Mrs. Sandra Edmonds, and Mrs. Rosemary Montoya for secretarial assistance. The authors also wish to thank Henry Hluang, Ph.D., for his assistance in preparing the appendix. Appendix The statistical uncertainty of the right ventricular ejection fraction calculated directly from the end-diastolic and end-systolic curve points can be estimated by the following equations: (EF)2 t ES2 [(ED)2 ED2 Where ED + (ES'] ES2 standard deviation; ED (1) J counts at end-diastole; ES counts end-systole; EF ejection fraction. Assuming the statistics follow Poisson's distribution, equation (1) becomes: at

7 1084 CIRCULATION VOL 57, No 6, JUNE 1978 I (EF)I ES2 ( ) (2) Using several cardiac beats for calculating the ejection fraction (EF) becomes: (EF).V/(EF) I+ (EF)1... (EF)2 (3) Employing a weighted five point curve smoothing,* the (EF) is estimated to be further reduced by-~1/2; for example: (EF,) (EF) (4) The statistical uncertainty of the calculated ejection fraction from a typical right ventricular ejection fraction and using three cardiac cycles is derived: End-diastolic and end-systolic counts for the first beat are 250 and 120, for the second beat 185 and 89; and for the third beat 150 and 72. The average ejection fraction is. Using equation (2), the standard deviation of the calculation for the first beat is 0.05, for the second beat is 0.06 and for the third beat is The standard deviation for the ejection fraction averaged for all three beats is (equation 3). According to equation (4) the weighted five point curve smoothing reduces the standard deviation further to about *Howell PG: Introduction to Mathematical Statistics. John Wiley and Sons, Inc. New York, References 1. Forrester JS, Diamond G, Chatterjee K, Swan HJC: Medical therapy of acute myocardial infarction by application of hemodynamic subsets. N EngI J Med 25: 1356, Cohn JN, Guiha NH, Broder MI, Limas CJ: Right ventricular infarction: Clinical and hemodynamic features. Am J Cardiol 33: 209, Schelbert HR, Henning H, Ashburn WL, Verba JW, Karliner JS, O'Rourke RA: Serial measurements of left ventricular ejection fraction by radionuclide angiography early and late after myocardial infarction. Am J Cardiol 38: 407, Righetti A, O'Rourke RA, Schelbert H, Henning H, Hardarson T, Daily PO, Ashburn W, Ross J Jr.: Usefulness of preoperative and postoperative Tc-99m (Sn)-pyrophosphate scans in patients with ischemic and valvular heart disease. Am J Cardiol 39: 43, Schelbert HR, Verba JW, Johnson AD, Brock GW, Alazraki NP, Rose FJ, Ashburn WL: Nontraumatic determination of left ventricular ejection fraction by radionuclide angiocardiography. Circulation 51: 902, Berger HJ, Natthay RA, Marshal RA, Gottschalk A, Cohen LS, Zaret BL: Noninvasive radionuclide technique for right ventricular ejection fraction in man. Circulation 53/54 (suppl II): , Sharpe N, Botvinick E, Shames D, Chatterjee K, Massie B, Schiller H, Parmley W: Noninvasive diagnosis of right ventricular infarction: A common clinical entity. Circulation 54 (suppl II):II-76, Steele P, Kirch D, Matthews M, Davies H: Measurement of left heart ejection fraction and end-diastolic volume by a computerized scintigraphic technique using a wedged pulmonary artery catheter. Am J Cardiol 34: 179, Hamilton GW, Weaver WD, Williams DL, Ritchie JL, Caldwell J, Kennedy JW: Calculation of ejection fraction from first transit radionuclide data: A comparison of several methods with contrast angiography. (abstr) J Nucl Med 17: 556, Henning H, Schelbert H, Crawford MH, Karliner JS, Ashburn W, O'Rourke RA: Left ventricular performance assessed by radionuclide angiocardiography and echocardiography in patients with previous myocardial infarction. Circulation 52: 1069, Gentzler RD, Briselli MF, Gault JH: Angiographic estimation of right ventricular volume in man. Circulation 50: 324, Fisher EA, DuBrow IW, Hastreiter AR: Right ventricular volume in congenital heart disease. Am J Cardiol 36: 67, Ferlinz J, Gorlin R, Cohn PF, Hermar. MV: Right ventricular performance in patients with coronary artery disease. Circulation 52: 608, Steele P, Kirch D, LeFree M, Battock D: Measurement of right and left ventricular ejection fractions by radionuclide angiocardiography in coronary artery disease. Chest 70: 51, Wartman WB, Hellerstein HK: The incidence of heart disease in 2,000 consecutive autopsies. Ann Intern Med 28: 41, Yater WM, Traum AH, Brown WG, Fitzgerald RP, Geisler MA, Wilcox BB: Coronary artery disease in men eighteen to thirty-nine years of age. Report of eight hundred sixty-six cases, four hundred fifty with necropsy examinations. Am Heart J 36: 683, Erhardt LR: Clinical and pathological observations in different types of acute myocardial infarction. A study of 84 patients deceased after treatment in a coronary care unit. Acta Med Scand [suppl] 560: 1, Rigo P, Murray M, Taylor DR, Weisfeldt ML, Kelly DT, Strauss HW, Pitt B: Right ventricular dysfunction detected by gated scintiphotography in patients with acute inferior myocardial infarction. Circulation 52: 268, 1975