- Email: [email protected]
Relationship between segmental abnormalities and global left ventricular function in coronary artery disease: Validation of a theoretical model
Relationship between segmental abnormalities and global left ventricular function in coronary artery disease: Validation of a theoretical model A theoretlcal model of the heart, which suggests a direct relatlonshlp between segmental abnormalities of left ventricular (LV) wall motion (WM) and LV ejection fraction (EF), was tested uslng twodlmenslonal echocardlography (2DE) and multiple-gated equtlibrlum blood pool sclntlgraphy (AGES) In a population of 25 coronary artery disease patients. MGES was used to determine EF, and 2DE was used to develop a method of analysis of LV segmental WM abnormalities. Two orthogonal apical 2DE vlews were analyzed. The length of the end-diastolic segments in which normal contraction occurred durfng systole were measured in each view, summed and divided by the sum of the end-diastolic silhouette lengths. The fraction thus created was multlplled by 100 and deflned as the percentage of normally contracting myocardlum (%NCM). %NCM correlated well with EF determined by MGES (I = 0.94). Determination of %NCM was highly reproducible for the same observer (r = 0.98), as well as for two observers (r = 0.98), and the standard error of estlmate was low in both cases (4%). These findings, in addition to confirming the theoretlcal model, provide a new technique to assess LV segmental WM abnormalities by 2DE. (AM HEART J 102330, 1981.)
Clain Beeder, B.A., Yzhar Charuzi, M.D., Irving K. Loh, M.D., Howard Staniloff, M.D., and H. J. C. Swan, M.D., Ph.D. Los Angeles,
Segmental abnormalities of left ventricular (LV) wall motion (WM) are associated with myocardial &hernia and infarction (MI).‘-” A conceptual model of the heart in acute myocardial infarction (AMI) has been proposed by Swan et al.” which suggests a direct relationship between segmental abnormalities and global LV function. Utilizing a two-component model consisting of a noncontractile infarct area and a residual non-infarcted area with normal contractility, the authors postulate a linear reduction in ejection fraction (EF) as the noncontractile infarct area increases. Support for this model can be found in several experimental and clinical studies using contrast ventriculography.“-R However, these studies share several limitations. The area-length method used to calculate EF assumes a symmetrically contracting ventricle. Further, the same tracing was used to assesssegmental abnormalities and calculate EF. From the Division of Cardiology, Medical Center, UCLA School Funded Received
in part
by the Save-A-Heart
for publication
Reprint requests: Clain Cedars-Sinai Medical 90048.
330
Department of Medicine.
Feb. Beeder, Center,
of Medicine,
Cedars-Sinai
Foundation.
27, 1981;
accepted
Apr.
B.A., c/o Publications 8700 Beverly Blvd.,
15, 1981. Office Los
(Cardiology), Angeles, CA
Calif.
Determination of EF by multiple-gated equilibrium blood pool scintigraphy makes no geometric assumptions, and is therefore not hampered by segmental abnormalities. Two-dimensional echocardiography (2DE) has been used to identify segmental abnormalities of LVWM.“-” Due to the combination of views available by 2DE, the extent of segmental abnormalities may be better appreciated than by other techniques. Thus segmental abnormalities and EF can be more satisfactorily assessedby these independent techniques. The present study was undertaken to test the conceptual model of Swan, using these improved, independent techniques in a group of patients with segmental abnormalities of LVWM due to coronary artery disease (CAD). METHODS Patient populatbn.
We studied 27 consecutive CAD patients. Age rangedfrom 45 to 77 years, with mean ageof 62 years ( + 7.7 SD). Twenty-two patients were male and five were female. Twenty-four patients were between 1 and 3 weeks post-AMI, and the remaining three patients had angiographically demonstrated CAD with no prior MI. All patients had both LVEF determined by Tc-99m multiple-gated equilibrium blood pool scintigraphy (MGES), and a 2DE study performed on the same day. 0002-8703/81/090330
+ 05$0050/O
0 1981
The C. V. Mosby Co.
Volume Number
102 3, part 1
Twenty-five patients had 2DE studies suitable for analysis. Two patients were excluded from analysis due to incomplete visualization of the LV silhouette in the apical long-axis view. We excluded from consideration any patient with clinically evident cardiomyopathy, hemodynamically significant valvular disease, complete left bundle branch block, ventricular pacemaker, or cardiac surgery, because of potentially confounding contraction abnormalities unrelated to CAD. EF determination. MGES was performed in the following manner: 10 ml of blood were withdrawn from the patient, incubated with 20 to 25 mCi of technetium-99m stannous phosphate, and injected intravenously.“, 12 Patients were studied using a portable scintillation camera (Ohio Nuclear Series 420 Mobile) in an approximate 40-degree left anterior oblique projection with lodegree caudal tilt.” The exact projection in each case was chosen to provide the best separation between ventricles. A shielded, parallel-hole, low energy collimator was used, and a total of 200,000 counts were obtained and recorded on a mobile minicomputer (Medical Data Systems). The signal was gated to the surface ECG and divided into I4 frames for the first two thirds of the cardiac style. EF was determined by the semiautomated MUGA program (Medical Data Systems) and expressed as a percentage. Measurement of EF by this technique is highly reproducible with variation less than 6% between two separate studies performed 1 hour apart.” 2DE evaluation. 2DE studies were performed using a Toshiba SSH 10-A 78-degree phased array ultrasonograph. All studies were done with the patient rotated 30 degrees into a left lateral decubitus position. 2DE views were obtained with the patient in held normal, unforced, end-expiration. This was done to provide a stable image in which the LV did not change position relative to the echo transducer due to respiration. Studies were recorded on an NEC sA inch video cassette recorder for subsequent analysis. Each 2DE study contained two apical views from which calculations were made. The apical four-chamber view was obtained by placing the transducer at the point of maximal cardiac impulse and angling the beam toward the right shoulder. The angle of the transducer was manually adjusted to provide an image of all four cardiac chambers, as well as portions of the mitral (MV) and tricuspid valves, in which maximal chamber size was obtained for the LV and right ventricle. The apical long-axis view was obtained from the same chest position by rotating the transducer approximately 90 degrees to obtain an image of the LV and left atrium, with the LV at maximum dimensions. In this way, two orthogonal views were obtained which allowed visualization of the cardiac apex, interventricular septum, and anterior, inferior, and lateral LV walls (Fig. 1). Thus information on WM was obtained for each major LV segment. No other views were considered for quantitative analysis. Data analysis. The 2DE studies were read independently and on different occasions by two observers, blinded to the identity of the patient as well as to the
LV segmental-global
functional
relations in CAD
331
Fig. 1. Diagrammatic representation of the apical imaging planes, illustrating the orthogonal relationship between the apical long-axis and apical four-chamber views. Ao = aorta; PA = pulmonary artery; LA = left atrium; LV = left ventricle; RA = right atrium; RV = right ventricle.
results of the MGES. A transparent plastic sheet was taped over the television monitor on which the study was displayed, and the inner border of the LV was traced in both end-diastole and end-systole with a marking pen, using stop-frame images gated to the ECG. End-diastole was defined as the peak of the R wave, and end-systole as the end of the T wave. The position of the inner border was confirmed on a moving image and visually “signal averaged” over five to seven serial cardiac cycles. The innermost border which formed signals that could be visually appreciated was traced in both end-diastole and end-systole on the same plastic transparency. A tracing was made in both the apical four-chamber and apical long-axis view. In’both views, the inner LV border was traced from MV ring to MV ring. The length of the end-diastolic LV outline was .measured in each view, using a Numonics Corporation model 277 Graphics Calculator. Abnormal contraction was defined as akinesis or dyskinesis.“-” Those end-diastolic segments demonstrating normal contraction during systole were measured in each view and added together. This value was divided by the sum of the length of both end-diastolic outlines. The fraction thus created was multiplied by 100 and defined as the percentage of normally contracting myocardium (%NCM). Fig. 2 illustrates the method of calculation of %NCM. The 2DE studies were reread on all 25 patients blindly and in random order, by one observer, 2 months ater the first reading. %NCM was recalculated and a linear regression was performed between the original reading and the second reading.
332
Beeder et al.
American
------
END DIASTOLE END SYSTOLE
A4C
% NCM = TOTAL
seplember, 1981 Heart Journal
ALA A+ B+C+D END DIASTOLIC LENGTH
( A4C + ALA)
x
100
Fig. 2. Representative example demonstrating the manner of calculating the percentage of normally
contracting myocardium (%NCM) using tracings from one of the patients in this study. A, B, C, and D indicate the length of the end-diastolic outline where normal contraction occurred during systole. A4C = apical four-chamber view; ALA = apical long-axis view. of the regression line, constrained to pass through zero, was y = 0.976 x, with SEE of 3.91% and r = 0.98. %NCM-2DE to LVEF-MGES relationships. Fig. 5 illustrates the relationship between %NCM by 2DE and EF by MGES. A linear relationship between %NCM and EF was demonstrated (r = 0.94, y = 0.68 X - 0.65, SEE = 5.06%). %NCM ranged from 29% to 100% (61% mean + 20% SD), and EF ranged from 17% to 65% (42% f 15%). Effect of akinesls
O2
20 40 60 80 % NCM (OBSERVER I)
100
Fig. 3. Interobserver agreementfor independent determination of %NCM by two observers.The continuous line is the regressionline, constrained to passthrough zero. The slope of the regressionline, r and p values, and standard error of estimate (SEE) are illustrated.
RESULTS interobserver
and intraobserver
%NCM-POE
variabili-
ty. Fig. 3 demonstrates the interobserver variability of %NCM in 25 patients. The equation of the regression line, constrained to pass through zero, was y = 0.997 x, with a standard error of estimate (SEE) of 4.40% and r = 0.98. Fig. 4 illustrates the intraobserver variability of %NCM. The equation
critwia
on %NCM
to LVEF relation.
While akinesis was used as the criterion for distinguishing a normal segment from an abnormal segment, it is clear that other criteria could be employed. For this reason, four criteria were investigated. Measurements were made using the following definitions separating normal from abnormal segments: akinesis, 25% of the maximum systolic inward motion, 50% of maximum, and 75% of maximum. The relationship between %NCM and EF varied in stepwise fashion from r = 0.94 using the akinetic criterion to r = 0.92 using the 75% of maximum criterion. The best results were obtained using akinesis as the dividing line between normal and abnormal contraction. DISCUSSION Linear relation between %NCM by 2DE and LVEF by MGES. Our study demonstrates a linear relationship
between %NCM by 2DE and LVEF by MGES. Determination of %NCM by two observers of varied experience was highly reproducible. Similarly, repeat determination of %NCM by one observer, 2
Volume
102
Number
3, part 1
;
LV segmental-global
loo
y= .997x
80
r= .98 (p<,OOi) SEE= 4.40 n=25
60
0 .a
/
20 40 60 80 %NCM (READING I)
100
the regressionline, constrained to passthrough zero. The slope of the regressionline, r and p values, and standard error of estimate (SEE) are illustrated. months after the original reading, was also highly reproducible. However, a limitation of the %NCM technique is that it may only be applicable to patients with segmental contraction abnormalities such as those associated with CAD. If the myocardium cannot be readily divided into contracting and noncontracting segments, then the relationship between %NCM and EF may be altered. of NCM-MI
concept
of LV function
relations
Y= .68x- -65 ~;~45(*gDoI) = n=25 ’
in CAD.
A theoretical, two-compartmental model of cardiac performance has been proposed by Swan et al.” to define abnormalities of ventricular function in semiquantitative terms. The variables considered included the proportion of perimeter which is noncontractile, the diastolic compliance of the abnormal myocardium, and the adverse effects of additional hemodynamic burdens such as mitral regurgitation or ventricular septal rupture. The proposed pressure-volume relationship has been validated by Maggs et al.‘j in an experimental preparation. One basic hypothesis of Swan’s model” was prediction of a linear relationship between EF and the proportion of noncontractile myocardium, or percent infarction. Our %NCM measurement is the compliment of percent infarction; that is, 100% minus percent infarction equals %NCM. By substituting %NCM for percent infarction, the theoretical model predicts an EF of 67% with %NCM of 100, and an EF of 0% with a %NCM of 0. Our results with CAD produced
OV
in CAD
0 0
333
:
3 /
t
Fig. 4. Intraobserver agreement for repeated measurement of %NCM by one observer. The continuous line is
Validation
I-
9
functional
I 20
I
I
I
40 60 80 %NCM (2DE)
I 100
Fig. 5. Relationship between the percentage of normally contracting myocardium @NCM) by two-dimensional echocardiography(2DE) and left ventricular ejection fraction (LVEF) by multiple-gated equilibrium scintigraphy (MGES) in 25 CAD patients. The-equation of the line, standard deviation, and r and p values are illustrated. a relationship extremely close to that predicted by the theoretical model, and suggest that systolic stiffness of these recent infarcts was relatively uniform. Consideration of LV compliance. Bogen et a1.16 properly commented that the infarcted and noninfarcted segments cannot be regarded as physically separate, and favor the more correct but complex mechanical model of Janz and Waldron.” However, our data support the simplified assumption that the heart can be treated as two functionally discrete components following MI without substantive error. One component is comprised of normal tissue with normal contractility and compliance, while the other component is infarcted tissue with no contractility and abnormal compliance. While the diseased myocardium may, and probably does, impose a mechanical disadvantage on the adjacent healthy myocardium, some geometric change may be imposed by the contractile or hypercontractile adjacent perfused muscle on the inert segments. Since this interface is a near maximal dimension, relatively small shape changes may generate significant compensating volume changes. Further studies will be required to establish the temporal changes in compliance of the myocardium during the early stages of AMI, as established between the ratio of
334
Beeder et al.
contractile and noncontractile myocardium and and to establish a valid ventricular pressures,lg diastolic mode1.16 Accuracy of noninvasive techniques in evaluating segmental to global LV function. Our findings confirm the observations of prior contrast ventriculographic studies, which established a linear relationship between regional myocardial performance and EF, when both are evaluated by the same technique.“-” These studies used contrast ventriculography to calculate both the proportion of akinetic LV silhouette length and EF, and reported linear relationships ranging from r = - 0.89 to r = - 0.97. Our correlation of 0.94 between %NCM by 2DE and EF by MGES falls within this range. Due to inherent biologic variability, as well as variability inherent in using a separate technique to determine EF, our correlation is as strong as can be reasonably expected. Compared to other techniques for calculating EF by 2DE, %NCM demonstrates a correlation at least as strong. as, or slightly better than, these other techniques.‘Y-“’ Clinical implications of SDE-determined %NCM in CAD. In addition to validating the physiologic constructs involved in the theoretical model,;’ the technique ,of %NCM may have significant clinical applications. The noninvasive nature and ease of application of 2DE may allow improved evaluation and reevaluation of patients with CAD, especially in AMI. 2DE does not expose the patient to the risks associated with contrast ventriculography or the radiation involved in MGES. Thus acutely ill patients can be rapidly reevaluated with no known risk at the bedside. In addition to serial applications in patients with AMI, %NCM can be readily applied to patients with chronic contraction abnormalities which represent the majority of CAD patients. The authors thank Lawrence R. O’Connor, M.D., James S. Forrester, M.D., Daniel S. Berman, M.D., and Samuel Meerbaum, Ph.D., for their expert assistance. We also express our gratitude to JoAnn Prause, MS., for assistance with statistical analysis, to Patricia Allen for editorial assistance, to Joye Nunn for secretarial assistance, to Lane Laforteza for graphics, and to Patricia Edwards for photography. REFERENCES
1. 2.
Tennant R, Wiggers C: The effect of coronary occlusion on myocardiil contraction. Am J Physiol 112:351, 1935. Corya BC, Rasmussen S, Knoebel SB, Fiegenbaum H: Echocardiography in acute myocardial infarction. Am J Cardiol38:l. 1975.
American
3. 4.
5.
6.
7.
8.
9.
10.
11.
12.
13. 14. 15.
16.
17.
18.
19.
20.
September, 1991 Heart Journal
Kerber RE, Abboud FM: Echocardiographic detection of regional myocardial infarction. Circulation 47:997, 1973. Theroux P, Franklin D, Ross J, Kemper WS: Regional myocardial function during acute coronary occlusion and its modification by pharmacologic agents in the dog. Circ Res 35:896, 1974. Swan HJC, Forrester JS, Diamond G, Chatterjee K, Parmley WW: Hemodynamic spectrum of myocardial infarction and cardiogenic shock: A conceptual model. Circulation 45:1097, 1972. Pairolero PC, Hallerman FJ, Ellis FH: The left ventriculogram in experimental myocardial infarction. Radiologv -. 95:311,1970: Field BJ, Russel RO Jr, Dowling JT, Rackley CE: Regional left ventricular performance in the year following myocardial infarction. Circulation 46:679, 1972. Rigaud M, Rocha P, Boschat J, Farcot JC, Bardet J, Boundarias JP: Regional left ventricular function assessed by contrast angiography in acute myocardial infarction. Circulation 60:130, 1979. Weyman AE, Peskoe SM, Williams ES, Dillon JC, Feigenbaum H: Detection of left ventricular aneurysms by crosssectional echocardiography. Circulation 54:936, 1976. Kisslo JA, Robertson D, Gilbert BW, Von Ramm 0, Behar VS: A comparison of real-time, two-dimensional echocardiography and cineangiography in detecting left ventricular asynergy. Circulation 55134, 1977. Heger JJ, Weyman AE, Wann LW, Dillon JC, Feigenbaum H: Cross-sectional echocardiography in acute myocardial infarction: Detection and localization of regional left ventricular asynergy. Circulation 60:531, 1979. Maddahi J, Berman DS, Diamond GA, Shah PK, Gray RJ, Forrester JS: Evaluation of left ventricular ejection fraction and segmental wall motion by multiple-gated equilibrium cardiac blood pool scintigraphy. In Cady L, editor: Computer techniques in cardiology. New York, 1978, Marcel Dekker, Inc, p 389. Eckelman W, Richards P, Hauser W, Atkins HL: Technetium-labelled red blood cells. J Nucl Med 12:22, 1971. Budinger TF, Rollo FD: Physics and instrumentation. Progr Cardiovasc Dis 20:19, 1977. Maggs PR, Nener TPE, Ellis FH: Effect of left ventricular akinesis on cardiac performance: Experimental study using a new model. Circulation 59:1019. 1979. Bogen DK, Rabinowitz SA, Needleman A, McMahon TA, Abelmann WHL: An analysis of the mechanical disadvantage of myocardial infarction in the canine left ventricle. Circ Res 47:728. 1980. Janz RF, Waldron RJ: Predicted effect of chronic apical aneurysms on the passive stiffness of the human left ventricle. C& Res 42:255, 1978. Folland ED, Parisi AF, Moynihan PF, Jones DR, Feldman CL, Tow DE: Assessment of left ventricular ejection fraction and volumes by real-time, two-dimensional echocardiography. Circulation 60:760, 1979. Schiller NB, Acquatella HA, Ports TA, Drew D, Goerke J, Ringertz H, Silverman NH, Brundage B, Botvinick EH, Boswell R, Carlson E, Parmley WW: Left ventricular volume from paired biplane to two-dimensional echocardiography. Circulation 60~547. 1979. C&r-KW, Engler RL, Forsythe JR, Johnson AD, Gosink B: Measurement of left ventricular ejection fraction by mechanical cross-sectional echocardiography. Circulation 59:1196, 1979.
Clain Beeder, B.A., Yzhar Charuzi, M.D., Irving K. Loh, M.D., Howard Staniloff, M.D., and H. J. C. Swan, M.D., Ph.D. Los Angeles,
Segmental abnormalities of left ventricular (LV) wall motion (WM) are associated with myocardial &hernia and infarction (MI).‘-” A conceptual model of the heart in acute myocardial infarction (AMI) has been proposed by Swan et al.” which suggests a direct relationship between segmental abnormalities and global LV function. Utilizing a two-component model consisting of a noncontractile infarct area and a residual non-infarcted area with normal contractility, the authors postulate a linear reduction in ejection fraction (EF) as the noncontractile infarct area increases. Support for this model can be found in several experimental and clinical studies using contrast ventriculography.“-R However, these studies share several limitations. The area-length method used to calculate EF assumes a symmetrically contracting ventricle. Further, the same tracing was used to assesssegmental abnormalities and calculate EF. From the Division of Cardiology, Medical Center, UCLA School Funded Received
in part
by the Save-A-Heart
for publication
Reprint requests: Clain Cedars-Sinai Medical 90048.
330
Department of Medicine.
Feb. Beeder, Center,
of Medicine,
Cedars-Sinai
Foundation.
27, 1981;
accepted
Apr.
B.A., c/o Publications 8700 Beverly Blvd.,
15, 1981. Office Los
(Cardiology), Angeles, CA
Calif.
Determination of EF by multiple-gated equilibrium blood pool scintigraphy makes no geometric assumptions, and is therefore not hampered by segmental abnormalities. Two-dimensional echocardiography (2DE) has been used to identify segmental abnormalities of LVWM.“-” Due to the combination of views available by 2DE, the extent of segmental abnormalities may be better appreciated than by other techniques. Thus segmental abnormalities and EF can be more satisfactorily assessedby these independent techniques. The present study was undertaken to test the conceptual model of Swan, using these improved, independent techniques in a group of patients with segmental abnormalities of LVWM due to coronary artery disease (CAD). METHODS Patient populatbn.
We studied 27 consecutive CAD patients. Age rangedfrom 45 to 77 years, with mean ageof 62 years ( + 7.7 SD). Twenty-two patients were male and five were female. Twenty-four patients were between 1 and 3 weeks post-AMI, and the remaining three patients had angiographically demonstrated CAD with no prior MI. All patients had both LVEF determined by Tc-99m multiple-gated equilibrium blood pool scintigraphy (MGES), and a 2DE study performed on the same day. 0002-8703/81/090330
+ 05$0050/O
0 1981
The C. V. Mosby Co.
Volume Number
102 3, part 1
Twenty-five patients had 2DE studies suitable for analysis. Two patients were excluded from analysis due to incomplete visualization of the LV silhouette in the apical long-axis view. We excluded from consideration any patient with clinically evident cardiomyopathy, hemodynamically significant valvular disease, complete left bundle branch block, ventricular pacemaker, or cardiac surgery, because of potentially confounding contraction abnormalities unrelated to CAD. EF determination. MGES was performed in the following manner: 10 ml of blood were withdrawn from the patient, incubated with 20 to 25 mCi of technetium-99m stannous phosphate, and injected intravenously.“, 12 Patients were studied using a portable scintillation camera (Ohio Nuclear Series 420 Mobile) in an approximate 40-degree left anterior oblique projection with lodegree caudal tilt.” The exact projection in each case was chosen to provide the best separation between ventricles. A shielded, parallel-hole, low energy collimator was used, and a total of 200,000 counts were obtained and recorded on a mobile minicomputer (Medical Data Systems). The signal was gated to the surface ECG and divided into I4 frames for the first two thirds of the cardiac style. EF was determined by the semiautomated MUGA program (Medical Data Systems) and expressed as a percentage. Measurement of EF by this technique is highly reproducible with variation less than 6% between two separate studies performed 1 hour apart.” 2DE evaluation. 2DE studies were performed using a Toshiba SSH 10-A 78-degree phased array ultrasonograph. All studies were done with the patient rotated 30 degrees into a left lateral decubitus position. 2DE views were obtained with the patient in held normal, unforced, end-expiration. This was done to provide a stable image in which the LV did not change position relative to the echo transducer due to respiration. Studies were recorded on an NEC sA inch video cassette recorder for subsequent analysis. Each 2DE study contained two apical views from which calculations were made. The apical four-chamber view was obtained by placing the transducer at the point of maximal cardiac impulse and angling the beam toward the right shoulder. The angle of the transducer was manually adjusted to provide an image of all four cardiac chambers, as well as portions of the mitral (MV) and tricuspid valves, in which maximal chamber size was obtained for the LV and right ventricle. The apical long-axis view was obtained from the same chest position by rotating the transducer approximately 90 degrees to obtain an image of the LV and left atrium, with the LV at maximum dimensions. In this way, two orthogonal views were obtained which allowed visualization of the cardiac apex, interventricular septum, and anterior, inferior, and lateral LV walls (Fig. 1). Thus information on WM was obtained for each major LV segment. No other views were considered for quantitative analysis. Data analysis. The 2DE studies were read independently and on different occasions by two observers, blinded to the identity of the patient as well as to the
LV segmental-global
functional
relations in CAD
331
Fig. 1. Diagrammatic representation of the apical imaging planes, illustrating the orthogonal relationship between the apical long-axis and apical four-chamber views. Ao = aorta; PA = pulmonary artery; LA = left atrium; LV = left ventricle; RA = right atrium; RV = right ventricle.
results of the MGES. A transparent plastic sheet was taped over the television monitor on which the study was displayed, and the inner border of the LV was traced in both end-diastole and end-systole with a marking pen, using stop-frame images gated to the ECG. End-diastole was defined as the peak of the R wave, and end-systole as the end of the T wave. The position of the inner border was confirmed on a moving image and visually “signal averaged” over five to seven serial cardiac cycles. The innermost border which formed signals that could be visually appreciated was traced in both end-diastole and end-systole on the same plastic transparency. A tracing was made in both the apical four-chamber and apical long-axis view. In’both views, the inner LV border was traced from MV ring to MV ring. The length of the end-diastolic LV outline was .measured in each view, using a Numonics Corporation model 277 Graphics Calculator. Abnormal contraction was defined as akinesis or dyskinesis.“-” Those end-diastolic segments demonstrating normal contraction during systole were measured in each view and added together. This value was divided by the sum of the length of both end-diastolic outlines. The fraction thus created was multiplied by 100 and defined as the percentage of normally contracting myocardium (%NCM). Fig. 2 illustrates the method of calculation of %NCM. The 2DE studies were reread on all 25 patients blindly and in random order, by one observer, 2 months ater the first reading. %NCM was recalculated and a linear regression was performed between the original reading and the second reading.
332
Beeder et al.
American
------
END DIASTOLE END SYSTOLE
A4C
% NCM = TOTAL
seplember, 1981 Heart Journal
ALA A+ B+C+D END DIASTOLIC LENGTH
( A4C + ALA)
x
100
Fig. 2. Representative example demonstrating the manner of calculating the percentage of normally
contracting myocardium (%NCM) using tracings from one of the patients in this study. A, B, C, and D indicate the length of the end-diastolic outline where normal contraction occurred during systole. A4C = apical four-chamber view; ALA = apical long-axis view. of the regression line, constrained to pass through zero, was y = 0.976 x, with SEE of 3.91% and r = 0.98. %NCM-2DE to LVEF-MGES relationships. Fig. 5 illustrates the relationship between %NCM by 2DE and EF by MGES. A linear relationship between %NCM and EF was demonstrated (r = 0.94, y = 0.68 X - 0.65, SEE = 5.06%). %NCM ranged from 29% to 100% (61% mean + 20% SD), and EF ranged from 17% to 65% (42% f 15%). Effect of akinesls
O2
20 40 60 80 % NCM (OBSERVER I)
100
Fig. 3. Interobserver agreementfor independent determination of %NCM by two observers.The continuous line is the regressionline, constrained to passthrough zero. The slope of the regressionline, r and p values, and standard error of estimate (SEE) are illustrated.
RESULTS interobserver
and intraobserver
%NCM-POE
variabili-
ty. Fig. 3 demonstrates the interobserver variability of %NCM in 25 patients. The equation of the regression line, constrained to pass through zero, was y = 0.997 x, with a standard error of estimate (SEE) of 4.40% and r = 0.98. Fig. 4 illustrates the intraobserver variability of %NCM. The equation
critwia
on %NCM
to LVEF relation.
While akinesis was used as the criterion for distinguishing a normal segment from an abnormal segment, it is clear that other criteria could be employed. For this reason, four criteria were investigated. Measurements were made using the following definitions separating normal from abnormal segments: akinesis, 25% of the maximum systolic inward motion, 50% of maximum, and 75% of maximum. The relationship between %NCM and EF varied in stepwise fashion from r = 0.94 using the akinetic criterion to r = 0.92 using the 75% of maximum criterion. The best results were obtained using akinesis as the dividing line between normal and abnormal contraction. DISCUSSION Linear relation between %NCM by 2DE and LVEF by MGES. Our study demonstrates a linear relationship
between %NCM by 2DE and LVEF by MGES. Determination of %NCM by two observers of varied experience was highly reproducible. Similarly, repeat determination of %NCM by one observer, 2
Volume
102
Number
3, part 1
;
LV segmental-global
loo
y= .997x
80
r= .98 (p<,OOi) SEE= 4.40 n=25
60
0 .a
/
20 40 60 80 %NCM (READING I)
100
the regressionline, constrained to passthrough zero. The slope of the regressionline, r and p values, and standard error of estimate (SEE) are illustrated. months after the original reading, was also highly reproducible. However, a limitation of the %NCM technique is that it may only be applicable to patients with segmental contraction abnormalities such as those associated with CAD. If the myocardium cannot be readily divided into contracting and noncontracting segments, then the relationship between %NCM and EF may be altered. of NCM-MI
concept
of LV function
relations
Y= .68x- -65 ~;~45(*gDoI) = n=25 ’
in CAD.
A theoretical, two-compartmental model of cardiac performance has been proposed by Swan et al.” to define abnormalities of ventricular function in semiquantitative terms. The variables considered included the proportion of perimeter which is noncontractile, the diastolic compliance of the abnormal myocardium, and the adverse effects of additional hemodynamic burdens such as mitral regurgitation or ventricular septal rupture. The proposed pressure-volume relationship has been validated by Maggs et al.‘j in an experimental preparation. One basic hypothesis of Swan’s model” was prediction of a linear relationship between EF and the proportion of noncontractile myocardium, or percent infarction. Our %NCM measurement is the compliment of percent infarction; that is, 100% minus percent infarction equals %NCM. By substituting %NCM for percent infarction, the theoretical model predicts an EF of 67% with %NCM of 100, and an EF of 0% with a %NCM of 0. Our results with CAD produced
OV
in CAD
0 0
333
:
3 /
t
Fig. 4. Intraobserver agreement for repeated measurement of %NCM by one observer. The continuous line is
Validation
I-
9
functional
I 20
I
I
I
40 60 80 %NCM (2DE)
I 100
Fig. 5. Relationship between the percentage of normally contracting myocardium @NCM) by two-dimensional echocardiography(2DE) and left ventricular ejection fraction (LVEF) by multiple-gated equilibrium scintigraphy (MGES) in 25 CAD patients. The-equation of the line, standard deviation, and r and p values are illustrated. a relationship extremely close to that predicted by the theoretical model, and suggest that systolic stiffness of these recent infarcts was relatively uniform. Consideration of LV compliance. Bogen et a1.16 properly commented that the infarcted and noninfarcted segments cannot be regarded as physically separate, and favor the more correct but complex mechanical model of Janz and Waldron.” However, our data support the simplified assumption that the heart can be treated as two functionally discrete components following MI without substantive error. One component is comprised of normal tissue with normal contractility and compliance, while the other component is infarcted tissue with no contractility and abnormal compliance. While the diseased myocardium may, and probably does, impose a mechanical disadvantage on the adjacent healthy myocardium, some geometric change may be imposed by the contractile or hypercontractile adjacent perfused muscle on the inert segments. Since this interface is a near maximal dimension, relatively small shape changes may generate significant compensating volume changes. Further studies will be required to establish the temporal changes in compliance of the myocardium during the early stages of AMI, as established between the ratio of
334
Beeder et al.
contractile and noncontractile myocardium and and to establish a valid ventricular pressures,lg diastolic mode1.16 Accuracy of noninvasive techniques in evaluating segmental to global LV function. Our findings confirm the observations of prior contrast ventriculographic studies, which established a linear relationship between regional myocardial performance and EF, when both are evaluated by the same technique.“-” These studies used contrast ventriculography to calculate both the proportion of akinetic LV silhouette length and EF, and reported linear relationships ranging from r = - 0.89 to r = - 0.97. Our correlation of 0.94 between %NCM by 2DE and EF by MGES falls within this range. Due to inherent biologic variability, as well as variability inherent in using a separate technique to determine EF, our correlation is as strong as can be reasonably expected. Compared to other techniques for calculating EF by 2DE, %NCM demonstrates a correlation at least as strong. as, or slightly better than, these other techniques.‘Y-“’ Clinical implications of SDE-determined %NCM in CAD. In addition to validating the physiologic constructs involved in the theoretical model,;’ the technique ,of %NCM may have significant clinical applications. The noninvasive nature and ease of application of 2DE may allow improved evaluation and reevaluation of patients with CAD, especially in AMI. 2DE does not expose the patient to the risks associated with contrast ventriculography or the radiation involved in MGES. Thus acutely ill patients can be rapidly reevaluated with no known risk at the bedside. In addition to serial applications in patients with AMI, %NCM can be readily applied to patients with chronic contraction abnormalities which represent the majority of CAD patients. The authors thank Lawrence R. O’Connor, M.D., James S. Forrester, M.D., Daniel S. Berman, M.D., and Samuel Meerbaum, Ph.D., for their expert assistance. We also express our gratitude to JoAnn Prause, MS., for assistance with statistical analysis, to Patricia Allen for editorial assistance, to Joye Nunn for secretarial assistance, to Lane Laforteza for graphics, and to Patricia Edwards for photography. REFERENCES
1. 2.
Tennant R, Wiggers C: The effect of coronary occlusion on myocardiil contraction. Am J Physiol 112:351, 1935. Corya BC, Rasmussen S, Knoebel SB, Fiegenbaum H: Echocardiography in acute myocardial infarction. Am J Cardiol38:l. 1975.
American
3. 4.
5.
6.
7.
8.
9.
10.
11.
12.
13. 14. 15.
16.
17.
18.
19.
20.
September, 1991 Heart Journal
Kerber RE, Abboud FM: Echocardiographic detection of regional myocardial infarction. Circulation 47:997, 1973. Theroux P, Franklin D, Ross J, Kemper WS: Regional myocardial function during acute coronary occlusion and its modification by pharmacologic agents in the dog. Circ Res 35:896, 1974. Swan HJC, Forrester JS, Diamond G, Chatterjee K, Parmley WW: Hemodynamic spectrum of myocardial infarction and cardiogenic shock: A conceptual model. Circulation 45:1097, 1972. Pairolero PC, Hallerman FJ, Ellis FH: The left ventriculogram in experimental myocardial infarction. Radiologv -. 95:311,1970: Field BJ, Russel RO Jr, Dowling JT, Rackley CE: Regional left ventricular performance in the year following myocardial infarction. Circulation 46:679, 1972. Rigaud M, Rocha P, Boschat J, Farcot JC, Bardet J, Boundarias JP: Regional left ventricular function assessed by contrast angiography in acute myocardial infarction. Circulation 60:130, 1979. Weyman AE, Peskoe SM, Williams ES, Dillon JC, Feigenbaum H: Detection of left ventricular aneurysms by crosssectional echocardiography. Circulation 54:936, 1976. Kisslo JA, Robertson D, Gilbert BW, Von Ramm 0, Behar VS: A comparison of real-time, two-dimensional echocardiography and cineangiography in detecting left ventricular asynergy. Circulation 55134, 1977. Heger JJ, Weyman AE, Wann LW, Dillon JC, Feigenbaum H: Cross-sectional echocardiography in acute myocardial infarction: Detection and localization of regional left ventricular asynergy. Circulation 60:531, 1979. Maddahi J, Berman DS, Diamond GA, Shah PK, Gray RJ, Forrester JS: Evaluation of left ventricular ejection fraction and segmental wall motion by multiple-gated equilibrium cardiac blood pool scintigraphy. In Cady L, editor: Computer techniques in cardiology. New York, 1978, Marcel Dekker, Inc, p 389. Eckelman W, Richards P, Hauser W, Atkins HL: Technetium-labelled red blood cells. J Nucl Med 12:22, 1971. Budinger TF, Rollo FD: Physics and instrumentation. Progr Cardiovasc Dis 20:19, 1977. Maggs PR, Nener TPE, Ellis FH: Effect of left ventricular akinesis on cardiac performance: Experimental study using a new model. Circulation 59:1019. 1979. Bogen DK, Rabinowitz SA, Needleman A, McMahon TA, Abelmann WHL: An analysis of the mechanical disadvantage of myocardial infarction in the canine left ventricle. Circ Res 47:728. 1980. Janz RF, Waldron RJ: Predicted effect of chronic apical aneurysms on the passive stiffness of the human left ventricle. C& Res 42:255, 1978. Folland ED, Parisi AF, Moynihan PF, Jones DR, Feldman CL, Tow DE: Assessment of left ventricular ejection fraction and volumes by real-time, two-dimensional echocardiography. Circulation 60:760, 1979. Schiller NB, Acquatella HA, Ports TA, Drew D, Goerke J, Ringertz H, Silverman NH, Brundage B, Botvinick EH, Boswell R, Carlson E, Parmley WW: Left ventricular volume from paired biplane to two-dimensional echocardiography. Circulation 60~547. 1979. C&r-KW, Engler RL, Forsythe JR, Johnson AD, Gosink B: Measurement of left ventricular ejection fraction by mechanical cross-sectional echocardiography. Circulation 59:1196, 1979.