- Email: [email protected]
Abnormalities of early depolarization in patients with remote anterior myocardial infarction and ventricular septal hypoperfusion
Abno~~ties of Early Depolarization in Patients With Remote Anterior Myocardial Infarction and Ventricular Septal Hypoperfusion Diagnosis of Septal A&Iby BSM
Isao Kubota MD, Michiyasu Yamaki, MD, Kozue Ikeda, MD, Ichiro Yamaguchi, MD, Ichiro Tonooka, MD; Kai Tsuiki, MD, and Shoji Yasui, MD
Abstract: The authors conducted this study to find the difference in body surface isopotential maps in 46 patients with previous anterior infarction with and without septal involvement. Thajiium-201 myocardial-pe~usion imaging identified 25 patients with septal infarction (group P) and 21 without (group N). In contrast to group N, group P had a prominent minimum on the anterior chest during the early phases of the QRS. According to the results obtained, the following criteria for identifying patients with septal infarction (group P) were proposed ( 1). Criterion 1: The absolute value of the voltage of the minimum is equal to or greater than that of the maximum at 5 ms after the onset of the QRS; (2) Criterion 2: During the early portion of the QRS the voltage of the minimum reaches - 0. IO mV at the same time or earlier than the maximum reaches 0.10 mV. Both criteria had higher sensitivities (100% and lOO%), specificities (71.4% and 90.5%), and predictive accuracies (87.0% and 95.7%) than either Franklead vectorcardiograms or standard 12 -lead electrocardiograms in the study popuIation. Thus, body surface isopotential mapping is considered to be useful for the diagnosis of septal involvement in patients with previous anterior myocardial infarction. Key words: body surface isopotential map, septal infarction.
The ve~t~cular septum is where ve~t~cular activation frrst occurs in the norma heart. In body surface isopotentiai maps, this septal activation is beFrom the First Depar~emt ofInternal Medicine and D~artment of ~ab~rato~ medicines Yamasata Uni~~i~ School o~~edicine, Yamagata, Japan.
Reprint requests: Isao Kubota, MD, The First Department of Internal Medicine, Yamagata University School of Medicine, 2-2-2 Iida-Nishi, Yamagata 990-23, Japan.
307
lieved to cause a positive potential area on the anterior chest. If2 In an experimental study, Sugiyama et ak3 reported that in dogs with septat infarction made by the ligation of the left anterior descending artery, a negative potential area was observed on the anterior chest from the beginning of ventricuiar activation. This suggested to us that septal involvement in anterior infarction may be diagnosed by the careful evaluation of body surface
308
Journal of Electrocardiology
Vol. 23 No. 4 October 1990
isopotential maps at the initial stage of ventricular activation. In this study, we assessed septal infarction by thallium-20 1 myocardial-per-fusion imaging and attempted to find the difference in body surface isopotential maps between patients with and without septal involvement in previous anterior infarction. We then proposed body surface mapping criteria for the diagnosis of septal involvement in anterior infarction.
Materials and Methods Subjects Forty-six patients (39 men and 7 women), aged between 35 and 73 years (mean, 55 years), were selected for this study by satisfying the following criteria: ( 1) clinical diagnosis of myocardial infarction established by typical chest pain and serum enzyme changes; (2) significant stenosis of 70% or more of the left anterior descending artery: (3) Q and QS patterns meeting the Minneosta Code criteria ( 1982)4 in leads V 1-V5 (anterior infarction), but not in leads II, III, aVF (inferior infarction) ; (4) no other heart disease, such as congenital heart disease, myocardial disease, or valvular heart disease; and (5) QRS duration less than 100 ms with no conduction disturbances, such as right bundle branch block, left bundle branch block, or Wolff-Parkinson-White syndrome. The time from onset of myocardial i~arction to this study ranged from 2 to 22 months (mean 5.3 months). According to the findings of the thallium-201 myocardial-perfusion imaging, the patients were divided into two groups: negative septal infarction group (group N, n = 21) and positive septal infarction group (group P, n = 25). Forty normal subjects (34 men and 6 women), aged between 34 and 71 years (mean, 48 years), were studied to evaluate the normal patterns of body surface isopotential maps (group C). All had normal findings on physical exa~nation and standard Itlead electrocardiograms. None had a history of heart disease or hypertension. Thallium-201 Myocardial Perfusion Imaging Five minutes after the injection of 2 mCi of thallium-201, imaging was started in the 30” right anterior oblique projection and was continued in the
anterior, 45” left anterior oblique, and left lateral projections. All images were recorded until a preset count of 1,000 counts/cm2 was reached. An Ohio Nuclear Sigman 4 1OS gamma imager equipped with a slant-hole collimator and a 20% window centered on the 80 keV X-ray peak was used for data recording. All images were stored by an interfaced computer system (DEC-gamma 11) on a floppy disk in a 64 x 64 word matrix. In this study, we used computer-assisted data processing of the 45” left anterior oblique image to quantitate septal uptake of thallium-20 1 by the following method.5,6 The image was first smoothed using a 9point smoothing algorithm and was then corrected by means of Goris’ background subtraction technique. Nine regions of interest (ROIs) were superimposed automatically on the left ventricular myocardial image. First, an ROI with a 7 x 7 matrix was placed in the center of the myocardial image. The remaining eight ROIs were placed at 45” intervals around the center ROI. Two ROIs that seemed to be located in a valvular structure and in the left atrium were excluded (Fig. 1) . In Figure 1, the black square indicates the septal ROI. In order to estimate the septal thallium-201 uptake, an uptake index (percent) was calculated by dividing the septal ROI counts by the maximal ROI counts among all of the seven ROIs. Mean and standard deviation (SD) of the uptake index determined
Fig. 1.
Schematic representation of the thallium-201 myocardial perfusion image in the 45’ left anterior oblique view. Seven regions of interest (ROIs) are superimposed on the left ventricular myocardial wall. A black square indicates the septal ROI.
Early Depolarization
from 18 control subjects who underwent coronary arteriography because of atypical chest pain but showed no stenotic lesions were 90.9% and 7.9%, respectively. The lower limit of the uptake index was defined as the mean minus 2SD of the control v&es. Therefore, septal infarction was considered to be present if the uptake index was Iess than 75.1%. Twenty-five (54%) of the 46 patients met this criterion of septal infarction (group P) and 21 did not (group N).
Group C
*
Kubota
et
al.
309
7 6 5 4 3 2 1
Body Surface lsopotential Maps Body surface mapping was performed with a body surface potential mapping system, ~F~-5~0~ unit (Chunichi Denshi Co., Nagoya). Because the procedure for data sampling and processing has been described in detail elsewhere,7 it will be reviewed here only briefly. Eighty-seven body surface leads were arranged in 13 columns of 7 leads each, except on the midaxil1ar-y lines that have 5 points, to cover the entire thoracic surface. In Figures 2 and 3, lead
Group C
7
Group P
ABCDEFGHIJKLM Front
6
Back
Fig. 3, The sites of rna~~rn~rn (open circle) and ~~irn~rn
(dosed circle) v&age when a maximum reaches 0.10 mV and a minimum reaches - 0.10 mV after the onset of QRS. 2
*.
0
l
* .*
1i
4
0 . . y: 0.
3 2
00
points Al-A5, El-E7, and 11-15 are on the right midaxillary, midsternaf and left ~daxilla~ lines, respectively, and lead points E6 and E4 are on the height of 2nd and 5th intercostal spaces, respectively. Electrocardiograms from 87 unipolar leads with Wilson’s central terminal as reference, Prank lead vectorcardiograms, and standard 12-lead electrocardiograms were sampled s~u~taneously. The flat portion of the FQ segment was chosen as the baseline. The onset of QRS was determined from the superimposed Prank X, Y, and Z leads and the spatial magnitude, The following items were examined in each subject and were compared among the three groups {groups C, N, and P) to identify the characteristic finding for group P.
08, a oa 00 D
1 0
00
ABCDEFGHI Front
40
LI
0
JKLM Back
maximum (open circle) and ~nirn~ (closed circle) at 5 ms after the onset of the QRS in each group. See text for details. I%& 2. The sites of
1, Maximum and minimum voltages at 5, IO, 15, and 20 ms after the onset of the QRS. 2. Sites of maximum and minimum voltages at 5, 10, 15, and 20 ms after the onset of the QRS. 3. Durations until a maximum (or minimum) reaches 0.10 (-O.lO), 0.15 (-0.15) and 0.20 ( - 0.20) mV after the onset of the QRS.
310
Journal
of Electrocardiology
Vol. 23 No. 4 October
1990
4. Sites of maximum and minimum voltages when a maximum (or minimum) reaches 0.10 (-O.lO), 0.15 (-0,15) and 0.20 (-0.20) mV after the onset of the QRS.
group. In group C, the maximum was significantly greater than the absolute value of the minimum from 5 to 20 ms (p < 0.01). In contrast, the absolute value of the minimum in group P was significantly greater than the maximum from 5 to 20 ms (p < 0.01). In group N, a significant difference in voltage between the maximum and the absolute value of the minimum was observed only at 5 ms. At this time, as in group C, the maximum was significantly greater than the absolute value of the minimum (p < 0,05). In group C, almost all of the maxima were located in a limited region of the middle to left anterior chest and most of the minima were located in the left lateral chest and the back 5-20 ms after the onset of the QRS. The difference in sites of maxima and minima between group N and group P was most clearly identified at 5 ms after the onset of the QRS. Figure 2 shows the sites of maximum voltage (open circles) and minimum voltages (cldsed circles) at 5 ms after the onset of the QRS. At this time, the distribution pattern of maxima and minima in group N resembled that of group C: the maxima were mainly located in the left anterior chest and the minima were mainly in the left lateral chest and the back. On the other hand, in group P the minima were localized in the region of the middle to left anterior chest where the maxima were situated in group C, and the maxima were scattered around the region. Table 2 shows the mean duration in milliseconds from the onset of the QRS to when the maximum (minimum) voltage reaches 0.10 (-O.lO), 0.15 (- 0.15), and 0.20 (-0.20) mV for each group. In group C, the maximum reaches 0.10,0.15, and 0.20 mV signi~cantly earlier than when the minimum reaches - 0.10, - 0.15, and - 0.20 mV, respectively (p < 0.0 I). In group P, in contrast, the minimum reaches - 0.10, - 0.15, and - 0.20 significantly earlier than when the maximum reaches 0 . 10, 0.15, and 0.20 mV, respectively (p < 0.01). In group N, significant difference between the time from the onset of the QRS until the maximum and minimum reach +O.lO, r0.15, and kO.20 mV was observed only at 5 0.10 mV. As in group C, the maximum reaches 0. IO mV significantly earlier than when the minimum reaches -0.10 mV (p < 0.05).
Standard 124ead Electrocardiograms and Frank Lead Vectorcardiograms Septal infarction was considered to be present if both the 10 ms and the 20 ms QRS vectors were situated in the left or right posterior quadrant in the horizontal plane of the Frank lead vectorcardiogram,’ or if there was no R or initial R wave meeting the Minnesota Code criteria4 in leads Vl and V2 in the standard 12-lead electrocardiogram. Statistical Analysis Quantitative data were expressed as mean + SD. Statistical difference was examined by the paired ftest, and p C 0.05 was considered significant. The sensitivity, specificity, and predictive accuracy of the body surface isopotential map, Frank lead vectorcardiogram, and standard 12-lead electrocardiogram for detecting septal infarction were calculated. Sensitivity was defined as the number of true-positive tests divided by the sum of the true-positive and false-negative tests. Specificity was defined as the number of true-negative tests divided by the sum of the true-negative and false-positive tests. Predictive accuracy was defined as the number of true-positive and true-negative tests divided by the total number of tests. Results Findings of Body Surface lsopotential Maps in the Three Groups Table 1 su~arizes the mean voltages of the maximum and the absolute value of the minimum at 5, 10, I5, and 20 ms after the onset of the QRS in each
Table 1. Voltages (100 /LV) of Maximum and Absolute Vahre of Minimum at Four n&r%& After the Onset of QRS Maximum 5 ms 10 ms 15 ms 20 ms
1.92 4.13 6.34 8.69
-c k + 2
0.57 1.38 2.31 3.23
Minimum 0.70 1.17 1.40 1.86
Group P
Group N
Group C + * It f
0.20.f 0.42t 0.40* o.szt
Minimum
Maximum 1.01 1.92 2.92 4.56
+ + f 2
0.42 1.01 1.68 2.93
0.75 1.54 3.00 4.40
* p < 0.05, t p < 0.01 maximum vs. minimum in each group.
+ f + +
0.36* 0.91 1.31 2.29
Maximum 0.67 1.09 2.13 3.64
f + f ”
0.24 0.58 1.27 2.30
Minimum 1.20 2.50 4.01 6.53
+ 2 -c +
0.54.f 1.14? 1.92f 3.84t
Early Depolarization Table
2. Length
of Time (ms) for a Maximum or Minimum to Reach r0.20 mV After the Onset of the QRS
-+O.lO mV 20.15 mV 2 0.20 mV
5 0.10,
Minimum
Minimum
Maximum
2.9 f 1.1 4.6 k 1.2 5.9 t 1.4
8.6 2 3.3t 15.6 2 5.lt 19.9 + 4.8t
5.7 k 4.7 10.3 ” 5.8 13.8 t 7.0
8.4 2 3.1* 10.9 * 3.1 13.0 2 4.1
Kubota et al.
k0.15,
311
and
Group P
Group N
Group C Maximum
l
Maximum
Minimum
10.5 k 5.0 14.4 2 4.9 17.0 I?I 5.6
4.8 ? 2.0t 8.0 rt 3.5+ 10.0 2 4.5t
* p < 0.05, t p < 0.01 maximum vs. minimum in each group.
In group C, most of the maxima were located in a limited region of the middle to left anterior chest from 0.10 to 0.20 mV and most of the minima were located in the left lateral chest and the back from - 0.10 to - 0.20 mV. In groups N and P, especially in group P, the minima were mainly located in the middle to left anterior chest and the maxima tended to be located around the minima from + 0.10 to 20.20 mV. Figure 3 shows the sites of maximum and minimum voltage when a maximum reaches 0.10 mV and a minimum reaches - 0.10 mV after the onset of the QRS.
Diagnosis of Septal MI by Body Surface lsopotential Maps According to the findings stated above, we defined the following criteria for the diagnosis by body surface isopotential maps of septal involvement in patients with anterior infarction: 1. Criterion 1. The voltage of the absolute value of the minimum is equal to or greater than that of the maximum at 5 ms after the onset of QRS. 2. Criterion 2. The voltage of the minimum reaches - 0.10 mV at the same time or earlier than the maximum reaches 0.10 mV after the onset of QRS.
Table 3 lists the number of patients in each group judged to have septal infarction by body surface isopotential maps (Criteria 1 and 2), Frank lead vectorcardiograms or standard 12 -lead electrocardioTable 3. Number of Patients Judged to Have Septal Infarction by MAP, VCG, and ECG
grams. None of subjects in group C met any of the criteria. All of subjects in group P were judged to have septal infarction by both body surface isopotential map criteria. Table 4 summarizes the sensitivity, specificity, and predictive accuracy for the diagnosis of septal infarction in patients with previous anterior infarction by our criteria of body surface isopotential maps, Frank lead vectorcardiograms, and standard 12 -lead electrocardiograms. Both body surface isopotential map criteria (Criteria 1 and 2) had higher sensitivities (100% and lOO%), specificities (7 1.4% and 90.5%), and predictive accuracies (87.0% and 95.7%) than either Frank lead vectrocardiograms or standard 12lead electrocardiograms in the study population. The Frank lead vectorcardiogram had a moderate predictive accuracy (67.4%), but the standard 12-lead electrocardiogram had a poor predictive accuracy (58.7%).
Discussion This study demonstrates the difference in body surface isopotential maps between patients with previous anterior myocardial infarction with and without involvement of the ventricular septum. The characteristic of group P in which predominant minimum appears on the anterior chest early in the QRS is probably due to the loss of the anterior electromotive force of the ventricular septum. As a simple and Table 4. Sensitivity, Specificity, and Predictive Accuracy for the Diagnosis of Septal Infarction in Patients With Anterior Infarction
MAP
Group C Group N Group P
MAP
n
Criterion 1
Criterion 2
VCG
ECG
40 21 25
0 6 25
0 2 25
0 7 17
0 12 18
MAP, body surface isopotential maps; VCG, Frank lead vectorcardiograms; ECG, standard 12-lead electrocardiograms.
Sensitivity Specificity Predictive accuracy
Criterion 1
Criterion 2
VCG
ECG
100% 71.4% 87.0%
100% 90.5% 95.7%
68.0% 66.7% 67.4%
72.0% 42.9% 58.7%
MAP, body surface isopotential maps; VCG, Frank lead vectorcardiograms; ECG, standard 12-lead electrocardiograms.
312
Journal of Electrocardiology
Vol. 23 No. 4 October 1990
quantitative standard, we proposed the two criteria for the diagnosis of septal infarction. Both criteria show good sensitivities, specificities, and predictive accuracies as compared to the Frank lead vectorcardiogram and the standard 12-lead electrocardiogram in the patients of this study. Many reports have shown the clinical usefulness of body surface mapping for the assessment of myocardial infarction,9-‘g but there have been few reports about the effect of the ventricular septum in patients with anterior infarction on body surface isopotential maps. Hirai et al.” recorded the body surface isopotential maps of patients with previous anterior infarction who had no evidence of infarction on standard 12-lead electrograms and correlated the location of the minimum potential with the asynergic site as proven by left ventriculography. Although the main purpose of their study was to describe the difference between normal and infarction groups in body surface isopotential maps, they reported that the minimum during the early phases of the QRS (S- 15 ms) was located on the anterior chest in patients with septal asynergy, on the left upper anterior chest in patients with only anterolateral asynergy, and on the left anterior chest in patients with only apical asynergy. We also found that the minima in group P tended to be located in a limited region of the anterior chest 5-20 ms after the onset of the QRS in contrast to group N. However, it seems difficult to clearly differentiate the two groups by means of the sites of the minimum because a considerable number of minima in group N were also located in the same region. The diagnostic accuracy of either Frank lead vectorcardiograms or standard 12-lead eIectrocardio-
grams was poorer than that of our criteria of body surface isopotential maps. It is suggested that these methods are not sensitive enough to evaluate septal activation in the presence of anterior infarction. The limitation of this study is that the diagnostic accuracy of our criteria was investigated only in the population used as a “training set.” It would be necessary to test these diagnostic criteria on another population (a “test set”) to confirm its value. In particular, the diagnostic accuracy for patients having conduction defects, such as left anterior hemiblock or septal fascicular block, should be evaluated. In addition to septal branches of the left anterior
descending artery, the ventricular septum is also supplied by septal branches of the posterior descending artery. Therefore, involvement of the septum may occur in patients with right ventricular infarction.t9 Because right ventricular infarction is usually associated with left ventricular inferior infarction, and we
exciuded patients with electrocardiographic evidence of inferior infarction, septal infarction due to the inte~ption of blood supply from the posterior descending artery is not considered to be present in our study population. Further study will be needed to determine difference in body surface isopotential maps between patients with inferior infarction with and without involvement of the ventricular septum. We concluded from this study that among patients with previous myocardial infarction, those with septal involvement had a characteristic pattern on body surface isopotential maps during the early phases of the QRS. In contrast to Frank lead vectorcardiograms and standard 12 -lead elec~ocardiogra~, body surface isopotential maps could be a reliable method for the diagnosis of the involvement of the ventricular septum in anterior infarction.
References I.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Taccardi B: Distribution of heart potentials on the thoracic surface of normal subjects. Circ Res 12: 344,1963 Boineau JP, Spach MS, Clinton TC, Barr RC: Relationship between body surface potential and ventricular excitation in the dog. Circ Res 19:489, I966 Sugiyama S, Wada M, Sugenoya J et al: Experimental study of myocardial infarction through the use of body surface isopotential maps: ligation of the anterior descending branch of the left coronary artery. Am Heart J 93:51, 1977 Prineas RJ, Crow RS, Blackbum H: The Minnesota Code Manual of Electrocardiographic Findings: standard and procedures for measurement and classification. John Wright, Boston, 1982 Suzuki H, Watanabe Y, Kubota I et al: Quantitative evaluation of stress gallium-201 myocardial perfusion imaging for detection of the severity of coronary artery disease. Jpn Circ J 47:170, 1983 Tonooka I, Satoh S, Kanaya T et al: Alterations in myocardiai perfusion during exercise after isosorbide din&rate infusion in patients with coronary artery disease: assessment by thallium-201 scintigraphy. Am Heart J 111:525, 1986 Watanabe T, Toyama J, Toyoshima H et ai: A practical microcomputer-based mapping system for body surface, precordium, and epicardium. Comput Biomed Res 14:341, 1981 Cooksey JD, Dunn M, Massie E: Clinical vectorcardiography and electrocardiography (2nd edition). Year Book Medical Publishers, Chicago, 1977 Kubota I, Ikeda K, Yamaki M et al: Determination of the left ventricular asynergic site by QRST isointegral mapping in patients with myocardial infarction. Jpn Heart J 25:311, I984 Kubota I, Ikeda K, Kanaya T et al: Noninvasive as-
Early Depolarization
11.
12.
13.
14.
15.
sessment of left ventricular wall motion abnormalities by QRS isointegral maps in previous anterior infarction. Am Heart J 109:464, 1985 Ikeda K, Kubota I, Igarashi A et al: Detection of local abnormalities in ventricular activation sequence by body surface isochrone mapping in patients with previous myocardial infarction. Circulation 72: 80 1, 1985 Flowers NC, Horan LG, Sohi GS et al: New evidence for inferoposterior myocardial infarction on surface potential maps. Am J Cardiol 38576, 1976 Flowers NC, Horan LG, Johnson JC: Anterior infarctional changes occurring during mild and late ventricular activation detectable by surface potential mapping techniques. Circulation 54:90 1, 1976 Vincent GM, Abildskov JA, Burgess MJ et al: Diagnosis of old inferior myocardial infarction by body surface isopotential mapping. Am J Cardiol39:5 10, 1977 Ohta T, Kinoshita A, Ohsugi J et al: Correlation between body surface isopotential maps and left ven-
16.
17.
18.
19.
l
Kubota et al.
313
triculograms in patients with old inferoposterior infarction. Am Heart J 104:1262, 1982 Montague TJ, Smith ER, Spencer CA et al: Body surface electrocardiographic mapping in inferior myocardial infarction: manifestation of left and right ventricular involvement. Circulation 67:665, 1983 Tonooka I, Kubota I, Watanabe Y et al: Isointegral analysis of body surface maps for the assessment of location and size of myocardial infarction. Am ,I Cardiol 52: 1174, 1983 Hirai M, Ohta T, Kinoshita A et al: Body surface isopotential maps in old anterior myocardial infarction undetectable by la-lead electrocardiograms. Am Heart J 108975, 1984 Mike11 FL, Asinger RW, Hodges M: Functional consequences of interventricular septal involvement in right ventricular infarction: echocardiographic, clinical, and hemodynamic observations. Am Heart J 105:393, 1983
Isao Kubota MD, Michiyasu Yamaki, MD, Kozue Ikeda, MD, Ichiro Yamaguchi, MD, Ichiro Tonooka, MD; Kai Tsuiki, MD, and Shoji Yasui, MD
Abstract: The authors conducted this study to find the difference in body surface isopotential maps in 46 patients with previous anterior infarction with and without septal involvement. Thajiium-201 myocardial-pe~usion imaging identified 25 patients with septal infarction (group P) and 21 without (group N). In contrast to group N, group P had a prominent minimum on the anterior chest during the early phases of the QRS. According to the results obtained, the following criteria for identifying patients with septal infarction (group P) were proposed ( 1). Criterion 1: The absolute value of the voltage of the minimum is equal to or greater than that of the maximum at 5 ms after the onset of the QRS; (2) Criterion 2: During the early portion of the QRS the voltage of the minimum reaches - 0. IO mV at the same time or earlier than the maximum reaches 0.10 mV. Both criteria had higher sensitivities (100% and lOO%), specificities (71.4% and 90.5%), and predictive accuracies (87.0% and 95.7%) than either Franklead vectorcardiograms or standard 12 -lead electrocardiograms in the study popuIation. Thus, body surface isopotential mapping is considered to be useful for the diagnosis of septal involvement in patients with previous anterior myocardial infarction. Key words: body surface isopotential map, septal infarction.
The ve~t~cular septum is where ve~t~cular activation frrst occurs in the norma heart. In body surface isopotentiai maps, this septal activation is beFrom the First Depar~emt ofInternal Medicine and D~artment of ~ab~rato~ medicines Yamasata Uni~~i~ School o~~edicine, Yamagata, Japan.
Reprint requests: Isao Kubota, MD, The First Department of Internal Medicine, Yamagata University School of Medicine, 2-2-2 Iida-Nishi, Yamagata 990-23, Japan.
307
lieved to cause a positive potential area on the anterior chest. If2 In an experimental study, Sugiyama et ak3 reported that in dogs with septat infarction made by the ligation of the left anterior descending artery, a negative potential area was observed on the anterior chest from the beginning of ventricuiar activation. This suggested to us that septal involvement in anterior infarction may be diagnosed by the careful evaluation of body surface
308
Journal of Electrocardiology
Vol. 23 No. 4 October 1990
isopotential maps at the initial stage of ventricular activation. In this study, we assessed septal infarction by thallium-20 1 myocardial-per-fusion imaging and attempted to find the difference in body surface isopotential maps between patients with and without septal involvement in previous anterior infarction. We then proposed body surface mapping criteria for the diagnosis of septal involvement in anterior infarction.
Materials and Methods Subjects Forty-six patients (39 men and 7 women), aged between 35 and 73 years (mean, 55 years), were selected for this study by satisfying the following criteria: ( 1) clinical diagnosis of myocardial infarction established by typical chest pain and serum enzyme changes; (2) significant stenosis of 70% or more of the left anterior descending artery: (3) Q and QS patterns meeting the Minneosta Code criteria ( 1982)4 in leads V 1-V5 (anterior infarction), but not in leads II, III, aVF (inferior infarction) ; (4) no other heart disease, such as congenital heart disease, myocardial disease, or valvular heart disease; and (5) QRS duration less than 100 ms with no conduction disturbances, such as right bundle branch block, left bundle branch block, or Wolff-Parkinson-White syndrome. The time from onset of myocardial i~arction to this study ranged from 2 to 22 months (mean 5.3 months). According to the findings of the thallium-201 myocardial-perfusion imaging, the patients were divided into two groups: negative septal infarction group (group N, n = 21) and positive septal infarction group (group P, n = 25). Forty normal subjects (34 men and 6 women), aged between 34 and 71 years (mean, 48 years), were studied to evaluate the normal patterns of body surface isopotential maps (group C). All had normal findings on physical exa~nation and standard Itlead electrocardiograms. None had a history of heart disease or hypertension. Thallium-201 Myocardial Perfusion Imaging Five minutes after the injection of 2 mCi of thallium-201, imaging was started in the 30” right anterior oblique projection and was continued in the
anterior, 45” left anterior oblique, and left lateral projections. All images were recorded until a preset count of 1,000 counts/cm2 was reached. An Ohio Nuclear Sigman 4 1OS gamma imager equipped with a slant-hole collimator and a 20% window centered on the 80 keV X-ray peak was used for data recording. All images were stored by an interfaced computer system (DEC-gamma 11) on a floppy disk in a 64 x 64 word matrix. In this study, we used computer-assisted data processing of the 45” left anterior oblique image to quantitate septal uptake of thallium-20 1 by the following method.5,6 The image was first smoothed using a 9point smoothing algorithm and was then corrected by means of Goris’ background subtraction technique. Nine regions of interest (ROIs) were superimposed automatically on the left ventricular myocardial image. First, an ROI with a 7 x 7 matrix was placed in the center of the myocardial image. The remaining eight ROIs were placed at 45” intervals around the center ROI. Two ROIs that seemed to be located in a valvular structure and in the left atrium were excluded (Fig. 1) . In Figure 1, the black square indicates the septal ROI. In order to estimate the septal thallium-201 uptake, an uptake index (percent) was calculated by dividing the septal ROI counts by the maximal ROI counts among all of the seven ROIs. Mean and standard deviation (SD) of the uptake index determined
Fig. 1.
Schematic representation of the thallium-201 myocardial perfusion image in the 45’ left anterior oblique view. Seven regions of interest (ROIs) are superimposed on the left ventricular myocardial wall. A black square indicates the septal ROI.
Early Depolarization
from 18 control subjects who underwent coronary arteriography because of atypical chest pain but showed no stenotic lesions were 90.9% and 7.9%, respectively. The lower limit of the uptake index was defined as the mean minus 2SD of the control v&es. Therefore, septal infarction was considered to be present if the uptake index was Iess than 75.1%. Twenty-five (54%) of the 46 patients met this criterion of septal infarction (group P) and 21 did not (group N).
Group C
*
Kubota
et
al.
309
7 6 5 4 3 2 1
Body Surface lsopotential Maps Body surface mapping was performed with a body surface potential mapping system, ~F~-5~0~ unit (Chunichi Denshi Co., Nagoya). Because the procedure for data sampling and processing has been described in detail elsewhere,7 it will be reviewed here only briefly. Eighty-seven body surface leads were arranged in 13 columns of 7 leads each, except on the midaxil1ar-y lines that have 5 points, to cover the entire thoracic surface. In Figures 2 and 3, lead
Group C
7
Group P
ABCDEFGHIJKLM Front
6
Back
Fig. 3, The sites of rna~~rn~rn (open circle) and ~~irn~rn
(dosed circle) v&age when a maximum reaches 0.10 mV and a minimum reaches - 0.10 mV after the onset of QRS. 2
*.
0
l
* .*
1i
4
0 . . y: 0.
3 2
00
points Al-A5, El-E7, and 11-15 are on the right midaxillary, midsternaf and left ~daxilla~ lines, respectively, and lead points E6 and E4 are on the height of 2nd and 5th intercostal spaces, respectively. Electrocardiograms from 87 unipolar leads with Wilson’s central terminal as reference, Prank lead vectorcardiograms, and standard 12-lead electrocardiograms were sampled s~u~taneously. The flat portion of the FQ segment was chosen as the baseline. The onset of QRS was determined from the superimposed Prank X, Y, and Z leads and the spatial magnitude, The following items were examined in each subject and were compared among the three groups {groups C, N, and P) to identify the characteristic finding for group P.
08, a oa 00 D
1 0
00
ABCDEFGHI Front
40
LI
0
JKLM Back
maximum (open circle) and ~nirn~ (closed circle) at 5 ms after the onset of the QRS in each group. See text for details. I%& 2. The sites of
1, Maximum and minimum voltages at 5, IO, 15, and 20 ms after the onset of the QRS. 2. Sites of maximum and minimum voltages at 5, 10, 15, and 20 ms after the onset of the QRS. 3. Durations until a maximum (or minimum) reaches 0.10 (-O.lO), 0.15 (-0.15) and 0.20 ( - 0.20) mV after the onset of the QRS.
310
Journal
of Electrocardiology
Vol. 23 No. 4 October
1990
4. Sites of maximum and minimum voltages when a maximum (or minimum) reaches 0.10 (-O.lO), 0.15 (-0,15) and 0.20 (-0.20) mV after the onset of the QRS.
group. In group C, the maximum was significantly greater than the absolute value of the minimum from 5 to 20 ms (p < 0.01). In contrast, the absolute value of the minimum in group P was significantly greater than the maximum from 5 to 20 ms (p < 0.01). In group N, a significant difference in voltage between the maximum and the absolute value of the minimum was observed only at 5 ms. At this time, as in group C, the maximum was significantly greater than the absolute value of the minimum (p < 0,05). In group C, almost all of the maxima were located in a limited region of the middle to left anterior chest and most of the minima were located in the left lateral chest and the back 5-20 ms after the onset of the QRS. The difference in sites of maxima and minima between group N and group P was most clearly identified at 5 ms after the onset of the QRS. Figure 2 shows the sites of maximum voltage (open circles) and minimum voltages (cldsed circles) at 5 ms after the onset of the QRS. At this time, the distribution pattern of maxima and minima in group N resembled that of group C: the maxima were mainly located in the left anterior chest and the minima were mainly in the left lateral chest and the back. On the other hand, in group P the minima were localized in the region of the middle to left anterior chest where the maxima were situated in group C, and the maxima were scattered around the region. Table 2 shows the mean duration in milliseconds from the onset of the QRS to when the maximum (minimum) voltage reaches 0.10 (-O.lO), 0.15 (- 0.15), and 0.20 (-0.20) mV for each group. In group C, the maximum reaches 0.10,0.15, and 0.20 mV signi~cantly earlier than when the minimum reaches - 0.10, - 0.15, and - 0.20 mV, respectively (p < 0.0 I). In group P, in contrast, the minimum reaches - 0.10, - 0.15, and - 0.20 significantly earlier than when the maximum reaches 0 . 10, 0.15, and 0.20 mV, respectively (p < 0.01). In group N, significant difference between the time from the onset of the QRS until the maximum and minimum reach +O.lO, r0.15, and kO.20 mV was observed only at 5 0.10 mV. As in group C, the maximum reaches 0. IO mV significantly earlier than when the minimum reaches -0.10 mV (p < 0.05).
Standard 124ead Electrocardiograms and Frank Lead Vectorcardiograms Septal infarction was considered to be present if both the 10 ms and the 20 ms QRS vectors were situated in the left or right posterior quadrant in the horizontal plane of the Frank lead vectorcardiogram,’ or if there was no R or initial R wave meeting the Minnesota Code criteria4 in leads Vl and V2 in the standard 12-lead electrocardiogram. Statistical Analysis Quantitative data were expressed as mean + SD. Statistical difference was examined by the paired ftest, and p C 0.05 was considered significant. The sensitivity, specificity, and predictive accuracy of the body surface isopotential map, Frank lead vectorcardiogram, and standard 12-lead electrocardiogram for detecting septal infarction were calculated. Sensitivity was defined as the number of true-positive tests divided by the sum of the true-positive and false-negative tests. Specificity was defined as the number of true-negative tests divided by the sum of the true-negative and false-positive tests. Predictive accuracy was defined as the number of true-positive and true-negative tests divided by the total number of tests. Results Findings of Body Surface lsopotential Maps in the Three Groups Table 1 su~arizes the mean voltages of the maximum and the absolute value of the minimum at 5, 10, I5, and 20 ms after the onset of the QRS in each
Table 1. Voltages (100 /LV) of Maximum and Absolute Vahre of Minimum at Four n&r%& After the Onset of QRS Maximum 5 ms 10 ms 15 ms 20 ms
1.92 4.13 6.34 8.69
-c k + 2
0.57 1.38 2.31 3.23
Minimum 0.70 1.17 1.40 1.86
Group P
Group N
Group C + * It f
0.20.f 0.42t 0.40* o.szt
Minimum
Maximum 1.01 1.92 2.92 4.56
+ + f 2
0.42 1.01 1.68 2.93
0.75 1.54 3.00 4.40
* p < 0.05, t p < 0.01 maximum vs. minimum in each group.
+ f + +
0.36* 0.91 1.31 2.29
Maximum 0.67 1.09 2.13 3.64
f + f ”
0.24 0.58 1.27 2.30
Minimum 1.20 2.50 4.01 6.53
+ 2 -c +
0.54.f 1.14? 1.92f 3.84t
Early Depolarization Table
2. Length
of Time (ms) for a Maximum or Minimum to Reach r0.20 mV After the Onset of the QRS
-+O.lO mV 20.15 mV 2 0.20 mV
5 0.10,
Minimum
Minimum
Maximum
2.9 f 1.1 4.6 k 1.2 5.9 t 1.4
8.6 2 3.3t 15.6 2 5.lt 19.9 + 4.8t
5.7 k 4.7 10.3 ” 5.8 13.8 t 7.0
8.4 2 3.1* 10.9 * 3.1 13.0 2 4.1
Kubota et al.
k0.15,
311
and
Group P
Group N
Group C Maximum
l
Maximum
Minimum
10.5 k 5.0 14.4 2 4.9 17.0 I?I 5.6
4.8 ? 2.0t 8.0 rt 3.5+ 10.0 2 4.5t
* p < 0.05, t p < 0.01 maximum vs. minimum in each group.
In group C, most of the maxima were located in a limited region of the middle to left anterior chest from 0.10 to 0.20 mV and most of the minima were located in the left lateral chest and the back from - 0.10 to - 0.20 mV. In groups N and P, especially in group P, the minima were mainly located in the middle to left anterior chest and the maxima tended to be located around the minima from + 0.10 to 20.20 mV. Figure 3 shows the sites of maximum and minimum voltage when a maximum reaches 0.10 mV and a minimum reaches - 0.10 mV after the onset of the QRS.
Diagnosis of Septal MI by Body Surface lsopotential Maps According to the findings stated above, we defined the following criteria for the diagnosis by body surface isopotential maps of septal involvement in patients with anterior infarction: 1. Criterion 1. The voltage of the absolute value of the minimum is equal to or greater than that of the maximum at 5 ms after the onset of QRS. 2. Criterion 2. The voltage of the minimum reaches - 0.10 mV at the same time or earlier than the maximum reaches 0.10 mV after the onset of QRS.
Table 3 lists the number of patients in each group judged to have septal infarction by body surface isopotential maps (Criteria 1 and 2), Frank lead vectorcardiograms or standard 12 -lead electrocardioTable 3. Number of Patients Judged to Have Septal Infarction by MAP, VCG, and ECG
grams. None of subjects in group C met any of the criteria. All of subjects in group P were judged to have septal infarction by both body surface isopotential map criteria. Table 4 summarizes the sensitivity, specificity, and predictive accuracy for the diagnosis of septal infarction in patients with previous anterior infarction by our criteria of body surface isopotential maps, Frank lead vectorcardiograms, and standard 12 -lead electrocardiograms. Both body surface isopotential map criteria (Criteria 1 and 2) had higher sensitivities (100% and lOO%), specificities (7 1.4% and 90.5%), and predictive accuracies (87.0% and 95.7%) than either Frank lead vectrocardiograms or standard 12lead electrocardiograms in the study population. The Frank lead vectorcardiogram had a moderate predictive accuracy (67.4%), but the standard 12-lead electrocardiogram had a poor predictive accuracy (58.7%).
Discussion This study demonstrates the difference in body surface isopotential maps between patients with previous anterior myocardial infarction with and without involvement of the ventricular septum. The characteristic of group P in which predominant minimum appears on the anterior chest early in the QRS is probably due to the loss of the anterior electromotive force of the ventricular septum. As a simple and Table 4. Sensitivity, Specificity, and Predictive Accuracy for the Diagnosis of Septal Infarction in Patients With Anterior Infarction
MAP
Group C Group N Group P
MAP
n
Criterion 1
Criterion 2
VCG
ECG
40 21 25
0 6 25
0 2 25
0 7 17
0 12 18
MAP, body surface isopotential maps; VCG, Frank lead vectorcardiograms; ECG, standard 12-lead electrocardiograms.
Sensitivity Specificity Predictive accuracy
Criterion 1
Criterion 2
VCG
ECG
100% 71.4% 87.0%
100% 90.5% 95.7%
68.0% 66.7% 67.4%
72.0% 42.9% 58.7%
MAP, body surface isopotential maps; VCG, Frank lead vectorcardiograms; ECG, standard 12-lead electrocardiograms.
312
Journal of Electrocardiology
Vol. 23 No. 4 October 1990
quantitative standard, we proposed the two criteria for the diagnosis of septal infarction. Both criteria show good sensitivities, specificities, and predictive accuracies as compared to the Frank lead vectorcardiogram and the standard 12-lead electrocardiogram in the patients of this study. Many reports have shown the clinical usefulness of body surface mapping for the assessment of myocardial infarction,9-‘g but there have been few reports about the effect of the ventricular septum in patients with anterior infarction on body surface isopotential maps. Hirai et al.” recorded the body surface isopotential maps of patients with previous anterior infarction who had no evidence of infarction on standard 12-lead electrograms and correlated the location of the minimum potential with the asynergic site as proven by left ventriculography. Although the main purpose of their study was to describe the difference between normal and infarction groups in body surface isopotential maps, they reported that the minimum during the early phases of the QRS (S- 15 ms) was located on the anterior chest in patients with septal asynergy, on the left upper anterior chest in patients with only anterolateral asynergy, and on the left anterior chest in patients with only apical asynergy. We also found that the minima in group P tended to be located in a limited region of the anterior chest 5-20 ms after the onset of the QRS in contrast to group N. However, it seems difficult to clearly differentiate the two groups by means of the sites of the minimum because a considerable number of minima in group N were also located in the same region. The diagnostic accuracy of either Frank lead vectorcardiograms or standard 12-lead eIectrocardio-
grams was poorer than that of our criteria of body surface isopotential maps. It is suggested that these methods are not sensitive enough to evaluate septal activation in the presence of anterior infarction. The limitation of this study is that the diagnostic accuracy of our criteria was investigated only in the population used as a “training set.” It would be necessary to test these diagnostic criteria on another population (a “test set”) to confirm its value. In particular, the diagnostic accuracy for patients having conduction defects, such as left anterior hemiblock or septal fascicular block, should be evaluated. In addition to septal branches of the left anterior
descending artery, the ventricular septum is also supplied by septal branches of the posterior descending artery. Therefore, involvement of the septum may occur in patients with right ventricular infarction.t9 Because right ventricular infarction is usually associated with left ventricular inferior infarction, and we
exciuded patients with electrocardiographic evidence of inferior infarction, septal infarction due to the inte~ption of blood supply from the posterior descending artery is not considered to be present in our study population. Further study will be needed to determine difference in body surface isopotential maps between patients with inferior infarction with and without involvement of the ventricular septum. We concluded from this study that among patients with previous myocardial infarction, those with septal involvement had a characteristic pattern on body surface isopotential maps during the early phases of the QRS. In contrast to Frank lead vectorcardiograms and standard 12 -lead elec~ocardiogra~, body surface isopotential maps could be a reliable method for the diagnosis of the involvement of the ventricular septum in anterior infarction.
References I.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Taccardi B: Distribution of heart potentials on the thoracic surface of normal subjects. Circ Res 12: 344,1963 Boineau JP, Spach MS, Clinton TC, Barr RC: Relationship between body surface potential and ventricular excitation in the dog. Circ Res 19:489, I966 Sugiyama S, Wada M, Sugenoya J et al: Experimental study of myocardial infarction through the use of body surface isopotential maps: ligation of the anterior descending branch of the left coronary artery. Am Heart J 93:51, 1977 Prineas RJ, Crow RS, Blackbum H: The Minnesota Code Manual of Electrocardiographic Findings: standard and procedures for measurement and classification. John Wright, Boston, 1982 Suzuki H, Watanabe Y, Kubota I et al: Quantitative evaluation of stress gallium-201 myocardial perfusion imaging for detection of the severity of coronary artery disease. Jpn Circ J 47:170, 1983 Tonooka I, Satoh S, Kanaya T et al: Alterations in myocardiai perfusion during exercise after isosorbide din&rate infusion in patients with coronary artery disease: assessment by thallium-201 scintigraphy. Am Heart J 111:525, 1986 Watanabe T, Toyama J, Toyoshima H et ai: A practical microcomputer-based mapping system for body surface, precordium, and epicardium. Comput Biomed Res 14:341, 1981 Cooksey JD, Dunn M, Massie E: Clinical vectorcardiography and electrocardiography (2nd edition). Year Book Medical Publishers, Chicago, 1977 Kubota I, Ikeda K, Yamaki M et al: Determination of the left ventricular asynergic site by QRST isointegral mapping in patients with myocardial infarction. Jpn Heart J 25:311, I984 Kubota I, Ikeda K, Kanaya T et al: Noninvasive as-
Early Depolarization
11.
12.
13.
14.
15.
sessment of left ventricular wall motion abnormalities by QRS isointegral maps in previous anterior infarction. Am Heart J 109:464, 1985 Ikeda K, Kubota I, Igarashi A et al: Detection of local abnormalities in ventricular activation sequence by body surface isochrone mapping in patients with previous myocardial infarction. Circulation 72: 80 1, 1985 Flowers NC, Horan LG, Sohi GS et al: New evidence for inferoposterior myocardial infarction on surface potential maps. Am J Cardiol 38576, 1976 Flowers NC, Horan LG, Johnson JC: Anterior infarctional changes occurring during mild and late ventricular activation detectable by surface potential mapping techniques. Circulation 54:90 1, 1976 Vincent GM, Abildskov JA, Burgess MJ et al: Diagnosis of old inferior myocardial infarction by body surface isopotential mapping. Am J Cardiol39:5 10, 1977 Ohta T, Kinoshita A, Ohsugi J et al: Correlation between body surface isopotential maps and left ven-
16.
17.
18.
19.
l
Kubota et al.
313
triculograms in patients with old inferoposterior infarction. Am Heart J 104:1262, 1982 Montague TJ, Smith ER, Spencer CA et al: Body surface electrocardiographic mapping in inferior myocardial infarction: manifestation of left and right ventricular involvement. Circulation 67:665, 1983 Tonooka I, Kubota I, Watanabe Y et al: Isointegral analysis of body surface maps for the assessment of location and size of myocardial infarction. Am ,I Cardiol 52: 1174, 1983 Hirai M, Ohta T, Kinoshita A et al: Body surface isopotential maps in old anterior myocardial infarction undetectable by la-lead electrocardiograms. Am Heart J 108975, 1984 Mike11 FL, Asinger RW, Hodges M: Functional consequences of interventricular septal involvement in right ventricular infarction: echocardiographic, clinical, and hemodynamic observations. Am Heart J 105:393, 1983