Clin. Cardiol. 5,51-61 (1982) 0G. Witzstrock Publishing House Inc.

Echo- and Electrocardiographic Findings in Relation to Autopsy in Myocardial Infarction K. LINDVALL, M.D., L. ERHARDT, M.D., A. S J O G R E N , M.D.

Department of Medicine, Serafimerlasarettet, Stockholm; Danderyds sjukhus, Danderyd, Sweden

Summary: Left ventricular (LV) wall motion studied by echocardiography (Echo) may give additional information of site and extension of myocardial damage in myocardial infarction (MI) compared to conventional ECG. We have therefore investigated patients with apparent poor short-term prognosis in the acute phase of an MI. Adequate autopsy data were available in 17 patients who died within a stipulated 2-month period. The left ventricle was divided into 16 segments and concomitant evaluation with ECG and Echo was possible in 164 (56%) of 272 (16 X 17) segments. The Echo parameter mean systolic wall velocity (V) was used with its percent deviation (PD-V) from that obtained in a healthy group. A reduction in PD-V by 30% was used for separating healthy from infarcted segments. This limit gives the method a sensitivity (S) of 92%, specificity (SP) 66%, and predictive accuracy (PA) of 79% with highest PA for anterior and posteroseptal segments (7847%) in contrast to inferior and lateral segments (73%). Sensitivity for single ECG leads is lower (76%) with S P and PA of 76 and 75%, respectively. PA was highest for anteroseptal and lateral segments (83%) as compared to 60% for inferoposterior segments. In 11 patients the MI was predominantly anterior which was adequately reflected by both Echo and ECG. Extension beyond this site was better evidenced by Echo than ECG. ECG was nondiagnostic in true posterior extension in contrast to Echo which was correct in all. In 6 patients the MI was

Address for reprints: K. Lindvall, M.D. Department of Medicine Danderyds sjukhus 182 88 Danderyd, Sweden Received: March 13, 198 1 Accepted: August 3 I , 198 1

predominantly inferior. Echo correctly identified this in 82% of segments and ECG in 71%. Again extensions be.yond this site were better evidenced by Echo. Extensions to the true posterior and anterolateral walls were indicated in 79-100% by Echo, while ECG indicated MI in two posterior segments only. We conclude that Echo mapping provides useful information of segmental LV wall function in MI. In comparison to ECG additional information is obtained of MI involvement of especially the basal segments as well as the apical-posteroseptal segments and also of extension beyond the conventional anterior and inferior locations.

Keywords: echocardiography, electrocardiography, autopsy, myocardial infarction, left ventricular wall motion

Introduction Several studies have been presented, comparing the location of pathological Q waves on the conventional 12-lead ECG with either left ventricular (LV) wall motion on angiography (Bodenheimer et al., 1975; Miller et al., 1974; Nieminen 1977; Williams et al., 1973) or infarction site according to autopsy (Savage et al., 1977; Woods et al., 1963). Accordingly, the site of dyssynergic wall motion at angiography agrees well with that of pathological Q waves, with the best correlations in patients with anterior myocardial infarction (MI) (Bodenheimer et al., 1975; Miller et al., 1974). Similarly, Horan et al. (1 97 1) in a postmortem study found that pathological Q waves on a conventional 12-lead ECG could correctly diagnose the site of necrosis in 79% of patients.

Clin. Cardiol. Vol. 5, January 1982

52

TABLE1 Summary of patients studied ~

Patient No.

No. of previous infarcts

Interval between investigation and death (d)

Site of the index MI

CKmax pkat/ liter

LDmax1

pkat/ liter

Sex

Age

1

F

71

1

Antero

33

30

2 3 4 5 6

M M

0

7

8

M F

70 61 75 69 83 68 87

0 23

Antero Antero Antero Antero Antero Antero Antero

23 37 18 29 26 35 33

16 67 54 51 37 50 51

F M

M

9

M

10

M

11

M F

12 13 14 15 16 17 Mean Range

M F M

M M

1

0 24 10

70 55

1 1

4 13

Antero Antero

26 31

45 45

78 54 46 78 64

1 1

17 1 26 52 16

Antero Infero Infero Infero Infero

23 8 40 7

31 13 6 53 10

0

Infero Infero

6 26 24 4-40

24 39 37 6-67

13 65 65 46-83

2

40 13.4 0-52

Echocardiographic (Echo) studies on wall motion have been compared with both ECG and autopsy findings (Heikkila and Nieminen, 1975, 1980; Lindvall, 1981 ;Nieminen and Heikkila, 1976; Rasmussen et al., 1978). Furthermore studies of acute coronary occlusion in dogs have shown reduced wall motion measured by Echo in the ischemic region (Stefan and Bing, 1972). Thus, studies of regional LV wall motion with Echo may give additional information regarding site as well as the functional effect of myocardial necrosis in the clinical situation. As only very few patients have been evaluated both by ECG and Echo, with subsequent analysis also of autopsy findings, we have studied M-mode echocardiograms in patients with severe MI and expected poor prognosis. Our purpose was to analyze the relationships between abnormal wall motion as reflected by Echo, necrosis of the myocardium as diagnosed on ECG, and findings at autopsy.

4

Terminal event

Reinfarction-cardiogenic shock Antero-apical rupture Cardiogenic shock Antero-apical rupture Sudden death Ventricular fibrillation Cardiogenic shock Reinfarction-apical rupture Cardiogenic shock Reinfarction-cardiogenic shock Cardiogenic shock Cardiogenic shock Sudden death Sudden death Reinfarction-cardiogenic shock Cardiogenic shock Sudden death

findings on auscultation or Echo suggesting any valvular abnormality. In order to obtain reliable autopsy data, only patients dying within two months from the index MI were included in the present study. Twenty-three patients fulfilled these criteria, but autopsy data were available in only 17 patients forming the study group. Salient data on these patients (12 men and 5 women) are given in Table I. Patients ranged in age from 46 to 87 years with a mean of 65 years. Seven patients died within the first 24 hours, five during hospitalization, and the remaining five after discharge. The mean interval between Echo registration and autopsy was 13 days with the range 0-52 days. Twelve patients were hospitalized for their first MI, four for their second, and one for the third. The cause of death was progressive heart failure in six patients, sudden death in four, LV wall rupture in three, and reinfarction with subsequent cardiogenic shock in three patients and ventricular fibrillation in one patient. Echocardiographic Equipment and Registration

Patients and Methods

Between November 1976 and May 1978,49 severely ill patients with coronary artery disease were examined soon after the acute phase of an MI. Apart from having no history of other cardiovascular disease, they had no

Echocardiograms were obtained with an Organon Teknika Echocardiovisor (generation 111) and recorded on a Honeywell 1856 fiberoptic line scan recorder by one of us. Two 2.25 MHz collimated transducers, focused on either 4 cm (for near field registrations) or 7.5 cm (for

K. Lindvall et al.: Echo- and electrocardiographic findings in MI

53

FIG. 1 Schematic presentation of Echo beam directions in relation to the 16 segments of the left ventricular wall. Basal segments (a) peripherally and apical segment (b) in the center. Sectors 1 and 2 represent the antero-septa1 wall; 3 and 4 the lateral wall; 5 and 6 the postero-inferior wall; and 7 and 8 the postero-septa1wall. Shaded segments are concomitantly reflected by ECG and Echo. In the panel below the ECG leads are given with their corresponding segment. The sectors marked with thick lines correspond to the conventionally registered segments.

registration of the posterior wall segments), were used. Each patient had one Echo examination performed within 5 days of the index MI with a mean delay of 2.3 days (range 0-5 days). Mean systolic wall velocity was calculated according to standard techniques from the point of maximal endocardia1 deviation in later diastole to that showing maximal inward movement in systole (Inoue et af.,1971; Kraunz and Kennedy, 1970), for each of 16 LV segments. Conventional Echo registrations were obtained parasternally according to standard techniques, and left ventricular end-systolic and end-diastolic diameters, left atrial dimension and the early closing velocity of the anterior mitral leaflet were calculated. The ECG registration positions Vz, V4, and V5 were marked on the chest with an inkpen, and were used by both Echo and ECG, to allow for optimal comparisons.

(v)

Together with a fourth position in the subxiphoid registration area, 14 out of the 16 LV segments were viewed (Fig. 1). Two segments, both basal anterolateral, were recorded with the transducer held one intercostal space above V4 and V5. From each registration position transverse and longitudinal sweep scans were performed in order to visualize and correctly classify the LV segments. Sectors 1 and 2 represent the anteroseptal, 3 and 4 the lateral, 5 and 6 the posteroinferior, and 7 and 8 the posteroseptal wall (Fig. 1). The anterior septum and the lateral aspect of the postero-diaphragmatic wall are registered from position VZ,also used for conventional ECG registrations (Segment la, 1b, 5a, and 5b) (Feigenbaum et af., 1976). Registrations from positions V4 and V5, perpendicular to the anterior heart wall, reflect the apical portions of the anterior LV wall (Rasmussen er af., 1978). This could be established by letting the

K. Lindvall et a/.: Echo- and electrocardiographic findings in MI TABLEI I Correct classified segments (CCS) by Echo with reference to autopsy Patient no. I 2 3 4

5 h 7 8 9 10

II 12 13 14

15 16 17

I n total

ccs

ICS

%

Yo

69 81 56 50 75 63 75 69 69 94 75 69 75 88 63 63 69 70.8

19 0 25 31 13 19 19 13 13 6 19 6 25 0 25 31 12 16.2

55

Echo, ECG, and autopsy. Accordingly the ECG reflects two of the four anteroseptal wall segments ( 1 a, 2b), three of the four lateral (3a, 3b, 4a) posteroinferior (5a, 5b, 6b), and posteroseptal (7b, 8a, 8b) segments. The criteria by Kramer et al. (1 978) for posterior MI ( R > 0.04 s or R/S ratio > 1 in V 2 ) were applied for segments 5a and 6a, but proved totally unsensitive for this purpose.

No. registrations possible %I 12 19 19 19 12 18

Autopsy

6 18 18 0 6 25 0 12 12 6 19

In 12 patients the hearts were autopsied with a transverse heart slice technique and stained with nitro-BT (Erhardt, 1974) at the department of pathology, Serafimerlasarettet (Dr. K. Erhardt) and direct attention was paid to the classification of segments as used in this study. In the remaining 5 patients (No. 5, 13, 14, 15, and 17) conventional clinical autopsy reports were used as enough information was available for differentiation of old from recent MI, transmural from subendocardial MI, as well as extension of the MI with reference to the 16 segments. Segments were classified as infarcted when

13.0

Abbreviation: ICS, incorrectly classified segments

TABLEI l l Combinations of ECG-leads and six reasonably located LV segments. AM1 distribution is evaluated by percent correct classified segments (CCS), predictive accuracy (PA), and sensitivity (S),and presented as CCS/PA/S. For segment classification, see Fig. I . Boldface represents the most accurate combinations 1.V segments

V4R Vl

V2 V4

la

Ib

2a

59/78/56 59/631.56 a2/a9/a0 76167 J I00

59/78/58 65/88/58 82/85192 82 179J I00

i9/11 J 100 65145 J 100 59/42/100

VS

2b

3b

3a

4a

4b

351-J54J - J -

h5js4ja8 591-J 13 41 /25/13 59/57 /5Q 47118/33 471331 I3 47/60/55

82/80/86

aa/a5/92 82/79/79

V7

JVL

S ~ J ~ ~ J I O O8 2 / 4 5 / 8 5 aajn~a5 59/36/100 61 1-1291 I7J 14 35/21 121 66 J S 0 J S O

‘iVF

ail-/591-J-

I

82/50/75

41/29/31

II

so/-/58J - J 53/44/51

Ill Non Q in I or Vs

5a

Sb

6a

LV segments 6b

la

lb

8a

8b

47/25/40 53/33/60

59/49/47 59/89/57

7 I /67/75 65/60/75 65/54/88

59/78p a 71 Jl00JS8

56/50/57

63/70 164

~

V4R VI V2

76/111 8 3

v4

Vs V7

611-J-

651-/-

aVL aVF

531-/62/57/62 65/14/33 82/60/75 7 1/44/60

351-165/57/51 53/20/29 6 5 161129 65/56/71

I II Ill Non Q in I or V S

59/57/59

65/56/7 I

76/50/75 7 1 /44/60

S9JSOJ 14 65/57/57 3 8 1 20140

50/20/67

56/40/50

56/61/6O

Clin. Cardiol. Vol. 5, January 1982

56

TABLEIV Echocardiographic data from conventional parasternal registration

Patient No. 1

2 3 4 5 6 7 8 9 10 11

12 13 14 15

16 17 MeanfSD Range Controls (n = 37) Meanf SD Range

LAD

EF-slope

LVESD

LVEDD

30 47 38 36 38 45 35 40 45 50 40 45 38 37 45 55 32

110

90 160 120

35 48 45 49 44 55 46 33 41 56 65 50 47 49 48 62 42

46 59 55 54 55 60 58 46 49 68 71 58 55 60 65 68 54

R

b

R

b

41 f 7 30-55

109f23 75-160

48f8 33-65

58f 7 46-7 1

30f5 22-40

142f31 88- I90

33f4 25-44

50f5 42-63

80 140 95 115 75 112 108 I I8 125 115

70 105 110

No. of infarcted areas hit by echobeam

2 i

1 2 1

2 1

2 1 1

2 2 1

1 2 -

Abbreviations: LAD, left atrial dimension; EF, early diastolic closing velocity of the mitral valve; LVESD, left ventricular end-systolic dimension; LVEDD, left ventricular end-diastolic dimension 0 p