- Email: [email protected]
Infarct expansion: Pathologic analysis of 204 patients with a single myocardial infarct
349
lACC Vol. 7, No 2 February 1986.349-54
MORPHOLOGIC STUDIES
Infarct Expansion: Pathologic Analysis of 204 Patients With a Single Myocardial Infarct JOHN S. PIROLO, BS, GROVER M. HUTCHINS, MD, G. WILLIAM MOORE, MD, PHD Baltimore, Maryland
The reasons for the marked variability in expansion of myocardial infarcts are unknown. To examine this ques• tion, the hearts in 204 patients with a single myocardial infarct, autopsied at The Johns Hopkins Hospital and studied after coronary arteriography and fixation in dis• tension, were reviewed. There were 58 (28 %) hearts with marked infarct expansion, 34 (17%) with moderate ex• pansion and 112 (55%) with no or minimal expansion. The degree of expansion was greater in larger, more transmural infarcts (p < 0.001). Infarcts with greater expansion had significantly more endocardial thrombus (p < 0.001) and endocardial fibroelastosis (p < 0.01). Larger heart weight and degree of left ventricular hy• pertrophy had a significant negative correlation with infarct expansion (p < 0.05). A markedly greater degree of expansion was noted in the 101 infarcts (50%) caused
Expansion of myocardial infarcts, a serious complication of acute infarction in humans, has been described pathologi• cally (I) and subsequently confirmed with two-dimensional echocardiography in a previous clinical investigation (2). Myocardial infarct expansion is defined as acute dilation and thinning of the necrotic ventricular wall segment unex• plained by new tissue necrosis, and has been attributed 0.3) to a progressive stretching or separation of necrotic muscle fibers within the myocardial wall. This intramural disrup• tion. occurring before the formation of more resilient scar tissue. may result in a significant alteration of the mor• phology of the left ventricle, and is sometimes marked by a distinct clinical event (4). usually 4 to 6 days after in• farction. Specific morphologic changes include an increase From the Autopsy Pathology DiVISIOn of the Department of Pathology. The Johns Hopkins Medical Institutions, Baltimore. Maryland ThiS study was supported by NatIOnal Institutes of Health Grant T35 AM-07384 from the National Institute for Arthritls, Diabetes, Digestive and Kidney DI,• eases. Grant HL-17655 from The Natlonal Heart, Lung, and Blood Institute and Grant LM-03651 from The NatIOnal Library of Medlcme, Bethesda. Maryland Manuscript received May 14, 1985: revised manuscript received August 6. 1985, accepted September 18. 1985 Address for reprints: Grover M. Hutchins. MD, Department of Pa• thology, The Johns Hopkms HospitaL Baltimore. Maryland 21205. © 1986 by the Amencan College of Cardiology
by lesions in the distribution of the left anterior descend• ing coronary artery as compared with the 57 infarcts (28%) secondary to right coronary lesions and the 46 infarcts (23%) in the distribution of the left circumflex coronary artery (p < 0.001). The results show that expansion is associated with large infarcts but is less marked in hearts with ventric• ular hypertrophy. Expansion occurs predominantly in infarcts in the left anterior descending coronary artery distribution, that is, regions of the left ventricular myo• cardium with the greatest curvature. These results sug• gest that the degree to which an infarct expands may be influenced by the preinfarction thickness of the ventric• ular wall. (J Am Coil CardioI1986;7:349-54)
in the regional radius of curvature of the expanded wall segment, an alteration in the spatial relation of the anterior and posterior papillary muscles, an increase in overall left ventricular diameter, and local wall thinning (1,2,4). Severe expansion has been associated with left ventricular aneu• rysm formation, as well as rupture of a myocardial infarct (5). Even in less severe cases, the increase in functional infarct size and of myocardial oxygen demands resulting from the morphologic alterations following expansion prob• ably contribute further to cardiac dysfunction (2). Although the pathologic and clinical consequences of infarct expansion have been described, the specific factors that govern the initiation of expansion of a given infarct, and subsequently regulate the extent to which it will pro• gress, are unknown. To determine some of the variables associated with the myocardial infarct expansion, we ex• amined the heart in 204 patients with a single myocardial infarct.
Methods Study cases. Adult patients were selected from among those listed in the autopsy files of The Johns Hopkins Hos• pital I) if they had a single myocardial infarct, defined as 0735-1097/86/$3.50
350
JACC Vol 7. No 2
PIROLO ET AL INFARCT EXPANSION
February 1986:349-54
an area of myocardial necrosis at least 3 cm in one dimension (6), documented at autopsy; 2) if the heart was available for gross and microscopic examination after postmortem coronary arteriography and fixation in distension (7); and 3) if no cardiac operation had been performed. Two hundred four patients met these criteria and were included in the study. Angiographic features of the hearts were evaluated on a semiquantitative scale (0 to 4 +), as previously described (8). A grade of 0 to 1 + was considered as no or minimal change in features, 3 + to 4 + was considered a marked change and an intermediate grade was considered a moderate change. The following features were evaluated: overall coronary artery disease, diffuseness of coronary artery dis• ease, degree of collateral development and degree of coro• nary artery tortuosity. The number of major vessels (left anterior descending, left circumflex and right coronary ar• teries) involved by atherosclerotic disease, with obstruction of more than 75% of the cross-sectional area of the vessel lumen, was also noted from postmortem arteriograms. Coro• nary artery lesions associated with myocardial infarction were studied from the arteriograms, by gross inspection after transverse sectioning, and by serial histologic sectioning to determine their age and nature (6). Morphologic study. The surface area of the left ven• tricle was determined by measuring on the arteriograms the apex to base length of the ventricle and the left ventricular outer diameter near the base. The latter measurement is taken as the average of several diameters because the cross section is not circular in many hearts. The ventricle was assumed to be half of a prolate spheroid, a figure generated by rotating an ellipse around its long axis, and the outer ventricular surface area was calculated from the formula: . Ventncular surface area
= 7r
where the eccentricity (e) =
[b 2 + (ab)
(a2 _ b2) -
(sin - I e) ], e 112
, b is the raa dius of the ventricle and a is the length of the ventricle (6). Infarct size was measured by the maximal vertical (A) and horizontal (B) dimensions. Infarct configuration was con• sidered to be approximated by an ellipse and the measured dimensions of the infarct were taken to be the axes (A and B) of the ellipse. Calculation of the total surface area of the infarct was performed using the formula:
Infarct surface area
(AB)
= 7r - - .
4
Normalization of infarct size was achieved by expressing the calculated infarct surface area as a percent of the cal• culated total left ventricular surface area. These normalized values were utilized in all correlation and regression anal• yses. No attempt was made to adjust for infarct size alter• ations occurring as a result of expansion.
Arteriograms and gross specimens were used in grading the following morphologic features on a scale of 0 to 4 + : dilation of cardiac chambers, hypertrophy of chamber walls and overall infarct expansion. Expansion of the infarct was evaluated as previously described (1); briefly, it was defined as distortion of the normal left ventricular topography, and the severity of expansion was graded according to the overall degree of distortion. A separate evaluation was made of the two major components of infarct expansion: internal (or local) deviation from the normal cavity contour in the region of the infarct (local left ventricular cavity dilation) and the degree of deviation from the normal external left ventricular silhouette (left ventricular contour break) assessed on a scale of 0 to 4+ (Fig. O. Microscopic study. Microscopic analysis was carried out using transmural myocardial tissue sections prepared with hematoxylin-eosin and Verhoeff-van Gieson elastic stains from infarcted and non infarcted regions of the left ventricle. Infarct sections included the center of the infarcted region, whereas control sections were taken from normal myocar• dium adjacent to the infarcted region. Histologic specimens were scored on the 0 to 4 + scale for both endocardial fibroelastosis and endocardial mural thrombus. Myocardial reperfusion injury was assessed for each lesion and results were expressed as a percent of the total cellular damage consisting of contraction band necrosis. Infarct transmural• ity was determined microscopically as the ratio of S (the cross-sectional thickness of the surviving myocardium in the infarct at the point of greatest gross thinning) to A (the thickness of the adjacent noninfarcted myocardial wall) (Sf A thickness ratio). All infarcts were dated microscopically and subdivided into age groups as follows: hours old «2 days of age), days old (2 to 14 days of age), weeks old (15 to 6D days of age), months old (61 to 365 days of age) and years old (>365 days of age). Data analysis. Data were typed and proofread on a Ray• theon VTl303 word processor with communications soft• ware and transmitted by dial-up or direct line to a Digital Equipment Corporation PDP-linD minicomputer with an American National Standards Institute MUMPS operating system and programming language in the Department of Laboratory Medicine of The Johns Hopkins Medical Insti• tutions. Frequency distributions, means, standard devia• tions, correlation coefficients (Pearson's r) and chi-square contingency tables were calculated. All variables that cor• related significantly with the degree of infarct expansion were entered as independent variables in a forward, stepwise multivariate linear regression (9).
Results Patient group. Of the 204 autopsy patients included in the study (Table 1), 72 (35%) were women and 70 (34%)
JACC Vol. 7. No.2 February 1986:349-54
PIROLO ET AL INFARCT EXPANSION
Figure 1. Hearts with (A) a single myocardial in• farct and marked expansion and (B) aneurysm for• mation with contour break in a previously expanded infarct. The hearts have been sectioned in an apex to base plane that passes through the outflow tract of the right ventricle (RV), aorta (AO), left atrium (LA), mitral valve and left ventricle (LV). In A, note the loss of trabecular muscle and thinning of the infarcted portion of the ventricle, including the interventricular septum. In B, there is extensive organized, organizing and fresh thrombus within the aneurysm. The contour break in the left ventricular silhouette is well seen at the lower right, where the thrombus-containing aneurysm extends under the surviving inferior left ventricular wall.
351
LV ,,;Wi.
A were black. Patients ranged in age from 22 to 89 years (mean ± SEM 65 ± I). No significant differences in infarct expansion were correlated with patient age, race or sex. The ages of the myocardial lesions ranged from several hours to older than 10 years. Of the 204 infarcts, 20 (10%) were hours old, 48 (24%) were days old, 27 (13%) were weeks old, 12 (6%) were months old and 97 (48%) were years old. Most resulted from atherosclerotic coronary le• sions, with embolic coronary lesions responsible for the myocardial infarct in 36 cases (18%). The frequency of embolic coronary lesions in each of these age groups was 3 (15%) of 20,4 (8%) of 48,5 (19%) of 27,2 (17%) of 12 and 22 (23%) of 97, respectively. Of the 204 myocardial lesions studied, 101 (50%) were in the distribution of the left anterior descending coronary artery, 57 (28%) in the distribution of the right coronary artery and 46 (23%) in the distribution of the left circumflex coronary artery. Infarct expansion and correlates. Among the 204 hearts studied, moderate to marked infarct expansion was seen in 92 (45%). Expansion occurred in 6 (30%) of 20 infarcts that were hours old, 22 (46%) of 48 that were days old, 17 (63%) of 27 that were weeks old, 6 (50%) of 12 that were months old and 41 (42%) of 97 that were years old. Mean severity of expansion ± SEM for each age group is given in Table 1. The degree of infarct expansion did not correlate with infarct age. Heart weight and hypertrophy. The degree of infarct expansion correlated negatively with both heart weight and the degree of left ventricular hypertrophy (p < 0.05). Local left ventricular cavity dilation in the infarct correlated neg• atively, and the degree of transmurality as assessed by the Sf A thickness ratio correlated directly, with the degree of left ventricular hypertrophy (p < 0.05). The degree of left ventricular hypertrophy correlated negatively with infarct size (p < C.005).
Size of infarct. The degree of expansion correlated with the size of the infarct (p < 0.00 I ). Infarcts resulting from left anterior descending coronary artery lesions were sig• nificantly larger (p < 0.001) than other infarcts and in• volved, on average, 39 ± 3% of the left ventricular surface area. In contrast, infarcts resulting from lesions in the left circumflex and right coronary arteries involved 28 ± 3 and 24 ± 2%, respectively, of the left ventricular surface area. Site of coronary lesion. The specific site of the coronary artery lesion responsible for the observed myocardial lesion was highly correlated with both the frequency and the se• verity of infarct expansion (p < 0.001). Lesions in the distribution of the left anterior descending coronary artery displayed moderate or marked expansion in 60 (59%) of 101 hearts, while lesions in the right coronary artery and left circumflex coronary artery distributions showed expan• sion in 19 (33%) of 57 and 13 (28%) of 46 hearts, respec• tively. The severity of expansion of lesions in the distri• bution of the left anterior descending coronary artery was greater, with 49 (82%) of 60 lesions having marked expan• sion. In contrast, only 7 (37%) of 19 infarcts in the right coronary artery distribution and 2 (18%) of 11 in the left circumflex coronary artery distribution showed marked ex• pansion (p < 0.001). SfA thickness ratio. A highly significant (p < 0.001) negative correlation between the degree of infarct expansion and the Sf A thickness ratio was observed in the study group. Thus, the degree of expansion of an infarct was inversely related to the cross-sectional thickness of the surviving mus• cle within the infarcted wall segment. Mean Sf A thickness ratio was 0.35 ± 0.02 for a nonexpanded infarct compared with 0.15 ± 0.02 for an infarct with moderate or marked expansion. Degree of left ventricular contour break. The degree of local left ventricular cavity dilation as well as overall ex-
352
PIROLO ET AL INFARCT EXPANSION
JACC Vol 7, No 2 February 1986.349-54
Table 1. Features of 204 Patients With a Single Myocardial Infarct Infarct Age
Cases (no.) Age (yr)* Race (W/B) Sex (M/F) Heart weight (g)* Grade LVH (0 to 4+)* ExpanSIOn (0 to 4 + )* Coronary lesion (A/E) LocatIOn (LAD/LCx/RCA) S/A ralIo* Grade EFE (0 to 4 + )* Infarct size (%)* Rupture
Hours
Day,
Weeh
Months
Years
Total
20 59 ± 3 17/3 15/5 471 ± 33 0.7 ± 0.2 1.1 ± 0.4 17/3 14/5/1 0.18 ± 0.04 0.2 ± 0.1 44 ± 5 6
48 65 ± I 35113 23/25 481 ± 16 0.8 ± 0.1 1.5 ± 0.2 44/4 22/1917 0.19 ± 0.03 0.4 ± 0.1 41 ± 3 19
27 62 ± I 17/10 18/9 498 ± 23 1.3 ± 0.2 2.2 ± 0.4 22/5 17/5/5 0.17 ± 004 1.5 ± 0.2 42 ± 6 4
12 66 ± 4 1111 7/5 545 ± 40 1.3 ± 0.3 1.8 ± 0.3 10/2 2/1/9 0.29 ± 0.03 1.3 ± 0.4 24 ± 6 I
97 67 ± I 54143 69/28 522 ± 17 1.4 ± 0.1 1.4 ± 0.2 75122 31/27/39 0.33 ± 0.02 1.5 ± 0.1 24 ± 2 I
204 65 ± I 134170 132/72 506 ± 10 1.2 ± 0.1 1.5 ± 0.1 168/36 46/57/101 0.26 ± 0.01 1.1 ± 0.1 32 ± 2 31
*Mean ± SEM. Coronary lesion is the tyPe of coronary artery occlusion, either atherosclerotic (A) or embolic (E); location of lesions is the left anterior descending (LAD), left circumflex (LCx) or right (RCA) coronary artery. Average severity of both left ventricular hypertrophy (LVH) and endocardial fibroelastosis (EFE) is based on semiquantitative grading of 0 to 4 +. S/ A ratio is the ratio of the cross-sectional thickness of surviving muscle in the infarct to the thickness of the adjacent noninvolved segment of myocardium. B = black; F = female; M = male; W = white.
pansion correlated with the degree of left ventricular contour break (p < O.DOl). The development of left ventricular contour break occurred only after regional infarct thinning, as indicated by local left ventricular cavity dilation. In 88 (96%) of 92 expanded infarcts, moderate or marked local left ventricular dilation was seen. Of these 88 infarcts, 6 (1DO%) of6 were hours old, 20 (91%) of22 were days old, 17 (100%) of 17 were weeks old, 6 (I DO%) of 6 were months old and 39 (95%) of 41 were years old. Moderate to marked left ventricular contour break was seen less frequently, oc• curring in only 7 (8%) of92 expanded infarcts (p < 0.001). Such contour break was found in 0 (0%) of 6 expanded infarcts that were hours old, 2 (9%) of 22 that were days old, 2 (12%) of 17 that were weeks old, 0 (0%) of 6 that were months old and 3 (7%) of 41 that were years old. In no case with expansion was left ventricular contour break seen in the absence of local left ventricular cavity dilation. The degree of left ventricular contour break was observed to correlate negatively with the SI A thickness ratio (p < O.DOl). Myocardial rupture. Myocardial rupture was observed in 6 (30%) of 20 infarcts that were hours old, 19 (40%) of 48 that were days old, 4 (15%) of 27 that were weeks old, I (8%) of 12 that were months old and I (1 %) of 97 that were years old. No correlation was seen between the degree of expansion and myocardial rupture. Although the degree of expansion did not correlate with myocardial rupture, a highly significant (p < O.DOl) negative correlation between myocardial rupture and the SIA thickness ratio was observed. Endocardial jibroelastosis and mural thrombus. The de• gree of infarct expansion was also directly related to the degree of both endocardial fibroelastosis (p < 0.01) and
endocardial mural thrombus (p < O.DOl). Both the fre• quency and severity of endocardial fibroelastosis increased with time in infarcts with moderate or marked expansion; and among these, endocardial fibroelastosis was not seen in hours old infarcts and its severity, graded on a semiquan• titative scale (mean ± SEM), was 0.37 ± 0.14 for days old infarcts, 1.65 ± 0.31 for weeks old infarcts, 1.25 ± 0.43 for months old infarcts and 1.89 ± 0.22 for years old infarcts. Internal left ventricular cavity dilation. In 15 (13%) of 112 hearts with no or minimal infarct expansion, moderate or marked internal left ventricular cavity dilation was seen. The 15 hearts fell into the infarct age groups as follows: 3 (15%) of 20 hours old, 4 (8%) of 48 days old, I (4%) of 27 weeks old, I (8%) of 12 months old and 6 (6%) of 97 years old. No significant difference was seen in frequency of internal left ventricular cavity dilation between any of the age groups. Thirteen (87%) of these 15 lesions resulted from fixed atherosclerotic coronary occlusion, whereas 2 (13%) were secondary to embolic coronary occlusion. The occurrence of local cavity dilation without infarct expansion was associated with the location of the coronary artery lesion responsible for the infarct. Of these 15 lesions, 10 were among the 57 (18%, p < 0.005) infarcts secondary to right coronary artery lesions, 3 were among the 101 (3%) infarcts secondary to left anterior descending coronary artery lesions and 2 were among the 46 infarcts secondary to left circum• flex coronary artery occlusion. Predictors of infarct expansion. Multivariate regression analysis showed that infarct size, S/A thickness ratio, lo• cation of the infarct in the distribution of the left anterior descending coronary artery, the degree of left ventricular
PIROLO ET AL INFARCT EXPANSION
JACC Vol. 7, No 2 February 1986 349-54
contour break and the degree of endocardial fibroelastosis were highly significant (p < 0.001) predictors of infarct expansion.
Discussion Expansion of infarcts. Previous studies (1) found myo• cardial infarct expansion to be a frequent and severe com• plication of acute myocardial infarction in humans. Infarct expansion has specifically been associated with a large, transmural myocardial infarct and myocardial aneurysm for• mation. The results of this study support these concepts. Moderate to marked infarct expansion was seen in almost half of all cases studied and was highly correlated with the development of left ventricular aneurysm, as indicated by the degree of left ventricular contour break. We also found a striking association between both the size of the myo• cardial infarct and the degree of transmurality of the myo• cardial infarct as indicated by the Sf A thickness ratio (ratio of the thickness of surviving [S] muscle in the center of the infarct to the thickness of the adjacent [A] noninvolved myocardium) and the development of infarct expansion. Infarct expansion has previously (5) been associated with rupture of the infarcted myocardial wall segment, and noted to occur most frequently within the first week after infarction (10,11). In our study, 25 (83%) of 30 cases with rupture occurred in infarcts that were hours to days old. Rupture was a relatively early event and it is possible that it is one early end point in the progression of expansion, and occurs in those infarcts that do not possess sufficient viable tissue to allow for both intramural fiber slippage during expansion and maintenance of wall integrity. The observation that infarct rupture was associated with a low Sf A thickness ratio supports this concept. Our results indicate that sequential changes in segmental wall configuration occur during infarct expansion, and that in the early phases of expansion the major change within the infarcted segment consists primarily of thinning, and that only later in the course of expansion do breaks in the left ventricular contour, suggestive of aneurysm develop• ment, occur. Our results also suggest that the degree of transmurality of the infarct influences the development of left ventricular aneurysm, with more transmural infarcts displaying a greater degree of aneurysmal change. It is pos• sible that infarct expansion may be influenced by lesion location. This finding is supported by the striking tendency for myocardial infarcts in the distribution of the left anterior descending coronary artery to display moderate to marked degrees of expansion. This agrees with another study (2), which found that anterior and anteroseptal myocardial in• farcts were at high risk for expansion. Although infarcts in the distribution of the left anterior descending coronary ar• tery were found to be significantly larger than infarcts in the distribution of either the left circumflex or right coronary
353
artery, and thus at greater risk for expansion, multivariate regression analysis showed that infarct location in the dis• tribution of the left anterior descending coronary artery was a distinct predictor of infarct expansion. Myocardial thickness protects against expansion. In examining the differences between expanded and nonex• panded myocardial infarcts, we found that both left ven• tricular hypertrophy and an increased heart weight were associated with significantly less infarct expansion. Left ventricular hypertrophy and heart weight in this study show a tendency to increase with greater age of the infarct; how• ever, the differences between groups are not significant and even those hearts with early infarcts are hypertrophied. Al• though some degree of compensatory hypertrophy under• standably develops in long survival after an infarct, a more likely interpretation of our results is that the hypertrophied heart is less likely to suffer from a larg@ infarct and the complications of infarct expansion and earlier death. The finding that left ventricular hypertrophy may mini• mize the development of infarct expansion has interesting implications as a possible explanation for both the prefer• ential association of expansion with myocardial infarcts in the distribution of the left anterior descending artery and the relative protection from expansion seen with lesions in the distribution of the right coronary artery. Increases in myocardial wall thickness occur in response to the elevation of intramural wall tension accompanying increased left ven• tricular pressure, as predicted by the Laplace pressure• tension relation (12). Accordingly, regions of the left ven• tricular myocardium with the greatest radii of curvature experience the greatest intramural tension and, in response, thicken to the greatest degree. The effect of different radii of curvature on regional myocardial thickness is easily ob• served in the normal heart, where the apical myocardial wall has the smallest radius of curvature and is also noted to have the least cross-sectional wall thickness (12-14). One would therefore expect that those wall segments in the dis• tribution of the left anterior descending coronary artery, which have relatively smaller radii of curvature (greatest degree of wall curvature), would be thinner than those myo• cardial segments in the distribution of the right coronary artery, which have relatively greater radii of curvature. These differences in the degree of normal segmental thickness may, in part, account for the increased tendency for infarct expansion observed between myocardial lesions in the dis• tribution of the left anterior descending coronary artery and lesions in the distribution of the right coronary artery. The precise mechanism by which segmental thickening may protect against infarct expansion is not immediately apparent from the results of this study. However, the ob• servation that the degree of left ventricular hypertrophy was correlated with the Sf A thickness ratio may provide an ex• planation. In the normal left ventricle, the free wall in the
354
PIROLO ET AL INFARCT EXPANSION
distribution of the left circumflex coronary artery has the greatest thickness, while the free wall in the distribution of the left anterior descending coronary artery is thinnest. Among our study group the thickness (mean ± SEM) of nonin• volved adjacent myocardium was 15.00 ± 0.44 mm in the distribution of the left circumflex coronary artery and 13.85 ± 0.31 mm in the distribution of the left anterior descending coronary artery (p < 0.05). The thickness of surviving mus• cle within the infarct for lesions in those distributions was, respectively, 4.74 ± 0.51 and3.31 ± 0.30mm(p < 0.05). For comparison, the thickness of adjacent myocardium was 13.79 ± 0.39 mm and of surviving myocardium was 3.33 ± 0.46 mm for infarcts in the distribution of the right coronary artery. Thus, in relatively thicker regions of myo• cardium in the left coronary artery distribution, infarcts tend to be less transmural. It should be noted that, although hypertrophic hearts had significantly fewer transmural in• farcts, these infarcts also involved significantly less of the total left ventricular surface area, and it is possible that some of the apparent protection from expansion in these hearts was due to smaller infarct size. Endocardial fibroelastosis and mural thrombus. The observation that the degree of endocardial fibroelastosis is directly related to the degree of infarct expansion may relate to local changes in the radius of curvature of the expanded infarct segment. Endocardial fibroelastosis has been found to develop in regions of myocardial infarction, and has been interpreted as a nonspecific cellular prolifer• ative response to increased endocardial tension (15). In ex• panded infarcts, the increase in the local radius of wall curvature and loss of myocardial support may result in in• creased endocardial tension. The correlation of endocardial mural thrombus and the degree of infarct expansion may be understood by consid• eration of the changes occurring on the endocardial surface during expansion. As the effective area of the infarct en• larges with expansion, the endocardial surface is subjected to increased stress. This probably results in microscopic endocardial injury and these injured sites serve as foci for thrombus formation. With greater degrees of expansion, more severe endocardial injury would occur and thrombus formation would be more marked. Conclusion. The results of this study support the as• sociation of infarct expansion with larger and more trans• mural myocardial lesions. Expansion was associated with infarcts in the distribution of the left anterior descending coronary artery, whereas those in the distribution of the right coronary artery seemed relatively protected from such expansion. In addition, left ventricular hypertrophy was as• sociated with less infarct expansion. A potential explanation for these observations is the concept that the differential curvature of various myocardial wall segments, occurring normally within the heart, gives rise to differences in seg-
JACC Vol 7. No.2
February 1986 349-54
mental thickness. This differential thickness serves to pro• mote differential degrees of transmurality of infarction, which in turn may be responsible for the different degrees of ex• pansion observed in myocardial infarcts in different regions of the heart. Our results suggest that severe infarction expansion is most often an event occurring in lesions in the distribution of the left anterior descending coronary artery, that is, in lesions that involve left ventricular wall segments that nor• mally have the greatest degree of curvature. In addition, our findings imply that increased left ventricular wall thickness may have a protective effect, inhibiting the development of severe infarct expansion.
References I. Hutchins OM, Bulkley BH. Infarct expansion versus extension: two different complication~ of acute myocardial infarction. Am J Cardiol 1978;41:1127-32. 2. Eaton LW, Weis~ lL, Bulkley BH, Oarrison lB, Welsfeldt ML. Re• gional cardiac dilatallon after acute myocardial infarction. N Engl 1 Med 1979;300:57-62. 3 Erlebacher lA, WeIss lL, Eaton LW, Kallman C, Weisfeldt ML, Bulkley BH. Late effects of acute mfarct dllatallon on heart size: a 2-dimensional echocardlographic study. Am J Cardiol 1982;49:1120-6. 4. Silverman Kl, Hutchin~ OM. Infarct expansion: the "second event" after acute myocardial mfarction Am Heart J 1980;100:230-8. 5. Shuster EH, Bulkley BH. Expansion of transmural myocardial in• farction: a pathophy~iologic factor in cardIac rupture. Circulation 1979:60: 1532-8. 6. Ridolfi RL, Hutchins OM. The relationshIp between coronary artery lesions and myocardial mfarcts. UlceratIOn of atherosclerotic placques precipItating coronary thrombosl~. Am Heart 1 1977;93:468-86. 7 Hutchins GM, Anaya ~A. Measurements of cardiac size, chamber volumes and valve onfices at autopsy. 10hns Hopkins Med 1 1973:133:96-106. 8. Yigonta YJ. Moore OW, Hutchin~ OM. Absence of correlation be• tween coronary arterial atherosclero~i~ and seventy or duration of diabetes mellitus of adult onset Am 1 Cardlol 1980;46:535-42. 9. Draper NR, Smith H. Multiple regression applied to analysis of var• iance problems. In: Bradley RA, Hunter lS, Kendall DO, Watson OS, eds. Applied Regression AnalysIs 2nd ed. New York: John Wiley & Sons, 1981.423-57. 10 Wei lY, Hutchm~ OM The pathogenesIs of papIllary muscle rupture compllcatmg myocardial infarction: hemorrhage accompanying con• traction band necrosi~. Lab Invest 1978;39:204-9. II. Hutchim OM. Rupture of the mterventncular ~eptum complicating myocardial infarcllon: pathologIcal analysis of 10 patients with c1m• ically diagno~ed perforations. Am Heart 1 1979;97: 165-73 12. Hutchin~ OM, Bulkley BH, Moore OW, Piasio MA, Lohr Fr. Shape of the human cardiac ventricles. Am 1 Cardiol 1978;41:646-54. 13 Orant C, Greene DO, Bunnell IL. Left ventricular enlargement and hypertrophy A clinical and anglocardiographlc study Am 1 Med 1965:39:895-904 14. Orossman W, Jones D. McLaurin LP. Wall stres~ and patterns of hypertrophy m the human left ventricle. J Clin Inve,t 1975,56:56-64. 15. Hutchins OM, Bannayan ~A. Development of endocardial fibroelas• tOSI~ following myocardial mfarction Arch Pathol 1971,91: 113-8.
lACC Vol. 7, No 2 February 1986.349-54
MORPHOLOGIC STUDIES
Infarct Expansion: Pathologic Analysis of 204 Patients With a Single Myocardial Infarct JOHN S. PIROLO, BS, GROVER M. HUTCHINS, MD, G. WILLIAM MOORE, MD, PHD Baltimore, Maryland
The reasons for the marked variability in expansion of myocardial infarcts are unknown. To examine this ques• tion, the hearts in 204 patients with a single myocardial infarct, autopsied at The Johns Hopkins Hospital and studied after coronary arteriography and fixation in dis• tension, were reviewed. There were 58 (28 %) hearts with marked infarct expansion, 34 (17%) with moderate ex• pansion and 112 (55%) with no or minimal expansion. The degree of expansion was greater in larger, more transmural infarcts (p < 0.001). Infarcts with greater expansion had significantly more endocardial thrombus (p < 0.001) and endocardial fibroelastosis (p < 0.01). Larger heart weight and degree of left ventricular hy• pertrophy had a significant negative correlation with infarct expansion (p < 0.05). A markedly greater degree of expansion was noted in the 101 infarcts (50%) caused
Expansion of myocardial infarcts, a serious complication of acute infarction in humans, has been described pathologi• cally (I) and subsequently confirmed with two-dimensional echocardiography in a previous clinical investigation (2). Myocardial infarct expansion is defined as acute dilation and thinning of the necrotic ventricular wall segment unex• plained by new tissue necrosis, and has been attributed 0.3) to a progressive stretching or separation of necrotic muscle fibers within the myocardial wall. This intramural disrup• tion. occurring before the formation of more resilient scar tissue. may result in a significant alteration of the mor• phology of the left ventricle, and is sometimes marked by a distinct clinical event (4). usually 4 to 6 days after in• farction. Specific morphologic changes include an increase From the Autopsy Pathology DiVISIOn of the Department of Pathology. The Johns Hopkins Medical Institutions, Baltimore. Maryland ThiS study was supported by NatIOnal Institutes of Health Grant T35 AM-07384 from the National Institute for Arthritls, Diabetes, Digestive and Kidney DI,• eases. Grant HL-17655 from The Natlonal Heart, Lung, and Blood Institute and Grant LM-03651 from The NatIOnal Library of Medlcme, Bethesda. Maryland Manuscript received May 14, 1985: revised manuscript received August 6. 1985, accepted September 18. 1985 Address for reprints: Grover M. Hutchins. MD, Department of Pa• thology, The Johns Hopkms HospitaL Baltimore. Maryland 21205. © 1986 by the Amencan College of Cardiology
by lesions in the distribution of the left anterior descend• ing coronary artery as compared with the 57 infarcts (28%) secondary to right coronary lesions and the 46 infarcts (23%) in the distribution of the left circumflex coronary artery (p < 0.001). The results show that expansion is associated with large infarcts but is less marked in hearts with ventric• ular hypertrophy. Expansion occurs predominantly in infarcts in the left anterior descending coronary artery distribution, that is, regions of the left ventricular myo• cardium with the greatest curvature. These results sug• gest that the degree to which an infarct expands may be influenced by the preinfarction thickness of the ventric• ular wall. (J Am Coil CardioI1986;7:349-54)
in the regional radius of curvature of the expanded wall segment, an alteration in the spatial relation of the anterior and posterior papillary muscles, an increase in overall left ventricular diameter, and local wall thinning (1,2,4). Severe expansion has been associated with left ventricular aneu• rysm formation, as well as rupture of a myocardial infarct (5). Even in less severe cases, the increase in functional infarct size and of myocardial oxygen demands resulting from the morphologic alterations following expansion prob• ably contribute further to cardiac dysfunction (2). Although the pathologic and clinical consequences of infarct expansion have been described, the specific factors that govern the initiation of expansion of a given infarct, and subsequently regulate the extent to which it will pro• gress, are unknown. To determine some of the variables associated with the myocardial infarct expansion, we ex• amined the heart in 204 patients with a single myocardial infarct.
Methods Study cases. Adult patients were selected from among those listed in the autopsy files of The Johns Hopkins Hos• pital I) if they had a single myocardial infarct, defined as 0735-1097/86/$3.50
350
JACC Vol 7. No 2
PIROLO ET AL INFARCT EXPANSION
February 1986:349-54
an area of myocardial necrosis at least 3 cm in one dimension (6), documented at autopsy; 2) if the heart was available for gross and microscopic examination after postmortem coronary arteriography and fixation in distension (7); and 3) if no cardiac operation had been performed. Two hundred four patients met these criteria and were included in the study. Angiographic features of the hearts were evaluated on a semiquantitative scale (0 to 4 +), as previously described (8). A grade of 0 to 1 + was considered as no or minimal change in features, 3 + to 4 + was considered a marked change and an intermediate grade was considered a moderate change. The following features were evaluated: overall coronary artery disease, diffuseness of coronary artery dis• ease, degree of collateral development and degree of coro• nary artery tortuosity. The number of major vessels (left anterior descending, left circumflex and right coronary ar• teries) involved by atherosclerotic disease, with obstruction of more than 75% of the cross-sectional area of the vessel lumen, was also noted from postmortem arteriograms. Coro• nary artery lesions associated with myocardial infarction were studied from the arteriograms, by gross inspection after transverse sectioning, and by serial histologic sectioning to determine their age and nature (6). Morphologic study. The surface area of the left ven• tricle was determined by measuring on the arteriograms the apex to base length of the ventricle and the left ventricular outer diameter near the base. The latter measurement is taken as the average of several diameters because the cross section is not circular in many hearts. The ventricle was assumed to be half of a prolate spheroid, a figure generated by rotating an ellipse around its long axis, and the outer ventricular surface area was calculated from the formula: . Ventncular surface area
= 7r
where the eccentricity (e) =
[b 2 + (ab)
(a2 _ b2) -
(sin - I e) ], e 112
, b is the raa dius of the ventricle and a is the length of the ventricle (6). Infarct size was measured by the maximal vertical (A) and horizontal (B) dimensions. Infarct configuration was con• sidered to be approximated by an ellipse and the measured dimensions of the infarct were taken to be the axes (A and B) of the ellipse. Calculation of the total surface area of the infarct was performed using the formula:
Infarct surface area
(AB)
= 7r - - .
4
Normalization of infarct size was achieved by expressing the calculated infarct surface area as a percent of the cal• culated total left ventricular surface area. These normalized values were utilized in all correlation and regression anal• yses. No attempt was made to adjust for infarct size alter• ations occurring as a result of expansion.
Arteriograms and gross specimens were used in grading the following morphologic features on a scale of 0 to 4 + : dilation of cardiac chambers, hypertrophy of chamber walls and overall infarct expansion. Expansion of the infarct was evaluated as previously described (1); briefly, it was defined as distortion of the normal left ventricular topography, and the severity of expansion was graded according to the overall degree of distortion. A separate evaluation was made of the two major components of infarct expansion: internal (or local) deviation from the normal cavity contour in the region of the infarct (local left ventricular cavity dilation) and the degree of deviation from the normal external left ventricular silhouette (left ventricular contour break) assessed on a scale of 0 to 4+ (Fig. O. Microscopic study. Microscopic analysis was carried out using transmural myocardial tissue sections prepared with hematoxylin-eosin and Verhoeff-van Gieson elastic stains from infarcted and non infarcted regions of the left ventricle. Infarct sections included the center of the infarcted region, whereas control sections were taken from normal myocar• dium adjacent to the infarcted region. Histologic specimens were scored on the 0 to 4 + scale for both endocardial fibroelastosis and endocardial mural thrombus. Myocardial reperfusion injury was assessed for each lesion and results were expressed as a percent of the total cellular damage consisting of contraction band necrosis. Infarct transmural• ity was determined microscopically as the ratio of S (the cross-sectional thickness of the surviving myocardium in the infarct at the point of greatest gross thinning) to A (the thickness of the adjacent noninfarcted myocardial wall) (Sf A thickness ratio). All infarcts were dated microscopically and subdivided into age groups as follows: hours old «2 days of age), days old (2 to 14 days of age), weeks old (15 to 6D days of age), months old (61 to 365 days of age) and years old (>365 days of age). Data analysis. Data were typed and proofread on a Ray• theon VTl303 word processor with communications soft• ware and transmitted by dial-up or direct line to a Digital Equipment Corporation PDP-linD minicomputer with an American National Standards Institute MUMPS operating system and programming language in the Department of Laboratory Medicine of The Johns Hopkins Medical Insti• tutions. Frequency distributions, means, standard devia• tions, correlation coefficients (Pearson's r) and chi-square contingency tables were calculated. All variables that cor• related significantly with the degree of infarct expansion were entered as independent variables in a forward, stepwise multivariate linear regression (9).
Results Patient group. Of the 204 autopsy patients included in the study (Table 1), 72 (35%) were women and 70 (34%)
JACC Vol. 7. No.2 February 1986:349-54
PIROLO ET AL INFARCT EXPANSION
Figure 1. Hearts with (A) a single myocardial in• farct and marked expansion and (B) aneurysm for• mation with contour break in a previously expanded infarct. The hearts have been sectioned in an apex to base plane that passes through the outflow tract of the right ventricle (RV), aorta (AO), left atrium (LA), mitral valve and left ventricle (LV). In A, note the loss of trabecular muscle and thinning of the infarcted portion of the ventricle, including the interventricular septum. In B, there is extensive organized, organizing and fresh thrombus within the aneurysm. The contour break in the left ventricular silhouette is well seen at the lower right, where the thrombus-containing aneurysm extends under the surviving inferior left ventricular wall.
351
LV ,,;Wi.
A were black. Patients ranged in age from 22 to 89 years (mean ± SEM 65 ± I). No significant differences in infarct expansion were correlated with patient age, race or sex. The ages of the myocardial lesions ranged from several hours to older than 10 years. Of the 204 infarcts, 20 (10%) were hours old, 48 (24%) were days old, 27 (13%) were weeks old, 12 (6%) were months old and 97 (48%) were years old. Most resulted from atherosclerotic coronary le• sions, with embolic coronary lesions responsible for the myocardial infarct in 36 cases (18%). The frequency of embolic coronary lesions in each of these age groups was 3 (15%) of 20,4 (8%) of 48,5 (19%) of 27,2 (17%) of 12 and 22 (23%) of 97, respectively. Of the 204 myocardial lesions studied, 101 (50%) were in the distribution of the left anterior descending coronary artery, 57 (28%) in the distribution of the right coronary artery and 46 (23%) in the distribution of the left circumflex coronary artery. Infarct expansion and correlates. Among the 204 hearts studied, moderate to marked infarct expansion was seen in 92 (45%). Expansion occurred in 6 (30%) of 20 infarcts that were hours old, 22 (46%) of 48 that were days old, 17 (63%) of 27 that were weeks old, 6 (50%) of 12 that were months old and 41 (42%) of 97 that were years old. Mean severity of expansion ± SEM for each age group is given in Table 1. The degree of infarct expansion did not correlate with infarct age. Heart weight and hypertrophy. The degree of infarct expansion correlated negatively with both heart weight and the degree of left ventricular hypertrophy (p < 0.05). Local left ventricular cavity dilation in the infarct correlated neg• atively, and the degree of transmurality as assessed by the Sf A thickness ratio correlated directly, with the degree of left ventricular hypertrophy (p < 0.05). The degree of left ventricular hypertrophy correlated negatively with infarct size (p < C.005).
Size of infarct. The degree of expansion correlated with the size of the infarct (p < 0.00 I ). Infarcts resulting from left anterior descending coronary artery lesions were sig• nificantly larger (p < 0.001) than other infarcts and in• volved, on average, 39 ± 3% of the left ventricular surface area. In contrast, infarcts resulting from lesions in the left circumflex and right coronary arteries involved 28 ± 3 and 24 ± 2%, respectively, of the left ventricular surface area. Site of coronary lesion. The specific site of the coronary artery lesion responsible for the observed myocardial lesion was highly correlated with both the frequency and the se• verity of infarct expansion (p < 0.001). Lesions in the distribution of the left anterior descending coronary artery displayed moderate or marked expansion in 60 (59%) of 101 hearts, while lesions in the right coronary artery and left circumflex coronary artery distributions showed expan• sion in 19 (33%) of 57 and 13 (28%) of 46 hearts, respec• tively. The severity of expansion of lesions in the distri• bution of the left anterior descending coronary artery was greater, with 49 (82%) of 60 lesions having marked expan• sion. In contrast, only 7 (37%) of 19 infarcts in the right coronary artery distribution and 2 (18%) of 11 in the left circumflex coronary artery distribution showed marked ex• pansion (p < 0.001). SfA thickness ratio. A highly significant (p < 0.001) negative correlation between the degree of infarct expansion and the Sf A thickness ratio was observed in the study group. Thus, the degree of expansion of an infarct was inversely related to the cross-sectional thickness of the surviving mus• cle within the infarcted wall segment. Mean Sf A thickness ratio was 0.35 ± 0.02 for a nonexpanded infarct compared with 0.15 ± 0.02 for an infarct with moderate or marked expansion. Degree of left ventricular contour break. The degree of local left ventricular cavity dilation as well as overall ex-
352
PIROLO ET AL INFARCT EXPANSION
JACC Vol 7, No 2 February 1986.349-54
Table 1. Features of 204 Patients With a Single Myocardial Infarct Infarct Age
Cases (no.) Age (yr)* Race (W/B) Sex (M/F) Heart weight (g)* Grade LVH (0 to 4+)* ExpanSIOn (0 to 4 + )* Coronary lesion (A/E) LocatIOn (LAD/LCx/RCA) S/A ralIo* Grade EFE (0 to 4 + )* Infarct size (%)* Rupture
Hours
Day,
Weeh
Months
Years
Total
20 59 ± 3 17/3 15/5 471 ± 33 0.7 ± 0.2 1.1 ± 0.4 17/3 14/5/1 0.18 ± 0.04 0.2 ± 0.1 44 ± 5 6
48 65 ± I 35113 23/25 481 ± 16 0.8 ± 0.1 1.5 ± 0.2 44/4 22/1917 0.19 ± 0.03 0.4 ± 0.1 41 ± 3 19
27 62 ± I 17/10 18/9 498 ± 23 1.3 ± 0.2 2.2 ± 0.4 22/5 17/5/5 0.17 ± 004 1.5 ± 0.2 42 ± 6 4
12 66 ± 4 1111 7/5 545 ± 40 1.3 ± 0.3 1.8 ± 0.3 10/2 2/1/9 0.29 ± 0.03 1.3 ± 0.4 24 ± 6 I
97 67 ± I 54143 69/28 522 ± 17 1.4 ± 0.1 1.4 ± 0.2 75122 31/27/39 0.33 ± 0.02 1.5 ± 0.1 24 ± 2 I
204 65 ± I 134170 132/72 506 ± 10 1.2 ± 0.1 1.5 ± 0.1 168/36 46/57/101 0.26 ± 0.01 1.1 ± 0.1 32 ± 2 31
*Mean ± SEM. Coronary lesion is the tyPe of coronary artery occlusion, either atherosclerotic (A) or embolic (E); location of lesions is the left anterior descending (LAD), left circumflex (LCx) or right (RCA) coronary artery. Average severity of both left ventricular hypertrophy (LVH) and endocardial fibroelastosis (EFE) is based on semiquantitative grading of 0 to 4 +. S/ A ratio is the ratio of the cross-sectional thickness of surviving muscle in the infarct to the thickness of the adjacent noninvolved segment of myocardium. B = black; F = female; M = male; W = white.
pansion correlated with the degree of left ventricular contour break (p < O.DOl). The development of left ventricular contour break occurred only after regional infarct thinning, as indicated by local left ventricular cavity dilation. In 88 (96%) of 92 expanded infarcts, moderate or marked local left ventricular dilation was seen. Of these 88 infarcts, 6 (1DO%) of6 were hours old, 20 (91%) of22 were days old, 17 (100%) of 17 were weeks old, 6 (I DO%) of 6 were months old and 39 (95%) of 41 were years old. Moderate to marked left ventricular contour break was seen less frequently, oc• curring in only 7 (8%) of92 expanded infarcts (p < 0.001). Such contour break was found in 0 (0%) of 6 expanded infarcts that were hours old, 2 (9%) of 22 that were days old, 2 (12%) of 17 that were weeks old, 0 (0%) of 6 that were months old and 3 (7%) of 41 that were years old. In no case with expansion was left ventricular contour break seen in the absence of local left ventricular cavity dilation. The degree of left ventricular contour break was observed to correlate negatively with the SI A thickness ratio (p < O.DOl). Myocardial rupture. Myocardial rupture was observed in 6 (30%) of 20 infarcts that were hours old, 19 (40%) of 48 that were days old, 4 (15%) of 27 that were weeks old, I (8%) of 12 that were months old and I (1 %) of 97 that were years old. No correlation was seen between the degree of expansion and myocardial rupture. Although the degree of expansion did not correlate with myocardial rupture, a highly significant (p < O.DOl) negative correlation between myocardial rupture and the SIA thickness ratio was observed. Endocardial jibroelastosis and mural thrombus. The de• gree of infarct expansion was also directly related to the degree of both endocardial fibroelastosis (p < 0.01) and
endocardial mural thrombus (p < O.DOl). Both the fre• quency and severity of endocardial fibroelastosis increased with time in infarcts with moderate or marked expansion; and among these, endocardial fibroelastosis was not seen in hours old infarcts and its severity, graded on a semiquan• titative scale (mean ± SEM), was 0.37 ± 0.14 for days old infarcts, 1.65 ± 0.31 for weeks old infarcts, 1.25 ± 0.43 for months old infarcts and 1.89 ± 0.22 for years old infarcts. Internal left ventricular cavity dilation. In 15 (13%) of 112 hearts with no or minimal infarct expansion, moderate or marked internal left ventricular cavity dilation was seen. The 15 hearts fell into the infarct age groups as follows: 3 (15%) of 20 hours old, 4 (8%) of 48 days old, I (4%) of 27 weeks old, I (8%) of 12 months old and 6 (6%) of 97 years old. No significant difference was seen in frequency of internal left ventricular cavity dilation between any of the age groups. Thirteen (87%) of these 15 lesions resulted from fixed atherosclerotic coronary occlusion, whereas 2 (13%) were secondary to embolic coronary occlusion. The occurrence of local cavity dilation without infarct expansion was associated with the location of the coronary artery lesion responsible for the infarct. Of these 15 lesions, 10 were among the 57 (18%, p < 0.005) infarcts secondary to right coronary artery lesions, 3 were among the 101 (3%) infarcts secondary to left anterior descending coronary artery lesions and 2 were among the 46 infarcts secondary to left circum• flex coronary artery occlusion. Predictors of infarct expansion. Multivariate regression analysis showed that infarct size, S/A thickness ratio, lo• cation of the infarct in the distribution of the left anterior descending coronary artery, the degree of left ventricular
PIROLO ET AL INFARCT EXPANSION
JACC Vol. 7, No 2 February 1986 349-54
contour break and the degree of endocardial fibroelastosis were highly significant (p < 0.001) predictors of infarct expansion.
Discussion Expansion of infarcts. Previous studies (1) found myo• cardial infarct expansion to be a frequent and severe com• plication of acute myocardial infarction in humans. Infarct expansion has specifically been associated with a large, transmural myocardial infarct and myocardial aneurysm for• mation. The results of this study support these concepts. Moderate to marked infarct expansion was seen in almost half of all cases studied and was highly correlated with the development of left ventricular aneurysm, as indicated by the degree of left ventricular contour break. We also found a striking association between both the size of the myo• cardial infarct and the degree of transmurality of the myo• cardial infarct as indicated by the Sf A thickness ratio (ratio of the thickness of surviving [S] muscle in the center of the infarct to the thickness of the adjacent [A] noninvolved myocardium) and the development of infarct expansion. Infarct expansion has previously (5) been associated with rupture of the infarcted myocardial wall segment, and noted to occur most frequently within the first week after infarction (10,11). In our study, 25 (83%) of 30 cases with rupture occurred in infarcts that were hours to days old. Rupture was a relatively early event and it is possible that it is one early end point in the progression of expansion, and occurs in those infarcts that do not possess sufficient viable tissue to allow for both intramural fiber slippage during expansion and maintenance of wall integrity. The observation that infarct rupture was associated with a low Sf A thickness ratio supports this concept. Our results indicate that sequential changes in segmental wall configuration occur during infarct expansion, and that in the early phases of expansion the major change within the infarcted segment consists primarily of thinning, and that only later in the course of expansion do breaks in the left ventricular contour, suggestive of aneurysm develop• ment, occur. Our results also suggest that the degree of transmurality of the infarct influences the development of left ventricular aneurysm, with more transmural infarcts displaying a greater degree of aneurysmal change. It is pos• sible that infarct expansion may be influenced by lesion location. This finding is supported by the striking tendency for myocardial infarcts in the distribution of the left anterior descending coronary artery to display moderate to marked degrees of expansion. This agrees with another study (2), which found that anterior and anteroseptal myocardial in• farcts were at high risk for expansion. Although infarcts in the distribution of the left anterior descending coronary ar• tery were found to be significantly larger than infarcts in the distribution of either the left circumflex or right coronary
353
artery, and thus at greater risk for expansion, multivariate regression analysis showed that infarct location in the dis• tribution of the left anterior descending coronary artery was a distinct predictor of infarct expansion. Myocardial thickness protects against expansion. In examining the differences between expanded and nonex• panded myocardial infarcts, we found that both left ven• tricular hypertrophy and an increased heart weight were associated with significantly less infarct expansion. Left ventricular hypertrophy and heart weight in this study show a tendency to increase with greater age of the infarct; how• ever, the differences between groups are not significant and even those hearts with early infarcts are hypertrophied. Al• though some degree of compensatory hypertrophy under• standably develops in long survival after an infarct, a more likely interpretation of our results is that the hypertrophied heart is less likely to suffer from a larg@ infarct and the complications of infarct expansion and earlier death. The finding that left ventricular hypertrophy may mini• mize the development of infarct expansion has interesting implications as a possible explanation for both the prefer• ential association of expansion with myocardial infarcts in the distribution of the left anterior descending artery and the relative protection from expansion seen with lesions in the distribution of the right coronary artery. Increases in myocardial wall thickness occur in response to the elevation of intramural wall tension accompanying increased left ven• tricular pressure, as predicted by the Laplace pressure• tension relation (12). Accordingly, regions of the left ven• tricular myocardium with the greatest radii of curvature experience the greatest intramural tension and, in response, thicken to the greatest degree. The effect of different radii of curvature on regional myocardial thickness is easily ob• served in the normal heart, where the apical myocardial wall has the smallest radius of curvature and is also noted to have the least cross-sectional wall thickness (12-14). One would therefore expect that those wall segments in the dis• tribution of the left anterior descending coronary artery, which have relatively smaller radii of curvature (greatest degree of wall curvature), would be thinner than those myo• cardial segments in the distribution of the right coronary artery, which have relatively greater radii of curvature. These differences in the degree of normal segmental thickness may, in part, account for the increased tendency for infarct expansion observed between myocardial lesions in the dis• tribution of the left anterior descending coronary artery and lesions in the distribution of the right coronary artery. The precise mechanism by which segmental thickening may protect against infarct expansion is not immediately apparent from the results of this study. However, the ob• servation that the degree of left ventricular hypertrophy was correlated with the Sf A thickness ratio may provide an ex• planation. In the normal left ventricle, the free wall in the
354
PIROLO ET AL INFARCT EXPANSION
distribution of the left circumflex coronary artery has the greatest thickness, while the free wall in the distribution of the left anterior descending coronary artery is thinnest. Among our study group the thickness (mean ± SEM) of nonin• volved adjacent myocardium was 15.00 ± 0.44 mm in the distribution of the left circumflex coronary artery and 13.85 ± 0.31 mm in the distribution of the left anterior descending coronary artery (p < 0.05). The thickness of surviving mus• cle within the infarct for lesions in those distributions was, respectively, 4.74 ± 0.51 and3.31 ± 0.30mm(p < 0.05). For comparison, the thickness of adjacent myocardium was 13.79 ± 0.39 mm and of surviving myocardium was 3.33 ± 0.46 mm for infarcts in the distribution of the right coronary artery. Thus, in relatively thicker regions of myo• cardium in the left coronary artery distribution, infarcts tend to be less transmural. It should be noted that, although hypertrophic hearts had significantly fewer transmural in• farcts, these infarcts also involved significantly less of the total left ventricular surface area, and it is possible that some of the apparent protection from expansion in these hearts was due to smaller infarct size. Endocardial fibroelastosis and mural thrombus. The observation that the degree of endocardial fibroelastosis is directly related to the degree of infarct expansion may relate to local changes in the radius of curvature of the expanded infarct segment. Endocardial fibroelastosis has been found to develop in regions of myocardial infarction, and has been interpreted as a nonspecific cellular prolifer• ative response to increased endocardial tension (15). In ex• panded infarcts, the increase in the local radius of wall curvature and loss of myocardial support may result in in• creased endocardial tension. The correlation of endocardial mural thrombus and the degree of infarct expansion may be understood by consid• eration of the changes occurring on the endocardial surface during expansion. As the effective area of the infarct en• larges with expansion, the endocardial surface is subjected to increased stress. This probably results in microscopic endocardial injury and these injured sites serve as foci for thrombus formation. With greater degrees of expansion, more severe endocardial injury would occur and thrombus formation would be more marked. Conclusion. The results of this study support the as• sociation of infarct expansion with larger and more trans• mural myocardial lesions. Expansion was associated with infarcts in the distribution of the left anterior descending coronary artery, whereas those in the distribution of the right coronary artery seemed relatively protected from such expansion. In addition, left ventricular hypertrophy was as• sociated with less infarct expansion. A potential explanation for these observations is the concept that the differential curvature of various myocardial wall segments, occurring normally within the heart, gives rise to differences in seg-
JACC Vol 7. No.2
February 1986 349-54
mental thickness. This differential thickness serves to pro• mote differential degrees of transmurality of infarction, which in turn may be responsible for the different degrees of ex• pansion observed in myocardial infarcts in different regions of the heart. Our results suggest that severe infarction expansion is most often an event occurring in lesions in the distribution of the left anterior descending coronary artery, that is, in lesions that involve left ventricular wall segments that nor• mally have the greatest degree of curvature. In addition, our findings imply that increased left ventricular wall thickness may have a protective effect, inhibiting the development of severe infarct expansion.
References I. Hutchins OM, Bulkley BH. Infarct expansion versus extension: two different complication~ of acute myocardial infarction. Am J Cardiol 1978;41:1127-32. 2. Eaton LW, Weis~ lL, Bulkley BH, Oarrison lB, Welsfeldt ML. Re• gional cardiac dilatallon after acute myocardial infarction. N Engl 1 Med 1979;300:57-62. 3 Erlebacher lA, WeIss lL, Eaton LW, Kallman C, Weisfeldt ML, Bulkley BH. Late effects of acute mfarct dllatallon on heart size: a 2-dimensional echocardlographic study. Am J Cardiol 1982;49:1120-6. 4. Silverman Kl, Hutchin~ OM. Infarct expansion: the "second event" after acute myocardial mfarction Am Heart J 1980;100:230-8. 5. Shuster EH, Bulkley BH. Expansion of transmural myocardial in• farction: a pathophy~iologic factor in cardIac rupture. Circulation 1979:60: 1532-8. 6. Ridolfi RL, Hutchins OM. The relationshIp between coronary artery lesions and myocardial mfarcts. UlceratIOn of atherosclerotic placques precipItating coronary thrombosl~. Am Heart 1 1977;93:468-86. 7 Hutchins GM, Anaya ~A. Measurements of cardiac size, chamber volumes and valve onfices at autopsy. 10hns Hopkins Med 1 1973:133:96-106. 8. Yigonta YJ. Moore OW, Hutchin~ OM. Absence of correlation be• tween coronary arterial atherosclero~i~ and seventy or duration of diabetes mellitus of adult onset Am 1 Cardlol 1980;46:535-42. 9. Draper NR, Smith H. Multiple regression applied to analysis of var• iance problems. In: Bradley RA, Hunter lS, Kendall DO, Watson OS, eds. Applied Regression AnalysIs 2nd ed. New York: John Wiley & Sons, 1981.423-57. 10 Wei lY, Hutchm~ OM The pathogenesIs of papIllary muscle rupture compllcatmg myocardial infarction: hemorrhage accompanying con• traction band necrosi~. Lab Invest 1978;39:204-9. II. Hutchim OM. Rupture of the mterventncular ~eptum complicating myocardial infarcllon: pathologIcal analysis of 10 patients with c1m• ically diagno~ed perforations. Am Heart 1 1979;97: 165-73 12. Hutchin~ OM, Bulkley BH, Moore OW, Piasio MA, Lohr Fr. Shape of the human cardiac ventricles. Am 1 Cardiol 1978;41:646-54. 13 Orant C, Greene DO, Bunnell IL. Left ventricular enlargement and hypertrophy A clinical and anglocardiographlc study Am 1 Med 1965:39:895-904 14. Orossman W, Jones D. McLaurin LP. Wall stres~ and patterns of hypertrophy m the human left ventricle. J Clin Inve,t 1975,56:56-64. 15. Hutchins OM, Bannayan ~A. Development of endocardial fibroelas• tOSI~ following myocardial mfarction Arch Pathol 1971,91: 113-8.