Predictors of left ventricular thrombus formation and disappearance after anterior wall myocardial infarction

Transcription

1 European Heart Journal (1998) 19, Predictors of left ventricular thrombus formation and disappearance after anterior wall myocardial infarction A. N. Neskovic, J. Marinkovic, M. Bojic and A. D. Popovic Cardiovascular Research Center, Dedinje Cardiovascular Institute, Belgrade University Medical School, Belgrade, Yugoslavia Aims This study sought to determine predictors of left ventricular thrombus formation and resolution after acute anterior wall myocardial infarction. Methods and Results We have analysed clinical, echocardiographic and angiographic data in 53 consecutive patients with anterior myocardial infarction. Twodimensional and Doppler echo cardiographic examinations were performed on days 1,2, 3 and 7, after 3 and 6 weeks, and 3, 6, and 12 months following infarction. Coronary angiography was performed in 44 patients before hospital discharge. Left ventricular thrombus was detected in 30/53 patients (29/30 in the first week after infarction). Univariate analysis showed that left ventricular thrombus formation was associated with a higher initial end-systolic volume index (beta=0 04, P=O'OOI), and end-diastolic volume index (beta=0'03, P=0'03), a larger infarct perimeter (beta=0'02, P=O'OI), a lower initial ejection fraction (beta= - 0'06, P=O'OOI), a higher initial wall motion score index (beta = 1 75, P= 0'023), a higher peak creatine kinase level (beta = 3 90, P=O'OI), Killip class >1 (beta = 1'11, P=Q'003), infarct expansion (beta=0'78, P=0 04), occluded infarct-related artery (beta= - 0'87, P=0'04) and non- thrombolytic therapy (beta= - 0'76, P=0 047). According to the Cox proportional regression model, independent predictors of thrombus formation after anterior myocardial infarction were high end-systolic volume index (beta=0'06, P=O'OOI) and high peak creatine kinase level (beta=5'17, P=0 046). Thrombus disappeared in (36'7%) patients during one-year echo cardiographic follow-up. The only independent predictor of thrombus disappearance after acute myocardial infarction was the absence of apical dyskinesis 6 weeks after infarction (beta= - 1'53, P=0 045). Conclusions Our data demonstrate that the best predictor of left ventricular thrombus formation after acute anterior myocardial infarction is a high initial end-systolic volume. Thrombus resolution is more likely to occur in patients without apical dyskinesis at the end of the healing phase of infarction. (Eur Heart J 1998; 19: ) Key Words: Myocardial infarction, left ventricular thrombosis. Introduction It is well known that the incidence of left ventricular thrombosis is high in patients with acute myocardial infarction[1-31. Although the vast majority of patients with left ventricular thrombi have large anterior infarcts with depressed left ventricular systolic function, thrombi are not invariably present in such patients. Thrombi can also be found in small apical infarcts, with good global systolic function[21. Left ventricular thrombosis occurs Revision submitted 1 December 1997, and accepted 20 December Correspondence: Aleksandar D. Popovic, MD, PhD, FESC, F ACC, Cardiovascular Research Center, Dedinje Cardiovascular Institute, Milana Tepica I, Belgrade, Yugoslavia X/98/ $18.00/0 hj rarely in patients with inferior infarcts[41. These facts indicate that the process of left ventricular thrombosis is complex and that factors other than infarct size and site may have an impact on the development of left ventricular thrombosis after myocardial infarction. Two-dimensional echocardiography is a highly accurate non-invasive technique for the detection of left ventricular thrombi[ Serial echocardiographic examinations allow reliable assessment of the incidence and natural history of left ventricular thrombi. In addition, other variables which may influence thrombus formation or dissolution can be evaluated using echocardiography, such as global and regional systolic function. There are many different, but important clinical consequences of left ventricular thrombosis after myocardial infarction[8-12 1, so prediction of thrombus formation and 1998 The European Society of Cardiology

2 Predictors of L V thrombus formation and disappearance 909 Table 1 Baseline characteristics of patients enrolled in the study Total Thrombus No thrombus (n=53) (n=30) (n=23) Age (years) 55 ± 9 54± 8 55 ±9 Sex (F/M) 17/36 10/ Hypertension 19/53 13/29 6/23 Diabetes 9/53 6/29 3/23 Smoking 28/53 12/29 16/23 Q-wave infarction 47/53 28/30 19/23 Peak CK level (IU. 1-1) 1428 ± ± ± 681 * Killip class> I 19/53 16/30 3/23* Thrombolysis 27/53 11/30 16/23* Patent IRA 28/44 12/ * Multivessel CAD 18/44 5/19 13/25 Mitral regurgitation 20/53 13/30 7/23 Infarct expansion 23/51 17/28 6/23* Died 6/53 4/30 2/23 CAD=coronary artery disease, CK=creatine kinase, IRA=infarct-related artery. * P<0'05 between the groups with and without thrombus. Values for age and CK level are expressed as mean ± SD. resolution may have a significant impact on patient management. The purpose of this study was to determine predictors of left ventricular thrombus formation and resolution after acute anterior wall myocardial infarction. Methods Patients We have prospectively evaluated 53 consecutive patients with a first acute anterior wall myocardial infarction who met the following criteria: (1) age ::;; 70 years, (2) chest pain lasting > 30 min, (3) ST segment elevation ~ 2 mm at least in two anterior electrocardiographic leads, (4) transient elevation of creatine kinase and/or MB isoenzyme, and (5) first echo cardiogram performed within 24 h of the onset of pain. These patients are a subgroup of 131 patients with acute myocardial infarction reported previously[l31. Since none of the 78 patients with inferior infarcts developed left ventricular thrombus during one year follow-up, they were not included in the study. Baseline characteristics of patients enrolled in the study are shown in Table 1. Clinical assessment of patients' cardiovascular status was made according to the Killip classification[l41. Six patients enrolled in the study died during the one-year follow-up period (two patients died on day 2, and one patient each after 3 days, 3 weeks, 6 weeks, and 6 months); their data were included in the study until death. Two patients had technically inadequate images for infarct quantitation. Therapy Thrombolytic therapy (intravenous streptokinase IU over min) was administered in 27/53 (51 %) patients. The remaining patients did not receive thrombolytic therapy due to late admission after the onset of the pain or some contraindication for thrombolysis. In this study, patients were not randomized to various anti thrombotic regimens. During hospitalization, 42/53 patients received intravenous heparin (1000 IV/h) for 2 days, followed with subcutaneous heparin for 5 days, and 45/53 received aspirin ( mg daily). Oral anticoagulation was started in 28/53 patients because of poor left ventricular function and/or presence of left ventricular thrombus detected by echocardiography (coumarin, target INR=2-3). During first 3 months after infarction, 26/49 patients received oral anticoagulants and 18/49 patients received aspirin. After 3 months, 40 patients received oral aspirin throughout the echocardiographic follow-up. Echocardiograms Patients were serially evaluated by two-dimensional and Doppler echo cardiography in the following sequence: on admission (day 1), and on days 2, 3 and 7 after 3 and 6 weeks, 3 and 6 months and 1 year after infarction. All examinations were performed with an Acuson 128 machine (Mountain View, CA, U.S.A.), using a 2 5 MHz transducer, and were stored on VHS video tapes for later analysis. The diagnosis of left ventricular thrombosis was made when an echo dense mass with a margin distinct from the left ventricular wall was detected within the left ventricular cavity and was visible throughout the cardiac cycle in at least two different echocardiographic views and associated with asynergy (akinesis or dyskinesis) of the adjacent myocardium[l]. During recording, particular care was taken to obtain all possible views and to optimize depth and gain settings in order to minimize the possibility of false-positive or false-negative

3 910 A. N. Neskovic et al. readings. B-colour was also used in equivocal situations. Each echocardiographic study was interpreted independently by two echocardiographers who were blinded to patients' clinical data. A total of 393 studies were reviewed. In 384 (97'7%) studies observers were in complete agreement on the presence or absence of thrombus. In nine (2'3%) studies in which disagreement occurred, consensus was reached after reviewing the studies again. Left ventricular end-diastolic and end-systolic volumes and ejection fraction were determined from apical two- and four-chamber views using the Simpson's biplane formula, according to the recommendations of the American Society of Echocardiography[15 1. Tracing of endocardial borders in end-diastole and end-systole was performed on the Acuson 128 machine in the technically best cardiac cycle. The volumes were normalized for body surface area and expressed as indexes. Infarct perimeter (defined as the length of the left ventricular endocardium with wall motion abnormalities) and total left ventricular perimeter were determined in end-diastole from apical four- and two-chamber views; the ratio of two perimeters was expressed as infarct percentage. Infarct expansion was defined as: (1) an increase of >5% in infarct percent and total perimeter on days 2 to 3 in either view, or (2) an initial infarct percentage > 50%, with an increase of > 5% of the total perimeter on days 2 to 3[161. Infarct and total left ventricular perimeters obtained in four- and twochamber views were averaged and mean perimeters were used in further analyses. In order to calculate the wall motion score index, the left ventricle was divided into 16 segments[151. Segmental wall motion was graded as follows: normal motion at rest (score = 1); hypokinetic - marked reduction in endocardial motion and systolic thickening (score=2); akinetic - virtual absence of inward motion and systolic thickening (score = 3); and dyskineticparadoxical wall motion away from the centre of the left ventricle in systole (score =4). The wall motion score index was calculated by summation of individual segment scores divided by the number of interpreted segments. Pulsed-wave and colour Doppler flow mapping were used for the detection and grading of mitral regurgitation. Severity of mitral regurgitation was assessed semiquantitatively according to the length of the turbulent flow jet into the left atrial cavity, using the scale 0-4+[171. Coronary angiography Forty-four patients enrolled in the study underwent coronary angiography before discharge; four patients died before the scheduled procedure and five patients did not agree to undergo coronary angiography. Perfusion of the infarct-related artery was assessed using TIMI criteria[ Successful reperfusion was defined as TIMI grade 3. Significant stenosis was defined as >70% area stenosis of the major epicardial coronary artery. Clinical follow-up Follow-up data regarding systemic embolism and death were obtained in all patients by regular visits or telephone contact with patients or their family. Statistical analysis The chi-square test was used for comparison of nonparametric data and the t-test was used to compare continuous variables. Follow-up echocardiographic experience was summarized by Kaplan-Meier life table estimates of thrombus-free survival. The Cox regression model was used to identify variables associated with thrombus formation and resolution (univariate analysis) and independent predictors of thrombus formation and thrombus resolution (multivariate analysis); P<0'05 was considered significant. Parametric censored regression was used to estimate the impact of each variable on the timing of the appearance and disappearance of left ventricular thrombi during echocardiographic followup. Variables examined included clinical (age, sex, Killip class, thrombolysis, peak serum creatine kinase level), echocardiographic (left ventricular end-diastolic volume index, left ventricular end-systolic volume index, ejection fraction, wall motion score index, severity of apical asynergy, infarct perimeter, infarct expansion, presence of mitral regurgitation) and angiographic (extent of coronary artery disease, patency of infarct related artery) descriptors. Results Left ventricular thrombi Left ventricular thrombi were detected in 30/53 (56'6%) patients. All thrombi were localized in the left ventricular apex, and almost all thrombi (29/30, 96'7%) were detected within the first week after infarction; only one thrombus was detected for the first time 3 weeks after infarction. In 19/30 (63'3%) patients, thrombus, once detected, was present during the entire echocardiographic follow-up. However, in 11/30 (36'7%) cases permanent or temporary thrombus disappearance was noted on follow-up echocardiographic studies. The prevalence of left ventricular thrombi during the follow-up period is shown in Fig. 1. When detected for the first time, thrombi were laminar in 25/30 (83'3%) patients, whereas a protruding shape was noted in 5/30 (16'7%) patients. However, in 6/30 (20'0%) patients a laminar thrombus changed its shape to protruding on subsequent echocardiographic

4 Predictors of L V thrombus formation and disappearance , I-- rn 25 I-- I'l -:!:l..., I Q)..0 a, t- ;:l z Days Figure 1 The prevalence of left ventricular thrombi over the follow-up period. = patients with left ventricular thrombus; 0 = patients without left ventricular thrombus. studies_ Mobile fragments were detected III only 1130 (3-3%) patient with protruding thrombus_ Embolic events None of the patients enrolled in the study had any symptom or sign of systemic embolism during hospitalization or during the follow-up period_ Predictors of left ventricular thrombus formation Univariate analysis (Table 2) showed that left ventricular thrombosis was associated with a higher initial end-systolic volume index, and end-diastolic volume index, a larger infarct perimeter, a lower initial ejection fraction, a higher initial wall motion score index, a higher peak creatine kinase level, Killip class > 1, infarct expansion, occluded infarct-related artery and nonthrombolytic therapy_ This analysis revealed that age, sex, anti thrombotic (heparin, aspirin, warfarin), betablocker and ACE inhibitor therapy, severity of apical wall motion asynergy on admission, presence of mitral regurgitation, and the extent of coronary artery disease were not predictive of thrombus formation_ According to the Cox proportional regression model, independent predictors of thrombus formation after acute myocardial infarction were a high endsystolic volume index and a high peak creatine kinase level (Table 3)_ Thrombi occurred less frequently in patients with an initial end-systolic volume index <35 ml _ m - 2, compared with those with an end-systolic volume index ~35 m!. m -2 (7/21 vs 23/32, P=0-006) (Fig_ 2)_ In addition, patients with thrombus formation progressively increased their end-systolic volume Table 2 Predictors of thrombus formation according to univariate Cox regression analysis Beta P value ESVi (day I) EF (day I) Killip class > I Peak CK level Infarct perimeter (day I) WMSi (day I) EDVi (day I) Infarct expansion Patent IRA Thrombolysis Beta-blockers 1-80 NS Warfarin 1-69 NS Heparin 1-67 NS Multivessel CAD I-51 NS Hypertension 1-44 NS ACE inhibitors 1-36 NS Q-wave infarction 0-87 NS Mitral regurgitation 0-27 NS Severity of apical asynergy (day 1) 0-08 NS Aspirin 0-07 NS Sex NS Age NS ACE=angiotensin convertase enzyme, CAD=coronary artery disease, CK = creatine kinase, EDVi = end-diastolic volume index, EF=ejection fraction, ESVi=end-systolic volume index, WMSi= wall-motion score index, IRA=infarct-related artery_ index during the follow-up period, while those without thrombus revealed no change (Fig_ 3)_ Censored regression showed that if thrombus appeared, it would appear earlier in the course of the disease if there were: an initial ejection fraction ::;;40% (parameter estimate = 164-1, t=7-05, P<O-OOl), an occluded infarct-related artery (parameter estimate = 126-9, t= 1-74, P=0-09), multivessel coronary artery disease (parameter estimate= , t= , P<O-OOI) and a high peak creatine kinase level (parameter estimate= , t= , P=O-OOl)_ Other variables were not related to the timing of thrombus appearance_ Predictors of left ventricular thrombus disappearance As mentioned above, in patients after thrombus was detected, its permanent or temporary disappearance was noted on following echocardiographic studies_ No difference was found regarding antithrombotic therapy (heparin, coumarin, aspirin) between patients without left ventricular thrombi, with permanent thrombi and with variable thrombi_ Demographic, clinical, angiographic and echocardiographic data of patients with and without thrombus resolution are shown in Table 4_ Apical dyskinesis 6 weeks after infarction was found more frequently in patients with permanent thrombus_ Univariate Cox

5 912 A. N. Ndkovic et al. Table 3 Significant independent predictors of thrombus formation according to Cox proportional regression model* Beta P value RR 95% CI End-systolic volume index (day I) Peak creatine kinase level (1'026-1'108) ( ) CI=confidence interval, RR=relative risk ratio. *Other variables included in the model were: initial end-diastolic volume index, ejection fraction and wall motion score index, infarct expansion and infarct perimeter, thrombolysis, therapy with heparin, aspirin, warfarin, beta-blockers and ACE inhibitors, Killip class, extent of coronary artery disease, patency of the infarct-related artery, severity of apical asynergy and presence of mitral regurgitation. The model was tested for inter-relations between end-systolic volume index and other parameters of left ventricular function. Cox analysis showed the same results when only end-systolic volume index was arbitrarily included in the model. analysis revealed that the only variable associated with thrombus resolution was less severe apical asynergy at 6 weeks (beta= - 1,64, P=Q Ql). Patients with permanent and variable thrombus were similar regarding all other variables shown in Table 4, including oral beta-blocker therapy, anticoagulant and antiplatelet therapy, during hospitalization and after discharge. According to the Cox proportional regression model, the only independent predictor of thrombus disappearance was the absence of apical dyskinesis at 6 weeks after infarction (Table 5). However, censored regression did not reveal any variable that might influence the timing of thrombus disappearance. Discussion Several issues should be considered regarding the clinical importance of left ventricular thrombosis after acute myocardial infarction. First, higher mortality has been reported in patients with left ventricular thrombi after infarction l81, especially when these develop within the first 48 h after infarction[9l. Second, left ventricular 25, , p = rn +' Q.~ +' oj <ll..c 10 S ::; Z 5 o ESVi < 35 ml.m- 2 ESVi :? 35 ml.m- 2 Figure 2 The incidence of left ventricular thrombus and initial end-systolic volume. ESVi, end-systolic volume index; = patients with left ventricular thrombus; D = patients without left ventricular thrombus. thrombi may have embolic potential[1o,11,19,201; however, the real incidence of embolic events and the best prevention and treatment strategy still remain controversial. Finally, lower early morbidity and mortality and improvement of functional class was reported in patients with mural thrombus after infarction Therefore, understanding the circumstances in which left ventricular thrombosis occurs may influence patient management. To our knowledge, there is no prospective study that has investigated predictors of thrombus formation and disappearance in the same group of consecutive patients with acute myocardial infarction using serial echocardiographic follow-up during the first year after infarction. In several previously reported studies dealing with this problem, only a few echocardiographic examinations were utilized to detect thrombusi In others, the first week after infarction was missed and a logistic regression model to determine predictors was used l201. The logistic regression model does not consider the time of thrombus appearance and/or disappearance and therefore is less powerful in this setting compared to the Cox proportional regression model. Predictors of left ventricular thrombus formation The incidence of left ventricular thrombi in our study was 56'6%, which is higher than previously published data 11,2,6-8,231, and is approaching the incidence reported in autopsy studiesi24,251. It is probably the result of particular efforts to obtain the best possible visualization of the entire left ventricular cavity and frequent echo cardiographic examinations during the follow-up period. It has been shown that thrombi may appear very early in the course of infarction, as early as within a few hours after experimental infarction l261. All but one thrombus in our study were detected for the first time during the first week after infarction. Therefore, frequent echocardiographic examinations are essential to determine the exact incidence of left ventricular thrombosis, particularly in the first week after infarction. Studies that miss the first week after infarction may lack substantial data regarding thrombus incidence and natural history.

6 Predictors of L V thrombus formation and disappearance "" I~ 55 ] ~ w. 45 ~ 35 * * * t t t t t t Days Figure 3 End-systolic volume change in patients with and without thrombus formation during follow-up period. ESVi, end-systolic volume index; = patients with left ventricular thrombus; =patients without left ventricular thrombus; *, P<O'01; t, P<O OOl. Our patients were not randomized to anti thrombotic therapy and the majority of them were treated with one or a combination of antithrombotic agents. However, no difference was found regarding the incidence of thrombus formation and resolution between patients with and without heparin, aspirin and oral anticoagulants, either during hospitalization or after discharge. In several previously published studies, a trend in favour of the beneficial effect of anticoagulant therapy was noted[ and in one of these studies this trend reached statistical significance[29 1. Antiplatelet therapy failed to demonstrate a significant decrease in left ventricular thrombus after infarction[31, Prospective echo cardiographic studies investigating the effect of thrombolysis on thrombus appearance after infarction revealed controversial results: some of them did not demonstrate a reduction of thrombus incidence following thrombolysis[33 1, while others showed that thrombolytic therapy, by preserving the left ventricular function, reduced its incidence[ Our data showed that left ventricular thrombosis was less frequent in patients treated with thrombolysis. Although a direct lytic effect may decrease the incidence of left ventricular thrombosis, an unchanged relationship between left ventricular function and thrombosis risk after thrombolysis indicates that improved global and regional left ventricular function have a predominant role in prevention of thrombus formation[ In the univariate model, we also found that thrombus formation is associated with larger infarcts, infarct expansion, failed reperfusion and/or more severe left ventricular dysfunction. However, the possibility that the absence of thrombolysis may be predictive of thrombus formation because of increased blood coagulation in respect to thrombolysed patients cannot be excluded. We found that the initial end-systolic volume, which has been previously reported as a major predictor of survival after acute myocardial infarction[38 1, was the strongest independent predictor of thrombus formation. This is the first study to demonstrate that lower end-systolic volume has a protective effect against thrombus formation, regardless of the presence of other, thrombus formation-prone patient characteristics. This may facilitate early detection of patients who, among the population with large anterior infarcts, have a particularly high probability of developing left ventricular thrombosis. Accordingly, modification of the management strategy regarding therapy, as well as the need for further diagnostic procedures may be suggested. For example, it could be recommended that the echo cardiographic examination be repeated if the initial echocardiogram revealed a large endsystolic volume, in order not to miss a left ventricular thrombosis. We showed that early thrombus formation should be expected in patients with large infarcts, more severe left ventricular dysfunction, more extensive coronary artery disease and failed infarct-related artery reperfusion. It is well known that these factors are associated with poor prognosis, so these findings may explain the increased mortality in patients in whom thrombus formation occurred within 48 h of infarction, compared to those in whom thrombi developed after that period[9 1. It has been demonstrated that mitral regurgitation prevents thrombus formation in patients with dilated cardiomyopathy[39 1 However, no study has investigated whether this observation can be extended to patients with acute myocardial infarction. The design of this study enabled us to assess the impact of mitral regurgitation on thrombus formation after acute myocardial infarction. The protective effect of mitral regurgitation may be the consequence of augmented early diastolic flow velocities at the mitral annulus

7 914 A. N. Ne kovic et al. Table 4 Demographic, clinical, angiographic and echo cardiographic data o/patients with permanent and variable thrombus Permanent thrombus (n= 19) Variable thrombus (n= II) Age, y Sex (F/M) Hypertension Diabetes Smoking Q-wave infarction Peak CK level (IV. 1-1) Killip class> I Thrombolysis Patent IRA Multivessel CAD EDVi (day I) EDVi (day 21) EDVi (day 42) ESVi (day I) ESVi (day 21) ESVi (day 42) EF (day I) EF (day 21) EF (day 42) WMSi (day I) WMSi (day 21) WMSi (day 42) Infarct perimeter (day I) Infarct perimeter (day 21) Infarct perimeter (day 42) Apical dyskinesis (day I) Apical dyskinesis (day 21) Apical dyskinesis (day 42) Mitral regurgitation Infarct expansion Died 54±9 6/ / ± 775 8/ ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± /17 13/15 7/ ± / ± /11 4/11 4/8 3/ ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± / * / CAD = coronary artery disease, CK = creatine kinase, EDVi = end-diastolic volume index, EF=ejection fraction, ESVi=end-systolic volume index, IRA=infarct-related artery. WMSi=wall motion score index, *P<0 05. Values for age, CK level, EDVi, ESVi, EF, infarct perimeter and WMSi are expressed as mean ± SD. Table 5 Significant independent predictor 0/ thrombus resolution according to Cox proportional regression model* Beta P value RR 95% CI Severity of apical asynergy (day 42) -1, (0-48-0'97) CI=confidence interval, RR=relative risk ratio. *Other variables included in the model were: end-diastolic volume index, ejection fraction and wall motion score index at 6 weeks after infarction, infarct expansion, infarct perimeter, thrombolysis, therapy with aspirin, warfarin, beta-blockers and ACE inhibitors, Killip class, extent of coronary artery disease, patency of the infarct-related artery, and presence of mitral regurgitation. The model was tested for inter-relations between end-systolic volume index and other parameters of left ventricular function. Cox analysis showed the same results when only end-systolic volume index was arbitrarily included in the model. level[ , as well as through the entire length of the left ventricle[39 1, protecting the left ventricular cavity from a stagnant, thrombogenic blood flow pattern. In addition, an aetiological link between low flow rates as well as an inverse relationship between flow velocities in the left ventricular apex and thrombus formation have been recently demonstrated[42, However, our data did not indicate any association between the presence of mitral regurgitation and left ventricular thrombus formation.

8 Predictors of L V thrombus formation and disappearance 915 Left ventricular thrombus disappearance Left ventricular thrombi can disappear either spontaneously or after anticoagulation. Thrombus disappearance may be associated with embolic events or may occur without any clinically detectable sign or symptom. Reported rates of thrombus resolution ranged from 20_71%[ , One third of thrombi detected in our study disappeared during the echocardiographic follow-up but no embolic episodes were detected during the follow-up period. This may reflect the low overall embolic potential of left ventricular thrombi after infarction, but the fact that the majority of studied patients received some anti thrombotic therapy should also be considered. However, pooled data showed that arterial embolic events occurred in 18% patients with left ventricular thrombus but only in 2% patients without thrombus and that the risk is especially high in those with mobile and protruding thrombi[ Univariate analysis showed that the only variable found to be associated with thrombus resolution was less severe apical asynergy at the end of the healing phase. In addition, the absence of apical dyskinesis was the only independent predictor of thrombus disappearance. However, no variables were identified to predict the time of thrombus disappearance. It is interesting that the severity of apical asynergy was not crucial for thrombus formation, but it seems to be important for thrombus resolution. Keren et azpol also analysed factors associated with thrombus resolution, and their results showed an association between thrombus resolution with anticoagulant therapy and apical akinesis instead of dyskinesis. However, echo cardiographic examinations were not performed during the first week after infarction, angiographic parameters were not included in the model and logistic regression, instead of the Cox proportional regression model, was used. Surprisingly, a recent report on better outcome in patients with left ventricular thrombi indicated that thrombi may be beneficial[lol; thrombus persistence in patients with extensive left ventricular damage and a left ventricular aneurysm might reflect the natural protective mechanism which is trying to minimize the impairment caused by infarction. Therefore, a different approach should be considered for a mobile, protruding left ventricular thrombus in the first hours or days after infarction, and for an adherent, flat thrombus few weeks or months later. Conclusions Our data demonstrate that a high initial end-systolic volume is the major predictor of left ventricular thrombus formation after acute anterior myocardial infarction. It appears that determination of end-systolic volume can differentiate patients with anterior infarction who would develop left ventricular thrombosis from those who would not. Thrombus resolution is more likely to occur in patients without left ventricular dyskinesis at the end of the healing phase. References [1] Asinger RW, Mikell FL, Elsperger 1, Hodges M. Incidence of left ventricular thrombosis after acute transmural myocardial infarction: serial evaluation by two-dimensional echocardiography. N Engl 1 Med 1981; 305: [2] Weinreich Dl, Burke lf, Pauletto Fl. Left ventricular mural thrombi complicating acute myocardial infarction: long-term follow-up with serial echocardiography. Ann Intern Med 1984; 100: [3] Keating EC, Gross SA, Schlamowitz RA. Mural thrombi in myocardial infarctions: prospective evaluation by twodimensional echocardiography. Am 1 Med 1983; 74: [4] Vuille C, Weyman AE. Left ventricle I: General considerations, assessment of chamber size and function. In: Weyman AE, ed. Principles and practice of echocardiography. Philadelphia: Lea & Febiger, 1994: [5] Stratton lr, Lighty GW lr, Pearlman AS et al. Detection of left ventricular thrombus by two-dimensional echocardiography: sensitivity, specificity, and causes of uncertainty. Circulation 1982; 66: [6] Visser CA, Kan G, David GK, Lie KI, Durrer D. Twodimensional echocardiography in the diagnosis of left ventricular thrombus: a prospective study of 67 patients with anatomic validation. Chest 1983; 83: [7] Pechacek L W, Lazar A V, Sonnemaker RE et al. Comparison of two-dimensional echocardiography, radionuclide ventriculography and cineangiography in detecting surgically documented left ventricular thrombi. Tex Heart Inst 1984; 11: [8] Friedman M, Carlson K, Marcus FI, Woolfenden 1M. Clinical correlations in patients with acute myocardial infarction and left ventricular thrombus detected by two-dimensional echocardiography. Am 1 Med 1982; 72: [9] Spirito P, Belloti P, Chiarella F, Domenicucci S, Sementa A, Vecchio C. Prognostic significance and natural history of left ventricular thrombi in patients with acute anterior myocardial infarction: a two-dimensional echocardiographic study. Circulation 1985; 72: [10] Nihoyannopoulos P, Smith GC, Maseri A, Foale RA. The natural history of left ventricular thrombus in myocardial infarction: a rational in support of masterly inactivity. 1 Am Coli Cardiol1989; 14: [11] lugdutt BI, Sivaram CA. Prospective two-dimensional echocardiographic evaluation of left ventricular thrombus and embolism after acute myocardial infarction. 1 Am Coli Cardiol 1989; 13: [12] Meltzer RS, Visser CA, Fuster V. Intracardiac thrombi and systemic embolization. Ann Intern Med 1986; 104: [13] Popovic AD, Neskovic AN, Babic R et al. Independent impact of thrombolytic therapy and vessel patency on left ventricular dilation after myocardial infarction: serial echocardiographic follow-up. Circulation 1994; 90: [14] Killip T 3rd, Kimball IT. Treatment of myocardial infarction in a coronary care unit. A two year experience with 250 patients. Am 1 Cardiol 1967; 20: [15] Schiller N, Shah PM, Crawford M et al. Recommendation for quantitation of the left ventricle by two-dimensional echocardiography. 1 Am Soc Echocardiogr 1989; 2: [16] PopoviC AD, Neskovic AN, Marinkovic 1, Thomas ld. Acute and long-term effects of thrombolysis after anterior wall acute myocardial infarction with serial assessment of infarct expansion and late ventricular remodeling. Am 1 Cardiol 1996; 77: [17] Miyatake K, Izumi S, Okamoto M et al. Semiquantitative grading of severity of mitral regurgitation by real-time twodimensional Doppler flow imaging technique. 1 Am Coli Cardiol 1986; 7: Eur Heart 1, Vol. 19, lune 1998

9 916 A. N. Ndkovic et al. [18] The TIMI study group. The thrombolysis in myocardial infarction (TIMI) trial. N Engl 1 Med 1985; 312: [19] Kupper AJ, Verheught FW, Peels CH, Galema TW, Roos IP. Left ventricular thrombus incidence and behavior studied by serial two-dimensional echocardiography in acute anterior myocardial infarction, ventricular wall motion, systemic embolism and oral anticoagulation. 1 Am Coli Cardiol 1989; 13: [20] Keren A, Goldberg S, Gottlieb S et al. Natural history of left ventricular thrombi, their appearance and resolution in the posthospitalization period of acute myocardial infarction. J Am Coli Cardiol 1990; 15: [21] Kontny F, Dale 1, Nesvold A, Lem P, Soberg T. Left ventricular thrombosis and arterial embolism in acute anterior myocardial infarction. 1 Intern Med 1993; 233: [22] Vecchio C, Chiarella F, Lupi G, Bellotti P, Domenicucci S. Left ventricular thrombus in anterior acute myocardial infarction after thrombolysis. A GISSI-2 connected study. Circulation 1991; 84: [23] Keating EC, Gross SA, Schlamowitz RA et al. Mural thrombi in myocardial infarctions: prospective evaluation by twodimensional echocardiography. Am J Med 1983; 74: [24] Hilden T, Inversen K, Raaschon F, Schwartz M. Anticoagulants in acute myocardial infarction. Lancet 1961; 2: [25] Veterans Administration Cooperative Study Group. Anticoagulants in acute myocardial infarction: results of cooperative clinical trial. JAMA 1973; 225: [26] Mikell FL, Asinger RW, Elspherger KI, Anderson WR, Hodges M. Tissue acoustic properties of fresh ventricular thrombi and visualization by two-dimensional echocardiography: experimental observations. Am J Cardiol 1982; 49: [27] Davis MI, Ireland MA. Effect of early anticoagulation on the frequency of the left ventricular thrombi after anterior wall acute myocardial infarction. Am J Cardiol 1986; 57: [28] Arvan S, Boscha K. Prophylactic anticoagulation for left ventricular thrombi after acute myocardial infarction: a prospective randomized trial. Am Heart J 1987; 113: [29] Nordrehaug IE, Johannessen KA, von der Lippe G. Usefulness of high-dose anticoagulants in preventing left ventricular thrombus in acute myocardial infarction. Am J Cardiol 1985; 55: [30] Turpie AG, Robinson JG, Doyle DJ et al. Comparison of high-dose with low-dose subcutaneous heparin to prevent left ventricular mural thrombosis in patients with acute transmural anterior myocardial infarction. N Engl J Med 1989; 320: [31] Funke Kupper AJ, Verheugt FW, Peels CH, Galema TW, den Hollander W, Roos JP. Effect of low dose acetylsalicylic acid on the frequency hematologic activity of left ventricular thrombus in anterior wall acute myocardial infarction. Am J Cardiol1989; 63: [32] Johannessen KA, Stratton JR, Taulow E, Osterud B, von der Lippe G. Usefulness of aspirin plus dipyridamole in reducing left ventricular thrombus formation in anterior wall acute myocardial infarction. Am J Cardiol 1989; 63: [33] Held AC, Gore 1M, Paraskos 1 et al. Impact of thrombolytic therapy on left ventricular mural thrombi in acute myocardial infarction. Am 1 Cardiol 1988; 62: [34] Eiger N, Maurer G, Shah PK. Effect of early systemic thrombolytic therapy on left ventricular mural thrombus formation in acute myocardial infarction. Am J Cardiol 1984; 54: [35] Bhatnagar SK, al-yusuf AR. Effects of intravenous recombinant tissue-type plasminogen activator therapy on the incidence and associations of left ventricular thrombus in patients with first acute Q wave anterior myocardial infarction. Am Heart J 1991; 122: [36] Lupi G, Domenicucci S, Chiarella F, Bellotti P, Vecchio C. Influence of thrombolytic treatment followed by full dose anticoagulation on the frequency ofleft ventricular thrombi in acute myocardial infarction. Am J Cardiol 1989; 64: [37] Natarajan D, Hotchandani RK, Nigam PD. Reduced incidence of left ventricular thrombi with intravenous streptokinase in acute myocardial infarction: prospective evaluation by cross-sectional echocardiography. Int J Cardiol 1988; 20: [38] White HD, Norris RM, Brown MA, Brandt PWT, Whitlock RML, Wild CJ. Left ventricular end-systolic volume as the major determinant of survival after recovery from myocardial infarction. Circulation 1987; 76: [39] Blondheim DS, Jacobs LE, Kotler MN, Costacurta GA, Parry WR. Dilated cardiomyopathy with mitral regurgitation: decreased survival despite a low frequency of left ventricular thrombus. Am Heart J 1991; 122: [40] Takenaka K, Dabestani A, Gardin 1M et al. Pulsed Doppler echocardiographic study of left ventricular filling in dilated cardiomyopathy. Am 1 Cardiol 1986; 58: [41] Shaikh MA, Levine SJ. Effect of mitral regurgitation on diastolic filling with left ventricular hypertrophy. Am J Cardiol 1988; 61: [42] Maze SS, Kotler MN, Parry WR. Flow characteristics in the dilated left ventricle with thrombus: qualitative and quantitative Doppler analysis. 1 Am Coli Cardiol 1989; 13: [43] Delemarre BJ, Visser CA, Bot H, Dunning AJ with the technical assistance of de Koning H. Prediction of apical thrombus formation in acute myocardial infarction based on left ventricular spatial flow pattern. J Am Coli Cardiol 1990; 15: [44] Visser CA, Kan G, Meltzer RS, Lie KI, Durrer D. Long-term follow-up of left ventricular thrombus after acute myocardial infarction: a two-dimensional echocardiographic study in 96 patients. Chest 1984; 86: [45] Meltzer RS, Guthaner D, Rakowski H, Popp RL, Martin RP. Diagnosis of left ventricular thrombi by two-dimensional echocardiography. Br Heart J 1979; 42: [46] Tramarin R, Pozzo1i M, Febo 0 et al. Two-dimensional echocardiographic assessment of anticoagulant therapy in left ventricular thrombosis early after acute myocardial infarction. Eur Heart J 1986; 7: [47] Stratton JR, Nemanich JW, Johannessen KA, Resnick AD. Fate of left ventricular thrombi in patients with remote myocardial infarction or idiopathic cardiomyopathy. Circulation 1988; 78: [48] Dantzig JMV, Delemarre BJ, Bot H, Visser CA. Left ventricular thrombus in acute myocardial infarction. Eur Heart ; 17: