Role of Echocardiography in the Assessment of Left Ventricular Thrombus Embolic Potential After Anterior Acute Myocardial Infarction
Abstract
The contribution of cardiac ultrasound in assessment of the embolic potential of left ventricular thrombi after anterior acute myocardial infarction was verified in a prospective study of serial echocardiograms (mean, 18.9 examinations per patient) obtained over a long-term period (1–72 months; mean, 38±12). The study population comprised 222 patients (162 men; age, 64±11 years) with a first anterior acute myocardial infarction, treated with thrombolysis (group A) or receiving no antithrombolic therapy (group B). Embolism occurred in a total of 12 patients (11 with a left ventricular thrombus; p<0.005) and was more frequent in group B (10 patients; p<0.04). Predictors of embolism were the absence of thrombolysis, detection of a left ventricular thrombus, protrusion or mobility of the thrombus, and morphologic changes in the thrombus over time. Patients in group A had a lower incidence of each of these predictors, and a higher thrombus resolution rate. An appropriate echocardiographic protocol is crucial to assessment of the embolic potential of left ventricular thrombi after anterior acute myocardial infarction and may help to identify candidates for aggressive antithrombotic therapy.
Left ventricular (LV) thrombi frequently form after anterior acute myocardial infarction (AMI) and carry a risk of embolism.1–6 Although the reported incidence of embolism after anterior AMI is quite low; its occurrence usually has a deleterious to lethal effect on the patient's clinical course. Accurate identification of LV thrombi and of the markers of their embolic potential are crucial to the assessment of AMI patients and implementation of effective antithrombotic treatment when necessary, as well as to reduction of the overall risk of hemorrhagic complications. Previous studies, primarily based on retrospective data, have shown that echocardiography provides a noninvasive, reliable assessment of both LV thrombus incidence and anatomic markers of embolism.7–10 However, striking differences in these two parameters have been reported, mainly because of differences in the echocardiographic study protocols employed.3,5,11–18 Using serial echocardiography, we prospectively evaluated the incidence of LV thrombi after anterior AMI, as well as their anatomic evolution over time, in order to define the parameters for a rational ultrasonic assessment of thrombus.
Methods
Patients. The study population comprised 222 patients (162 men, 60 women; age range, 23–92 years; mean age, 64±11) with a first anterior AMI characterized by ST elevation on electrocardiography. Five patients had suffered a previous inferior AMI. Ninety-seven patients (group A) were treated with thrombolysis. Group B consisted of 125 patients who were enrolled before fibrinolytic treatment of AMI had become routine in our department, and they did not receive antithrombotic treatment during the study period. Thus, the two groups of patients underwent the same echocardiographic study protocol at different time periods. Detailed characteristics of the study population and of the two subgroups are listed in Table I.
| Total Population(n=222) | Group A(n=97) | Group B(n=125) | |
|---|---|---|---|
| Age (years) | 64±11 | 63±11 | 65±12 |
| Men/women | 162/60 | 72/25 | 90/35 |
| Previous inferior AMI | 5 | 1 | 4 |
| Killip class on admission | |||
| I–II | 204 (92%) | 91 (94%) | 113 (90%) |
| III–IV | 18 (8%) | 6 (6%) | 12 (10%) |
| Antithrombotic drugs | |||
| During hospitalization | |||
| Streptokinase | 59 (61%) | ||
| Alteplase | 38 (39%) | ||
| Calcium heparin (12,500 UI twice/day) | 29 (30%) | ||
| Acetylsalicylic acid (325 mg/day) | 89 (92%) | ||
| After hospital discharge | |||
| Acetylsalicylic acid (325 mg/day) | 78 (80%) | ||
| Group A=patients treated with thrombolysis; Group B=patients not receiving any antithrombotic therapy during the study period; AMI=acute myocardial infarction | |||
Echocardiography. Serial echocardiograms were obtained during hospitalization (first examination within 24 hours of admission), every month during the first year, then every 6 months; follow-up ranged from 1–72 months (mean, 38±12). Examinations were performed and recorded by three experienced echocardiographers, two of whom blindly analyzed all videotaped echocardiograms in real-time, slow motion, and stop-frame modes. The mean number of echocardiograms obtained for each patient was 18.9.
LV Thrombus. Echocardiographic diagnosis of LV thrombus was based on standard criteria:3–5 an echo-dense mass with defined margins adjacent to asynergic myocardium, identifiable throughout the cardiac cycle and distinguishable from other structures, such as chordae, muscle trabeculae, and false masses. In order to minimize false-positive diagnoses, doubtful cases were considered negative for thrombus. The sensitivity and specificity of the diagnosis of LV thrombus in our laboratory have been reported previously.6
According to standard criteria defined in previous publications,3–5,19 a thrombus was morphologically defined as protrudent when the intraventricular mass predominantly projected into the LV cavity, and mural if it appeared flat and parallel to the contiguous endocardium. Mobility was defined as a portion of thrombus exhibiting motion independent of that of the adjacent endocardium. The diagnosis and morphologic description of LV thrombi required a blinded consensus of two observers. Changes in configuration, i.e., transformation from protrudent to mural or the reverse, were recorded, as were variations in mobility, i.e., the appearance of mobility patterns that were absent in earlier studies or disappearance of previously noted motion.
LV wall motion was assessed in all serial echocardiograms performed during the study period, and a wall motion index was calculated according to a method established previously.20,21 LV aneurysm was defined as a demarcated bulging of the LV wall contour, visible in both diastole and systole, and demonstrating akinesia or dyskinesia.
Embolism. For diagnosis of central nervous system embolism, the algorithm proposed by Hart et al.22 was applied in all cases. Diagnosis of a peripheral embolus required surgical or autopsy documentation.
Statistical Analysis. Continuous data are expressed as means±standard deviation. Differences between means were analyzed with Student's t test for unpaired data, and differences between percentages were assessed with the chi-square or the Fisher exact test. Stepwise logistic regression analysis was used to identify variables predictive for the risk of embolic events. A p value of <0.05 was considered statistically significant.
Results
LV Thrombus. LV thrombus was detected at least once in 97 of 222 patients (44%)—26/97 of group A (27%) and 71/125 of group B (57%) (p<0.005). The time of LV thrombus development varied from the first hours after AMI to 362 days after infarction (mean, 12±47 days). Figure 1 shows the distribution over time of the first detection of thrombus during the entire study period: 58 thrombi (60%) were detected within the first 48 hours of AMI, and 30 (31%) before hospital discharge. LV thrombus was first diagnosed after hospitalization in nine patients (9%): four (4%) within 1 month, four (4%) during the second and third months, and one (1%) between 3 months and 1 year after AMI. LV thrombi tended to develop earlier in patients treated with thrombolysis (Figure 1), and none was first detected at more than 1 month in this group.
Distribution over time of first echocardiographic detection of left ventricular (LV) thrombus after anterior acute myocardial infarction (AMI). Patients are subdivided according to the presence (Group A) or the absence (Group B) of thrombolytic treatment. Pts=patients; 48 hr=within 48 hours of AMI; Hosp=from the third day after AMI until hospital discharge; 1 mth=from hospital discharge until 1 month after AMI; 3 mth=from the end of the first month until 3 months after AMI; 1 yr=from the end of the third month until 1 year after AMI
Resolution of thrombi was more frequent and rapid in group A than in group B (Figure 2): in more than 70% of thrombolysis-treated patients with a diagnosis of thrombus, the intracardiac mass disappeared within the first month after AMI, and in another 8%, it resolved within the third month.
Percentage distribution over time of resolution of left ventricular (LV) thrombus after anterior acute myocardial infarction (AMI). Patients are subdivided according to the presence (Group A) or the absence (Group B) of thrombolytic treatment. Pts=patients; Hosp=during hospitalization; 1 mth=from hospital discharge until 1 month after AMI; 3 mth=from the end of the first month until 3 months after AMI; 1 yr=from the end of the third month until 1 year after AMI
Protrusion of the thrombus on initial detection was more frequent in patients in group B (49/71, or 69% vs. 8/26, or 31% in group A; p<0.002) (Table II). Group B patients also had a higher percentage of mobile thrombi, but the difference did not reach statistical significance (13/71, or 18% vs. 2/26, or 8%). During follow-up, thrombi in group B more frequently changed morphologically (31/71, or 44% vs. 2/26, or 8%; p<0.003) (Table II) and in mobility patterns (27/71, or 38% vs. 1/26, or 4%; p<0.003). Changes in shape and/or mobility, considered together, were noted in 34 group B and two group A patients (48% vs. 8%; p<0.002).
| (Group A vs. B) | Total Population | Group A | Group B | P Value |
|---|---|---|---|---|
| Incidence of LV thrombus | 97 (44%) | 26 (27%) | 71 (57%) | <0.005 |
| Day of LV thrombus development | 12±47 | 11±39 | 12±47 | NS |
| In-hospital development | 89 (92%) | 24 (92%) | 65 (92%) | NS |
| Anatomic patterns at first detection | ||||
| Protrudent | 56 (58%) | 8 (31%) | 49 (69%) | <0.002 |
| Mural | 41 (42%) | 18 (69%) | 22 (31%) | <0.002 |
| Mobile | 14 (14%) | 2 (8%) | 13 (18%) | NS |
| Resolution | ||||
| Total resolution | 62 (64%) | 22 (85%) | 40 (56%) | <0.002 |
| In-hospital resolution | 39 (40%) | 19 (73%) | 20 (28%) | <0.001 |
| Changes in morphologic patterns | 36 (37%) | 2 (8%) | 34 (48%) | <0.002 |
| Changes in shape | 33 (34%) | 2 (8%) | 31 (44%) | <0.003 |
| Changes in mobility | 28 (29%) | 1 (4%) | 27 (38%) | <0.003 |
| LV wall motion score index21 | ||||
| All LV thrombi | 1.52±0.21 | 1.58±0.21 | 1.53±0.19 | NS |
| Thrombi with morphologic changes | 1.51±0.18 | 1.59±0.23 | 1.52±0.22 | NS |
| LV=left ventricular; Group A=patients treated with thrombolysis; Group B=patients who received no antithrombotic therapy during the study period | ||||
The general, consistent tendency for protrusion and mobility to decrease over time was independent of thrombolytic therapy (Figure 3). At more than 3 months after AMI, persistence of protrusion was observed in only one patient in group A (4%) and eight in group B (11%), and thrombus motion was absent in all cases.
Percentage distribution over time of protrusion and mobility of left ventricular thrombus after anterior acute myocardial infarction (AMI). Patients are subdivided according to the presence (Group A) or the absence (Group B) of thrombolytic treatment. Hosp=during hospitalization; 1 mth=from hospital discharge until 1 month after AMI; 3 mth=from the end of the first month until 3 months after AMI; 1 yr=from the end of the third month until 1 year after AMI
In the presence of LV thrombus, the degree of LV wall motion abnormalities did not differ in relation to thrombolysis, the number of thrombi, or morphologic changes (Table II).
Embolism. During the study period, embolism occurred in 12 of 222 patients (5.4%): 11 were in the central nervous system, and one in a lower limb. LV thrombi had been detected in 11 cases (p<0.005); in the only patient without a diagnosis of thrombus, a nonuniform, hypoechogenic plaque was observed in the left internal carotid artery, lateral to the cerebral lesion.
Analysis of embolic events by group revealed one occurrence in group A (1%) and 11 in group B (8.8%) (p<0.04). In 10 of the 11 group B patients, echocardiography had previously shown a LV thrombus (p<0.02). A thrombus was also noted in the single group A patient. The timing of embolic events (Figure 4) was as follows: during hospitalization in four cases, after hospital discharge and within 3 months of AMI in seven, and after 3 months in one. As Table III shows, on univariate and multivariate analysis, the parameters that identified patients at higher risk for embolism were the absence of thrombolysis, the presence of LV thrombus, the anatomic characteristics of the thrombus (specifically, a protrudent form and mobility), and the occurrence of morphologic changes in the thrombus over time. Age, sex, and the degree of LV wall motion abnormalities were not predictive.
Distribution over time of embolism after anterior acute myocardial infarction (AMI). Patients are subdivided according to the presence (Group A) or the absence (Group B) of thrombolytic treatment. Pts=patients; Hosp=during hospitalization; 1 mth=from hospital discharge until 1 month after AMI; 3 mth=from the end of the first month until 3 months after AMI; >3 mth=from the third month until the end of follow-up
| Univariate Analysis | Multivariate Analysis | |
|---|---|---|
| Age | NS | NS |
| Sex | NS | NS |
| Thrombolysis | <0.02 | NS |
| Presence of LV thrombus | <0.001 | <0.05 |
| Protruding LV thrombus | <0.0001 | <0.01 |
| Mobile LV thrombus | <0.00001 | <0.02 |
| Morphologic changes of LV thrombus | <0.00001 | <0.001 |
| LV=left ventricular; AMI=acute myocardial infarction. The p values were obtained from logistic regression analysis. | ||
Discussion
This study demonstrates that the echocardiographic protocol employed after anterior AMI is crucial to reliable detection of LV thrombi and their anatomic evolution, which accurately predicts the embolic potential and identifies candidates for aggressive antithrombotic treatment.23–25
The incidence of LV thrombi in our study population is one of the highest among those reported from similar studies. Our prospective ultrasonic study protocol, in which echocardiography is performed frequently at short time intervals, has conceivably increased the diagnostic accuracy of the echocardiographers and permitted better evaluation of the overall incidence of thrombi. This has been corroborated in similar studies, which have also revealed a positive relationship between the number of ultrasonic examinations performed and the reported incidence of LV thrombus.12,13,16–19,25–29
Early thrombus development, within 24–48 hours of AMI, occurs in a high percentage of cases (60% in our sample); a large proportion of these thrombi have morphologic characteristics consistent with a high embolic potential, which indicates the need for anticoagulation. Therefore, an early echocardiographic search for thrombi is recommended, especially in patients with a large anterior AMI and extensive wall motion abnormalities. A second ultrasonic examination should be performed at the time of hospital discharge, in order to identify thrombi that may have developed after the first examination or document changes in, or resolution of, a previously detected LV thrombus. A third ultrasonic evaluation should be performed approximately 1 month after anterior AMI; at this time, it is possible either to detect the late development of thrombi or to assess the anatomic evolution of thrombi. Since embolism most often occurs in the early phase of AMI, frequently while the patient is hospitalized, ultrasonic detection and assessment of LV thrombi must be focused on this interval. Most changes in thrombus configuration occur during the first month after AMI; after this point, initial detection of a thrombus with high embolic potential is rare. In addition, most thrombus resolution occurs within the first month, and remaining thrombi tend to evolve to a more benign anatomic configuration, i.e., mural, and to lose mobility. Our study thus provides an explanation for the reported progressive decrease in the embolic potential of LV thrombi after the first month following an AMI. However, it must be stressed that whenever the anatomy of LV thrombus appears consistent with a higher embolic risk, additional ultrasonic studies, performed at even shorter time intervals, must be scheduled.
This prospective, long-term evaluation has confirmed the capability of cardiac ultrasound to identify LV thrombi prone to embolize. In fact, echocardiographic detection of a protrudent and/or mobile LV thrombus was recorded at least once within the first month after AMI in all patients who subsequently had an embolic event attributable to the presence of a thrombus.
Late embolism due to chronic LV thrombus is possible, and has been thoroughly described by Stratton and Resnick.30 In our population, there was only one instance of thrombus-related embolism more than 3 months after AMI, and in that case, serial echocardiography had previously shown, at least once, a LV thrombus with anatomic characteristics consistent with a high embolic potential.
Our data indicate that the lower incidence of embolism in patients who undergo thrombolysis can be attributed to three factors: first, the incidence of LV thrombus was lower in this group than in those not receiving antithrombotics. Second, the prevalence of thrombus characteristics consistent with a higher embolic potential, such as protrusion, mobility, and anatomic instability, was lower in these patients. Third, resolution of LV thrombi occurred earlier and more frequently in patients treated with thrombolysis.
Our data do not address the efficacy of the different antithrombotic drugs employed to reduce the development and the embolic potential of thrombi. It is possible that the use of new antithrombotic drugs, such as glycoprotein IIb/IIIa receptor antagonists, will further reduce the thrombus-related embolic risk after anterior AMI.
