(Chest. 2000;117:314-320.)
© 2000
American College of Chest Physicians
Left Ventricular Thrombus and Subsequent Thromboembolism in Patients With Severe Systolic Dysfunction*
Nagaraja D. Sharma, MD;
Peter A. McCullough, MD, MPH;
Edward F. Philbin, MD and
W. Douglas Weaver, MD
*
From the Henry Ford Heart and Vascular Institute, Henry Ford Health System, Detroit, MI.
Correspondence to: Peter A. McCullough, MD, MPH, Henry Ford Health System, Henry Ford Heart and Vascular Institute, 6525 Second Ave, Detroit, MI 48202; e-mail: pmc{at}mich.com
 |
Abstract
|
|---|
Study objectives: To determine the frequency of left
ventricular (LV) thrombi by echocardiography and to define the
predictors of LV thrombus and subsequent thromboembolism.
Design: Retrospective case-control design.
Setting: Single tertiary care center.
Patients: Twenty-eight patients with LV thrombus in a
consecutive series of 144 patients with severe LV dysfunction and
follow-up period for a mean of 27.6 months.
Measurements and
results: Thirty-five clinical and echocardiographic variables
were evaluated. The mean age of patients with (n = 28) vs patients
without (n = 116) LV thrombus was 50.3 ± 11.0 years vs
54.2 ± 11.1 years (p = 0.09), with 22 patients (78.6%) and 78
patients (67.2%) being male (p = 0.24), respectively. The mean
ejection fraction (EF) for those with vs those without LV thrombus was
17.5 ± 5.5 vs 20.0 ± 6.9 (p = 0.08), with 16 patients (57.1%)
and 42 patients (36.2%) having an EF < 20% (p = 0.04),
respectively. The groups were similar with respect to other baseline
characteristics, comorbid illnesses, and drug therapies other
than anticoagulants. All 28 patients with LV thrombus (100%) and 54 of
those without LV thrombus (46.6%) were treated with warfarin. Ischemic
etiology of the cardiomyopathy (odds ratio, 4.78; 95% confidence
interval, 1.51 to 15.11; p = 0.008) and increased LV internal
diastolic dimension (LVIDD; odds ratio, 1.10; 95% confidence interval,
1.03 to 1.18; p = 0.004) were found to be independent predictors of
thrombus formation. Peripheral embolism occurred in 5 patients (17.9%)
vs 13 patients (11.2%) of those with and without LV thrombi,
respectively (p = 0.35). Ischemic etiology of the cardiomyopathy
(odds ratio, 3.79; 95% confidence interval, 1.13 to 12.64; p = 0.03)
and EF (odds ratio, 0.91; 95% confidence interval, 0.82 to 1.00;
p = 0.04) were found to be independent predictors of systemic
embolization. The patients with an embolic event suffered a
significantly higher mortality (7 of 18 patients; 38.9%) during the
follow-up period when compared to those without an embolic event (13 of
126 patients; 10.3%; p < 0.0001).
Conclusions: We conclude that ischemic cardiomyopathy and
dilated LV chamber sizes (LVIDD > 60 mm) are independently associated
with LV thrombi. A peripheral embolic event is related to poor
long-term survival in this patient group.
Key Words: congestive heart failure • left ventricular thrombus • systemic embolism
|
Introduction
|
|---|
Congestive
heart failure (CHF) is a major public health problem affecting > 3
million people in the United States.1
There is limited
information in the medical literature that addresses the risks of
thromboembolism among patients with severe left ventricular (LV)
dysfunction.2
3
4
5
6
7
8
9
10
11
12
The need to develop preventive and
therapeutic evidence-based approaches for chronic anticoagulation in
these patients is apparent.
In this study, we sought to do the following: (1) identify the
prevalence of LV thrombus by two-dimensional echocardiography among
patients undergoing evaluation for heart transplant; (2) describe the
characteristics of patients with LV thrombus; (3) determine the
predictors of thromboembolic events; and (4) compare the survival for
those with and without LV thrombus and thromboembolism during the
follow-up period.
|
Materials and Methods
|
|---|
Study Patients
We reviewed the records of 144 consecutive patients who were
evaluated in the CHF and transplantation program at Henry Ford Hospital
from 1991 to 1996 with follow-up available for a mean of 27.6 months.
Patients who had LV thrombus (n = 28) formed the cases, and those who
did not have LV thrombus (n = 116) formed the control group. Patients
with prosthetic valves and/or an ejection fraction (EF) > 35% were
excluded. Comprehensive histories, physical examinations, laboratory
investigations, and two-dimensional echocardiograms were performed on
all patients during the initial evaluation and subsequent regular
follow-up every 3 to 6 months. Ischemic etiology of cardiomyopathy was
defined by the presence of significant epicardial coronary artery
disease, confirmed by coronary angiography. The initial echocardiograms
were reviewed by an experienced echocardiographer who was masked to
subsequent clinical outcomes. Thrombus was defined by the presence of a
distinct echogenic mass, identified in at least two different views and
associated with regional or global wall-motion abnormality. The
decision regarding anticoagulation was left to the discretion of the
attending physician. A cerebrovascular accident was diagnosed by the
clinical evidence of acute stroke, confirmed by neurologic
consultation. In all cases of stroke, intracranial bleeding was
excluded by brain imaging with CT or MRI. The occurrence of peripheral
arterial embolization was diagnosed by the development of acute
extremity arterial occlusion, with confirmation by angiography or
findings at the time of vascular surgery or autopsy. All patients were
followed up in the Henry Ford Congestive Heart Failure (CHF) and
Transplant Clinic through the study period, and no patients were lost
to follow-up.
Statistical Analysis
Univariate statistics including baseline characteristics are
reported in means ± SD and proportions with 95% confidence intervals
as appropriate. Comparisons are made using analysis of variance or the
Kruskal-Wallis H test for continuous variables and
2 or Fisher’s Exact Test for dichotomous
variables. Odds ratios are reported with 95% confidence intervals.
Multiple logistic regression was used to evaluate the independent
predictors of LV thrombus formation and subsequent thromboembolism.
Results from the logistic regression models are reported with the
ß-coefficient for each significant predictor and the Wald statistic
(ß-coefficient/SE), which indicates the relative importance of the
predictor in the model. All models were tested for interaction terms.
The variables of aspirin use, warfarin use, LV thrombus, and left
atrial thrombus were excluded because of multicollinearity
(ie, all patients with LV thrombus were taking warfarin).
The predictive model for LV thrombus was then used to generate the
probability of LV thrombus given a clinical scenario, with ischemic vs
nonischemic etiology and LV internal diastolic dimension (LVIDD) being
the modifiable variables. Survival analysis was performed using the
Kaplan-Meier method with the log rank statistic for differences between
groups. The Cox model with adjustment for covariates was not used
because of sparse data and the small number of absolute events over
time. All hypothesis testing was two-tailed, and p < 0.05 was
considered statistically significant.
|
Results
|
|---|
Baseline Characteristics
Baseline characteristics were similar for patients with and
without LV thrombi, as shown in Table 1
. Ischemic cardiomyopathy and history of prior myocardial infarction
were more frequent in patients with LV thrombus, whereas nonischemic
cardiomyopathy was more common in patients without LV thrombus. No
patient had a prior diagnosis of stroke or peripheral embolic event.
Medication use was similar among the groups, with the exception of
warfarin, as shown in Table 2
. Of note, all 28 of the patients were prescribed warfarin, and 5 of 28
(17.9%) were taking aspirin and warfarin together at the time of
echocardiographic evaluation. Echocardiographic variables are given in
Table 3
. The frequency distribution of LVIDD among all patients is shown in
Figure 1 .
Patients with LV thrombi had larger LV chamber sizes with LVIDD
measurements than patients without LV thrombus (74.0 ± 7.7 mm vs
66.4 ± 9.9 mm; p = 0.0005) in addition to more frequently
anteroapical aneurysms (14.3% vs 3.5%; p = 0.05) and a demonstrated
trend toward lower EF (17.5 ± 5.5% vs 20.0 ± 6.9%; p = 0.07),
respectively.
Embolic Events and Survival
Systemic arterial embolism occurred in a total of 18 patients
(12.5%; Table 4
). These events occurred in 5 patients (17.9%) vs 13 patients (11.2%)
with and without visualized LV thrombi, respectively (p = 0.34).
Figure 2
shows the distribution of embolic events by LVIDD decile. Two of the 18
patients were taking aspirin and warfarin, 10 were taking warfarin
alone, and 6 were taking aspirin alone. The aspirin and warfarin
prescriptions were made by referring physicians, presumably based on
clinical grounds and echocardiograms, prior to referral to our center
for cardiac transplantation evaluation. Of these 18 patients, 10
patients (55.6%) subsequently had transplant or death during the
follow-up period, compared to 20 of 126 (15.9%) in those without an
embolic event (p = 0.001). Also, heart transplantation occurred in 3
of 18 patients (16.7%) with an embolic event vs 7 of 126 patients
(5.6%) without an embolic event (p = 0.11), whereas death occurred
in 7 of 18 patients (38.9%) with an embolic event vs 14 of 126
patients (11.1%) of those without an embolic event (p = 0.004),
during follow-up.
The mean follow-up periods were 22.9 ± 10.8 months and
28.7 ± 15.6 months for patients with and without LV thrombi,
respectively (p = 0.06). Cardiac transplantation or death occurred in
7 patients (25.0%) and 23 control subjects (19.8%; p = 0.61).
Kaplan-Meier analysis, shown in Figure 3
, revealed no difference in freedom from transplant or death between
patients with and without LV thrombi (p = 0.35 by log rank). However,
when stratified by the occurrence of stroke or peripheral embolic event
during the follow up, patients with embolic events had significantly
worsened freedom from death or transplant (p < 0.0001 by log rank;
Fig 4 ). None of the embolic events were listed as the primary cause of death.
View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3.. Survival analysis for the end point of transplant
or death in the follow-up period stratified by the presence of LV
thrombi. Hashed lines represent patients with LV thrombus (cases), and
solid lines represent patients without LV thrombus (control subjects;
p = 0.35 by log rank).
|
|
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4.. Survival analysis for the end point of transplant
or death in the follow-up period stratified by the occurrence of stroke
or peripheral embolic event during the follow-up period. Hashed lines
represent patients with stroke or embolic event, and solid lines
represent patients without stroke or embolic event (p < 0.0001 by
log rank).
|
|
Associations With LV Thrombus
Multiple logistic regression for the outcome of LV thrombus (Table 5
) revealed that an ischemic etiology (odds ratio, 4.78; 95% confidence
interval, 1.51 to 15.11; p = 0.008) and LV chamber size (LVIDD
millimeter unit increment above the referent; odds ratio, 1.10; 95%
confidence interval, 1.03 to 1.18; p = 0.004) were independent risk
factors. There was a trend toward significance in the association
between apical aneurysm and the presence of LV thrombus. However, age
was found to be an independent negative predictor for the outcome of LV
thrombus. The model could not adjust for aspirin and warfarin use due
to multicollinearity. Factors that were found not to be independently
associated with LV thrombus included gender, functional class, atrial
fibrillation, mitral regurgitation (MR), and measured EF. No LV thrombi
were found in patients with LVIDD chamber dimensions < 60 mm. There
was a graded relationship between the predicted probability LV thrombi
and the LVIDD irrespective of the etiology of cardiomyopathy, as shown
in Figure 5
.
Systemic Thromboembolism
Multiple logistic regression for the outcome of systemic
embolization (Table 6
) revealed that an ischemic etiology (odds ratio, 3.79; 95% confidence
interval, 1.13 to 12.64; p = 0.03) and lower EF (odds ratio, 0.91;
95% confidence interval 0.82 to 1.00; p = 0.04) to be an independent
risk factors. There was a trend toward significance in the association
between increasing chamber size and peripheral thromboembolism
irrespective of the presence or absence of LV thrombus, as shown in
Figure 2
. Only 4% of patients with LVIDD < 60 mm had systemic
embolism when compared to 14.3% of those with LVIDD > 60 mm
(p < 0.05).
|
Discussion
|
|---|
The incidence of LV thrombus in patients with end-stage
cardiomyopathy was 19.4%, similar to rates of 11 to 44% published in
the literature.11
13
14
A poorly contracting ventricle
allows blood stasis, which can lead to thrombus formation and
subsequent embolization. The annual risk of systemic embolization in
patients with dilated cardiomyopathy is 1.4 to 12.0%.2
15
Mobile and protruding thrombi are thought to carry the highest
risk.16
Anticoagulation may reduce the chance of embolic
events, but there is controversy about the necessity of routine
anticoagulation in all patients with dilated
cardiomyopathy.17
Several case-control studies and small
prospective cohorts have produced mixed results and recommendations
regarding the use of chronic anticoagulation in these
patients.11
15
18
19
20
21
Most of the studies that have examined the incidence of LV thrombus and
thromboembolism in dilated cardiomyopathy involved small numbers of
patients. Earlier autopsy studies reported a very high frequency of
thromboembolic events, ranging from 37 to 50%.14
22
Fuster et al,11
in a retrospective study of 104 patients
with nonischemic dilated cardiomyopathy, reported an 18% frequency of
thromboembolic events in patients not taking warfarin and an incidence
of four clinically apparent events per 100 patient-years. More recent
studies reported a lower incidence of thromboembolic events ranging
from 1.7/100 to 3.2/100 patient-years in patients with severe LV
dysfunction.19
21
In the Vasodilators in Heart Failure Trial, the incidence of
thromboembolism was 2.5/100 patient-years and did not differ between
patients receiving anticoagulation and those not receiving
anticoagulation.23
Ciaccheri et al15
, found
an 11% prevalence of LV thrombus in patients with nonischemic dilated
cardiomyopathy, with no relation between the presence of intracavitary
thrombus and systemic embolization. In the Survival and Ventricular
Enlargement Trial, the incidence of stroke was 1.5/100 patient-years
and the risk was higher in older patients with lower
EF.10
24
In the Studies of Left Ventricular Dysfunction
Trial, the results were similar, except for the fact that women were
found to be at increased risk for thromboembolic events when compared
to men.25
We found that LVIDD and ischemic cardiomyopathy were independent
predictors of LV thrombus formation. None of the patients with LVIDD
< 60 mm demonstrated LV thrombus. There was a consistent, graded
relationship between increasing chamber size and the presence of LV
thrombus. Also, the risk of LV thrombus was higher in patients with
apical aneurysm. Age was found to be an independent negative predictor
for the outcome of LV thrombus. This finding may in part be secondary
to the relatively young age of the cases when compared to the control
group (ie, the patients studied were not age matched). There
was no significant association between LV thrombus and MR. Kalaria et
al26
have reported an inverse relationship between
severity of MR and the formation of LV thrombus. However, that report
did not control for aspirin use and did not report any association
between LV chamber size and LV thrombus.
In our study, an ischemic etiology and lower EF were both independent
predictors of systemic embolism in patients with and without LV
thrombi. This supports the concept that these predictors are operative
in patients in whom an LV thrombus cannot yet be identified. These
findings are consistent with subgroup analysis from the Survival and
Ventricular Enlargement Trial, in that patients with lower EF were
reported to have higher rates of subsequent stroke during
follow-up.10
24
There was a trend toward significance in
the association between increasing chamber size and peripheral
thromboembolism irrespective of the presence or absence of LV thrombus.
Overall, 4% of patients with LVIDD < 60 mm suffered a peripheral
embolic event compared to 14.3% of patients with LVIDD > 60 mm.
We found that patients with an embolic event (12.5% over 27.6 months)
suffered a significantly higher mortality (38.9%) during follow-up
when compared to those without an embolic event (10.3%). In addition,
the occurrence of an embolic event, rather than the echocardiographic
finding of LV thrombus, had an important bearing on survival (freedom
from death or cardiac transplant). This implies that any anticoagulant
therapy, if it is to impact on survival, probably should be given
preemptively, before the appearance of a thrombus on echocardiography.
We acknowledge the limitations of this study in its retrospective,
case-control design that cannot control for unmeasured confounders. The
magnitude of the measures of association, including the multivariate
odds ratios for ischemic cardiomyopathy and LVIDD, make it extremely
unlikely that these associations are spurious, despite the case-control
design. Ideally, serial echocardiography and a protocol for
anticoagulation with aspirin or warfarin would have provided additional
inferences with respect to the temporal development of visualized
thrombi and possibly the effect of warfarin. We expect these inferences
will be more appropriately made in future randomized trials of
anticoagulation in cardiomyopathy patients.
In summary, patients with ischemic cardiomyopathy, dilated LV chamber
size (LVIDD > 60 mm), lower EF, and apical aneurysm appear to be at
particularly high risk for LV thrombus formation and subsequent
thromboembolism. In addition, the occurrence of an embolic event is
associated with worsened long-term survival. The small size of our
series and wide confidence intervals around point estimates, however,
call for confirmatory studies.
|
Acknowledgements
|
|---|
We are indebted to Arlene Levine, MD, T. Barry
Levine, MD, and Suzanne L. Havstad, MA, for participation in data
collection and management. We would also like to express gratitude to
Keisha R. Marks, BS, who assisted in the final preparation of this
manuscript. Finally, we are indebted to Paul D. Stein, MD, who provided
critical review of the article.
|
Footnotes
|
|---|
Abbreviations: CHF = congestive heart
failure; EF = ejection fraction; LV = left ventricular;
LVIDD = LV internal diastolic dimension; MR = mitral regurgitation
Presented in part at the 47th Annual Scientific Session of the American
College of Cardiology, Atlanta, GA, March 29, 1998.
Received for publication March 22, 1999.
Accepted for publication July 15, 1999.
|
References
|
|---|
-
Ho, KK, Pinsky, JL, Kannel, WB, et al (1993) The epidemiology of heart failure: the Framingham Study. J Am Coll Cardiol 22(Suppl A),6-13[ISI][Medline]
-
Gottdiener, JS, Gay, JS, Van Voorhees, L, et al (1983) Frequency and embolic potential of left ventricular thrombus in dilated cardiomyopathy: assessment by two-dimensional echocardiography. Am J Cardiol 52,1281-1285[CrossRef][ISI][Medline]
-
Segal, JP, Stapleton, JF, McClellan, JR, et al (1978) Idiopathic cardiomyopathy: clinical features, prognosis and therapy. Curr Probl Cardiol 3,1-48
-
Reeder, GS, Tajik, AJ, Seward, JB (1981) Left ventricular mural thrombus: two dimensional echocardiographic diagnosis. Mayo Clin Proc 45,82-86
-
Al-Khadra, AS, Salem, DN, Rand, WM, et al (1998) Warfarin anticoagulation and survival: a cohort analysis from the studies of left ventricular dysfunction. J Am Coll Cardiol 31,749-753[Abstract/Free Full Text]
-
Williams, JF, Jr, Bristow, MR, Fowler, MB, et al (1995) Guidelines for the evaluation and management of heart failure: report of the American College of Cardiology/American Heart Association Task Force on Practical Guidelines (Committee on Evaluation and Management of Heart Failure). J Am Coll Cardiol 26,1376-1398[ISI][Medline]
-
Laupacis, A, Albers, G, Dunn, M, et al (1992) Antithrombotic therapy in atrial fibrillation. Chest 102(Suppl),S426-S433
-
Falk, RH (1990) A plea for a clinical trial of anticoagulation in dilated cardiomyopathy. Am J Cardiol 65,914-915[CrossRef][ISI][Medline]
-
Koniaris, LS, Goldhaber, SZ (1998) Anticoagulation in dilated cardiomyopathy. J Am Coll Cardiol 31,745-748[Abstract/Free Full Text]
-
Loh, E, Sutton, MS, Wun, CC, et al (1997) Ventricular dysfunction predicts stroke following myocardial infarction. N Engl J Med 336,251-257[Abstract/Free Full Text]
-
Fuster, V, Gersh, BJ, Giuliani, ER, et al (1981) The natural history of idiopathic dilated cardiomyopathy. Am J Cardiol 47,525-530[CrossRef][ISI][Medline]
-
Cheng, JW, Spinler, SA (1994) Should all patients with dilated cardiomyopathy receive chronic anticoagulation? Ann Pharmacother 28,604-609[Abstract]
-
Wilensky, RL, Jung, SC (1995) Thromboembolism in patients with decreased left ventricular function: incidence, risk, and treatment. J Cardiovasc Risk 2,91-96[CrossRef][Medline]
-
Roberts, WC, Siegel, RJ, McManus, BM (1987) Idiopathic dilated cardiomyopathy: analysis of 152 necropsy patients. Am J Cardiol 60,1340-1355[CrossRef][ISI][Medline]
-
Ciaccheri, M, Castelli, G, Cecchi, F, et al (1989) Lack of correlation between intracavitary thrombosis detected by cross-sectional echocardiography and systemic emboli in patients with dilated cardiomyopathy. Br Heart J 62,26-29[Abstract/Free Full Text]
-
Fuster, V, Halperin, JL (1989) Left ventricular thrombi and cerebral embolism. N Engl J Med 320,392-394[ISI][Medline]
-
Stevenson, LW, Perloff, JK (1988) The dilated cardiomyopathies: clinical aspects. Cardiol Clin 6,197-218
-
Stratton, JR, Nemanich, JW, Johannessen, KA, et al (1988) Fate of left ventricular thrombi in patients with remote myocardial infarction or idiopathic cardiomyopathy. Circulation 78,1388-1393[Abstract/Free Full Text]
-
Natterson, PD, Stevenson, WG, Saxon, LA, et al (1995) Risk of arterial embolization in 224 patients awaiting cardiac transplantation. Am Heart J 129,564-570[CrossRef][ISI][Medline]
-
Falk, RH, Foster, E, Coats, MH (1992) Ventricular thrombi and thromboembolism in dilated cardiomyopathy: a prospective follow-up study. Am Heart J 123,136-142[CrossRef][ISI][Medline]
-
Katz, SD, Marantz, PR, Biasucci, L, et al (1993) Low incidence of stroke in ambulatory patients with heart failure: a prospective study. Am Heart J 126,141-146[ISI][Medline]
-
Spodick, DH, Littmann, D (1958) Idiopathic myocardial hypertrophy. Am J Cardiol 1,610-623[CrossRef][ISI][Medline]
-
Dunkman, WB, Johnson, GR, Carson, PE, et al (1993) Incidence of thromboembolic events in congestive heart failure: the V-HeFT VA Cooperative Studies Group. Circulation 87(6Suppl),VI94-VI101
-
Pfeffer, MA, Braunwald, E, Moye, LA, et al (1992) Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. N Engl J Med 327,669-677[Abstract]
-
Dries, DL, Rosenberg, Y, Waclawiw, M, et al (1997) Ejection fraction and risk of thromboembolic events in patients with systolic dysfunction and sinus rhythm: evidence for gender differences in the studies of left ventricular dysfunction trials. J Am Coll Cardiol 29,1074-1080[Abstract]
-
Kalaria, VG, Passannante, MR, Shah, T, et al (1998) Effect of mitral regurgitation on left ventricular thrombus formation in dilated cardiomyopathy. Am Heart J 135,215-220[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
|
|
 
C. C. Trabuco, P. Antonio Pereira de Jesus, A. S. Bacellar, and J. Oliveira-Filho
Successful thrombolysis in cardioembolic stroke from Chagas disease
Neurology,
January 11, 2005;
64(1):
170 - 171.
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. A. Harrington, R. C. Becker, M. Ezekowitz, T. W. Meade, C. M. O'Connor, D. A. Vorchheimer, and G. H. Guyatt
Antithrombotic Therapy for Coronary Artery Disease: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy
Chest,
September 1, 2004;
126(3_suppl):
513S - 548S.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. H. Shuhaiber, M. Massad, and A. Geha
Study design in valve surgery and outcome
J. Thorac. Cardiovasc. Surg.,
November 1, 2003;
126(5):
1660 - 1661.
[Full Text]
[PDF]
|
|
|
TOP
Abstract
Introduction
