Low incidence of stroke in ambulatory patients with heart failure: A prospective study The current study was undertaken to determine prospectively the risk of cerebral thromboembolism and the prognostic significance of left ventricular thrombus in ambulatory patients with chronic congestive heart failure. A total of 264 ambulatory patients (mean age 62 years, mean left ventricular ejection fraction 27%) were followed prospectively for 24 + 9 months to determine the incidence of nonhemorrhagic stroke, transient ischemic attack, and mortality. Two-dimensional echocardiographic studies, performed for clinical indications other than previous systemic thromboembolism in 109 patients, were analyzed to relate the presence of left ventricular thrombus to subsequent outcome. Nine cerebral thromboembolic events occurred in 264 patients during the two-year mean follow-up period, yielding a rate of 1.7 thromboembolic events per 100 patient-years of follow-up. Known risk factors for stroke (hypertension, diabetes mellitus, and/or atrial fibrillation) were present in all nine patients with cerebral thromboembolic events. The 109 patients with echocardiographic studies had more severe heart failure than patients without echocardiographic studies (functional class 2.6 vs 2.1, p < 0.01), greater risk of a thromboembolic event (2.4 vs 1.4 events/100 patient-years of follow-up, p < 0.01), and higher mortality (21.3 vs 5.5 deaths/100 patient-years, p < 0.01). Left ventricular thrombus, detected in 54 (50%) of these patients, was associated with a greater risk of thromboembolism in a univariate model (p = 0.03) and was an independent predictor of mortality in a Cox analysis (relative risk 2.2, p = 0.03). Thromboembolic stroke and transient ischemic attack occurred infrequently in ambulatory patients with heart failure during a 2-year follow-up period. In selected patients with echocardiographic studies, left ventricular thrombus was associated with increased mortality. A controlled trial of oral anticoagulation therapy is warranted. (AM HEART J 1993;126:141-146.)
Stuart D. Katz, MD, Paul R. Marantz, MD, MPH, Luigi Biasucci, MD, Guillaume Jondeau, MD, Kevin Lee, MD, Claudia Brennan, MS, and Thierry H. LeJemtel, MD Bronx, N.Y.
The risk of systemic thromboembolism in patients with congestive heart failure has been estimated previously from longitudinal observational studies. Published rates of systemic thromboembolic events have varied almost 10-fold, ranging from 1.4 to 12 events per 100 patient-years of follow-up.'^'^ This wide variability observed among previous studies, in which thromboembolic event rates were determined either retrospectively or prospectively in selected patients with heart failure referred for echocardiographic studies, leaves the clinician uncertain of the risk of thromboembolism in the general population of
From the Department of Medicine, Division of Cardiology, The Albert Einstein College of Medicine. Received for publication Nov. 12, 1992; accepted Dec. 28, 1992. Reprint requests: Stuart D. Katz, MD, Columbia-Presbyterian Medical Center, Division of Circulatory Physiology-Milstein 5,177 Fort Washington Ave., New York, NY 10032. Copyright ® 1993 by Mosby-Year Book, Inc. 0002-8703/93/$1.00 + .10 4 / 1 / 4 6 4 0 4
patients with congestive heart failure. Moreover, the association between left ventricular thrombus in patients with congestive heart failure and the subsequent risk of thromboembolic stroke has not been firmly established. Surprisingly the presence of echocardiographically detected left ventricular thrombus has not been linked consistently in previous reports to increased thromboembolic risk.^'^ Prospective studies in unselected patients with congestive heart failure are needed to better assess the risk of thromboembolism and assist in the planning of controlled trials of chronic oral anticoagulation therapy in this patient population. Accordingly the present study was undertaken to determine prospectively the incidence of cerebral thromboembolic events in a general population of ambulatory patients with chronic congestive heart failure. In addition, the prognostic significance of left ventricular thrombus was evaluated in the subset of patients from the study population who were referred for echocardiographic studies. 141
142
Katz et al.
METHODS Patient population. A total of 264 ambulatory patients (179 men and 85 women; mean age 61.8 ± 10.9 years) with chronic congestive heart failure, attending an outpatient heart failure clinic at the Albert Einstein College of Medicine for at least three visits between September 1,1988 and December 31, 1989, were enrolled in the study. This subspecialty clinic provides primary and tertiary care services for an urban population with a principle diagnosis of chronic congestive heart failure. The cause of congestive heart failure was classified as ischemic heart disease when a history of remote (>3 months) myocardial infarction or coronary artery narrowing >50% determined by coronary angiography was documented and otherwise was classified as nonischemic. Patients with recent ( < 3 months) myocardial infarction were excluded from the study. Functional classification at study entry was determined according to the criteria of the New York Heart Association.* Resting left ventricular ejection fraction, determined by radionuclide angiography or contrast ventriculography within 12 months of entry into the study, ranged from 10% to 46% (normal >55%). Medical therapy for congestive heart failure was tailored to the individual patient and was not standardized during the follow-up period. The majority of patients with symptomatic congestive heart failure were treated with digoxin, diuretics, and angiotensin-converting-enzyme inhibitors. Some patients with severely symptomatic heart failure despite optimal conventional therapy were treated with investigational oral phosphodiesterase-inhibiting agents. Aspirin (160 to 325 mg daily) was prescribed in 36% of patients, most commonly those with ischemic cardiomyopathy after myocardial infarction. For the majority of patients, routine chronic oral anticoagulation with warfarin was not prescribed. Patients with atrial fibrillation were treated with aspirin or warfarin unless specific contraindications to these forms of therapy were present. Patients referred to the clinic who were already receiving warfarin therapy were maintained on a chronic oral anticoagulation regimen. In patients receiving chronic oral anticoagulation therapy with warfarin, the dosage was adjusted to maintain the prothrombin time at 1.3 to 1.5 times the laboratory control value. Clinical end points. The clinical end points of the study were cerebral thromboembolism (stroke or transient ischemic attack), as classified by the National Institute of Neurologic Disorders and Stroke,^ and mortality. Clinical evidence of stroke or transient ischemic attack was determined by detailed interviews conducted at 6-month intervals through direct contact with the patient or with the patient's family or primary physician. At the completion of the prospective follow-up period in July 1991, an attempt was made to contact all patients directly for interview and all patient charts were reviewed. Three patients (1 %) were lost to follow-up. Their available clinical follow-up data were eliminated from the final analysis. Classification of the cause of stroke as cardiogenic embolism as opposed to cerebrovascular disease was not attempted, since these disorders are difficult to reliably distinguish from each other on clinical grounds.'^*' Thus all nonhemorrhagic
July 1993 American Heart Journal
strokes and transient ischemic attacks, as determined by computed axial tomography, were included as clinical end points. For each stroke or transient ischemic attack identified, detailed information concerning the event was obtained from the patient's physician or hospital record. Causes of death were determined by interview with the patient's family or physician, or by review of hospital records or death certificate, and were classified as sudden cardiac (death within <24 hours of the onset of new severe symptoms), nonsudden cardiac (progressive heart failure), noncardiac, or unknown when details of the circumstances of death were not available. Determination of clinical end points was made by a physician blinded to the echocardiographic findings. Two-dimensional echocardiography. Two-dimensional echocardiographic studies, which were obtained for clinical indications other than previous thromboembolism, were available for analysis in 109 patients. Studies were requested most often as part of an evaluation of a tertiary referral patient or a patient with a recent deterioration in clinical status. Two-dimensional echocardiographic studies were performed with a 2.5 MHz mechanical transducer connected to a Hewlett-Packard Sonos 100 ultrasound system (Hewlett-Packard Co., Medical Products Group, Andover, Mass.). Standard views from the parasternal and apical windows were obtained by one of us (L. B. or G. J.) and recorded on one-half inch videotape for later analysis. Particular attention was given to study of the left ventricular apical region in both parasternal short-axis and apical two- and 4-chamber views. A left ventricular thrombus was diagnosed by twodimensional echocardiography when an ecohgenic mass adjacent to but distinguishable from the left ventricular endocardium was identified.^'^ All echocardiograms were reviewed by two independent observers (L. B. and G. J.) who were blinded to the clinical status of the patient. In the event of a disagreement among observers, the study was reviewed a second time to reach a consensus. Data analysis. All data were entered into a computerized data base and analyzed on a personal computer with the SAS statistical software package (SAS Institue, Gary, N.C.). Continuous variables were presented as means ± standard deviation; comparisons between groups were made by means of Student's t test. Categorical variables were compared by means of chi-square tests with continuity correction. All tests were two tailed with a p value of <0.05 considered statistically significant. The occurrence of clinical end points is described as the incidence rate per patient-year. These rates are compared by means of the log-rank test (SAS proc lifetest). Multiple confounding variables were simultaneously controlled for by means of the Cox proportional hazards model (SAS proc phreg).
RESULTS All patients. Six strokes a n d t h r e e t r a n s i e n t isc h e m i c a t t a c k s occurred in 264 p a t i e n t s over a m e a n follow-up d u r a t i o n of 24 ± 9 m o n t h s ( T a b l e I), yielding a n overall r a t e of cerebral t h r o m b o e m b o l i c
Volume 126, Number 1 American Heart Journal
Katzetal.
143
Table I. Clinical characteristics in nine patients with cerebral thromboembolic events LVEF
Cause ofCHF
HTN
DM
Afih
Oral AC
ASA
26
Yes
No
No
Yes
Yes
N/A
3
28
Yes
Yes
No
No
Yes
N/A
M
4
22
Yes
No
No
No
No
N/A
65
M
3
27
Yes
Yes
No
No
No
N/A
5
73
M
1
35
No
No
Yes
Yes
Yes
N/A
6
62
F
3
20
Yes
No
No
Yes
No
LV thrombus
7
59
M
3
20
Yes
Yes
No
No
Yes
LV thrombus
8
65
F
3
45
NX
Yes
No
No
Yes
No
LV thrombus
9
58
F
2
21
NI
No
Yes
Yes
Yes
No
LV thrombus
NYHA class
Age (yr)
Sex
1
52
M
1
2
61
M
3
53
4
Patient
Mean ± SD 61 ± 7
(%)
2DE
Clinical diagnosis/ CT findings Vertebrobasilar infarct, CT normal Basal ganglia infarct, CT negative Basal ganglia with hemorrhage on CT Occipital infarct, CT positive TIA without residua, no CT done TIA without residua, CT normal Occipital infarct, CT positive Anterior cerebral infarct, CT positive TIA, transient stuttering, CT normal
2.5 ± 0.3 27 ± 6
AC, Anticoagulation therapy; A fib, atrial fibrillation; ASA, aspirin therapy; CHF, congestive heart failure; CT, computed axial tomography; DM, diabetes mellitus; HTN, hypertension; 7, ischemic cardiomyopathy; LV, left ventricle; LVEF, left ventricular ejection fraction; N/A, not available; NI, nonischemic cardiomyopathy; NYHA, New York Heart Association; TIA, transient ischemic attack; 2DE, two-dimensional echocardiography.
events of 1.7 per 100 patient-years of follow-up. The mean age, functional class, and left ventricular ejection fraction of the nine patients with cerebral thromboembolic events were similar to those values in the remainder of the study population. All nine patients had other known risk factors for cerebral thromboembolism (hypertension, diabetes, and/or atrial fibrillation). Five patients, including two with atrial fibrillation, were receiving chronic oral anticoagulation therapy with warfarin before the thromboembolic event. Two patients were receiving aspirin and two were receiving neither aspirin nor warfarin. Permanent neurologic disability was present in five patients after the cerebral thromboembolic event. Clinical risk factors for cerebral thromboembolism were also prevalent in the 255 patients without clinical evidence of stroke or transient ischemic attack. A history of hypertension alone was present in 89 patients (35%); diabetes mellitus alone was present in 28 (11 %). The combination of a history of hypertension and diabetes mellitus was present in 54 patients (21%). The cause of congestive heart failure was ischemic heart disease in 151 patients (59%). Seventyfive of these patients (50%) were receiving aspirin therapy. Atrial fibrillation was present in 31 patients; 24 of them (77%) were receiving therapy with warfarin (n = 11), aspirin (n = 12), or both (n = 1). Fifty-six patients (21 %) died during the follow-up period, yielding an overall mortality rate of 10.5 per
100 patient-years of follow-up. Death was classified as sudden cardiac in 14 patients, nonsudden cardiac in 21, noncardiac in 10, and was unclassified in 11 patients. Stroke was determined to be the cause of death in only one patient, a 73-year-old man with a history of hypertension who sustained a fatal subarachnoid hemorrhage while receiving chronic warfarin therapy. Patients referred for echocardiographic studies. The clinical characteristics of patients with and without echocardiographic studies are compared in Table II. Patients with echocardiographic studies were more likely to have a diagnosis of nonischemic cardiomyopathy. The prevalence of a history of hypertension and diabetes mellitus and the frequency of warfarin therapy were similar in the two groups, but aspirin was used less commonly in the patients with echocardiographic studies. Patients with echocardiographic studies had more severe congestive heart failure (mean NYHA functional class 2.6 vs 2.1, p < 0.01) and consequently were more likely to be treated with diuretics, digoxin, and investigational phosphodiesterase inhibitors. Two strokes and two transient ischemic attacks occurred among the 109 patients with echocardiographic studies over a mean follow-up duration of 18.6 ± 9.0 months. The rate of cerebral thromboembolism in these 109 patients was nearly double the rate observed in the 155 patients without echocar-
July 1993
144
Katzetal.
American Heart Journal
Table II. Clinical characteristics of study population with and without echocardiographic studies Clinical characteristics Age (yr, mean ± SD) Sex (No., % male) Cause (No., % ischemic) L V E F ( % , m e a n + SD) NYHA class I II III IV History of H T N DM Atrial fibrillation Warfarin Aspirin Diuretics Digoxin ACE inhibitors PDE inhibitors Follow-up duration (mo) Thromboembolism Mortality
With echo (N = 109)
Without echo (N = 155)
All patients (N = 264)
61.2 ± 12.2 69 (63) 53 (49)* 26.1 ± 8.8 9(8)* 36 (33) 56 (51) 8(7) 64 (59) 33 (30) 14 (13) 18 (17) 27 (25)* 94 (86)* 82 (75)* 80 (73) 22 (20)* 18.6 + 9.0* 4(4) 36 (33)*
62.1 ± 9.9 110(71) 105 (68) 27.6 ± 6.9 38 (25) 61 (39) 55 (35) 1(1) 87 (56) 53 (34) 19 (12) 17 (11) 68 (44) 117 (75) 89 (57) 110(71) 9(6) 28.1 ± 6.5 5(3) 20 (13)
61.8 ± 10.9 179 (68) 158 (60) 27.0 ± 7.8 47 (18) 97 (37) 111(42) 9(3) 151 (57) 86 (33) 33 (13) 35 (13) 95 (36) 211 (80) 171 (65) 190 (72) 31 (12) 24.2 ± 8.9 9(3) 56 (21)
ACE, Angiotensin-converting enzyme; PDE, phosphodiesterase; other abbreviations as in Table I. All values are stated as number of patients (% of total in each group) unless otherwise indicated. *Denotes p < 0.05 versus patients without echocardiography.
diographic studies (2.4 vs 1.4/100 patient-years of follow-up, p < 0.01). Thirty-six deaths occurred in the 109 patients referred for echocardiographic studies. The overall mortality rate in these patients was significantly higher than the mortality rate in the 155 patients without echocardiographic studies (21.3 vs 5.5/100 patient-years of follow-up, p < 0.0001; Table II). To determine whether this increased mortality was related to differences in baseline clinical variables in those patients referred for echocardiographic study, multivariate analysis was performed with a Cox proportional hazards model controlling for age, sex, ejection fraction, NYHA functional class, the presence of ischemic heart disease or atrial fibrillation, and the use of diuretics, aspirin, or investigational oral phosphodiesterase inhibitors. Referral for echocardiographic study remained an independent predictor of increased mortality (relative risk 2.7, 95% confidence interval 1.5 to 4.9); other significant predictors were male sex (relative risk 1.9, 95% confidence interval 1.02 to 3.4), NYHA functional class (relative risk 2.2, 95% confidence interval 1.4 to 3.4), use of a phosphodiesterase inhibitor (relative risk 2.2, 95% confidence interval 1.2 to 4.2), and use of aspirin (relative risk 0.4, 95 % confidence interval 0.2 to 0.8). Prognostic value of left ventricular thrombus. In the 109 patients in whom echocardiographic studies were performed, left ventricular thrombus was detected in 54 (50%). None of the baseline clinical variables
listed in Table II was significantly different when patients with and without left ventricular thrombus were compared. Patients with left ventricular thrombus had a significantly shorter follow-up duration (16.8 ± 10.2 vs 20.5 ± 7.3 months, p = 0.03). Three strokes and one transient ischemic attack occurred among these 109 patients, all in patients with left ventricular thrombus detected by echocardiography. In a univariate model patients with left ventricular thrombus had a significantly increased risk of thromboembolism, compared with patients without thrombus (thromboembolic event rate 5.3 vs 0/100 patientyears of follow-up, p = 0.03). There were too few thromboembolic events in the study population to determine whether left ventricular thrombus would remain an independent predictor of thromboembolic events in a multivariate analysis. Among the 36 deaths that occurred in the 109 patients with echocardiographic studies, 23 occurred in 54 patients with left ventricular thrombus and 13 occurred in 55 patients without left ventricular thrombus (mortality rate 30.4 vs 13.8 deaths/100 patientyears of follow-up, respectively; p = 0.02). The presence of left ventricular thrombus was an independent predictor of mortality in this population in a Cox proportional hazards model (Table III). DISCUSSION
The findings in the present study demonstrate that thromboembolic stroke and transient ischemic at-
Volume 126, Number 1 American Heart Journal
tack occurred at a rate of 1.7 events per 100 patientyears in ambulatory patients with congestive heart failure during a mean 2-year follow-up period. The incidence of stroke in the current study is remarkably similar to the findings published in a preliminary report from the first Vasodilators in Heart Failure Trial (V-HeFT) in which the incidence of cerebral thromboembolism in 642 ambulatory patients with class II and III congestive heart failure was prospectively determined to be 1.8 events per 100 patient-years of follow-up.^^ In contrast, a smaller study of 25 patients with idiopathic dilated cardiomyopathy recruited from both inpatient and outpatient settings reported a thromboembolic event rate of 12 events per 100 patient-years of follow-up.^ Several clinical features of the study population may have contributed to the low rate of thromboembolic events observed in the current study. Angiotensin-converting enzyme inhibitors, which were administered to the majority of patients in the current study, have been shown to reduce the incidence of stroke in experimental hypertension models by mechanisms independent of changes in arterial pressure.^^ Warfarin and aspirin, although used in a minority of patients, may have reduced the incidence of cerebral thromboembolism, particularly in patients with atrial fibrillation.^^' ^^ The clinical criteria used for the diagnosis of cerebral thromboembolism in the current study may have failed to detect all such events. However, embolic strokes most typically are first seen with profound neurologic deficits that are readily evident in a detailed clinical evaluation.^ All clinically detected strokes in the current study were assumed to be of potential embolic origin. This method of reporting may have actually overestimated the true risk of embolic stroke in the study population, since primary thrombotic cerebrovascular disease may have been the cause of stroke in some of the study patients with a history of hypertension, diabetes mellitus, or atherosclerotic cardiovascular disease. The findings in the subset of patients referred for echocardiographic studies are comparable to those in previous prospective studies of thromboembolic risk in patients with dilated cardiomyopathy referred for echocardiographic studies, which have reported a prevalence of left ventricular thrombus ranging 11 % to 44% and rates of systemic thromboembolism ranging 1.4%-5.8%.^"^ Of note, the incidence of cerebral thromboembolism in our patients referred for echocardiographic study was significantly greater than the incidence in patients not referred for echocardiographic studies. Nonischemic dilated cardiomyopathy, which was more prevalent in patients with echocardiographic studies, may be associated
Katzetal.
145
Table III. Clinical predictors of mortality in 109 patients with echocardiographic studies (Cox proportional hazards model) Clinical variaible LV thrombus NYHA class PDE Aspirin Male sex Atrial fibrillation Age Ejection fraction Diuretic Ischemic cardiom;yopathy
Relative 2.2 2.9 2.2 0.2 2.1 2.4 1.01 1.0 2.1 0.78
risk
95% CI
p Value
1.08-4.6 1.6-5.3 1.09-4.5 0.06-0.8 0.97-4.5 0,98-6.1 0.98-1.04 0.95-1.04 0.26-17.4 0.38-1.6
0.03 0.0008 0.03 0.03 0.06 0.06 0.41 0.84 0.49 0.78
CI, Confidence interval; other abbreviations as in Tables I and II.
with a greater risk of thromboembolic events than ischemic cardiomyopathy.^^'^^ In addition, aspirin therapy, which was used less commonly in patients with echocardiographic studies, has been demonstrated to decrease the risk of thromboembolic events in patients with atrial fibrillation and may mediate similar benefits in patients with congestive heart failure.^^' ^"^ Last, since the indications for referral to echocardiographic studies were not controlled in the current study, it is possible that the higher rate of cerebral thromboembolic events in patients with echocardiographic studies is related to some unidentified bias inherent in the test referral process. Our findings suggest that the risk of stroke and transient ischemic attack determined in the selected population referred for echocardiographic studies for clinical indications other than previous thromboembolism may overestimate the risk of cerebral thromboembolism in the general population of ambulatory patients with congestive heart failure. Left ventricular thrombus was an independent predictor of mortality in the patients referred for echocardiography. A similar trend, which did not reach statistical significance, was reported by Gottdiener et al.^ in a smaller population of patients with dilated cardiomyopathy referred for echocardiographic studies.^ The mechanism by which left ventricular thrombus is associated with increased mortality is not readily apparent from the current study. Although the presence of left ventricular thrombus was associated with an increased risk of cerebral thromboembolism, it is unlikely that the increased mortality in patients with left ventricular thrombus was due to occult fatal strokes. A hemorrhagic stroke was implicated as the cause of death in only one patient in the current study. In a recent large clinical trial of vasodilator therapy in ambulatory patients with congestive heart failure (the SOLVD investiga-
146
Katz et al.
tion^^), stroke was an uncommon cause of cardiovascular death, occurring in less than 1 % of the study population. It is more likely that left ventricular thrombus may serve as a clinical marker for severe left ventricular systolic dysfunction independent of ejection fraction, inasmuch as thrombus formation may occur more readily in the presence of a low resting cardiac output in association with increased stasis of blood in the cardiac chambers. ^^ Current therapeutic recommendations suggest that chronic oral anticoagulation therapy be administered to all patients with dilated cardiomyopathy to reduce the risk of thromboembolism.^"' ^^ Nevertheless, in two large multicenter trials of vasodilator therapy in patients with congestive heart failure, less than 25 % of enrolled patients were receiving oral anticoagulation therapy.^*' ^^ The thromboembolic event rate in the current study population of ambulatory patients with congestive heart failure, which is less than the reported risk of major bleeding associated with longterm chronic oral anticoagulation therapy,2^ suggests the need for a controlled clinical trial to determine whether warfarin therapy can safely reduce the risk of stroke in patients with congestive heart failure. Chronic oral anticoagulation therapy with warfarin reduced the risk of stroke by 64 % to 67 % compared with placebo in patients with atrial fibrillation.^^' ^'^ A prospective trial with a power of 80% to detect a similar risk reduction in the current study population would require approximately 1500 patients per treatment group. Until the results of such a large trial are available, the merits of empiric oral anticoagulation therapy must be considered on an individual basis. REFERENCES
1. Fuster V, Gersh BJ, Giuliani ER, Tajik AJ, Brandenburg RO, Frye RL. The natural history of idiopathic dilated cardiomyopathy. Am J Cardiol 1981;47:525-31. 2. Stratton JR, Resnick AD. Increased emoblic risk in patients with left ventricular thrombi. Circulation 1987;75:1004-11. 3. Gottdiener JS, Gay JA, Van Voorhees L, DiBianco R, Fletcher RD. Frequency and embolic potential of left ventricular thrombus in dilated cardiomyopathy: assessment by 2-dimensional echocardiography. Am J Cardiol 1983;52:1281-5. 4. Stratton JR, Nemanich JW, Johannessen KA, Resnick AD. Fate of left ventricular thrombi in patients with remote myocardial infarction or idiopathic cardiomyopathy. Circulation 1988;78:1388-93. 5. Ciaccheri M, Castelli G, Cecchi F, Nannini M, Santoro G, Troiani V, Zuppiroli A, Dolara A. Lack of correlation between intracavitary thrombosis detected by cross sectional echocardiography and systemic emboli in patients with dilated cardiomyopathy. Br Heart J 1989;62:26-9.
July 1993 American Heart Journal
6. Blondheim DS, Hacobs LE, Kotler MN, Costacurta GA, Parry WR. Dilated cardiomyopathy with mitral regurgitation: decreased survival despite a low frequency of left ventricular thrombus. A M HEAET J 1991;122:763-71. 7. Falk RH, Foster E, Coats MH. Ventricular thrombi and thromboembolism in dilated cardiomyopathy: a prospective follow-up study. AM HEART J 1992;123:136-42. 8. The Criteria Committee of the New York Heart Association. Diseases of the heart and blood vessels. Nomenclature and criteria for diagnosis. 6th ed. Boston: Little, Brown & Company, 1964. 9. Whisnant J P , Basford JR, Bernstein EF, Cooper ES, Dyken ML, Baston JD, Little JR, Marler JR, Millikan CH, Petito CK, Price TR, Raichle ME, Robertson J T , Thiele B, Walker MD, Zimmerman RA. Classification of cerebrovascular diseases IIL Stroke 1990;21:653-69. 10. Hart RG, CouU BM, Hart D. Early recurrent embolism associated with nonvalvular atrial fibrillation: a retrospective study. Stroke 1983;14:688-93. 11. Meltzer RS, Guthaner D, Rakowski H, Popp RL, Martin RP. Diagnosis of left ventricular thrombi by two-dimensional echocardiography. Br Heart J 1979;42:261-5. 12. Dunkman WB, Johnson GR, Cohn JN. Incidence of thromboembolic events in congestive heart failure: the V-HeFT study [Abstract]. Circulation 1988;78(suppl II):II-617. 13. Stier CT Jr, Renter IF, Ahmad S, Zuo H, Selig N, Roethel S, Levine S, Iskovitz HD. Enalapril prevents stroke and kidney dysfunction in salt-loaded stroke-prone spontaneously hypertensive rats. Hypertension 1989;13:115-21. 14. Petersen P, Boysen G, Godtfredsen J, Andersen ED, Andersen B. Placebo-controlled, randomised trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. The Copenhagen AFASAK study. Lancet 1989;1:175-8. 15. LaPeyre III AC, Steele PM, Kazmier FJ, Chesebro J H , Vlietstra RE, Fuster V. Systemic embolism in chronic left ventricular aneurysm: incidence and the role of anticoagulation. J Am Coll Cardiol 1985;6:534-8. 16. Nihoyannopoulos P, Smith GC, Maseri A, Foale RA. The natural history of left ventricular thrombus in myocardial infarction: a rationale in support of masterly inactivity. J Am Coll Cardiol 1989;14:903-11. 17. Stroke Prevention in Atrial Fibrillation Study Group Investigators. Preliminary report of the stroke prevention in atrial fibrillation study. N Engl J Med 1990;322:863-8. 18. The SOLVD investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. N Engl J Med 1991;325:293-302. 19. Maze SS, Kotler MN, Parry WR. Flow characteristics in the dilated left ventricle with thrombus: qualitative and quantitative Doppler analysis. J Am Coll Cardiol 1989;13:873-81. 20. Meltzer RS, Visser CA, Fuster V. Intracardiac thrombi and systemic embolization. Ann Intern Med 1986;104;689-98. 21. Johnson RA, Palacios I. Dilated cardiomyopathies of the adult (first of two parts). N Engl J Med 1982;307:1051-8. 22. Cohn J N , Johnson G, Ziesche S, Cobb F, Francis G, Tristani F, Smith R, Dunkman WB, Loeb H, Wong M, Bhat G, Goldman S, Fletcher RD, Doherty J, Hughes CV, Carson P, Cintron G, Shabetai R, Haakenson C. A comparison of enalapril with hydralazine-isosorbide dinitrate in the treatment of congestive heart failure. N Engl J Med 1991;325:303-10. 23. Levine MN, Raskob G, Hirsch J. Hemorrhagic complications of long-term anticoagulant therapy. Chest 1989;95(suppl):25S36S.