433

Progress in Cerebrovascular Disease

Management of Cerebral Embolism of Cardiac Origin

J. DONALD EASTON, M.D., AND DAVID G. SHERMAN, M.D.

SUMMARY Tbe cardiac conditions most commonly associated with cerebral embolism are rheumatic heart
disease (RHD), atherosclerotic heart disease (myocardlal infarction and atrial arrhythmias) and other kinds of
nonvalvular atrial fibrillation (AF). Tbe natural history of cerebral embolism from these cardiac sources is
reviewed. Virtually all rheumatic hearts producing emboll hare mitral stenosis, but not all of them are in AF.
Of patients with RHD, 10-20% will experience a systemic embolus, and approximately half will hare a recur-
rence, usually early. Of patients with a myocardial infarction, 5-12% will have a clinically apparent systemic
embolus, and one-third to one-half have a recurrence, usually early. As many as 10-20% of patients with non-
rheumatic AF have a systemic embolus. Anticoagulation reduces systemic embolism to 10-20% of the natural
incidence in RHD, and it reduces embollc recurrences to 10-20% of tbe natural recurrence rate. Anticoagula-
tion diminishes tbe incidence of emboll In myocardial Infarction to 25% of the natural Incidence. It is not
known what effect anticoagulation has on the incidence of embolism hi nonrbeumatic AF. Tbe data regarding
the effect of valvulotomy and prosthetic valve placement in RHD are briefly reviewed. Recommendations are
made for tbe use and timing of anticoagulation based on the available data.
Stroke, Vol 11, No 5, 1980

I. Introduction II. Natural History

Cerebral embolism is a common cause of stroke yet
it is usually not clinically possible to diagnose it with
certainty. The diagnosis is frequently considered but a
number of fundamental questions about cerebral em-
Downloaded from http://ahajournals.org by on April 4, 2022
A. Cardiac Source of Emboll
The criteria used for making a diagnosis of cerebral
embolism vary. Clinically, the diagnosis is usually
made when a patient with RHD, especially with atrial

bolism remain. Often a definite source for the embolus
cannot be easily confirmed. There is uncertainty about
the incidence of embolism in patients with non-
fibrillating rheumatic heart disease (RHD), and in
patients with fibrillating non-rheumatic heart disease.
It is also not clear how likely a stroke is embolic in
origin when it occurs in a patient with coronary artery
disease but with no obvious evidence of a recent
myocardial infarction. When the diagnosis of cerebral
embolism is made, it is not certain whether anti-
coagulant therapy should be used and, if so, whether it
should be given immediately or after a few weeks
delay.
It has become clear that atherosclerotic disease in
the extracranial cerebral arteries may act as a site of
origin for material which can become emboli, and
reach the retina and brain. While this kind of artery-
to-artery embolism produces amaurosis fugax and,
almost certainly, transient cerebral ischemic attacks,
its role in producing sudden, major, completed strokes
is less clear.
This review will examine the natural history of
cerebral embolism associated with RHD and non-
rheumatic heart disease and will review the effects of
anticoagulant therapy and cardiovascular surgery on
the morbidity and mortality in patients with cerebral
embolism. Artery-to-artery embolism has separate
treatment implications.1
From the Department of Neurology, University of Missouri-
Columbia School of Medicine, Columbia, MO 65212.
Reprints: Dr. Easton, at above address.
fibrillation (AF), or coronary artery disease with
either AF or a recent myocardial infarction, has a
stroke with an abrupt onset but without subarachnoid
hemorrhage. More recently, the diagnosis is often
considered in patients with chronic sinoatrial disorder*
and a prolapsing mitral valve.8-4 These criteria may
exclude elderly patients with ventricular mural throm-
bi in whom a recent myocardial infarction cannot be
demonstrated and they may include some elderly
patients with non-rheumatic AF who actually have a
thrombotic cerebral infarct. The criteria for diagnos-
ing cerebral embolism in autopsy series usually in-
cludes the finding of an embolic source, the tendency
to hemorrhagic infarction and multiple arterial
occlusions.
The clinical features of cerebral embolism will be
reviewed along with the autopsy, angiographic and
surgical findings in the brain and in other organs.
/. Clinical Presentation
While the usual cerebral embolus manifests itself
with the abrupt onset of focal cerebral dysfunction,
approximately 13% of patients have vague prodromal
symptoms in the minutes or hours preceding the em-
bolism.8"" These symptoms may include a "stuttering"
focal neurologic deficit related to the cerebral territory
of the subsequent infarct and may be due to one or
more small preliminary emboli. Fisher suggested that
the focal stuttering of symptoms may result from an
incomplete vascular occlusion resulting in the
 434
prodrome. 9 This is followed later by total occlusion as
the embolus becomes firmly lodged. It also seems
possible that there is an initial complete occlusion of
the vessel at the site of the lodged embolus but that ini-
tially the collateral flow is marginally capable of
maintaining perfusion to the distal territory of the
artery. In the subsequent minutes or hours a thrombus
may propagate from the lodged embolus down the en-
tire distal artery producing the complete stroke.
While thrombotic cerebral infarctions frequently
occur or begin during sleep, Fisher10 reported that
only 22% of his patients with cerebral embolism had
their stroke during the night or on arising in the morn-
ing.
Approximately 12% of patients have a seizure at the
onset of an embolic stroke, with one-fourth of them
being focal and three-fourths generalized. This seizure
incidence is about 4 times greater than occurs with the
onset of a thrombotic cerebral infarction.11 Fisher,
however, reports only 1% of patients with cerebral em-
boli have a seizure at onset.10
Although headache at the onset of embolic cerebral
infarction is usually not severe, it occurs in ap-
proximately 25% of patients. As many as 30% of
patients may have a brief lapse of consciousness with
the onset of their ictus.7
2. Neurologic Findings
Most patient's symptoms suggest middle cerebral
artery distribution ischemia with hemiparesis,
Downloaded from http://ahajournals.org by on April 4, 2022
hemisensory symptoms, dysphasia or hemianoptic
visual loss. Rarely does transient monocular blindness
occur suggesting ophthalmic artery emboli. Clinical
and pathologic evidence confirm that anterior cerebral
artery emboli are uncommon compared to middle
cerebral artery emboli. Vertebrobasilar circulation
emboli may be associated with diplopia, vertigo, atax-
ia, crossed or bilateral sensory or motor findings and
coma. Homonymous visual field defects may result
from posterior cerebral artery emboli. One series7
reports twice as many emboli to the right as to the left
cerebral hemisphere and only 10% to the brainstem
(although this excludes the posterior cerebral artery
distribution) but others report twice as many to the
left cerebral hemisphere as to the right.11'13 A small
number are multifocal or of indeterminate location. A
few patients (10%) will have transient neurologic find-
ings with resolution in a few hours to a few days.9 The
appearance of multifocal or diffuse symptoms and
signs makes bacterial endocarditis a more likely con-
sideration.
3. Laboratory and Radiographic Findings
Lumbar puncture performed shortly following a
cerebral embolism is nearly always normal.10 Wells7
states that of 21 patients, except for the patient with
bacterial endocarditis, only 2 patients had more than 5
RBC or 5 WBC/cu mm in the cerebrospinal fluid
(CSF). While the CSF is usually normal following a
cerebral embolism, except when associated with
bacterial endocarditis, Fisher and Adams14 state,
STROKE VOL 11, No 5, SEPTEMBER-OCTOBER 1980
"hemorrhage in the meninges in some, though not in
every case, may be so extensive as to cause a
moderately sanguinous cerebrospinal fluid (therefore,
brain embolism should be included in the apoplectic
disorders causing subarachnoid hemorrhage)."
Angiography performed in the first one or two days
may show occlusion of the internal carotid artery in its
supraclinoid portion with retrograde clot extension
giving a "tailing off" appearance to the column of
contrast material in the more proximal artery. Repeat
angiography, or angiography delayed several days,
may show lysis of the carotid clot with occlusions in
more terminal branches. Complete lysis of the em-
bolus may subsequently occur and angiography per-
formed several days following the embolism may be
normal. 15 ' 1S
Computed tomographic (CT) scanning may
visualize the location and extent of cerebral infarction.
The CT scan can support the clinical diagnosis of
cerebral embolism by demonstrating hemorrhagic or
multiple infarcts. However, embolic injury may
appear as a single low density area compatible with
pale infarction. Petechial hemorrhages within the area
may not visualize as increased density tissue because
when superimposed on the low density infarct the
average tissue density is near that of normal brain.17
4. Neuropathology
Fisher and Adams examined the pathology in 373
cerebral infarcts "with vascular occlusion" and 18% of
them were hemorrhagic.14 They state that nearly 100%
of the hemorrhagic infarcts were embolic in origin.
Conversely, only 50% of embolic infarcts were hemor-
rhagic (if that many, since the etiology of the infarct
was undetermined in one-third of cases). Adams and
Vander Eecken18 later reported that 65% of embolic
infarcts were hemorrhagic. Thus, nearly all
hemorrhagic infarcts and many non-hemorrhagic in-
farcts are embolic in origin. The hemorrhages are
usually petechial, involving primarily the grey matter
of the cerebral hemisphere, but sometimes there is
frank hemorrhage with hematoma formation extend-
ing into the white matter.
5. Source of Cerebral Emboli
Patients with RHD or atherosclerotic heart disease
constitute the overwhelming majority of adults with
cerebral emboli arising from the heart. Atrial myx-
oma," fat emboli,20 septic material, 81 -" tumor,"
venous clots with a right-to-left cardiac shunt, 2426 con-
genital heart disease, cardiomyopathy,27 non-bacterial
thrombotic endocarditis,28'81 mitral annulus calcifica-
tion32' M and prolapsing mitral valves3'4 make up a
small group. Idiopathic AF as an etiology for arterial
embolization will be discussed with atherosclerotic
AF. Emboli from prosthetic heart valves and in-
traoperative emboli during cardiac surgery are special
considerations."**
a) Rheumatic Heart Disease. RHD has been reported
as the predominant cause of cerebral embolism in the
past.7- 40~47 Most patients with rheumatic hearts have
 CEREBRAL EMBOLISM OF CARDIAC ORIGIN/for/o/j and Sherman
mitral stenosis and some also have mitral insufficiency
and aortic valve disease. The thrombi form in the left
atrium, which is usually enlarged, but only half of the
clots are in the atrial appendage. 5 ' "• 48~61 Oc-
casionally, thrombi form on the valve itself.
Several authors report that only 55-82% of
rheumatic hearts are fibrillating (or obviously chang-
ing rhythm) at the time of the embolization.7- "• " Of
Wells' 53 patients with cerebral embolism, half had
normal sinus rhythm. However, 7 of these 25 had
bacterial endocarditis as their embolic source, rather
than a primary left atrial thrombus. Even so, AF is
not the "sine qua non" for cerebral embolism. While
as few as 55-82% of rheumatic hearts that produce
emboli are fibrillating, a fibrillating heart has a several
times greater chance of releasing emboli than one in
normal sinus rhythm.*7' **• "• M These observations
reflect the fact that the majority of rheumatic hearts
with mitral stenosis are in normal sinus rhythm. The
presence of AF and low cardiac output increase the
risk of systemic emboli whereas the severity of mitral
stenosis, as judged by valve area, may not correlate
with the occurrence of emboli."
Systemic emboli occur in 9-49% of RHD patients
followed over several years, with the average being
15-20%. 3 S 3 9 ' < 1 "' 4 6 4 8 ' 5 a - M - 6 7 The higher figures
come from autopsy series in which a careful search
was made for embolic infarcts. 48 ' 68 Twenty to 70% of
these systemic emboli are cerebral, 40% being the
average, with the higher figures coming from clinical
series.42' "• 6 9 Though many of the series reported are
Downloaded from http://ahajournals.org by on April 4, 2022
small and retrospective, it is clear that cerebral em-
bolism is common in patients with RHD, even when
they appear to be in a stable sinus rhythm.
b) Atherosclerotic Heart Disease. 1) Myocardial Infarction.
It is well known that myocardial infarction causes
ventricular mural thrombi to form which may then
break away to become emboli in the arterial circula-
tion. Cumulative data from a number of reports in-
dicate that approximately 45% (with a range of
17-83%) of lethal myocardial infarctions have
associated mural thrombi.60"*" These thrombi are most
common when the infarct is large, when it involves the
septum (particularly a through and through infarct)
and when there is associated congestive heart failure.
Juergens67 states that three-fourths of myocardial in-
farcts with ventricular aneurysm formation have
thrombi in the aneurysm.
Though some authors report the incidence of
peripheral emboli following acute myocardial infarc-
tion as low as 2-3%, and cerebral emboli in 10-90% of
them, other studies from autopsied patients with
myocardial infarction have shown peripheral infarcts,
presumably embolic, in from 45% M to 60%M with ap-
proximately half of them being cerebral. Although the
very high incidence reported by some authors oc-
curred in patients who died and in whom the diagnosis
was made by a careful search for infarcts at autopsy, it
is clear that the incidence is substantial in this very
common cardiac disease. Asymptomatic emboli must
be common, perhaps 2-3 times as common as symp-
tomatic ones.
435
In a recently reported clinical series of 783 con-
secutive patients with myocardial infarction, 1.7% had
a stroke.68 However, the stroke rate was 4.7% for the
third of patients with the largest infarcts as evidenced
by cardiac enzyme elevation in the plasma. Fulton and
Duckett70 suggest identifying patients at risk for
thromboemboli by monitoring plasma fibrinogen con-
centration. Additionally, the use of 2-dimensional
echocardiography to demonstrate ventricular mural
thrombi in patients following myocardial infarction
may prove useful in identifying patients at risk for
systemic emboli.71
Of considerable importance is the time of oc-
currence of cerebral emboli after an acute myocardial
infarction. Bean90 found that 11% occur in the first
week after the infarction, 33% in the second week and
16% in the third week. The fact that nearly two-thirds
occur in the first 3 weeks has clear therapeutic im-
plications. Another 24% occur in the fourth week, 6%
in the second month and 8% in the third month. This
latter group of patients with "delayed" emboli may
have little or no residual evidence of their myocardial
infarction even with a history suggesting an embolic
event.
2) Atrial Fibrillation. A matter of some importance and
controversy is the issue of whether thrombi are prone
to form in the left atrium of patients with AF not due
to RHD or other identifiable valvular disease.
Friedberg72 states that emboli are rarely caused by a
fibrillating heart without RHD and he advises against
anticoagulating such patients. Beer7' reported 295
autopsies of patients with a clinical diagnosis of
atherosclerotic heart disease without obvious myocar-
dial infarction. He found evidence of peripheral em-
boli in 0.8% of those patients known to have been in
normal sinus rhythm and in 2.0% in those known to
have been in AF (i.e., 2 emboli in 243 patients without
fibrillation and one embolus in 52 with fibrillation).
He concluded that peripheral embolization is uncom-
mon in patients with atherosclerotic heart disease
without myocardial infarction, regardless of the car-
diac rhythm. This author excluded another 8 emboli
from an atrial source because he discovered atrial in-
farction or abnormal heart valves. These patients
should not be discarded from the overall statistics
when one is managing patients only with clinical data,
that is, with a clinical diagnosis of non-rheumatic AF.
There is other evidence to suggest that "arterio-
sclerotic" and idiopathic AF may cause systemic em-
boli. Aberg6 reported an autopsy study of 506 patients
with AF but no valvular disease or congenital heart
lesions. Half of them had myocardial infarction and
the other half had other manifestations of
atherosclerotic or hypertensive heart disease, or some
other illness. Of these patients 13.7% had left atrial,
not ventricular, thrombi. Systemic emboli were found
in 41.7%, half of these to the brain. Thus, 20% of these
patients with AF and no valvular disease had cerebral
emboli. Darling et al.43 reported a clinical series of 260
consecutive patients with arterial emboli of whom 97
had arteriosclerotic AF. One-third of these 97 patients
had associated myocardial infarction and two-thirds
 436
had only atherosclerotic heart disease and fibrillation.
Thus, nearly 20% of all systemic emboli occurred in
patients with atherosclerotic heart disease with AF but
without myocardial infarction. They noted that about
70% of their patients had AF, with atherosclerosis be-
ing the most common cause. It is surprising that they
had no patients with RHD and normal sinus rhythm.
This may mean that some of their "arteriosclerotic
fibrillators" in fact had RHD with left atrial thrombi.
Nevertheless, these patients were in the ath-
erosclerotic age group (average age 72 years) and
clinically they had arteriosclerotic AF.
In an autopsy study of 333 patients with AF, Hin-
ton and co-workers" found embolism to be nearly as
common without RHD (59 of 171 or 35%) as with
mitral valve disease (29 of 70 or 41%). In 20% of the
non-RHD patients they found thrombus in the left
atrium.
Wolf and colleagues7* recently provided valuable
prospective epidemiological data from the Framing-
ham Study assessing the risk of embolic stroke in
patients with chronic AF. While AF with RHD had a
17-fold increased risk of stroke, compared to the
general population, chronic idiopathic AF had a 5.6-
fold increase.
Fogarty" found that among 300 consecutive
patients with systemic emboli, 183 had arteriosclerotic
AF as the source. That is, 61% of all emboli were due
to arteriosclerotic AF. The basis for his statement is
not clear, however.
Fisher10 reported 100 patients with AF and cerebral
Downloaded from http://ahajournals.org by on April 4, 2022
embolism. The underlying heart disease was non-
valvular in 83. He also described another group of 48
unselected patients with AF, though it is not clear how
many had non-valvular disease, who were followed
prospectively for 4 years. Thirty-five percent of them
suffered a typical embolic stroke. Friedman and
associates" also showed that ". . . atrial fibrillation in
the absence of myocardial infarction, congestive heart
failure, or rheumatic heart disease was strongly
associated with stroke."
Fairfax and associates2 described 100 patients with
chronic sinoatrial disorder. Sixteen of them had
evidence of systemic emboli and multiple episodes oc-
curred in 6. Fifteen of the 16 patients had the
bradycardia-tachycardia syndrome. An additional 41
patients with chronic ventricular bradycardia and
atrial flutter or fibrillation had an embolic prevalence
of 7.3%. These authors believe that "impaired atrial
function appears to be a key factor in predisposing to
intracardiac thrombosis, and paroxysmal supraven-
tricular tachycardia increases the risk of subsequent
embolization."
All things considered, AF without obvious RHD is
a major source of cerebral emboli.
Systemic embolism is a particular risk when the
atrial rhythm is changing. One to two per cent of
patients undergoing cardioversion for AF will ex-
perience an embolus in the first few days." The risk
exists whether the AF is rheumatic or idiopathic in
origin. Bjerkelund and Orning7" noted fewer emboli in
those patients anticoagulated at the time of cardiover-
STROKE VOL 11, No 5, SEPTEMBER-OCTOBER 1980
sion. Other investigators cite the low incidence of car-
dioversion associated emboli and recommend an-
ticoagulating only high risk patients.™1"
c) Mitral Valve Prolapse. Recently, several workers
described the neurological disturbances in several
patients with a prolapsing mitral valve and the data to
support their belief that the abnormal valve was the
source of emboli.8-4 Whether this common disorder is
a substantial cause of cerebral embolism is not yet
known but it is worth consideration, particularly in
young patients without another obvious source of
ischemic stroke.
d) Cardiomyopathy. Cardiomyopathy may be
associated with systemic emboli particularly if there
is associated congestive heart failure or cardiac
arrhythmia. Emboli are more common in those
patients with idiopathic or alcoholic cardiomyopathy
where from 35-100% of patients at autopsy have
mural thrombi present.27- "•7" Because of the risk of
systemic emboli, long-term anticoagulation should be
considered in patients with moderate or severe car-
diomyopathy."0
6. Evaluation
The nature and extent of the cardiac evaluation in a
patient with cerebral infarction is largely determined
by the degree of suspicion that cardiogenic embolism
may have occurred. Every patient with a stroke should
be carefully questioned for a history suggesting RHD,
cardiac arrhythmia and myocardial infarction. Ex-
amination of the heart should concentrate on
auscultation for abnormal heart sounds. A chest x-ray
will determine the heart size and configuration and an
electrocardiogram will identify rhythm and conduc-
tion abnormalities.
A more extensive cardiac evaluation should be un-
dertaken in patients in whom the suspicion of em-
bolism is high. They include the young, those with a
history of heart disease, those with evidence of mul-
ti focal or hemorrhagic infarcts and those with seizures
at onset. Continuous electrocardiographic monitor-
ing may demonstrate an intermittent arrhythmia and
an echocardiogram may disclose mitral valve disease,
an intracardiac thrombus or tumor and akinetic
segments of myocardium.
In patients with myocardial infarction, the clinical
and electrocardiographic findings, along with the
results of cardiac enzyme determinations, fluoroscopy
and radionuclide scanning, may identify those patients
with large infarcts, ventricular aneurysms, congestive
heart failure and other factors which predispose to
peripheral embolism.
7. Prognosis and Embolism Recurrence Rate
The immediate mortality from cerebral embolism is
about 25-35% and the immediate morbidity in sur-
vivors is approximately the same.7-10' *°- **• ** Thus,
approximately 50-70% of patients die or suffer major
morbidity immediately. The need to identify and treat
prophylactically those patients at risk for cerebral em-
bolism is obvious. Though it is often stated that one-
 CEREBRAL EMBOLISM OF CARDIAC ORIGIN/£aj/on and Sherman
half to two-thirds of all patients with cerebral em-
bolism will die within one year, largely because of
their serious primary cardiac disease, this estimate is
not well-documented. However, in Daley's long-term
follow up study,41 half of the patients with RHD
without bacterial endocarditis ultimately died of
cerebral embolism. Wells noted that advanced age,
prodromal symptoms, loss of consciousness and/or
seizures at onset, and failure to improve within 48
hours, were factors that correlated with a "poor out-
come." 7
a) Recurrence Rate of Embolization in RHD. The
natural recurrence rate of cerebral and systemic em-
bolization is from 30-75% in patients followed for
varying periods of time.7' M- "• "• "• "• **• "• M The 75%
comes from Carter's clinical series with the longest
follow up, from 6 to 12 years.*0 Approximately one-
half to two-thirds of recurrent emboli occur in the first
year and Szekely noted that 40% occurred in the first
month." Darling,43 in his 10 year follow up study,
noted a 56% recurrence rate for systemic embolism:
16% in 2-4 days and 18% in 5—14 days. Thus, one-
third of recurrences occurred within 2 weeks of the ini-
tial embolism.
Daley42 reported 194 patients with RHD without
bacterial endocarditis and 115 of them had a total of
201 recurrent emboli. Thirty-two occurred in the first
week and 41 occurred between the second and fourth
week. Thus, one-third of the total recurrences oc-
curred in the first month.
b) Recurrence Rate of Emboli in Atherosclerotic Heart
Downloaded from http://ahajournals.org by on April 4, 2022
Disease. Davies and Pomerance" reported 74 patients
at autopsy with chronic AF, 62% of whom had throm-
bi in the left atrial appendage. While they didn't
analyze this group of 46 patients by specific diagnosis,
they did state that only 18 of the 74 patients had
RHD. That is, most of the patients with atrial appen-
dage thrombi had some form of heart disease with AF
other than RHD.
I) Myocardial Infarction. As noted previously, Bean™
reviewed a number of reports in the literature which
showed that approximately 45% of patients who die
following a myocardial infarction have thrombi in the
left ventricle. Of the 300 patients in Bean's series, 60
had 78 systemic emboli, of which only 8% were
cerebral. That is, 20% of his patients with myocardial
infarction had a systemic embolus and one-third of
those had a recurrence. He states that an additional 38
instances of "thrombotic" infarction occurred. Forty-
four percent of the emboli occurred in the first 2 weeks
following the myocardial infarction and the number
had risen to 60% by the end of the third week and to
84% by the end of the first month.
Darling43 reported that in 28 patients with systemic
emboli following myocardial infarction, in 7 (25%)
subsequent emboli occurred, with 5 of the 7 (71%)
recurrences within 14 days of the preceding embolism.
If the patients who had AF with their myocardial in-
farction were added to the myocardial infarction only
group, the recurrence rate for the total group was 21
of 52 patients, or 40%. Half of these recurrences oc-
curred within 14 days of the initial embolus.
437
2) Arteriosclerotic Atrial Fibrillation (without myocardial
infarction). Darling41 reported that in 59 patients with
systemic emboli, twenty-five (42%) had subsequent
emboli. Of the recurrences, 48% occurred within 2
weeks, 24% between 2 weeks and 4 months, and 28%
between 4 months and 5 years.
Therefore, not only do non-rheumatic fibrillators
embolize, but they have a 42% chance of recurrent em-
bolization, usually early.
III. Treatment
A. Anticoagulation
Anticoagulation has been found to be a highly
beneficial treatment in the prevention of systemic, in-
cluding cerebral, emboli in patients with a cardiac
source for emboli.7' *"x- "• **- M-88 Some of the data
supporting this treatment are derived from prospec-
tive studies in patients without a previous embolus and
some from the effect of anticoagulation on prevention
of recurrence of emboli. In RHD, the natural
recurrence rate of approximately 50% is reduced to
5-25%.47- M In patients with lethal myocardial infarc-
tion, the 45% natural occurrence of mural thrombi is
halved by anticoagulation and the incidence of clinical
cerebral embolism is diminished to approximately
one-fourth of the natural incidence.87' *9"91 Although
there is still controversy about the overall efficacy of
anticoagulation in the acute phase of myocardial in-
farction," there is general agreement that the risk of
cerebral embolism is greatly reduced by this therapy.
While the effect of anticoagulation in patients with
non-rheumatic AF is not known, in part because these
patients are usually not given anticoagulants, 40% of
Darling's patients with arteriosclerotic AF and
systemic emboli, had atrial thrombi present at autop-
sy.43 This observation suggests that these patients also
may benefit from anticoagulation. While there is little
question about the efficacy of anticoagulation in the
prevention of cerebral embolism in "at risk patients,"
there is considerable debate about when anticoagula-
tion should be initiated and how long it should be con-
tinued. Because it was recognized that cerebral em-
bolism often produced hemorrhagic infarcts, many
clinicians feared that early treatment with anti-
coagulant would result in grossly hemorrhagic infarcts
with worsening of the morbidity and mortality. In
1956, Symonds stated that, "anticoagulant drugs
should not be used in case of cerebral embolism, for in
these cases hemorrhage into the area of infarction
often occurs, probably as the result of displacement or
fragmentation of the embolus . . ."."* He gave no evi-
dence, however, to support his view.
Moyes et al."4 and Wood et al.M have presented ex-
perimental evidence related to this problem. Both
studies involved the injection of vinyl acetate into the
carotid artery of dogs followed by anticoagulation
with warfarin derivatives. Moyes reported that 25% of
control animals, and 50% of anticoagulated animals,
die in the first 30 hours. Total occlusion of the internal
carotid artery was common and the control animals
had no grossly hemorrhagic infarcts. Even though
 438 STROKE
these animals were excessively anticoagulated, this
report is evidence for avoiding anticoagulant in acute
cerebral infarction but the pathogenesis clearly relates
more to cerebral thrombosis than cerebral embolism.
Wood's findings were very similar. In an experiment
perhaps more analogous to the human embolism
situation, Sibley*9 produced embolic cerebral infarcts
in 36 dogs by injecting homologous clot fragments
into the internal carotid artery. Six animals from both
the control and anticoagulated groups died in the first
24 hours following the embolism and the infarcts were
pathologically identical in the 2 groups. Only one of
the 12 survivors in the control series died more than 24
hours after operation while 4 of the 12 survivors in the
dicumarol-treated group died more than 24 hours
later. Most of these animals were anticoagulated 24
hours prior to producing embolism. Also, the
prothrombin time was more than 3 times the control
value in 3 of the 4 late deaths and approximately 5
times control in the fourth. Sibley concluded that an-
ticoagulant should be avoided in the treatment of
patients who have suffered recent cerebral infarction
due to embolism. He also suggested that the
prothrombin time should be returned to normal as
soon as possible in patients who have a cerebral em-
bolism on anticoagulant. Wright and McDevitt,90 on
the other hand, reported no excess of hemorrhage or
mortality with anticoagulation.
Carter, in 1957, reported on 29 patients with RHD
and showed that anticoagulation was clearly beneficial
and that early treatment was desirable." He also
Downloaded from http://ahajournals.org by on April 4, 2022
showed that those patients on anticoagulant who died
had "no sign of intracranial bleeding, and the in-
farcted areas were surrounded by the usual small area
of hemorrhage." In 1965, Carter updated his findings
in a report of 130 patients with 176 episodes of
cerebral embolism.40 He confirmed his previous view
that "treatment with anticoagulants has significantly
improved the prognosis of cerebral embolism in rela-
tion to immediate outcome, late survival and
recurrence". Also, Fleming and Bailey48 noted only 5
systemic emboli in their 217 patients treated with
long-term anticoagulation, for an embolus rate of
0.8% per patient-treatment year.
Our interpretation of these data is that anticoagu-
lant in experimental animals probably makes cere-
bral infarcts more hemorrhagic but probably doesn't
affect mortality; meager human studies appear to
show a beneficial effect on both morbidity and mor-
tality when anticoagulant is used immediately follow-
ing the cerebral embolization, in spite of anecdotal
reports of patients on anticoagulant dying with grossly
hemorrhagic infarcts.
However, on the issue of immediacy of anticoagula-
tion, Carter concluded that "the results show no
significant gain from the immediate use of anti-
coagulants in patients whose neurological lesion is
persistently complete. If anticoagulants are to be used
in these patients an interval of three weeks should
elapse before commencing treatment." 40 This view
does not seem well founded and is of considerable
clinical importance. Carter appears to have based his
VOL 11, No 5, SEPTEMBER-OCTOBER 1980
view on the following: "Wells (1959) found an in-
creased number of red cells in the CSF of some
patients treated with anticoagulants within 48 hours of
their cerebral embolism and noted that in these the
results 'were bad.' Similar reports of serious bleeding
after anticoagulation came from de Morsier and
Tissot and Ushiro and Schaller; it seemed likely that,
if a major infarction of the brain had occurred after
the embolic episode, it was probably of the hemor-
rhagic type, and unlikely to be helped by anticoagula-
tion." In fact, Wells' data support immediate an-
ticoagulation and only suggest that anticoagulants
might be avoided if the CSF is hemoirhagic. Ushiro
and Schaller97 supply no evidence that anticoagulation
was detrimental for cerebral embolism. In addition,
because of this concern Carter avoided immediate an-
ticoagulation whenever it seemed likely that a major
infarction of the brain had occurred and he therefore
supplied no evidence of his own to support or refute
his view.
Since Carter found "the survival rate improved
significantly in patients with incomplete or fluctuat-
ing lesions (in whom ischemia is more likely than in-
farction) when anticoagulants were used in the early
stages," and "in all patients the recurrence rate was
high immediately after the initial cerebral embolism,"
we conclude, as did Marshall,88 that immediate anti-
coagulation should be initiated in all patients with
cerebral embolism in whom the CSF is not
hemorrhagic and the neurologic deficit is not massive
and persistent beyond 24 hours.
Aside from the concern about converting a white in-
farct into a hemorrhagic one, bleeding is a risk and oc-
curs in as many as 20% of patients on prolonged
therapy. 99101 While it is usually minor or transient, es-
pecially with short-term therapy,87 several studies
report a mortality rate of 0.5-1.0%.101'105
B. Cardiac Surgery
The role of cardiac surgery in the prevention of
arterial embolization is becoming clearer. Several
authors'4- "• 41- "• 10*"ni provide evidence to indicate
that valvulotomy protects against later arterial em-
bolization and in a follow-up period of "a few months
to four years," Belcher34 reported late emboli in only
1% of his 430 patients after operation. Other inves-
tigators,58' 47-11S-l13 on the other hand, present data
indicating that valvulotomy provides no protection
against subsequent emboli. The operative mortality
for the valvulotomy itself is in the range of
3_9%39. iiriu a n d anot h e r 6 - 2 0 % " - "• "• *>•41- '"• 116
or
so of patients suffer an intraoperative arterial em-
bolism. Also of note is the belief of some that atrial
appendagectomy at the time of valvulotomy is
beneficial in preventing subsequent emboli, at least if
the appendage is large.34- M Others believe that appen-
dagectomy is of limited or no value.41-42 It therefore
seems that when the operative morbidity and mor-
tality are added to the equivocal benefit of the sur-
gery, mitral valvulotomy cannot be generally recom-
mended for the prevention of arterial embolization.
 CEREBRAL EMBOLISM OF CARDIAC ORIGIN/£ar/on and Sherman
While newer valves and techniques have resulted in
low operative mortality,10'- u e recent studies of mitral
valve replacement showed an overall hospital mor-
tality of 5-22%. m - 118 In addition to the operative
mortality there is a significant incidence of arterial
emboli in the late post-operative period. Some studies
show that the rate of occurrence of emboli is highest in
the early months and years after valve placement.
Starr reported an occurrence rate of 7% in a 3 year
follow up, with half in the first month and the other
half in the first year.118 Friedli" reported a 16% oc-
currence of emboli in 170 patients followed a mean of
26 months after valve placement with a decreasing in-
cidence of emboli with time. Others have shown that
the incidence is significantly higher in the longer term
follow up studies. Fishman, for example, reported that
22% of patients had an embolus in 5 years of follow
up.119 The use of new model valves results in embolic
rates of 5% over 3 years.120 Thus, it appears that mitral
valve replacement cannot be generally recommended
as a treatment for the prevention of arterial emboli
even though recent technical improvements are
diminishing the operative and postoperative mortality
and morbidity associated with this surgery.
IV. Conclusions and Recommendations
Cerebral embolism producing the abrupt onset of a
major, focal neurological deficit usually has its throm-
bus source in a diseased left heart. Headache, seizures
Downloaded from http://ahajournals.org by on April 4, 2022
and transient loss of consciousness in this setting in-
crease the likelihood of cerebral embolism over
cerebral thrombosis. There is usually no blood in the
CSF as occurs with ruptured aneurysms and vascular
malformations and with most intracerebral
hemorrhages.
The most common kinds of heart disease associated
with cerebral emboli are rheumatic and ath-
erosclerotic (i.e., coronary artery disease). The
rheumatic heart with a left atrial thrombus invariably
has mitral stenosis, the atrium is commonly enlarged
and 1/2 to 2/3 are fibrillating or changing rhythm.
The chances of emboli are several times higher with
AF than with normal sinus rhythm. Approximately
15-20% of patients with RHD will experience a
systemic embolus, and many will be cerebral. One-
third to 3/4 of those who have emboli will have a
recurrence. One-fourth to 1/3 of the recurrences will
occur in the first 2-4 weeks, with 1/2 to 2/3 in the first
year.
In the "arteriosclerotic" heart the abnormalities
associated with mural thrombus are largely infarction
and arrhythmias. About 45% of hearts following lethal
myocardial infarction have mural thrombi and
12-60% of these produce systemic emboli. A 5-12%
clinically apparent embolism rate is probably ac-
curate, with many of them being cerebral.68 Ap-
proximately 2/3 occur in the first 2-3 weeks after the
infarct, and 1/4 to 1/3 will recur shortly thereafter.
Arteriosclerotic or idiopathic AF is, in our view, a
clear cause of emboli. Limited evidence indicates that
439
15-20% of patients with non-rheumatic AF will have
emboli. Additionally, risk factors may include atrial
enlargement, the bradycardia-tachycardia syndrome
and congestive heart failure. Among those patients
who have emboli, 1/3 to 1/2 will have a recurrence,
half of them within the first 2 weeks. We believe that
the senile and coronary artery diseased heart without
valvular disease is underestimated as a cause of
cerebral embolism.
Anticoagulation reduces systemic embolism to
10-20% of the natural incidence in RHD. It also
reduces embolic recurrences to about 10-20% of the
natural recurrence rate.
Anticoagulation diminishes the incidence of emboli
in myocardial infarction to 25% of the natural in-
cidence.
It is not known what effect anticoagulation has on
the incidence of embolism in non-rheumatic AF.
We agree with Marshal,9* based on the evidence
available, that any patient with a probable cerebral
embolism from any cardiac source should be anti-
coagulated immediately, provided bacterial endocar-
ditis is unlikely and there is no blood in the CSF and
no hematoma on CT scan. This recommendation is
based on the high probability of a second embolus oc-
curring, very likely in the next 1-3 weeks. This can be
accomplished by immediate heparin or warfarin ad-
ministration. We give 7,500 units of heparin in-
travenously followed by 1,000-2,000 units per hour by
intravenous drip infusion monitored to maintain the
activated partial thromboplastin time at twice normal.
This regimen is maintained for the first several days
during which the patient is monitored for hemorrhagic
complications and evaluated. Then, a warfarin
derivative can be administered and when the
prothrombin time is 2 to 2-1/2 times normal, the
heparin is discontinued.
If the embolic source is self-limited or treatable,
anticoagulation should be discontinued when the high
risk period has passed. For myocardial infarction this
is after about 3 months. For a treatable car-
diomyopathy it may be several weeks after cardiac
function is normal. Following conversion of non-
valvular AF it may be a few weeks after sinus rhythm
appears to be successfully maintained.
If the embolic source cannot be eliminated, anti-
coagulation should generally be continued indefi-
nitely. This includes any patient with RHD, even if the
mitral valve is replaced or the AF is successfully con-
verted to normal sinus rhythm. It also includes the
patient with non-valvular AF, persistent car-
diomyopathy and possibly mitral valve prolapse, who
has had one or more systemic emboli.
The role of anticoagulation in the prevention of em-
boli in the at risk patient with no prior history of em-
bolism is less clear. In myocardial infarction, the risk
of systemic embolism increases with increasing infarct
size, congestive failure and ventricular aneurysm for-
mation. Consequently, patients with large infarcts
should be started on warfarin and anticoagulation
should be maintained at least during the 3 month high
risk period. Those with echocardiographically dem-
 440 STROKE
onstrated mural thrombi should be hcparinized and
then converted to warfarin.
All patients with prominent RHD are at risk for
systemic embolism and should be anticoagulated with
warfarin indefinitely. This includes patients with ob-
vious mitral stenosis, a large left atrium, congestive
heart failure, AF, and echocardiographically
demonstrated atrial thrombi. Whether the benefit of
long-term anticoagulation outweighs the risk in
patients with lesser degrees of RHD is unclear.121 The
role of long-term anticoagulation in patients with non-
valvular AF with no evidence of systemic embolism
also is unclear. However, these patients are clearly
embolus prone and if there are no significant contrain-
dications we favor anticoagulation for an indefinite
period.
The best treatment for cerebral embolism is preven-
tion and, therefore, we believe that more vigorous
treatment of patients with thrombus-prone heart dis-
ease is indicated. Virtually all patients with moderate
to severe RHD, large myocardial infarcts and many
patients with nonvalvular AF should be prophylac-
tically anticoagulated.
References
1. Byer JA, Easton JD: Controversies in therapy of ischemic
cerebral vascular disease. Ann Intern Med (in press)
2. Fairfax AJ, Lambert CD, Leatham A: Systemic embolism in
chronic sinoatrial disorder. N Engl J Med 295: 190-192, 1976
3. Barnett HJM, Jones MW, Boughner DR, Kostuk WJ:
Cerebral ischemic events associated with prolapsing mitral
Downloaded from http://ahajournals.org by on April 4, 2022
valve. Arch Ncurol 33: 777-782, 1976
4. Kostuk WJ, Boughner DR, Barnett HJM, Silver MD:
Strokes: a complication of mitral-leaflet prolapse? Lancet 2:
313-316, 1977
5. Aberg H: Atrial fibrillation. Ada Med Scand 185: 373-379,
1969
6. Achor RWP, Futch WD, Burchell HB, Edwards JE: The fate
of patients surviving acute myocardial infarction. Arch Intern
Med 98: 162-174, 1956
7. Wells CE: Cerebral embolism. Arch Neurol Psychiat 81:
667-677, 1959
8. Wells CE: Premonitory symptoms of cerebral embolism. Arch
Neurol 5: 44-50, 1961
9. Fisher CM, Pearlman A: The nonsudden onset of cerebral em-
bolism. Neurology (Minneap) 17: 1025-1032, 1967
10. Fisher CM: Reducing risks of cerebral embolism. Geriatrics
34: 59-66 (Feb) 1979
11. Louis S, McDowell FH: Epileptic seizures in nonembolic
cerebral infarction. Arch Neurol 17: 414-418, 1967
12. Keen G, Leveaux VM: Prognosis of cerebral embolism in
rheumatic heart disease. Br Med J 2: 91-92, 1958
13. Meyer JS, Charney JZ, Rivera VM, Mathew NT: Cerebral
embolization: prospective clinical analysis of 42 cases. Stroke
2: 541-554, 1971
14. Fisher M, Adams RD: Observations on brain embolism with
special reference to the mechanism of hemorrhagic infarction.
J Neuropathol Exp Neurol 10: 92-93, 1950
15. Bladin PF: A radiologic and pathologic study of embolism of
the internal carotid-middle cerebral arterial axis. Radiology
82: 614-624, 1964
16. Taveras JM, Wood EH: Diagnostic Neuroradiology. 2nd ed.
Baltimore, Williams and Wilkins, 1976: 186-188
17. Davis KR, Ackerman RH, Kistler JP, Mohr JP: Computed
tomography of cerebral infarction: hemorrhagic, contrast
enhancement and time of appearance. Comput Tomogr 1:
71-86, 1977
18. Adams RD, vander Eecken HM: Vascular disease of the
brain. Ann Rev Med 4: 213-252, 1953
VOL 11, N o 5, SEPTEMBER-OCTOBER 1980
19. Joynt RJ, Zimmerman G, Khalifett R: Cerebral emboli from
cardiac tumors. Arch Neurol 12: 84-91, 1965
20. McDowell FH: Cerebral embolism. In Vinken PJ, Broyn GW
(eds) Handbook of Clinical Neurology, Vascular Diseases of
the Nervous System, Part I, New York, American Elsevier
Publishing Co, 1972: 386^*14
21. Harrison MJG, Hampton JR: Neurologic presentation of
bacterial endocarditis. Br Med J 2: 148-151, 1967
22. Robinson MJ, Rudy J: Sequelae of bacterial endocarditis. Am
J Med 32: 922-928, 1962
23. Madow L, Alpers BJ: Cerebral vascular complications of
metastatic carcinoma. J Neuropathol Exp Neurol 11:
137-148, 1952
24. Gazzaniga AB, Dalen JE: Paradoxical embolism: its
pathophysiology and clinical recognition. Ann Surg 171:
137-142, 1970
25. Gleysteen JJ, Silver D: Paradoxical arterial embolism: collec-
tive review. Am Surg 36: 47-54, 1970
26. Meister SG, Grossman W, Dexter L, Dalen JE: Paradoxical
embolism. Am J Med 53: 292-298, 1972
27. Vost A, Wolochow DA, Howell DA: Incidence of infarcts of
the brain in heart disease. J Pathol Bacteriol 88: 463-470,
1964
28. Barron KD, Siqueira E, Hirano A: Cerebral embolism caused
by nonbacterial thrombotic endocarditis. Neurology
(Minneap) 10: 391-397, 1960
29. Kooiker JC, MacLean JM, Sumi SM: Cerebral embolism,
marantic endocarditis, and cancer. Arch Neurol 33: 260-264,
1976
30. Neufeld HN, Cadman NL, Miller AW, Edwards JE: Em-
bolism from marantic endocarditis as a manifestation of oc-
cult carcinoma. Proc Mayo Clin 35: 292-299, 1960
31. Remillard GM, Carpenter S: Cerebral embolism due to non-
bacterial thrombotic endocarditis. Can Med Assoc J 106:
573-576, 1972
32. Kirk RS, Russell JGB: Subvalvular calcification of mitral
valve. Br Heart J 3 1 : 684-692, 1969
33. Traill TA, Fortuin NJ: Mitral annulus calcification and
cerebral ischemia. Lancet 2: 860, 1979
34. Belcher JR, Somerville W: Systemic embolization and left
auricular thrombosis in relation to mitral valvulotomy. Br
Med J 2: 1000-1003, 1955
35. Ellis LB, Harken DE: Arterial embolization in relation to
mitral valvuloplasty. Am Heart J 62: 611-620, 1961
36. Fricdli B, Aerichide N, Grondin P, Campeau L: Thromboem-
bolic complications of heart valve prostheses. Am Heart J 81:
702-708, 1971
37. Vidne B, Erdman S, Levy MJ: Thromboembolism following
heart valve replacement by prosthesis: survey among 365 con-
secutive patients. Chest 63: 713-717, 1973
38. Wang Y: Evaluation of mitral commissurotomy for the
prevention of systemic embolism in mitral stenosis. Circula-
tion 22: 829, 1960
39. Wood P: An appreciation of mitral stenosis. Br Med J 1:
1051-1063, 1113-1124, 1954
40. Carter AB: Prognosis of cerebral embolism. Lancet 2:
514-519, 1965
41. Coulshcd N, Epstein EJ, McKendrick CS, Galloway RW,
Walker E: Systemic embolism in mitral valve disease. Br
Heart J 32: 26-34, 1970
42. Daley R, Mattingly TW, Holt CL, Bland EF, White PD:
Systemic arterial embolism in rheumatic heart disease. Am
Heart J 42: 566-581, 1951
43. Darling RC, Austen WG, Linton RR: Arterial embolism.
Surg Gynecol Obstet 124: 106-114, 1967
44. Harris AW, Levine SA: Cerebral embolism in mitral stenosis.
Ann Intern Med 15: 637-643, 1941
45. Hinton RC, Kistler JP, Fallon JT, Friedlich AL, Fisher CM:
Influence of etiology of atrial fibrillation on incidence of
systemic embolism. Am J Cardiol 40: 509-513, 1977
46. Loew DE, Harken DE, Ellis LB: Valvular heart disease: un-
diagnosed valvular involvement, concomitant coronary artery
disease and systemic embolization. Am J Cardiol 30:222-228,
1972
47. Szekely P: Systemic embolism and anticoagulant prophylaxis
 CEREBRAL EMBOLISM OF CARDIAC ORIGIN/fajfon and Sherman
in rheumatic heart disease. Br Med J 1: 1209-1212, 1964
48. Fleming HA, Bailey SM: Mitral valve disease, systemic em-
bolism and anticoagulants. Postgrad Med J 47: 599-604, 1971
49. Jordan RA, Scheifley CH, Edwards JE: Mural thrombosis
and arterial embolism in mitral stenosis. Circulation 3 :
363-367, 1951
50. Soderstrom N: Myocardial infarction and mural thrombosis
in the atria of the heart. Acta Med Scand Suppl 217: 1-114,
1948
51. Somerville W, Chambers RJ: Systemic embolism in mitral
stenosis: relation to the size of the left atrial appendix. Br Med
J 2: 1167-1169, 1964
52. Neilson GH, Galea EG, Hossack KF: Thromboembolic com-
plications of mitral valve disease. Aust NZ J Med 8: 372-376,
1978
53. Friedman GD, Loveland DB, Ehrlich SP: Relationship of
stroke to other cardiovascular disease. Circulation 38:
533-541, 1968
54. Rowe JC, Bland EF, Sprague HB, White PD: The course of
mitral stenosis without surgery: Ten- and twenty-year perspec-
tives. Ann Intern Med 52: 741-749, 1960
55. Casella L, Abelmann WH, Ellis LB: Patients with mitral
stenosis and systemic emboli. Arch Intern Med 114: 773-781,
1964
56. Olesen KH, Hansen JF, Lauridsen P: Systemic arterial em-
bolism after mitral valvulotomy. Scand J Thorac Cardiovasc
Surg 6: 52-56, 1972
57. Wilson JK, Greenwood WF: The natural history of mitral
stenosis. Can Med Assoc J 71: 323-331, 1954
58. Graham GK, Taylor JA, Ellis LB, Grcenberg DJ, Robbins
SL: Studies in mitral stenosis. II. A correlation of post-
mortem findings with the clinical course of the disease in one
hundred and one cases. Arch Intern Med 88: 532-547, 1951
59. Abcrnathy WS, Willis PW: Thromboembolic complications
of rheumatic heart disease. Cardiovasc Clin 5: 131-175, 1973
60. Bean WB: Infarction of the heart. III. Clinical course and
morphological findings. Ann Intern Med 12: 71-94, 1938
Downloaded from http://ahajournals.org by on April 4, 2022
61. Blackwood W, Hallpike JF, Kocen RS, Mair WGP: Athero-
matous disease of the carotid arterial system and embolism
from the heart in cerebral infarction: a morbid anatomical
study. Brain 92: 897-910, 1969
62. Garvin CF: Infarction in heart disease. Am J Med Sci 203:
473^76, 1942
63. Jordan RA, Miller RD, Edwards JE, Parker RL: Throm-
boembolism in acute and in healed myocardial infarction. Cir-
culation 6: 1-6, 1952
64. Levine J, Swanson PD: Nonatherosclerotic causes of stroke.
Ann Intern Med 70: 807-816, 1969
65. McDonald GSA, McCaughey WTE: Mural thrombosis and
myocardial infarction. J Irish Med Assoc 67: 173-176, 1974
66. Towbin A: Recurrent cerebral embolism. Arch Neurol
Psychiatry 73: 173-192, 1955
67. Juergens JL, Edwards JE, Achor RWP, Burchell HB: Prog-
nosis of patients surviving first clinically diagnosed myocardial
infarction. Arch Intern Med 105: 444-450, 1960
68. Hellerstein HK, Martin JW: Incidence of thromboembolic
lesions accompanying myocardial infarction. Am Heart J 33:
443^152, 1947
69. Thompson PL, Robinson JS: Stroke after acute myocardial
infarction: relation to infarct size. Br Med J 2: 457-459, 1978
70. Fulton RM, Duckett K: Plasma-fibrinogen and thromboem-
boli after myocardial infarction. Lancet 2: 1161-1164, 1976
71. DeMaria AN, Bommer W, Neumann A et al: Left ventricular
thrombi identified by cross-sectional echocardiography. Ann
Intern Med 90: 14-18, 1979
72. Friedberg CK: Diseases of the Heart. 3rd ed. Philadelphia,
WB Saunders, 1966: 546
73. Beer DT, Ghitman B: Embolization from the atria in
arteriosclerotic heart disease. JAMA 177: 287-291, 1961
74. Wolf PA, Dawber TR, Thomas HE, Kannel WB:
Epidemiologic assessment of chronic atrial fibrillation and risk
of stroke: the Framingham Study. Neurology (Minneap) 28:
973-977, 1978
75. Fogarty TJ: Sudden arterial occlusion: surgical aspects. Car-
diovasc Clin 3: 173-182, 1971
441
76. Bjerkelund CJ, Orning OM: The efficacy of anticoagulant
therapy in preventing embolism related to D.C. electrical con-
version of atrial fibrillation. Am J Cardiol 23: 208-216, 1969
77. Lown B: Electrical reversion of cardiac arrhythmias. Br Heart
J 29: 469^»89, 1967
78. McDonald CD, Burch GE, Walsh JJ: Prolonged bedrest in the
treatment of idiopathic cardiomyopathy. Am J Med 52:
41-50, 1972
79. Demakis JG, Proskey A, Rahimtoola SH et al: The natural
course of alcoholic cardiomyopathy. Ann Intern Med 80:
293-297, 1974
80. Rogers PL, Sherry S: Current status of antithrombotic
therapy in cardiovascular disease. Prog Cardiovasc Dis 19:
235-253, 1976
81. Davies MJ, Pomerance A: Pathology of atrial fibrillation in
man. Br Heart J 34: 520-525, 1972
82. Adams GF, Merrett JD, Hutchinson WM, Pollock AM:
Cerebral embolism and mitral stenosis: survival with and
without anticoagulants. J Neurol Neurosurg Psychiatry 37:
378-383, 1974
83. Askey JM, Cherry CB: Thromboembolism associated with
auricular fibrillation. JAMA 144: 97-100, 1950
84. Carter AB: The immediate treatment of cerebral embolism. Q
J Med 26: 335-348, 1957
85. Cosgriff SW: Prophylaxis of recurrent embolism of intracar-
diac origin. JAMA 143: 870-872, 1950
86. Cosgriff SW: Chronic anticoagulant therapy in recurrent em-
bolism of cardiac origin. Ann Intern Med 38: 278-287, 1953
87. Ebert RV: Anticoagulants in acute myocardial infarction.
Results of a cooperative clinical trial. JAMA 225: 724-729,
1973
88. Owren PA: The results of anticoagulant therapy in Norway.
Arch Intern Med 111: 158-165, 1963
89. Wright IS, Marple CD, Beck DF: Report of the committee for
the evaluation of anticoagulants in the treatment of coronary
thrombosis with myocardial infarction. Am Heart J 36:
801-815, 1948
90. Wright IS, McDevitt E: Cerebral vascular diseases: their
significance, diagnosis and present treatment, including the
selective use of anticoagulant substances. Lancet 2: 825-830,
1954
91. Medical Research Council Report: Assessment of short-term
anticoagulant administration after cardiac infarction. Br Med
J 1: 335-342, 1969
92. Chalmers TC, Matta RJ, Smith H, Kunzler AM: Evidence
favoring the use of anticoagulants in the hospital phase of
acute myocardial infarction. N Engl J Med 297: 1091-1096,
1977
93. Symonds C: Cerebrovascular accidents. Practitioner 176:
130-136, 1956
94. Moycs PD, Millikan CH, Wakim KG, Sayre GP, Whisnant
JP: Influence of anticoagulants on experimental canine
cerebral infarcts. Proceedings of the Staff Meetings of the
Mayo Clinic 32: 124-130, 1957
95. Wood MW, Wakim KG, Sayre GP, Millikan CH, Whisnant
JP: Relationship between anticoagulants and hemorrhagic
cerebral infarction in experimental animals. Arch Neurol
Psychiatry 79: 390-396, 1958
96. Sibley WA, Morledge JH, Lapham LW: Experimental
cerebral infarction: the effect of Dicumarol. Am J Med Sci
234: 663-677, 1957
97. Ushiro CS, Schaller WF: Anticoagulation therapy in cerebral
thrombosis and embolism. Neurology (Minneap) 7: 253-258,
1957
98. Marshall J: The Management of Cerebrovascular Disease. 3rd
ed. Oxford, Blackwell Scientific Publications, 1976: 112
99. Pastore BH, Resnick ME, Rodman T: Serious hemorrhagic
complications of anticoagulant therapy. JAMA 180: 747-751,
1962
100. Wright IS, Foley WT: Use of anticoagulants in the treatment
of heart disease. Am J Med 3: 718-739, 1947
101. Loeliger EA, Hensen A, Kroes F et al: A double-blind trial of
long-term anticoagulant treatment after myocardial infarc-
tion. Acta Med Scand 182: 549-566, 1967
102. Hilden T, Iversen K, Raaschow F, Schwartz M: Anti-
 442 STROKE
coagulants in acute myocardial infarction. Lancet 2: 327-331,
1961
103. Lieberman A, Hass WK, Pinto R et al: Intracranial
hemorrhage and infarction in anticoagulated patients with
prosthetic heart valves. Stroke 9: 18-24, 1978
104. Nichol ES, Keyes JN, Borg JF et al: Long-term anti-
coagulant therapy in coronary atherosclerosis. Am Heart J 55:
142-152, 1958
105. Wasserman AJ, Guttman LA, Yoe KB: Anticoagulants in
acute myocardial infarction. Am Heart J 7 1 : 43-49, 1966
106. Bannister RG: Risk of deferring valvulotomy in patients with
moderate mitral stenosis. Lancet 2: 329-332, 1960
107. Estes D: The effect of mitral commissurotomy on the natural
history of rheumatic heart disease. Am J Med 30: 449-463,
1961
108. Greenwood WF, Aldridge HE, McKelvey AD: Effect of mitral
commissurotomy on duration of life, functional capacity,
hemoptysis and systemic embolism. Am J Cardiol 11:
349-356, 1963
109. Larson DG, Estes D: Effect of mitral commissurotomy on
natural history of rheumatic heart disease. Am J Med 30:
449^63, 1961
110. Lowther CP, Turner RWD: Deterioration after mitral
valvulotomy. Br Med J 1: 1027-1036, 1102-1107, 1962
111. Ricordeau G, Coblentz B, Lenegre J: Embolics arterielles du
retrecissement mitral et commissurotomie. Arch Mai Coeur
Summary of 12th Conference on Cerebrovascular
Disease
Downloaded from http://ahajournals.org by on April 4, 2022
Williamsburg, Virginia, March 2-4, 1980
FLETCHER H. MCDOWELL, M.D.,
THE 12TH CONFERENCE on Cerebrovascular
Disease was held in Williamsburg, Virginia, March
2-4, 1980. The first session of the Conference was
devoted to a review of functional imaging of the brain
with in vivo biochemistry and perfusion studies.
The initial presentation was "Scanning Systems and
Analytical Methods using Positron Emission
Tomography" given by Dr. Marcus E. Raichle of the
Washington University School of Medicine, St. Louis
MO. He described current detector isotope strategy
mentioning that oxygen-15, nitrogen-13, carbon-11
and fluoride-18, which have half-lives of 2, 10, 20 and
110 minutes respectively, were the radioactive sub-
stances currently in use. The detector resolution
capability is about 6.0 mm at present and the
theoretical lower limit of resolution is about 1.5 mm.
Sodium iodide detectors are commonly used but more
sensitive detectors, such as those made with bismuth
germanate and cesium fluoride, are being developed.
Dr. Raichle reported that brain blood volume,
cerebral metabolism, tissue chemical composition,
cerebral blood flow and the permeability of the blood-
brain barrier can be measured by positron emission
tomography. He reviewed the details of measurement
of cerebral blood flow with a PET scanner and showed
examples of scans using H2O1S in patients with stroke.
VOL 11, No 5, SEPTEMBER-OCTOBER 1980
50: 112-125, 1957
112. Kellogg F, Liu CK, Fishman IW, Larson R: Systemic and
pulmonary cmboli before and after mitral commissurotomy.
Circulation 24: 263-266, 1961
113. Tabcr RE, Lam CR: Significance of atrial fibrillation and
arteriai embolization in rheumatic mitral valve disease. Cir-
culation 22: 821, 1960
114. Selzer A, Cohn KE: Natural history of mitral stenosis: a
review. Circulation 45: 878-890, 1972
115. Bailey CP, Olsen AK, Keown KK, Nichols HT, Jamison WL:
Commissurotomy for mitral stenosis. Technique for preven-
tion of cerebral complications. JAMA 149: 1085-1091, 1952
116. Starr A: Mitral valve replacement with ball valve prosthesis.
Br Heart J 33 (suppl): 47-55, 1971
117. Kittle CF, Dye WS, Gerbode F et al: Factors influencing risk
in cardiac surgical patients: cooperative study. Circulation 40
(suppl I): 169-171, 1969
118. Starr A, cited by Kirklin JW, Pacifico AD: Surgery for ac-
quired valvular heart disease. N Engl J Med 228: 133-140,
194-199, 1973
119. Fishman NH, Edmunds LH, Hutchinson JC, Rowe BB: Five-
year experience with the Smeloff-Cuttcr mitral prosthesis. J
Thorac Cardiovasc Surg 62: 345-356, 1971
120. Pluth JR, McGoon WC: Current status of heart valve replace-
ment. Mod Concepts Cardiovasc Dis 43: 65-70, 1974
121. Anticoagulants in mitral valve disease. Br Med J 1: 641, 1972
AND CLARK H. MILLIKAN, M.D.
He reported that there are now available many radio-
labeled Pharmaceuticals whose action and localization
can be traced through the central nervous system.
Utilization and consumption of metabolites can be
measured using radioactive oxygen in water or in car-
bon dioxide.
Functional mapping of metabolism in the brain
using positron emission tomography in animals was
discussed by Dr. Martin Reivich of the University
of Pennsylvania. He commented that there are very
tight spatial links between function and anatomy in
the rat; stimulating one whisker can cause a localized
increase in blood flow and cerebral metabolism as
demonstrated by the autoradiographic appearance of
the rat brain. Dr. Reivich discussed his experience
using fluorodeoxyglucose to study visual, tactile and
auditory stimuli and their effect on local cerebral
metabolism in man. In 6 human subjects, visual
stimuli were applied to the left visual field which im-
mediately produced a higher metabolic rate in the
right visual cortex and when the right visual field was
stimulated, the left visual cortex showed increased
blood flow and metabolism. Tactile stimuli to one
hand was accompanied by an increase of blood flow
and metabolism in localized areas of the cerebral cor-
tex in the contralateral sensory areas. Auditory