close

Вход

Забыли?

вход по аккаунту

?

Carotid and vertebral-basilar transient ischemic attacks Effect of anticoagulants hypertension and cardiac disorders on survival and stroke occurrenceЧ A population study.

код для вставкиСкачать
ORIGINAL ARTICLES
Carotid and Verte bral-Basilar Transient
Ischemic Attacks: Effect of Anticoagulants,
Hypertension, and Cardiac Disorders on
Survival and Stroke OccurrenceA Population Study
Jack P. Whisnant, MD, Niall E. F. Cartlidge, MB, and Lila R. Elveback, PhD
A population of 199 patients from Rochester, MN, was followed from the time of their first carotid or vertebralbasilar transient ischemic attack (TIA). Patients treated with anticoagulants had no significant difference in survival
from untreated patients. Among patients with carotid TIA who received anticoagulants, the net probability of stroke
was slightly but not significantly lower than in untreated patients. The difference favoring treated patients with
vertebral-basilar TIA was significant starting at three months.
The rate of intracranial hemorrhage was higher among all patients receiving anticoagulant treatment than among
untreated patients and was significantly higher among those 55 to 74 years old. Almost all the hemorrhages occurred
after a year or more of anticoagulant treatment and in patients more than 65 years old.
Patients with high diastolic blood pressure had a significantly higher net probability of stroke than did patients
with lower blood pressure and those receiving antihypertensive drugs. By implication, treatment of hypertension was
effective in preventing stroke in patients with TIA.
Linear discriminant analysis and actuarial analysis indicated that diastolic blood pressure and anticoagulant
therapy were the only factors that influenced stroke occurrence. There was no suggestion that previous myocardial
infarction, anginapectoris, valvular heart disease, cardiac arrhythmia, or congestive heart failureindividually or in
combination-influenced the occurrence of stroke or survival.
Whisnant JP, Cartlidge NEF, Elveback LR: Carotid and vertebral-basilar transient ischemic attacks: effect of
anticoagulants, hypertension, and cardiac disorders on survival and stroke occurrence-a population study.
Ann Neurol 3:107-115, 1978
The population of the community of Rochester, M N ,
represents a unique resource for study of the full
spectrum of disease processes [ 5 ] . Since the early part
of this century, most of the medical care for the
residents of Rochester has been provided by physicians of the Mayo Clinic. Since the early days of the
clinic, a unit record system has existed so that all of the
care of an individual-whether outpatient care, hospital care, home care, or emergency room services-is
available in a single record. The record is a comprehensive one, and details on the history of illnesses
and on examinations are almost always well recorded.
Autopsies in the community are performed by Mayo
Clinic pathologists, and the reports complement the
clinical medical record.
In the mid-1950s another group practice of
medicine began in Rochester, largely for the general
practice of medicine. In recent years a record linkage
system with the Mayo record system has been implemented for that group practice and its associated
small community hospital; the central file of medical
records compiled for the Rochester Epidemiologic
Project thus includes those of the Mayo Clinic and the
Olmsted Medical and Surgical Group as well as those
of other medical institutions in and around Rochester
[ 5 ] . Therefore, all of the medical information continues to be available for each person in the community who receives medical care.
The records are coded by diagnosis and may be
retrieved by diagnosis. After the diagnostic categories
that would encompass a particular illness under study
have been determined, records can be reviewed to see
whether an individual patient meets the criteria established by the investigators. With this kind of informa-
From the Cerebrovascular Clinical Research Center and the Section
of Medical Research Statistics, Mayo Clinic and Mayo Foundation,
Rochester, MN.
Accepted for publication Aug 8, 1977.
Address reprint requests to Dr Whisnant, Mayo Clinic, Rochester,
MN 55901.
0364-5 134/78/0003-0203$01.25 @ 1978 by the American Neurological Association 107
tion available, it is as if one had set up a prospective
study i n any year a n d k e p t records from t h a t p o i n t o n
as a part o f t h e regular practice of medicine b u t without a prescribed plan of management.
We have used this c o m m u n i t y r e s o u r c e for the
study of stroke and transient ischemic attacks (TIA).
Patients were selected from the total Mayo Clinic
practice o n the basis of residence i n the c o m m u n i t y of
Rochester; the study d o e s n o t include M a y o Clinic
patients with TIA who are n b t residents of t h e city.
Selection did n o t depend o n the m e m o r y of patients
for transient episodes, because the e v e n t s had b e e n
recorded i n t h e medical record if t h e patient was s e e n
by a physician or if t h e episode was r e p o r t e d at the
t i m e of a s u b s e q u e n t visit. Identification t h u s was
assured o f practically all residents who were s e e n by a
physician and for whom a diagnosis of TIA was made
or who r e p o r t e d such an e p i s o d e at t h e time of a later
visit.
Among t h o s e p e r s o n s i n the c o m m u n i t y who had a
s t r o k e and d i e d from 1955 through 1969,50% had a n
autopsy, including examination of the head. The autopsy rate is a little h i g h e r for deaths from other causes
and for y o u n g e r persons in the community.
We have previously n o t e d t h e incidence of first
episodes o f T I A [lo]. We h a v e also n o t e d that patients
with TIA treated with oral anticoagulants have less
probability of s t r o k e t h a n u n t r e a t e d patients, with
most of the difference b e i n g achieved i n t h e first few
m o n t h s of t r e a t m e n t [9]. T h e p u r p o s e s of t h e p r e s e n t
study were to define t h e effect and risk of anticoagulant therapy i n carotid and vertebral-basilar TIA separately as well as t o define t h e effect of hypertension
a n d cardiac disorders o n survival a n d probability of
s t r o k e in patients with TIA in this c o m m u n i t y .
Methods
The study included all patients who had their first episode of
TIA while residents of Rochester, MN. during the fifteenyear period 1955 through 1969.T h e diagnostic criteriawere
established by the investigators. All of o u r medical records
for patients with a diagnosis of TIA were reviewed for the
period 1955 through 1969. Records were also reviewed for
patients who had evidence of other transient neurological
symptoms that might have represented TIA. The records
of patients with the diagnosis of stroke were reviewed for
those who might have had transient symptoms before the
onset o f t h e stroke. The study does not include patients who
had a completed stroke before the first TIA occurred; nor
does it include patients with TIA whose first episode was
before 1955.
Any one or combination of the following symptoms was
considered to represent transient cerebral ischemia in the
carotid system: (1) motor or sensory symptoms limited to
one side of the body, (2) aphasia or dysphasia, and (3) retinal
ischemia (amaurosis fugax).
The vertebral-basilar system was considered the source of
108 Annals of Neurology Vol 3 No 2
February 1978
the TIA when any of the following symptoms occurred: (1)
motor or sensory symptoms (or both) occurring bilaterally
in the same attack, (2) ataxia of gait or clumsiness of the
extremities on both sides, (3) diplopia, (4) dysphagia, and (5)
bilateral homonymous hemianopia. Although TIA in the
vertebral-basilar system may be manifested by unilateral
motor or sensory symptoms, we did not consider such
symptoms indicative of TIA unless there were associated
symptoms indicating brainstem or cerebellar ischemia.
Vertigo was not considered to represent TIA when it
occurred alone, but the symptom was judged to arise from
the vertebral-basilar system if it occurred with other appropriate symptoms. Diplopia alone was not considered TIA,
nor was unexplained unconsciousness, syncope, or focal
symptoms associated with migraine. Patients who complained simply of transient “dizziness” were not included in
the study.
Either dysarthria or homonymous hemianopia was
judged to be compatible with either arterial system when
associated with other appropriate symptoms. When dysarthriaoccurred alone, the symptom was considered to be of
uncertain origin. When homonymous hemianopia occurred alone, the disorder was judged to be in the
vertebral-basilar system. When symptoms occurred from
both the carotid and the vertebral-basilar distributions, they
were considered to represent disorder of both systems, and
this group was combined for analysis with the group having
uncertain distribution.
Follow-up for all patients was carried out to July, 1975, for
determination of mortality or occurrence of stroke. Survival
was determined by actuarial analysis [2],starting with the
date of the first TIA. When expected survival is being
analyzed, it is for a population of the same age and sex as the
sample and is based on the death rates for the Minnesota
white population for 1960.
For actuarial estimation of net probability of stroke, when
death occurred from a cause other than stroke, the patient
was no longer included among the patients under observation. Therefore, in the actuarial analysis, stroke is the end
point, analogous to death in the usual survival curve. The net
probability of stroke is simply the complement of that. This
allows evaluation of stroke exclusive of other considerations
in the same sample of patients. Expected stroke rates are for
a population of the same age and sex as the sample and are
based on stroke incidence rates for Rochester, MN, from
1955 through 1969 [6].
Patients receiving anticoagulant therapy must have been
started on anticoagulants, either intravenous heparin or oral
coumarin drugs, within two months of the date of the first
TIA. If the medication was later stopped, the patient was
dropped from observation at that point. For patients not on
anticoagul%nts,anyone who was later given long-term anticoagulant therapy for any reason was also dropped from
further observation at the time treatment was begun. These
two groups represent a total of 12 patients.
When oral anticoagulants were administered, the aim was
to maintain the prothrombin time at about 2 times the
control value, in seconds, with the therapeutic range considered to be lY2 to 2?4 times the control value. It was unusual
for a patient to have a prothrombin time outside this range
when it was checked about once every four weeks for longterm therapy. Treated and untreated patients were scattered
throughout the study period.
When blood pressure was considered, the value used was
the diastolic pressure at the time the patient was initially
seen after the first TIA. The patients noted to be receiving
antihypertensive therapy were being given this medication
at the time of their visit after the first TIA or within three
months. There may be other patients in whom antihypertensive therapy was subsequently started. The patients treated
with antihypertensivemedication all received various drugs
or combinations of drugs known to cause a reduction in
blood pressure. However, they were considered treated
regardless of how much reduction in blood pressure occurred. Each of the patients treated with antihypertensive
medicanon is also represented in one of the three blood
pressure groups.
The cardiac disorders considered were as follows: (1)
documented history of myocardial infarction; ( 2 ) angina
pectoris; ( 3 ) valvular heart disease of any type; ( 4 )
documented history of cardiac arrhythmia or its presence
either at the time the patient was seen for the episode or
within the preceding six months (premature contractions
were not considered arrhythmia); and ( 5 ) documented
history or presence of congestive heart failure. Cardiac
hypertrophy, as judged by chest roentgenogram or electrocardiogram, was not included in this group because
hypertension was considered separately.
Approximately 90% of the patients were seen for diagnosis and management by an internist o r neurologist (or
both) at the Mayo Clinic. The others were seen by ageneral
physician or internist (or both) of the Olmsted Medical and
Surgical Group. The laboratory facilities used for control of
prothrombin times were almost always those of the Mayo
Clinic.
Results
During the period 1955 through 1969, 199 patients
were detected who had their first TIA. There were
122 patients with carotid TIA, 64 patients with
vertebral-basilar TIA, and 13 patients who had symp-
toms indicating that both systems were involved or in
whom it was uncertain which system was implicated.
When carotid and vertebral-basilar T I A were compared, the 13patients with either uncertain location or
both systems involved were excluded.
T h e probability of survival with and without anticoagulant treatment after the onset of TIA was compared with expected survival for each group (Table 1).
There was n o significant difference between survival
of treated and untreated patients or between survival
of treated patients and expected survival. However,
survival was significantly lower than expected for the
untreated patients (p < 0.05 at one year and < 0.01 at
five years).
Patients with carotid T I A who were not treated with
anticoagulants also had significantly lower survival
than expected through five years of observation.
Treated patients with carotid T I A and treated and
untreated patients with vertebral-basilar T I A had no
significant differences from the corresponding expected survival.
T h e net probability of stroke occurrence after the
onset of TIA in patients treated with anticoagulants
and those not treated is shown in Table 2. These data
are further compared with the expected probability of
stroke based o n the age-specific incidence rates for
stroke for Rochester, MN [6]. At three months there
was a significantly lower probability of stroke in the
patients treated with anticoagulants than in the untreated patients (p < 0.01), and the difference was still
evident at five years of observation. However, starting
with only those patients who were alive and had not
had a stroke at six months, the difference favoring
treated patients for the remainder of the observation
period was not significant. Both treated and untreated
patients had significantly higher stroke rates than the
expected rate for the same age and sex distribution.
Patients with carotid TIA who were treated with
Table 1 . Actuarial Estimation for Suwival of Patients with T I A with and without Anticoagulant Treatmenta
No. of Persons Alive
at Start of Interval
Survival Rate
Standard
Error of Rate
Expected
Survival Rateh
Years
after
No
Onset
Treatment
Treatment
Treatment
Treatment
0
1/2
1
2
130
109
101
89
80
73
61
69
62
59
48
40
1.000
0.906
0.872
0.802
0.729
0.674
0.598
0.926
0.896
0.83 1
0.742
0.703
0.660
3
4
5
No
35
28
1.000
No
Treatment
Treatment
No
Treatment
Treatment
0
0.026
0.030
0.036
0.041
0.044
0.046
0
0.032
0.037
0.047
0.056
0.060
0.063
0.943
0.889
0.837
0.787
0.738
0.954
0.907
0.861
0.815
0.770
=Average age of treated patients, 69.5 years; average age of untreated patients, 69.8 years.
"Corresponding population of same age and sex (Minnesota white population, 1960).
Whisnant et al: Carotid and Vertebral-Basilar Transient Ischemic Attacks
103
Table 2. Actuarial Estimation of Net Probability of Stroke Occurrence in Patient1 with TIA
with and without Anticoagulant Treatmenta
~
No. of Persons Alive
and Free of Stroke
at Start of Interval
Estimated
Net Probability
of Stroke
~~
Standard
Error of Estimate
Expected
Years after
First TIA
No
Treatment
0
130
1/12
111
101
3/12
6/12
1
2
94
82
67
59
52
44
3
4
5
Treatment
No
Treatment
69
65
64
62
58
46
39
34
27
0
0.132
0.156
0.181
0.235
0.313
0.376
0.409
0.433
Treatment
No
Treatment
Treatment
Rateh
0
0.044
0.044
0.074
0.089
0.173
0.192
0.214
0.214
0.030
0.032
0.034
0.038
0.043
0.046
0.048
0.049
0.025
0.025
0.032
0.035
0.048
0.050
0.054
0.054
0.02
b.03
0.04
0.05
Stroke
0.01
'Average age of treated patients, 69.5 years; average age of untreated patients, 69.8 years.
hFrom stroke incidence rates for Rochester, MN, 1955-1969 [ 6 ] .
anticoagulants had a slightly lower probability of
stroke occurrence than did untreated patients, but the
differences were not significant (Fig 1A). For patients
with vertebral-basilar TIA, the observed differences
favoring treated patients, noted in Figure lB, were
significant (s < 0.01) at three months and through five
years of observation. In the latter group, however, if
one starts the observation with those patients who
were alive and free of stroke at six months after the
first TIA, the observed difference between treated
and untreated patients is not significant.
Patients whose TIA was less than one hour in duration had no significant difference in stroke occurrence
from those whose episode lasted an hour or longer,
whether they had been treated with anticoagulants or
not.
The type of first stroke for each category of TIA is
shown in Table 3. Subdural hematoma is considered
stroke in this table and the next because of its likely
relationship to treatment. Nine patients had intracranial hemorrhage as their first stroke after TIA; 6 of
these were receiving oral anticoagulant therapy at the
time of the hemorrhage.
To assess the risk of intracranial hemorrhage from
anticoagulant treatment, we determined the number
of patients times the number of years of observation
(person-years) for each one receiving anticoagulants
and each not receiving them. The number of hemorrhages per 100 person-years for each age group on and
off anticoagulants was determined (Table 4 ) . The
age-adjusted rate for the time the patients were o n
anticoagulants was 1.14 hemorrhages per 100 person-years of observation, compared with 0.31 for
the time off anticoagulants. Among patients aged 55
to 7 4 years there were 1.69 hemorrhages per 100
Fig 1 . Net probability of occurrence ofstroke afterfirst TIA
in (A) carotid arterial distribution and (B)
certebral-basilararterial distribution in patients u'ho
receivedanticoagulants (A/C)and those who did not.
P
$
'*I
2 80
8
* ......
.....
40
....
*....-------
A/C trwtmeflt
n.47
4
6
8
90
ICorotid 1
A
Vol 3
...------
..........
No 2 February 1978
No treatment
n-47
* .......*
A I C treatment
1747
....
Q
.
2
Years after first T I A
110 Annals of Neurology
........
n= 75
..
......
.......
6
.
..-...... ..+ No treatment
.,
2
4
6
8
Years after first T I A
I0
Table 3 . Type of First Stroke after T I A
Type of Stroke
Cerebral infarct
Brainstem infarct
Cerebral, unknown
Brainstem, unknown
Cerebral hemorrhage
Brainstem hemorrhage
Subarachnoid hemorrhage
Subdural hematoma
Uncertain site or type
Total no. of strokes
Carotid
(N = 122)
Vertebral-Basilar
(N = 64)
Mixed or Uncertain
(N = 13)
29
6
14
11
1
3
0
1
6
0
3
2
0
0
0
0
1
1
0
1
2
48
0
0
0
1
6
0
3
31
Table 4 . Anticoagulants and lntracranial Hemorrhage in Rochester Patients with T I A
On Anticoagulant Therapy
Off Anticoagulant Therapy
Intracranial
Hemorrhage
Intracranial
Hemorrhage
Age (yr)
PY
<35
35-44
45-54
55-64
65-74
275
. . .
2.5
9.1
7 5.0
220.4
189.6
No.
1
4
1
Ageadjusted
rate
PY
=
No./100 PY
PY
No.
No./100 PY
1.33
1.81
0.53
251.8
91.4
125.6
145.2
334.7
489.2
1
2
0.30
0.41
1.14
0.31
person-years (number of patients times number of years of observation).
lants but had stopped many months before the hemorrhage occurred.
Actuarial estimation of survival and stroke occurrence among patients with TIA was calculated for
subgroups on the basis of diastolic blood pressure
determined at the time of onset of TIA. The groups
compared were: (1) diastolic pressures less than 85
mm Hg, (2) 85 lo 104 mm Hg, (3) 105 mm Hg or
more, and (4)those among the other three groups who
were being treated with antihypertensive medication.
person-years with anticoagulants and 0.2 1 hemorrhage per 100 person-years without. This difference is
significant (p < 0.03).
Among the 14 patients with intracranial hemorrhage described in Table 5 , the hemorrhage was the
second stroke in 5 ; 9 were receiving anticoagulants at
the time the hemorrhage occurred. In all but 1 patient
the anticoagulant had been taken for more than a year,
and in the exception it was for eleven months. Two
other patients had previously taken oral anticoagu-
TabLe 5 . lntracranial Hemorrhage in Patients with TIA"
Sex
Type of Hemorrhage
No.
Age (yr)
Intracerebral
Brainstern
Subarachnoid
Subdural
Total
9
2
2
1
14
60-87
65-70
82-90
65
60-90
.
M
F
2
2
0
1
5
7
0
2
0
9
Treatment at Time of Hemorrhage
None
Anticoagulant
(Duration)
0
2
0
3
6
2
0
1
5
9
(11-43 mo)
(27-57 mo)
(24 mo)
(1 1-57 mo)
"Blood pressure: systolic 120-190, >160 in 3 patients; diastolic 70-110, >90 in 2 patients.
Whisnant et al: Carotid and Vertebral-Basilar Transient Ischemic Attacks
111
loot
1
1
I
1
sure might have contributed to the difference in effect
of anticoagulant treatment, the net probability of
stroke occurrence was determined in patients who
received anticoagulants and those who did not among
those with a diastolic blood pressure of less than 105
mm Hg at onset (Table 6). There was a significantly
lower rate of stroke occurrence in treated patients
than in the untreated group (J < 0.001) starting at one
month and still evident at five years of observation.
For the entire group of patients with TIA, multivariate methods were used to search for risk factors
for the occurrence of stroke within one year after the
first TIA. Stepwise linear discriminant analysis was
used to select those factors that gave the best discrimination between the groups, namely, those who had a
stroke within one year and those who survived one
year without a stroke. The following variables were
included: (1) anticoagulant therapy begun less than
60 days after the first TIA; (2) diastolic blood pressure
at the time of TIA (as a continuous variable); and the
following diagnoses made prior to or at the time of the
TIA: ( 3 ) myocardial infarction, ( 4 )angina pectoris, (5)
valvular heart disease, (6) cardiac arrhythmia, or (7)
congestive heart failure.
The most significant variable selected by this process was diastolic blood pressure (p < 0.01). With diastolic pressure selected, anticoagulant therapy was also
a highly significant variable (p < 0.01). There was no
suggestion that the other variables made any further
contribution to the discrimination of the groups.
Stepwise linear discriminant analysis was then performed, excluding those patients with diastolic blood
pressure of 105 mm Hg or greater. With this analysis,
diastolic pressure was of no importance at all, and the
only significant variable selected was anticoagulant
therapy (p < 0.001).
I
1
2
3
4
Years after 1st T I A
5
Fig 2 . Net probability of occurrence of stroke according to
lei,elofdiastolic bloodpressureat onset of TIA: 5 I05 m m
H g (N = 32),85 to 104 n m Hg ( N = 82),and < 85
rnni Hg (N= 8 5 ) , and in those receiving
antihypertensive treatment (NHRx) ( N = 30).
(Froni Whisnant JP: A population study of stroke and TIA:
Rochester, Minnesota, in Gillingham FJ. Mawdsley C ,
Williams AE (eds):Stroke. Edinburgh and London,
Churchillllizingstone. 1976, p p 21 -39. By
permission.)
There was no significant difference in survival among
these groups through five years of observation.
The net probability of stroke occurrence among the
patients with TIA according to the status of their
diastolic blood pressure is shown in Figure 2. Patients
with diastolic pressures of 105 mm Hg or greater had a
significantly higher net probability of stroke occurrence than did the other three groups. The difference
between the highest and the lowest blood pressure
groups was significant at a p value of less than 0.01 at
six months and through three years of observation.
Because of the possibility that elevated blood pres-
Table 6. Actuarial Estimation of Net Probability of Stroke Occurrence with and without Anticoagulant
Treatment in Patients with TIA Who Had Diastolic Blood Pressures less than 105 mm Hg'
No. of Persons Alive
without Stroke at
Start of Interval
Estimated
Net Probability
o f Stroke
Standard
Error of Estimate
Years
after
Onset
Treatment
No Treatment
Treatment
No Treatment
Treatment
No Treatment
Expected
Stroke
Rateh
0
1/12
3112
6/12
1
2
60
60
59
58
54
42
107
92
84
79
69
57
50
43
0
0
0
0
0.018
0.115
0.138
0.164
0.164
0
0.123
0.142
0.163
0.207
0.280
0.344
0.385
0.416
0
0
0
0
0.018
0.044
0.049
0.054
0.054
0
0.032
0.034
0.036
0.041
0.046
0.050
0.053
0.054
0.01
0.02
0.03
0.04
0.05
3
35
4
5
30
25
35
HAverageage of treated patients, 69.6 years; average age of untreated patients, 70 years.
'Expected rates derived from age-specific rates for Rochester, MN [6].
112
Annals of Neurology
Vol 3
No 2 February 1978
given medications, when indicated, for other disorders. None of these patients underwent carotid endarterectomy.
Among the patients in this community, on the basis
of actuarial analysis, those who were treated with anticoagulants showed no significant difference in survival from untreated patients (see Table 1). This is not
unexpected, since the primary cause of death in patients with either carotid or vertebral-basilar T I A is
cardiac disease [ 1,101. Because untreated patients had
a slightly higher proportion of deaths due to stroke in
addition to the deaths from cardiac disease, survival in
that group is significantly lower than expected. A similar significant difference was demonstrated among untreated patients with carotid TIA.
In the whole group of patients with TIA, those
given anticoagulant treatment had a significantly lower
probability of stroke than untreated patients (see Table
2), as previously noted [91. A significant difference
favoring treated patients was also demonstrated for
patients with vertebral-basilar TIA (see Fig 1B).
Among those with carotid TIA, however, treated patients had a slightly lower probability of stroke, but
the differences from untreated patients were not significant (see Fig 1A).
The most imporrant risk of anticoagulant therapy is
intracranial hemorrhage. In our study the ageadjusted rate of this complication per 100 personyears of observation on anticoagulant treatment was
1.14. This is 3.7 times the rate (0.31) for patients not
on anticoagulants. For the age group 5 5 to 74 years the
rate was 1.69 per 100 person-years on anticoagulants
and 0.21 off (p < 0.03). The rate is more than 8 times
greater in treated than in untreated patients. There
was little difference in the number of hemorrhages per
100 person-years over age 75 (see Table 4).
O f the 14 patients who suffered intracranial hemorrhages after the first TIA (see Table 5 ) , 5 had the
hemorrhage as a second stroke. Among these 14 patients, 9 were receiving anticoagulants at the time of
the hemorrhage, but in only 1 patient did the hemorrhage occur in the first year of treatment, and in that
Actuarial estimation of survival was accomplished
by comparing two groups of patients relative to cardiac disorders: (1) patients with history or presence of
myocardial infarction, angina pectoris, valvular heart
disease, arrhythmia, or congestive heart failure-any
one or any combination-and (2) patients with none
of these factors (Fig 3A). There was no significant
difference in survival between these two groups; however, the patients with cardiac disorders had significantly lower survival than expected starting at one
year and throughout five years of observation.
Actuarial estimation of the net probability of stroke
occurrence was also calculated for these two groups
relative to cardiac disorders (Fig 3B). There was no
significant difference in stroke occurrence for the two
groups through five years of observation.
Discussion
Since Millikan and associates r7] advocated the use of
anticoagulant therapy in patients with vertebralbasilar and carotid arterial distribution of transient
cerebral ischemia, there has been controversy in the
literature concerning whether possible benefits of
treatment are sufficient to justify the risk, primarily of
intracranial hemorrhage.
The present study is not a case-control randomized
evaluation of treatment but represents a group extracted from our total practice, selected only on the
basis of residence. Whether anticoagulant or antihypertensive treatment was given was determined
by the internist or neurologist managing the patient. If
there was a collective bias favoring treatment o r lack
of treatment in a particular portion of the sample, it
could not be identified. No other definitive treatment
was given for patients with TIA, though they were
F i g 3. (A) Probability of surcivaland IB) netprobubility of
stroke occurrence i n patients with TIA who have any k n o w 1
cardiac disorders (myocardial infarction. ungina pectoris,
valz!ular heart dixeu.wsarrhythmia, or congestiveheart
fuilrrre?singly or any conzbination) compared with
patients with TIA who haw none of these factors.
100 L
z
y
?
.
._
c
D
c
0
Z
.--
40
n
n
m
E
L
80
- _---_No
60
-
L
20
0
- ----_
No cardiac disorders n -110
- -Cardiac disorders n 4 9
I
I
I
1
cardiac disorders n -110
-Cardiac
disorders n -89
)
.
I
._
-
I
Whisnanr er al: Carotid and Vertebral-Basilar Transienr Ischemic Attacks
1 I3
patient it was at eleven months after treatment had
begun. All of the intracranial hemorrhages occurred in
patients aged 60 years or older; all but one were in
patients more than 65. Hypertension did not appear
to be an influential factor in the patients who experienced intracranial hemorrhage.
The excessive number of intracranial hemorrhages
occurred among patients receiving anticoagulant
treatment who were being rather rigidly controlled in
a setting in which physicians were experienced in
long-term management of anticoagulant therapy and
laboratory resources were highly reliable. Other
studies, such as that by Heyman and associates [ 3 ] ,
have previously demonstrated a significant risk of intracranial hemorrhage among male veterans in the
United States who were on anticoagulant therapy for
any reason, but there was clear evidence of excess
anticoagulation in those patients.
Elevated blood pressure has long been considered
one of the risk factors for stroke. Only in recent years
has pharmacological control of elevated blood pressure been shown to have reduced the occurrence of
strokes [ 8 ] . However, another study has shown that
after a stroke has occurred, the recurrence rate is not
lowered by antihypertensive therapy [4]. Therefore,
the risk of recurrence is great enough that it is not
further influenced by control of elevated blood pressure.
In our study, actuarial analysis showed that patients
who had a diastolic blood pressure of 105 mm H g or
greater at the time of their initial visit after the first
TIA had a significantly higher probability of stroke
than patients with lower blood pressure and those
receiving antihypertensive agents. This is an implied
benefit of antihypertensive therapy, but what the
blood pressure would have been without treatment is
unknown, and the duration of hypertension may not
have been the same in the two groups.
Stepwise linear discriminant analysis selected diastolic blood pressure as the most important variable
relating to stroke occurrence in the first year of observation after the onset of TIA. Anticoagulant therapy
was also asignificantvariable. There was no suggestion
that previous myocardial infarction, angina pectoris,
valvular heart disease, cardiac arrhythmia, or history
of congestive heart failure-ither
individually or in
combination-helped
to distinguish patients with a
propensity for stroke during the first year of observation. The lack of effect of these cardiac disorders on
survival and stroke occurrence is also shown by the
actuarial analyses (see Fig 3).
The Framingham study has shown that persons with
cardiac disorders, including cardiac hypertrophy, are
at increased risk for ischemic stroke [ 111. Our observations may seem to be at variance with that finding.
However, it is likely that a large proportion of patients
114 Annals of Neurology
Vol 3
No 2
February 1978
with TIA have cardiac disease that may not have been
manifested clinically, and so the two groups might not
be as different as it appears from their cardiac symptoms.
When linear discriminant analysis was performed,
excluding the patients with diastolic blood pressure of
105 mm H g or greater, anticoagulant therapy was then
selected as the only significant variable. Diastolic blood
pressure at the remaining lower levels was of no importance as a determinant of stroke occurrence in the
first year.
Because elevated diastolic blood pressure was
shown to be such an important factor in stroke occurrence, the question arose whether the excessive
number of strokes in patients with elevated blood
pressure accounted for the differences between all
patients treated and all not treated with anticoagulants. We therefore did the actuarial analysis of net
probability of stroke among all the patients with blood
pressure less than 105 mm Hg diastolic. The difference favoring the patients treated with anticoagulants
was still highly significant by this analysis (p < 0.001;
see Table 6), implying that among treated patients
there was a relatively greater proportion of strokes in
those with high diastolic pressures than in those with
lower pressures. Only 9 patients with diastolic pressure of 105 mm Hg or greater were being treated with
anticoagulants, and 7 had strokes during our observation; 3 survived the stroke. None of the strokes was
due to intracranial hemorrhage. Seven of these 9 patients died during the period of observation; 4 of the
deaths were from ischemic stroke.
Linear discriminant analysis and actuarial analysis
each indicated that diastolic blood pressure of 105 mm
Hg or more and anticoagulant therapy were the only
factors that influenced stroke occurrence in this sample
of patients with TIA. Among those with high blood
pressure, anticoagulant therapy appeared to have no
beneficial effect; but no patients in that group had
intracranial hemorrhage. Among the combined
groups with diastolic blood pressure less than 105 mm
Hg, the level of diastolic pressure had no effect, but
anticoagulant therapy was a favorable factor with regard to lower stroke rates.
Supported in part by Research Grant NS-6663 from the National
Institutes of Health, US Public Health Service.
References
1. Cartlidge NEF, Whisnant JP, Elveback LR: Carotid and
vertebral-basilar transient cerebral ischemic attacks: a community study, Rochester, Minnesota. Mayo Clin Proc 52:117120, 1977
2. Elveback LR:Estimation of survivorship in chronic disease: the
"actuarial" method. J Am Stat Assoc 53:420-440, 1958
3. Heyman A, Nefzger MD, Acheson RM:Epidemiologic study
of the relationship of anticoagulant therapy to mortality from
4.
5.
6.
7.
intracranial hemorrhage. Trans Am Neurol Assoc 97:152154, 1972
Hypertension-Stroke Cooperative Study Group: Effect of antihypertensive treatment on stroke recurrence. JAMA
229:409-418, 1974
Kurland LT, Elveback LR, Nobrega ET:Population studies in
Rochester and Olmsted County, Minnesota, 1900-1968, in
Kessler 11, Levin ML (eds): The Community as an
Epidemiologic Laboratory: A Casebook of Community
Studies. Baltimore, The Johns Hopkins University Press,
1970, pp 47-69
Matsumoto N, Whisnant JP, Kurland LT, et el: Natural history
of stroke in Rochester, Minnesota, 1955 through 1969: an
extension of a previous study, 1945 through 1954. Stroke
4:20-29, 1973
Millikan CH, Sieken RG, Shick RM: Studies in cerebrovascular disease: 111. The use of anticoagulant drugs in the treatment
of insufficiency or thrombosis within the basilararterial system:
X.The use of anticoagulant drugs in the treatment of intermit-
tent insufficiency of the internal carotid arterial system. Proc
Staff Meet Mayo Clin 30:116-126, 578-586, 1955
8. Veterans Administration Cooperative Study Group on Antihypertensive Agents: Effects of treatment on morbidity in
hypertension: results in patients wirh diastolic blood pressures
averaging 115 through 129 mm Hg: 11. Results in patients with
diastolic blood pressure averaging 90 through 114 mm Hg.
JAMA 202~1028-1034, 1967; 213:1143-1152, 1970
9. Whisnant JP, Matsumoto N, Elveback LR: The effect of anticoagulant therapy on the prognosis of patients with transient
cerebral ischemic attacks in a community: Rochester, Minnesota, 1955 through 1969. Mayo Clin Proc 48:844-848,
1973
10. Whisnant JP, Matsumoto N, Elveback LR: Transient cerebral
ischemic attacks in a community: Rochester, Minnesota, 1955
through 1969. Mayo Clin Proc 48:194-198, 1973
1 1. Wolf PA, Dawber TR, Thomas HE, et al: The epidemiology of
atherothrombotic brain infarction (ABI): the Framingham
study (abstract). Stroke 8:4, 1977
Whisnant e t al: Carotid and Vertebral-Basilar Transient Ischemic Attacks
115
Документ
Категория
Без категории
Просмотров
2
Размер файла
716 Кб
Теги
population, carotid, hypertension, disorder, stroki, effect, transiente, ischemia, survival, occurrence, stud, vertebrate, attack, basilar, anticoagulant, cardiaca
1/--страниц
Пожаловаться на содержимое документа