Prestroke Anticoagulation and Paroxysmal Type Are Correlated

With Favorable Outcome in Ischemic Stroke Patients With Atrial

Fibrillation

Haruhiko Hoshino, MD,* Yoshiaki Itoh, MD,* Hiroaki Kimura, MD,*

Koichi Miyaki, MD,† and Norihiro Suzuki, MD*

Paroxysmal atrial fibrillation (AF), which often precedes permanent AF, is reported
to be a risk factor for milder ischemic stroke. We assessed whether the type of AF and
prestroke treatment with an anticoagulant were associated with physical disabilities
in patients with AF-related acute ischemic stroke. We identified 162 consecutive
acute ischemic stroke patients with AF who were admitted to our hospital over
a 3-year period. Disability was measured using the modified Rankin Scale (mRS)
at the time of discharge and was categorized according to favorable clinical outcome
(mRS score 0-2). Of the 162 patients, 71 (43.8%) had paroxysmal AF and 91 had
permanent AF. Fifty-six patients (34.6%) had been treated with a prophylactic
anticoagulant. A total of 103 patients (63.6%) had a favorable outcome. Multivariate
logistic analysis revealed that paroxysmal AF (odds ratio [OR], 1.58; P 5 .0187),
prestroke anticoagulation treatment (OR, 1.95; P 5.0019), and noncardiogenic embo-
lism (OR, 2.20; P 5 .0073) were independent factors associated with a favorable
clinical outcome. Our data indicate that paroxysmal AF and prestroke anticoagula-
tion treatment are independently associated with favorable clinical outcome at the
time of hospital discharge in patients with AF. Key Words: Permanent atrial
fibrillation—antithrombotic treatment—disability.
Ó 2012 by National Stroke Association

As the proportion of elderly individuals increases in
a population, so does the prevalence of atrial fibrillation
(AF). Because AF is closely associated with ischemic
stroke, the incidence of AF-related stroke has increased
recently. AF-related strokes are generally severe, with an
estimated 1-year mortality rate of 50%.1
Paroxysmal AF, which often precedes permanent AF,
progresses to a permanent subtype slowly but steadily
From the *Department of Neurology, Keio University School of
Medicine, Tokyo, Japan; and †Division of Genomic Epidemiology,
Department of Clinical Research and Informatics, International
Medical Center of Japan, Tokyo, Japan.
Received December 16, 2009; accepted March 10, 2010.
The authors have no conflicts of interest to disclose.
Address correspondence to Haruhiko Hoshino, MD, Department
of Neurology, Keio University School of Medicine, 35 Shinanomachi,
Shinjuku, Tokyo, 160-8582, Japan. E-mail: hhoshino-keio@umin.ac.jp.
1052-3057/$ - see front matter
Ó 2012 by National Stroke Association
doi:10.1016/j.jstrokecerebrovasdis.2010.03.014
(24.7% at 5 years).2 Although the risk of stroke is compa-
rable in patients with paroxysmal AF and those with
permanent AF,3,4 a few previous reports have suggested
that strokes arising from paroxysmal AF are less severe
than those arising from permanent AF.5,6
Many randomized controlled studies have shown that
anticoagulation treatment is effective in reducing the inci-
dence of ischemic stroke in patients with AF, irrespective
of the type of AF.4 Analysis of pooled data from 5 random-
ized trials shows that warfarin consistently decreased the
risk of stroke in patients with AF by 68%, with virtually
no increase in the frequency of major bleeding.7 In addi-
tion, anticoagulation therapy that results in a prothrombin
time International Normalized Ratio (PT-INR) of $2.0 at
the time of stroke onset reportedly reduces not only the
frequency of ischemic stroke, but also its severity, as well
as mortality from stroke once it occurs.8,9 However, the
Japanese Guidelines for the Management of Stroke 2009
recommend a target PT-INR for anticoagulation in elderly
patients with AF of 1.6-2.6.10 Thus, the optimum PT-INR

Journal of Stroke and Cerebrovascular Diseases, Vol. 21, No. 1 (January), 2012: pp 11-17 11
12
for Japanese patients with AF might differ from that for
Western patients.
We first assessed whether the type of AF was a con-
founding clinical background factor, and then evaluated
the essential efficacy of prestroke anticoagulation treat-
ment. We also evaluated the effect of the type of AF on
physical disability at the time of hospital discharge, which
was considered an index of the severity of ischemic stroke
in Japanese patients with AF.
Methods
We enrolled all the patients admitted to Keio University
Hospital due to a transient ischemic attack (TIA) or an
ischemic stroke accompanied by AF (known previously
or diagnosed during admission) between January 1,
2005, and December 31, 2008. TIA is thought to be the
mildest clinical outcome of ischemic stroke. The inclusion
of TIA was considered appropriate in an evaluation of the
effect of preadmission anticoagulants. Patients receiving
thrombolytic treatment were excluded.
The presence of comorbid conditions was determined
based on the patients’ medical records, including the
records of regular medical checkups performed in the
Japanese general population. The presence of hyperten-
sion, hypercholesterolemia, diabetes mellitus, smoking,
history of TIA/stroke, history of congestive heart failure,
coronary artery disease (CAD), and preadmission use of
antihypertensive or antithrombotic agents were investi-
gated. Hypertension was defined blood pressure $140/
90 mm Hg or the use of antihypertensive medication.
Hypercholesterolemia was defined according to the crite-
ria established by the Japan Atherosclerosis Society (low-
density lipoprotein $140 mg/dL),11 as well as a medical
history of hypercholesterolemia. Diabetes mellitus was
diagnosed according to the criteria defined by the Japan
Diabetes Society: fasting plasma glucose level $126
mg/dL and/or plasma glucose level $200 mg/dL at 2
hours after a 75-g oral glucose tolerance test. A casual
plasma glucose level .200 mg/dL also was considered
indicative of diabetes mellitus.12 In addition, patients
with a history of treatment for diabetes mellitus were
considered to have diabetes. AF was diagnosed in each
patient when previous electrocardiography (ECG) or
ECG monitoring performed on admission revealed AF.
Patients were categorized based on their clinical data
(medical history, admission, and other available ECG
results, including the results of Holter ECG) according
to whether AF was paroxysmal (recurrent AF with spon-
taneous conversion to sinus rhythm) or permanent. Most
cases of paroxysmal AF were diagnosed based on previ-
ous ECG results or ECG monitoring during admission.
Smoking habit was classified as current or noncurrent
smoker. CAD was defined as a history of angina pectoris
or myocardial infarction, with or without coronary artery
bypass surgery or percutaneous transluminal coronary
H. HOSHINO ET AL.
angioplasty. The physician in charge of the patient’s
care was responsible for diagnosing the presence of
congestive heart failure. The CHADS2 score was calcu-
lated based on these risk factors.13 Blood samples were
obtained on admission, and prothrombin time, INR, and
D-dimer and fibrin degradation product (FDP) levels
were determined. Transthoracic ultracardiography
(UCG) was performed whenever possible. A PT-INR
value of 2.0-3.0 (or 1.6-2.6 in patients age .75 years)
was considered therapeutic. The ischemic stroke subtype
was determined based on the criteria used in the analysis
of Stroke Prevention in Atrial Fibrillation (SPAF).14 The
modified Rankin Scale (mRS) score at the time of hospital
discharge was used as an index of clinical outcome.
Clinical outcome was classified a priori as minimal or
no dependency (mRS score 0-2) or dependency or death
(mRS score 3-5 and death). Patients with a preadmission
mRS score .2 were excluded from the study.
Statistical Analysis
We evaluated the associations of AF types, preadmis-
sion anticoagulation treatment, and clinical outcome at
the time of discharge with the presence of comorbid con-
ditions and other clinical characteristics using analysis of
variance, Fisher’s exact test, Pearson’s c2 test, and
Wilcoxon’s rank-sum test for continuous and categorical
variables, as appropriate. We then applied a multivariate-
adjusted logistic regression model to adjust for sequential
confounders. Factors found to contribute to the clinical
outcome at the time of discharge in initial univariate anal-
yses (P value ,.10, because of the risk of type II errors
arising from the low statistical power of such an analysis)
were included in the multivariate model as candidate
variables and then removed using a backward stepwise
selection procedure. All of the P values were 2-sided,
and a P value ,.05 was considered statistically signifi-
cant. All of the analyses were performed using a commer-
cially available software package (JMP version 6.0).
Ethics
The data used in this study were collected from our
hospital’s prospective clinical protocols, which comply
with the local ethics guidelines.
Results
During the study period, 639 patients were admitted
with TIA or ischemic stroke. Of these, 198 had paroxys-
mal or permanent AF. Thirty-seven patients who had an
mRS score .2 before admission were excluded from the
analysis. Of the 162 patients included, 71 (43.8%) were
diagnosed with paroxysmal AF and 91 (56.2%) were diag-
nosed with permanent AF. Table 1 summarizes the base-
line characteristics for the study population. Fifty-six
patients (34.6%) were receiving anticoagulation therapy,
PAROXYSMAL AF AND PRESTROKE ANTICOAGULATION 13

Table 1. Patient characteristics of overall and paroxysmal/permanent AF groups

Atrial fibrillation

Overall Paroxysmal Permanent P value

Number of patients 162 71 91

Age, years, mean 6 SD 74.6 6 9.7 73.7 6 1.2 75.3 6 1.0 .2869
Male sex, n (%) 98 (60.5%) 42 (59.2%) 56 (64.5%) .8714
Hypertension, n (%) 104 (64.2%) 46 (64.8%) 58 (63.7%) 1.0000
Hypercholesterolemia, n (%) 50 (30.9%) 25 (35.2%) 25 (27.5%) .3082
Diabetes mellitus, n (%) 39 (24.1%) 17 (23.9%) 22 (24.2%) 1.0000
Smoking, n (%) 21 (13.0%) 8 (11.3%) 13 (14.3%) .6424
Coronary heart disease, n (%) 21 (13.0%) 12 (16.9%) 9 (9.9%) .2397
Congestive heart failure, n (%) 18 (11.1%) 5 (7.0%) 13 (14.3%) .2079
Previous stroke/TIA, n (%) 54 (33.3%) 21 (29.6%) 33 (36.3%) .4044
CHADS2 score, n (%) .8788
0 17 (10.5%) 8 (11.3%) 9 (9.9%)
1 37 (22.8%) 17 (23.9%) 20 (22.0%)
2 51 (31.5%) 24 (33.8%) 27 (29.7%)
3 27 (16.7%) 12 (16.9%) 15 (16.5%)
4 22 (13.6%) 7 (9.9%) 15 (16.5%)
5 8 (4.9%) 3 (4.2%) 5 (5.5%)
6 0 (0.0%) 0 (0.0%) 0 (0.0%)
Preadmission medication, n (%)
Antihypertensives 91 (56.2%) 43 (60.6%) 48 (52.8%) .3421
ARB/ACEI 48 (29.6%) 23 (32.4%) 25 (27.5%) .6032
CCB 50 (30.9%) 24 (33.8%) 26 (28.6%) .4971
Statins 30 (18.5%) 17 (23.9%) 13 (14.3%) .1535
Antithrombotic therapy
Anticoagulants 56 (34.6%) 16 (22.5%) 40 (44.0%) .0048
Antiplatelets 70 (43.2%) 34 (47.9%) 36 (39.6%) .3384
NIHSS score at admission, mean 6 SD 8.0 6 9.4 6.89 6 0.11 8.85 6 0.98 .2111
Median 4 3 5 .0690
PT-INR, mean 6 SD (n)
With anticoagulant 1.53 6 0.45 (56) 1.47 6 0.11 (16) 1.56 6 0.07 (40) .5073
Without anticoagulant 0.97 6 0.07 (104) 0.97 6 0.01 (54) 0.97 6 0.01 (50) .9461
D-dimer, mean 6 SD (n)
With anticoagulant 3.79 6 10.40 (34) 2.19 6 3.32 (10) 4.46 6 2.14 (24) .5703
Without anticoagulant 4.36 6 11.88 (74) 5.64 6 1.95 (37) 3.08 6 1.95 (37) .3565
FDP, mean 6 SD (n)
With anticoagulant 52.6 6 78.6 (40) 47.6 6 22.1 (12) 54.7 6 14.5 (28) .7915
Without anticoagulant 78.4 6 213.4 (78) 53.0 6 32.5 (43) 109.6 6 36.0 (35) .2461
Left atrial size on UCG, cm, mean 6 SD (n) 4.3 6 0.9 (123) 3.8 6 0.1 (57) 4.6 6 0.1 (66) ,.0001
TIA/cerebral infarction subtype, n (%) .2397
Cardioembolism 128 (79.0%) 54 (76.1%) 74 (81.3%)
Large artery atherosclerosis 5 (3.1%) 1 (1.4%) 4 (4.4%)
Other cerebral infarction 4 (2.5%) 3 (4.2%) 1 (1.1%)
Small vessel occlusion 11 (6.8%) 4 (5.6%) 7 (7.7%)
TIA 14 (8.6%) 9 (12.7%) 5 (5.5%)
ADL on discharge, n (%) .1563
mRS 0 38 (23.5%) 23 (32.4%) 15 (16.5%)
mRS 1 41 (25.3%) 17 (23.9%) 24 (26.4%)
mRS 2 24 (14.8%) 11 (15.5%) 13 (14.3%)
mRS 3 11 (6.8%) 4 (5.6%) 7 (7.7%)
mRS 4 20 (12.3%) 9 (12.7%) 11 (12.1%)
mRS 5 12 (7%) 2 (2.8%) 10 (11.0%)
Death 16 (9.9%) 5 (7.0%) 11 (12.1%)
mRS 0-2 103 (63.6%) 51 (71.8%) 52 (57.1%) .0702

Abbreviations: ARB, angiotensin receptor blocker; ACEI, angiotensin-converting enzyme inhibitor; CCB, Ca channel blocker.
14
and 70 patients (43.2%) were taking an antiplatelet agent.
The mean PT-INR was 1.53 6 0.45 in the 56 patients
receiving anticoagulation treatment. Eighteen of these
patients (32.1%) had a therapeutic INR at the time of
admission for ischemic stroke. The most common ische-
mic stroke clinical subtype was cardioembolism (79.0%).
Paroxysmal and Permanent AF
Table 1 compares clinical features of the patients with
paroxysmal AF and those with permanent AF. The patients
with paroxysmal AF had a lower rate of preadmission anti-
coagulation treatment (22.5% vs 44.0%; P 5 .0048). The
prevalences of typical risk factors, including hypertension,
dyslipidemia, diabetes mellitus, and smoking, were similar
in the 2 groups. Although data were not available for all
patients, the PT-INR and D-dimer and FDP levels were
similar in the 2 groups, regardless of prestroke anticoagula-
tion treatment. Echocardiography demonstrated a smaller
left atrium in the patients with paroxysmal AF (3.8 cm vs 4.6
cm; P , .0001). The patients with paroxysmal AF had
a slightly lower National Institutes of Health Stroke Scale
(NIHSS) score on admission (P 5 .0690) and a more
frequent favorable clinical outcome (mRS score 0-2) on
discharge (71.8% vs 57.1%; P 5 .0702), although these
differences were not statistically significant.
Clinical Outcome at Hospital Discharge
A total of 103 patients (63.6%) had a good clinical out-
come (mRS score 0-2) at the time of hospital discharge.
Clinical background information and laboratory results
are summarized in Table 2. Younger age (73.4 years vs
76.7 years: P 5 .0349), paroxysmal AF type (49.5% vs
33.9%; P 5 .0702), preadmission use of antihypertensives
(62.1% vs 45.8%; P 5 .0495), including a calcium channel
blocker (CCB) (36.9% vs 20.3%; P 5 .0340), and prestroke
anticoagulation treatment (42.7% vs 20.3%; P 5 .0057)
were correlated with good clinical outcome. The D-dimer,
FDP, and C-reactive protein levels were significantly
lower in patients with a favorable clinical outcome. In ad-
dition, the frequency of cardioembolism was significantly
lower in the good clinical outcome group (70.9% vs 93.2%;
P 5 .0006). The admission periods were significantly lon-
ger in the severe clinical outcome group (20.3 vs 53.0 days;
P , .0001). Age, type of AF, preadmission CCB treatment,
preadmission anticoagulation treatment, and subtype of
cerebral infarction were found to contribute to clinical
outcome at the time of discharge using initial univariate
analyses (P , .10); these variables were then removed
using a backward stepwise selection procedure. Multivar-
iate logistic analysis disclosed that paroxysmal AF (odds
ratio [OR], 1.58; 95% CI, 1.09-2.34: P 5 .0187), preadmis-
sion anticoagulation treatment (OR, 1.95; 95% CI,
1.30-3.02: P 5 .0019), and noncardiogenic embolism (OR,
2.20; 95% CI, 1.30-4.21; P 5 .0073) were independent
H. HOSHINO ET AL.
factors associated with a good clinical outcome at the
time of hospital discharge (Table 3).
Discussion
In this study, paroxysmal AF and preadmission anticoa-
gulation treatment were independently associated with
a favorable clinical outcome at the time of hospital dis-
charge in patients with AF. Paroxysmal AF is thought to
precede permanent AF. A Japanese study found that
5.5% of patients with paroxysmal AF developed perma-
nent AF while receiving conventional antiarrhythmic ther-
apy.15 The prevalence of paroxysmal AF in the present
study was relatively high compared with previous reports
(6.3%-35%).1,6,16 This difference might reflect the more
common usage of prophylactic anticoagulation treatment
in patients with permanent AF than in those with
paroxysmal AF, resulting in a reduced stroke rate in
patients with permanent AF and a relatively higher
incidence of stroke in those with paroxysmal AF. Our
data may support this explanation, given that 44.0% of
permanent AF patients, but only 22.5% of paroxysmal
AF patients, were receiving anticoagulation treatment at
the onset of stroke. The possibility that paroxysmal AF is
associated with a greater risk of stroke than permanent
AF has been explored and shown to be untrue.4
Paroxysmal AF also was significantly associated with
a favorable clinical outcome in this study. One explanation
for the more favorable stroke outcome in patients with par-
oxysmal AF is the lower incidence of noncardioembolic
strokes, which are usually less severe than strokes of
cardiac origin.5 Staszewski et al6 reported more favorable
stroke outcomes in patients with paroxysmal AF than in
patients with permanent AF, because the strokes in the for-
mer were more frequently categorized as noncardioem-
bolic (35% vs 18%). In the present study, however, the
clinical subtypes of ischemic stroke were similar in
patients with paroxysmal AF and those with permanent
AF. Another possible explanation is the size of the throm-
botic emboli. Emboli might be smaller in patients with par-
oxysmal AF than in those with permanent AF. Li-Saw-Hee
reported a significantly higher soluble P-selectin level
(an index of platelet activation) in patients with permanent
AF compared with those with paroxysmal AF,17 indicating
a more severe prothrombotic state in the former. Although
no clinical background differences were observed in the
present study, in the study of Staszewski,6 prevalences of
chronic heart failure and diabetes mellitus were higher
in the group with permanent AF. We found a significantly
larger left atrial size in the permanent AF group, but
our routine laboratory coagulation tests revealed no
differences between the 2 groups.
In this study, preadmission oral anticoagulation treat-
ment was significantly associated with favorable clinical
outcome in patients with AF. Consistent with previous
reports, anticoagulation treatment resulting in an INR
PAROXYSMAL AF AND PRESTROKE ANTICOAGULATION 15

Table 2. Characteristics of patients with favorable clinical outcome

Clinical outcome at hospital discharge

mRS 0-2 mRS 3-6 P value

Number of patients 103 59

Age, years, mean 6 SD 73.4 6 0.9 76.7 6 1.3 .0349
Male sex, n (%) 63 (61.8%) 35 (59.3%) .8670
Hypertension, n (%) 70 (68.0%) 34 (57.6%) .2334
Hypercholesterolemia, n (%) 32 (31.7%) 18 (30.5%) 1.0000
Diabetes mellitus, n (%) 24 (23.3%) 15 (25.4%) .8493
Smoking, n (%) 13 (12.6%) 8 (13.6%) 1.0000
Coronary heart disease, n (%) 15 (14.6%) 6 (10.2%) .4761
Congestive heart failure, n (%) 12 (11.7%) 6 (10.2%) 1.0000
Previous stroke/TIA, n (%) 39 (37.9%) 15 (25.4%) .1212
CHADS2 score, n (%)
0 8 (7.8%) 9 (15.3%) .7254
1 25 (24.3%) 12 (20.3%)
2 32 (31.1%) 19 (32.2%)
3 18 (17.5%) 9 (15.3%)
4 14 (13.6%) 8 (13.6%)
5 6 (5.8%) 2 (3.4%)
6 0 (0.0%) 0 (0.0%)
Atrial fibrillation, n (%)
Paroxysmal 51 (49.5%) 20 (33.9%) .0702
Permanent 52 (50.5%) 39 (66.1%)
Preadmission medication, n (%)
Antihypertensives 64 (62.1%) 27 (45.8%) .0495
ARB/ACEI 33 (32.0%) 15 (25.4%) .4748
CCB 38 (36.9%) 12 (20.3%) .0340
Statin 19 (18.5%) 11 (18.6%) 1.0000
Antithrombotic therapy
Anticoagulants 44 (42.7%) 12 (20.3%) .0057
Antiplatelets 43 (41.8%) 27 (45.8%) .6255
NIHSS score at admission, median 2 17 ,.0001
PT-INR, mean 6 SD (n)
With anticoagulant 1.54 6 0.07 (44) 1.53 6 0.13 (12) .9730
Without anticoagulant 0.96 6 0.01 (57) 0.98 6 0.01 (47) .1174
D-dimer, mean 6 SD (n)
With anticoagulant 2.19 6 2.04 (25) 8.24 6 3.40 (9) .1363
Without anticoagulant 1.68 6 1.81 (41) 7.69 6 2.01 (33) .0296
FDP, mean 6 SD (n)
With anticoagulant 35.3 6 13.2 (29) 98.1 6 21.4 (11) .0168
Without anticoagulant 41.0 6 32.5 (42) 122.0 6 35.1 (36) .0947
CRP, mean 6 SD (n) 0.82 6 0.41 (95) 2.30 6 0.55 (52) .0325
Left atrial size on UCG, cm, mean 6 SD (n) 4.4 6 0.1 (82) 4.0 6 0.1 (41) .0091
TIA/cerebral infarction subtype, n (%)
Cardioembolism 73 (70.9%) 55 (93.2%) .0006
Large artery atherosclerosis 3 (2.9%) 2 (3.4%)
Other cerebral infarction 2 (1.9%) 2 (3.4%)
Small vessel occlusion 11 (10.7%) 0
TIA 14 (13.6%) 0
Admission period, days, mean 6 SD 20.3 6 2.9 53.0 6 3.8 ,.0001

of $2.0 reduced not only the frequency of ischemic stroke, stroke severity. First, anticoagulants prevent thrombus
but also its severity and the risk of death from stroke.6,8,9,18 formation and reduce the size of thrombi in the left
There may be several reasons for the correlation between atrium. Smaller thrombi cause smaller infarctions,
preadmission anticoagulation treatment and the lower possibly leading to a more favorable clinical outcome.
16
Table 3. Multivariate-adjusted logistic regression model
assessing predictors of favorable clinical outcome
OR (95% CI) P value
Age (10-year increments) 0.82 (0.56-1.17) .2717
Paroxysmal AF (vs 1.58 (1.09-2.34) .0187
permanent AF)
Preadmission anticoagulant 1.95 (1.30-3.02) .0019
Preadmission CCB 1.45 (0.97-2.22) .0733
Noncardiogenic embolism 2.20 (1.30-4.21) .0073
Second, although the presence of AF could have biased
observers to classify the events as cardioembolisms,
anticoagulants reduce the incidence of cardioembolism,
resulting in a lower proportional rate of cardioembolism
in patients receiving preadmission anticoagulation
treatment. Third, anticoagulants promote the lysis of
thrombi in occluded vessels, promoting collateral
formation and resulting in a favorable clinical outcome.
We found that anticoagulation treatment resulting in an
INR ,2 was also effective in reducing the severity of
ischemic stroke in patients with AF. However, because
the INR values were measured at the time of hospital
admission, the control of anticoagulation immediately
before stroke onset might have been higher (eg, an INR
of 2-3). Alternatively, the effective INR value for
the Japanese population might be lower than typical
(eg, 1.6-2.6), especially in elderly persons.19
This study has some limitations. First, the retrospec-
tive, observational nature of the study design might
have affected the validity of our findings. Second, the
registry included only patients treated at one hospital,
providing a potential selection bias. Third, physical dis-
ability was evaluated at the time of hospital discharge,
and thus the timing of the evaluation varied. But because
the patients with worse clinical outcome had significantly
longer admission periods, if those patients had been eval-
uated after a follow-up period equivalent to that for the
patients with favorable clinical outcome, the physical
disabilities of the former likely would have been worse
than the final score. Fourth, the number of patients in
this study was rather small. A larger number of subjects
and a prospective protocol are needed to confirm our
findings.
In conclusion, strokes caused by paroxysmal AF are
less severe than those caused by permanent AF. Pread-
mission anticoagulation treatment might allow for
a marked reduction in deaths or severe disabilities at
the time of hospital discharge after ischemic stroke. Our
results emphasize the importance of prophylactic antico-
agulation treatment in patients with paroxysmal AF as
well as those with permanent AF.
Acknowledgments: We thank Htay Lwin, MD, Division
of Genomic Epidemiology, Department of Clinical Research
H. HOSHINO ET AL.
and Informatics, International Medical Center of Japan, for
his advice on the statistical analysis.
References
1. Marini C, De Santis F, Sacco S, et al. Contribution of atrial
fibrillation to incidence and outcome of ischemic stroke:
Results from a population-based study. Stroke 2005;
36:1115-1119.
2. Kerr CR, Humphries KH, Talajic M, et al. Progression to
chronic atrial fibrillation after the initial diagnosis of par-
oxysmal atrial fibrillation: Results from the Canadian
Registry of Atrial Fibrillation. Am Heart J 2005;
149:489-496.
3. Hart RG, Pearce LA, Rothbart RM, et al. Stroke Preven-
tion in Atrial Fibrillation Investigators. Stroke with inter-
mittent atrial fibrillation: Incidence and predictors during
aspirin therapy. J Am Coll Cardiol 2000;35:183-187.
4. Hohnloser SH, Pajitnev D, Pogue J, et al. Incidence of
stroke in paroxysmal versus sustained atrial fibrillation
in patients taking oral anticoagulation or combined anti-
platelet therapy: An ACTIVE-W substudy. J Am Coll Car-
diol 2007;50:2156-2161.
5. Staszewski J, Staszewska M, Kotowicz J, et al. Cerebro-
vascular incidents differ in patients with permanent
and paroxysmal atrial fibrillation. Cerebrovasc Dis 2006;
22:446-449.
6. Staszewski J, Brodacki B, Tomczykiewicz K, et al. Strokes
in paroxysmal atrial fibrillation have more favorable
outcome than in permanent atrial fibrillation. Acta Neu-
rol Scand 2009;119:325-331.
7. Risk factors for stroke and efficacy of antithrombotic ther-
apy in atrial fibrillation. Analysis of pooled data from five
randomized controlled trials. Arch Intern Med 1994;
154:1449-1457.
8. Hylek EM, Go AS, Chang Y, et al. Effect of intensity of
oral anticoagulation on stroke severity and mortality in
atrial fibrillation. N Engl J Med 2003;349:1019-1026.
9. O’Donnell M, Oczkowski W, Fang J, et al. Preadmission
antithrombotic treatment and stroke severity in patients
with atrial fibrillation and acute ischaemic stroke: An ob-
servational study. Lancet Neurol 2006;5:749-754.
10. Japanese Guidelines for the Management of Stroke, 2009.
Tokyo: Kyouwa Kikaku, 2009.
11. Teramoto T, Sasaki J, Ueshima H, et al. Diagnostic criteria
for dyslipidemia: Executive summary of Japan Athero-
sclerosis Society (JAS) guideline for diagnosis and pre-
vention of atherosclerotic cardiovascular diseases for
Japanese. J Atheroscler Thromb 2007;14:155-158.
12. Kuzuya T, Nakagawa S, Satoh J, et al. Report of the Ccom-
mittee on the Classification and Diagnostic Criteria of Di-
abetes Mellitus. Diabetes Res Clin Pract 2002;55:65-85.
13. Gage BF, Waterman AD, Shannon W, et al. Validation of
clinical classification schemes for predicting stroke:
Results from the National Registry of Atrial Fibrillation.
JAMA 2001;285:2864-2870.
14. Hart RG, Pearce LA, Miller VT, et al. Cardioembolic vs
noncardioembolic strokes in atrial fibrillation: Frequency
and effect of antithrombotic agents in the Stroke Preven-
tion in Atrial Fibrillation studies. Cerebrovasc Dis 2000;
10:39-43.
15. Kato T, Yamashita T, Sagara K, et al. Progressive nature of
paroxysmal atrial fibrillation: Observations from
a 14-year follow-up study. Circ J 2004;68:568-572.
16. Hart RG, Pearce LA, McBride R, et al. Stroke Prevention
in Atrial Fibrillation (SPAF) Investigators. Factors
PAROXYSMAL AF AND PRESTROKE ANTICOAGULATION
associated with ischemic stroke during aspirin therapy in
atrial fibrillation: Analysis of 2012 participants in the
SPAF I-II clinical trials. Stroke 1999;30:1223-1229.
17. Li-Saw-Hee FL, Blann AD, Gurney D, et al. Plasma von
Willebrand factor, fibrinogen and soluble P-selectin levels
in paroxysmal, persistent and permanent atrial fibrilla-
tion: Effects of cardioversion and return of left atrial
function. Eur Heart J 2001;22:1741-1747.
17
18. Reynolds MW, Fahrbach K, Hauch O, et al. Warfarin anti-
coagulation and outcomes in patients with atrial fibrilla-
tion: A systematic review and meta-analysis. Chest 2004;
126:1938-1945.
19. Yasaka M, Minematsu K, Yamaguchi T. Optimal intensity
of international normalized ratio in warfarin therapy for
secondary prevention of stroke in patients with non-
valvular atrial fibrillation. Intern Med 2001;40:1183-1188.