Validation of the ABCD3-I score for 90 day prediction of early stroke risk after transient ischemic attack or minor ischemic stroke

Patients with transient ischemic attack or minor ischemic stroke are at high risk of early subsequent stroke. It might be confusing to determine the exposition to stroke recurrence as a consequence of the aggressive treatment and useful prevention. The objectives were to determine the incidence of early subsequent stroke and evaluate the ABCD3-I score for recurrent stroke risk.

VALIDATION OF THE ABCD3-I SCORE FOR 90-DAY

PREDICTION OF EARLY STROKE RISK AFTER TRANSIENT ISCHEMIC ATTACK OR MINOR ISCHEMIC STROKE

Tran Van Tu 1 , Nguyen Van Lieu 2 , Nguyen Huy Thang 1

1 Pham Ngoc Thach University of Medicine,

2 Hanoi Medical University.

Patients with transient ischemic attack or minor ischemic stroke are at high risk of early subsequent stroke It might be confusing to determine the exposition to stroke recurrence as a consequence of the aggressive treatment and useful prevention The objectives were to determine the incidence of early subsequent stroke and evaluate the ABCD3-I score for recurrent stroke risk A prospective cohort study in

203 patients with transient ischemic attack or minor ischemic stroke was at Department of Cerebrovascular disease, People’s Hospital 115 (HCMC) from Jan 1st 2016 to Jun 30th 2018 All patients were assessed by stroke specialists, performed brain imaging, classified according to ABCD2, ABCD3, ABCD3-I Scores and followed up for 90 days There were 15/203 patients (7.39%) experienced subsequent stroke by 90 days The incidence of stroke recurrence after 2 days, 7 days and 30 days were 1.97% (4/203), 3.94% (8/203) and 4.43% (9/203), respectively The high-risk group (ABCD3-I ≥ 8) was more likely to have recurrent stroke than the low-risk group (ABCD3-I < 8), significantly: within 7 days 10,2% vs 1,95% (p = 0.0097), 30 days 10.2% vs 2.6% (p = 0.0244) and 90 days 16.33% vs 4.55% (p = 0.0061).Furthermore, the ABCD3-I Score had a predictive value with area under the curve ROC with 0.67 The risk of recurrent stroke after TIA or MIS will increase by the time and can be predicted by applying the ABCD3-I Score in clinical practice

I INTRODUCTION

Key words: transient ischemic attack, minor ischemic stroke, subsequent stroke, ABCD3-I Score

Patients who have been diagnosed with

transient ischemic attack (TIA) or minor ischemic

stroke (MIS) are at high risk of early subsequent

stroke, with the rate ranging between 10

– 20% [1] This risk is usually variable and

depending on clinical symptoms, pathology,

and urgent medical treatment Determination

of the likelihood of stroke recurrence is

necessary for determining the aggressiveness

of treatment and useful prevention There have been various risk scores developed to identify high-risk subgroup patients which are applied widely by clinical specialists In 2005, Rothwell

et al proposed the ABCD Score based on age, blood pressure, clinical features, duration

of symptoms [2] After that, Johnston et al combined this score with a diabetic factor to establish the ABCD2 Score [3] Merwick et al

in anew consolidated the score with factors of dual TIA and features on imaging to introduce the ABCD3and ABCD3-I Score [4]

The number of stroke patients is increasing [5] Additionally, in Vietnam, the incidence

Corresponding author: Tran Van Tu, Pham Ngọc

Thach University of Medicine

Email: bstu71@gmai.com

Received: 27/11/2018

Accepted: 12/03/2019

of recurrent stroke is also increasing at an

alarming rate [6] Nevertheless, there are a

few studies about this issue on Vietnamese

population to determine high risk patients

This is the motivation creating this study with

the following objectives: (1) determine the

incidence of early subsequent stroke, and

(2) evaluate the ABCD3-I score for recurrent

stroke risk

II METHODS

This is a prospective cohort study and

random sampling, performed at the Department

of Cerebrovascular disease in People’s

Hospital 115 (HCMC) from January 1st 2016

to June 30th 2018 The procedure of the study

is indicated in Figure 1 TIA is defined as “a

transient episode of neurologic dysfunction

within 24 hours, caused by focal brain, spinal

cord or retinal ischemia without acute infarction”

by the criteria of WHO [7] MIS is confirmed unless there is intracerebral hemorrhage and stroke severity by National Institutes

of Health Stroke Scale (NIHSS) > 5 [8] The patients whose symptoms of TIA or MIS had occurred within 48 hours were selected for this study After being admitted to the emergency department, they were assessed independently

by one stroke specialist as well as undergoing brain imaging in within 10 minutes, and subsequently classified according to ABCD2, ABCD3, ABCD3-I Scores Before discharge, all participants were asked to be followed up in the clinic for out-patient each month during 90 days In cases of recurrent stroke, TIA or death, the patients or their relatives were asked to inform the research team by telephone number immediately

III RESULTS

Between January 1st 2016 to June 30th 2018, 203 patients who had been diagnosed with TIA

or MIS admitted to emergency department (TIA = 34, MIS = 169)

Table 1 Clinical features

Duration, min

TOAST, %

At first, there were 260 patients diagnosed

with TIA or MIS (TIA = 50, MIS = 210) After

that, 57 patients (TIA = 16, MIS = 41) were

excluded because of hypoglycemia (blood

glucose below 70 mg%), migraine headache

(have pain one side of head and/or have

been diagnosed), epilepsy (have convulsions

and/or have been diagnosed), brain tumor

(have detected on imaging and/or have been

diagnosed), had significant comorbidities (i.e

heart failure, liver failure or end-stage kidney

disease), their symptoms had occurred more

than 48 hours before admission to hospital, or

had NIHSS > 5

Statistical analysis was done with SPSS

(version 24.0) Parametric and non-parametric comparisons of categorical and continuous variables were done with χ² test, t-test, and Mann-Whitney U test, as appropriate A p-value of 0.05 or less was deemed significant Predictive values for ABCD3-I Scores was expressed as the area under the receiver operating characteristic curve (AUC) Ideal discrimination produces an AUC of 1.0, whereas discrimination that is no better than chance producesan AUC of 0.5

The ethics committee of People’s Hospital

115 approved this study Written informed consent was obtained from all participants

Population characteristics:

The mean age of all patients was 60.56 ± 13.5 years The age of female was higher than male significantly (64.71 vs 58.4; p = 0.0017) The sex ratio male:female was 1.57:1

Comorbidities:

Hypertension 62.07%, diabetes mellitus 14.26%, hypercholesterolemia 63.6%, atrial fibrillation 5.91% and previous TIA 5.4%

Clinical features (Table 1) and Classification base on the ABCD3-I Score (Figure 2)

5 26

46 42 35 40

8

1 0

5

10

15

20

25

30

35

40

45

50

ABCD3-I

Figure 2 Classification base on the ABCD3-I

Imaging features:

CT scan or CTA was done in 187/203 patients The percentage of patients with acute ischemic lesions exposed on CT were 21.9% (41/187) One-hundred and thirty-seven patients had been indicated for MRI, where 62.78% (86/137) of them had early infarct signs on DWI; 8.76% (12/137) had MRA stenosis ≥ 50% or completely occlusion Patients underwent carotid artery ultrasound screening; 7.88% (16/203) of cases had carotid artery stenosis > 50% Specifically, the percentages

of carotid artery stenosis 50 - 69% and stenosis ≥ 70% were 6.4% and 1.48% of patients, respectively The first diagnosis of atrial fibrillation on ECG was made at 5.91% (12/203) of patients No patient had cardiac thrombosis detected by transthoracic cardiac ultrasound

Treatment:

All patients were recommended to undergo healthy life-style changes and were treated with a statin Antiplatelet medications (aspirin or clopidogrel or dipyridamole + aspirin) were described in 96% of cases, 3.94% of patients were prescribed anticoagulation medications 71.43% patients needed to use antihypertensive drugs 13.79% of patients had been under pharmacologic therapy for diabetes

Recurrent stroke and other event (TIA, death):

There were 15 patients (7.39%) who experienced subsequent stroke within 90 days The incidence of stroke recurrence after 2 days, 7 days and 30 days was 1.97% (4/203), 3.94% (8/203) and 4.43% (9/203), respectively Risk subgroup classification based on the ABCD3-I Score is seen

in Figure 3 The risk percentages between two subgroups were different significantly: within 7 days 10.2% vs 1.95% (p = 0.0097), 30 days 10.2% vs 2.6% (p = 0.0244) and 90 days 16.33% vs 4.55% (p = 0.0061)

2.6

4.55 4.08

16.33

0

2

4

6

8

10

12

14

16

18

Low-risk (ABCD3-I = 0-7) High-risk (ABCD3-I=8-13)

P=0,0061

P = 0,0244

P = 0,0097

Figure 3 Risk subgroup classification base on the ABCD3-I Score

The risk estimated by Kaplan - Meier of predictive events within 90 days after TIA or MIS in the high-risk group (ABCD3-I ≥ 8) was more superior than in the low-risk group (ABCD3-I < 8) (Log rank test=7.84, df=1, p=0.0051) Furthermore, the ABCD3-I Score had a predictive value with area under the curve ROC of 0.67 (Figure 4)

Figure 4 [4a] The risk of recurrent stroke, TIA or deathe estimated by Kaplan Meier, [4b]

The predictive value of ABCD3-I score

IV DISCUSSION

The average patient age in our study was

60.56 ± 14.09 It seemed lower than results of

the studies of Kiyohara et al in 2014 [9] and

Kelly et al in 2016 [10] with 69 and 68 years old

The reason of this difference may be due to the

higher life expectancy in developed countries

and also their more-developed healthcare

systems The percentage of male patients

was 61.2%, similar to the results of 62.2% in

Kiyohara’s [9] and 59% in Kelly’s [10]

The number of patients with hypertension,

diabetes mellitus and hypercholesterolemia

was similar to other studies in Asian populations,

such as the study of Qiliang Dai et al (China)

in 2015 [11];owever, the rate of atrial fibrillation

was lower, 5.91%, versus 13 to 18.3% [9], [10],

[11] This could be explained that the average

patient age in our study was younger so the

rate of atrial fibrillation was lower than other

populations Furthermore, there were more

patients with previous stroke or TIA in other

foreign countries, from 10 - 21% [9], [10], [11]

whereas ours was only 5.4% That might be

because citizens in high-income nations had

the higher average age than our country

Most of these patients had unilateral

weakness, a few had unique speech impairment

(6.4%) More than a half of our patients were

admitted to emergency department after 60

minutes (69.4%) while Qiliang Dai et al [11]

had published 82.7% of cases were taken

to stroke centers The most popular cause

of stroke base on TOAST classification was

small vessel disease (47%), similar to the

investigation by Purroy et al in 2013 [12] On

the other hand, it was dissimilar to the study of

Qiliang Dai et al [11] where the most prevalent

cause was cardio-embolism (39,4%) Thereby,

the treatment would be suitable for each case,

depend on the pathology

The incidence of subsequent stroke within

90 days in our study was 7.39%, compared to 3.1% of Purroy’s [12], 10.4% of Kiyohara’s [9] and 10.6% of Qiliang Dai’s [11] Additionally, the risk would be increasing parallel by the score of ABCD3-I [9], [10] The area under the curve ROC of this score was 0.67, equivalent

to the finding of Purroy et al 0,69 [12] On the other hand, Qiliang Dai et al [11] had declared that the value of ABCD3-I Score could be enhanced in case dual TIA factor replaced by DWI on MRI (AUC 0,759 vs 0,729 ; p = 0,035)

V CONCLUSION

The risk of recurrent stroke after TIA or MIS increases with time and can be predicted by applying the ABCD3-I Score in clinical practice

Acknowledgements

Special thanks should be given

my research project supervisors for their professional guidance and valuable support

I also would like to thank my colleagues in Neurology Department, People’s Hospital 115 for their assistance with the collection of my data

REFERENCES

1 Giles MF, Rothwell PM (2007), “Risk of

stroke early after transient ischemic attack: a

systematic review and meta-analysis”, Lancet Neurol, 6 ,1063 – 72.

2 Rothwell PM, Giles MF, Chandratheva

A, Marquardt L, Geraghty O, Redgrave JN,

et al (2007), “Early use of Existing Preventive

Strategies for Stroke (EXPRESS) study Effect

of urgent treatment of transient isch- aemic attack and minor stroke on early recurrent stroke (EXPRESS study): a prospective population-based sequential comparison”,

Lancet, 370, 1432 – 1442

3 Johnston, S C., P M Rothwell, M

N Nguyen-Huynh, et al (2007),“Validation

and refinement of scores to predict very early

stroke risk after transient ischemic attack”,The

Lancet, 369(9558), 283 - 292.

4 Merwick A, Albers GW, Amarenco P,

Arsava EM, Ay H, Calvet D, et al (2010),

“Addition of brain and carotid imaging to the

ABCD2 score to identify patients at early risk of

stroke after transient ischemic attack: a multi-

centre observational study”, Lancet Neurol, 9,

1060 – 1069

5 Institute for Health Metrics and

Evaluation, “Global burden of diseases,

injuries, and risk factors study 2010”, 2301

Fifth Ave., Suite 600 Seattle, WA 98121 USA

6 Ngo Ba Minh, Cao Phi Phong (2011),

“Identify the patients with high risk of late

stroke after transient ischemic attack or minor

ischemic stroke by ABCD2 score”, Y Hoc TP

Ho Chi Minh, Vol 15 - Supplement of No 1 –

2011: 603 - 608

7 Ross, M A., S Compton, P Medado

et al (2007), “An Emergency Department

Diagnostic Protocol for Patients With Transient

Ischemic Attack: A Randomized Controlled

Trial’, Annals of emergency medicine, 50(2),

109 - 119

8 National Institute of Health, National

Institute of Neurological Disorders and Stroke

Stroke Scale, https://www.ninds.nih.gov/sites/ default/files/NIH_Stroke_Scale_Booklet.pdf

9 Takuya Kiyohara, Masahiro Kamouchi, Yasuhiro Kumai et al (2014), “ABCD3 and

ABCD3-I Scores Are Superior to ABCD2 Score

in the Prediction of Short- and Long-Term Risks

of Stroke After Transient Ischemic Attack”,

Stroke, 45, 418 - 425

10 Peter J Kelly, Gregory W Albers, AnastasiosChatzikonstatinou, Gian Marco

De Marchis, et al (2016), “Validation and

comparison of imaging-based scores for prediction of early stroke risk after transient ischemic attack: a pooled analysis of individual-patient data from cohort studies”,www thelancet.com/neurology, Vol 15 November 2016.

11 Dai Q, Sun W, Xiong Y et al (2015),

“From clinical to tissue-based dual TIA: Validation and refinement of ABCD3-I score”,

American Academy of Neurology, 84(14),1426

- 32

12 Purroy F, P E Jimenez-Caballero, G Mauri-Capdevila, M J Torres et al (2013),

“Predictive value of brain and vascular imaging including intracranial vessels in transient ischemic attack patients: external validation

of the ABCD3-I score”, European Journal of Neurology, 20, 1088 – 1093.