Prediction of Ischemic Stroke in Patients with Tissue-Defined Transient Ischemic Attack Takeshi Hayashi, MD, PhD, Yuji Kato, MD, PhD, Harumitsu Nagoya, MD, PhD, Yasuko Ohe, MD, PhD, Ichiro Deguchi, MD, PhD, Takuya Fukuoka, MD, PhD, Hajime Maruyama, MD, PhD, Yohsuke Horiuchi, MD, Yuito Nagamine, MD, Hiroyasu Sano, MD, and Norio Tanahashi, MD, PhD

Background: The risk of future stroke after transient ischemic attack (TIA) has been widely studied, but most findings were obtained for classically defined TIA (timedefined TIA). A new definition of TIA, that is, tissue-defined TIA, which requires the absence of fresh brain infarction on magnetic resonance imaging, could change stroke risk assessments. We, therefore, aimed to evaluate the risk of future stroke in patients with tissue-defined TIA. Methods: We retrospectively reviewed 74 patients with tissue-defined TIA, who could be followed for 2 years. Clinical, laboratory, and radiological data were collected and compared between groups that did and did not develop ischemic stroke within the 2-year period. Results: Ischemic stroke occurred in 11 patients (14.9%). Increased age, hemiparesis, and/or dysarthria during the TIA, old cerebral infarction revealed by magnetic resonance imaging, and large-artery stenosis detected by magnetic resonance angiography and/or ultrasonography tended to increase the risk of future stroke, but no individual factor showed statistically significant effect. TIA etiology did not significantly affect the risk. ABCD2 score, an established score for predicting stroke after time-defined TIA, showed only a weak association with future stroke. In contrast, new scores that we created reliably predicted future stroke; these included the APO (age, paresis, and old cerebral infarction) and APOL (age, paresis, old cerebral infarction, and largeartery stenosis) scores. The areas under the receiver operating characteristic curves were .662, .737, and .807 for ABCD2, APO, and APOL, respectively. Conclusions: Compared with the established measures, our newly created scores could predict future stroke for tissue-defined TIA more reliably. Key Words: Clinical score— large-artery disease—old cerebral infarction—risk factor—stroke prediction— transient ischemic attack. Ó 2013 by National Stroke Association

Introduction From the Department of Neurology and Cerebrovascular Medicine, Saitama Medical University International Medical Center, Hidaka, Japan. Received June 11, 2013; revision received September 9, 2013; accepted November 20, 2013. Address correspondence to Takeshi Hayashi, MD, PhD, Department of Neurology and Cerebrovascular Medicine, Saitama Medical University International Medical Center, 1397-1 Hidaka, Saitama Prefecture, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2013 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2013.11.020

Transient ischemic attack (TIA) is a medical emergency. Approximately 5% of the patients with TIA develop stroke within 48 hours, and 10%-20% of patients do so within 3 months.1-3 Conversely, 7%-40% of patients with ischemic stroke have experienced a previous TIA episode.3,4 The prompt evaluation of patients with TIA and appropriate medical intervention are, thus, mandatory. Subsequent stroke is a matter of serious concern, and physicians are often asked to evaluate the probability of future stroke after TIA. As this issue is of great

Journal of Stroke and Cerebrovascular Diseases, Vol. -, No. - (---), 2013: pp 1-6

1

T. HAYASHI ET AL.

2

importance in a clinical setting, it has been extensively investigated. Many factors, such as age, symptoms and their duration, comorbidities, and radiological features have been shown to influence the risk of stroke.1,5 To this end, the ABCD2 score, a risk stratification tool, was established and is now widely used as an aid for making medical decisions.6 In the meantime, a new definition of TIA was proposed by the American Heart Association/American Stroke Association as a Scientific Statement in 2009.3 TIA had previously been operationally defined as a focal cerebral ischemic event that lasts less than 24 hours (time-defined TIA).7 However, this definition does not reflect the pathobiological basis of the disease, and the time limit of 24 hours was considered too arbitrary. The absence of new cerebral infarction on diffusion-weighted image (DWI) in magnetic resonance imaging (MRI) is critical in discriminating between TIA and ischemic stroke, and this information was considered in the new definition of TIA (tissuedefined TIA).3,8 This new definition better contributes to our understanding of the pathophysiology of TIA and aids clinicians in making medical decisions.3,9 Knowledge of future stroke risk, and the ABCD2 score system, has been based on findings obtained for timedefined TIA.5,6,10 As the standard definition of TIA has changed from time defined to tissue defined, the risk evaluation system may need to be changed. In the present study, we aimed to evaluate the risks of future stroke in patients with tissue-defined TIA. Furthermore, we aimed to assess the reliability of the ABCD2 scoring system for tissue-defined TIA and, if necessary, develop new stratification tools.

Methods This study was approved by our institutional ethics committee (Institutional Review Board at Saitama Medical University International Medical Center). All patients who were admitted to our hospital with transient focal neurologic deficits underwent MRI, unless it was contraindicated. When the first MRI revealed no fresh lesions on DWI, a second MRI was performed 37 days later. Those without high-intensity lesions revealed by DWI on both scans were diagnosed with tissue-defined TIA. Patients with lasting focal neurologic deficits on admission that disappeared within 24 hours were also included, if repeated MRI after the neurologic deficits resolved demonstrated no new lesions. Patients with transient neurologic deficits and no DWI lesions by the first MRI, but caused cerebral infarction before the second MRI examination, were included. In total, there were 124 patents with tissue-defined TIA from April 2007 to December 2010, and among them, 74 patients could be followed for 2 years; these 74 patients were included in this study. Patients who did not undergo MRI were not included in the study.

We retrospectively reviewed patients’ profiles (age and sex), risk factors for stroke (hypertension, diabetes mellitus, dyslipidemia, and smoking), symptoms (presence of hemiparesis and/or dysarthria during the transient attack), and symptom duration. Patients with blood pressure of 140/90 mm Hg or more or who were receiving antihypertensive medication were said to have hypertension.11 Diabetes mellitus was defined as blood glucose level on admission of 200 mg/dL or more and HbA1c of 6.5% or more or if the patient was being treated with antidiabetic medication.11 Dyslipidemia was diagnosed when the patient had any of the following: low-density lipoprotein cholesterol of 140 mg/dL or more, high-density lipoprotein cholesterol of 40 mg/dL or less, triglycerides of 150 mg/dL or more, or was being treated with lipidlowering medication.11 We also assessed the presence of any old cerebral infarction and large-artery stenosis, regardless of their relevance to the transient symptoms. Old cerebral infarction was investigated by MRI, and high-intensity lesions on T2 and fluid-attenuated inversion recovery images were considered old infarctions. Large-artery stenosis was investigated by magnetic resonance angiography and carotid ultrasonography and more than 50% stenosis as assessed according to the North American Symptomatic Carotid Endarterectomy Trial criteria was considered to be stenotic.12,13 TIA etiology was categorized according to the Trial of Org 10172 in Acute Stroke Treatment.14,15 Any subsequent use of antiplatelets, anticoagulants, statins, and stenting or bypass formation was also reviewed. ABCD2 score,6 a stroke risk stratification tool developed for timedefined TIA, was assessed in all patients. The above-mentioned variables were compared between the groups with and without subsequent ischemic stroke occurring within 2 years. Statistical analysis was performed using the PASW Statistics software (version 18; SPSS Inc., Chicago, IL). Differences in age between the groups were analyzed using the Wilcoxon test, and differences in other variables were assessed using the chi-square test. Based on our results, we developed 2 new scales to predict future stroke risk. Factors with small P value by Wilcoxon or chi-square test were first selected, but to further confirm the pertinence, we carried out multivariate logistic regression analysis and determined the factors for the new scales. The predictive values of these scales, and that of ABCD2, were quantified using the area under the receiver operating characteristics curve (AUC) with a 95% confidence interval (CI).

Results Of the 124 patients who fulfilled the criteria for tissuedefined TIA, 74 patients were followed-up for 2 years and included in our analyses.

STROKE AFTER TISSUE-DEFINED TIA

3

The clinical characteristics of these patients are presented in Table 1. The mean age of the patients was 66.6 6 11.0 years. Fifty-one, 9, 29, and 16 patients had hypertension, diabetes mellitus, dyslipidemia, and smoking, respectively. Fifty-five patients had experienced hemiparesis and 32 patients had experienced dysarthria. The duration of symptoms was variable as shown in Table 1. Radiological investigation revealed old cerebral infarction in 39 patients and large-artery stenosis in 25 patients. TIA etiology was as follows: cardioembolism in 20 patients, large-artery atherosclerosis in 18 patients, smallvessel occlusion in 24 patients, 2 identified causes in 6 patients, and undetermined in 6 patients. Eleven patients experienced an ischemic stroke within 2 years. Age tended to be higher in the future stroke group, but this tendency did not reach statistical significance. There was also a tendency for hemiparesis and dysarthria to have occurred more frequently in the future stroke group. The presence of old cerebral infarction and large-artery stenosis were also more frequent in the future stroke group. TIA caused by large-artery atherosclerosis

more frequently led to subsequent ischemic stroke than those with other etiologies. ABCD2 score,6 a well-known stroke risk stratification scale, was assessed for all patients. As shown in Table 2, this score’s predictive value for future stroke was modest. As shown in Figure 1, AUC of this scale was .662 (95% CI .500-.824). Therefore, we developed new scales to predict future stroke, based on our results. As shown in Table 1, age, dysarthria, hemiparesis, old cerebral infarction, and large-artery stenosis all had a mild association with future stroke by Wilcoxon or chi-square test. Multivariate logistic regression analysis revealed that age, dysarthria, hemiparesis, old cerebral infarction, and large-artery stenosis had P values of .096, .263, .196, .210, and .144, respectively. Other factors showed larger P values (data not shown). We, thus, selected 3 or 4 factors with the smallest P value, which were age, hemiparesis, old cerebral infarction, and large-artery stenosis. As for age, we made receiver operating characteristics curve and found that cut-off age as 65 years made the largest Youden index and meant this cut-off age was most appropriate.

Table 1. Clinical characteristics of patients with tissue-defined TIA

Age, mean 6 SD Sex, male Clinical features Hemiparesis Dysarthria Risk factors Hypertension Diabetes mellitus Dyslipidemia Current smoking Atrial fibrillation Duration of symptoms ,10 min 10-60 min 60 min to 24 h Imaging features Old cerebral infarction Large-artery stenosis Etiology Cardioembolism Large-artery atherosclerosis Small-vessel occlusion Two causes identified Undetermined Subsequent medication Antiplatelets Anticoagulants Statins Stent or bypass formation

Total (N 5 74)

No ischemic stroke within 2 y (n 5 63)

Ischemic stroke within 2 y (n 5 11)

P value

66.6 6 11.0 44 (59.5%)

65.7 6 10.8 38 (60.3%)

72.1 6 10.3 6 (54.5%)

.090 .719

55 (74.3%) 32 (43.2%)

45 (71.4%) 25 (39.7%)

10 (90.9%) 7 (63.6%)

.172 .139

51 (68.9%) 9 (12.2%) 29 (39.2%) 16 (21.6%) 16 (21.6%)

43 (68.3%) 8 (12.7%) 26 (41.3%) 15 (23.8%) 14 (22.2%)

8 (72.7%) 1 (9.1%) 3 (27.3%) 1 (9.1%) 2 (18.2%)

.767 .736 .380 .274 .764

23 (31.1%) 32 (43.2%) 19 (25.7%)

20 (31.7%) 26 (41.3%) 17 (27.0%)

3 (27.3%) 6 (54.5%) 2 (18.2%)

.767 .721 .537

39 (52.7%) 25 (33.8%)

31 (49.2%) 19 (30.2%)

8 (72.7%) 6 (54.5%)

.149 .115

20 (27.0%) 18 (24.3%) 24 (32.4%) 6 (8.1%) 6 (8.1%)

19 (30.2%) 13 (20.6%) 21 (33.3%) 5 (7.9%) 5 (7.9%)

1 (9.1%) 5 (45.5%) 3 (27.3%) 1 (9.1%) 1 (9.1%)

.147 .077 .692 .897 .897

59 (79.7%) 16 (21.6%) 17 (23.0%) 7 (.9%)

50 (79.4%) 14 (22.2%) 16 (25.4%) 7 (11.1%)

9 (81.8%) 2 (18.2%) 1 (9.1%) 0 (.0%)

.852 .764 .236 .245

Abbreviations: TIA, transient ischemic attack.

T. HAYASHI ET AL.

4

Table 2. ABCD2 score and occurrence of ischemic stroke within 2 y of TIA

ABCD2 score

Total (N 5 74)

No ischemic stroke within 2 y (n 5 63)

Ischemic stroke within 2 y (n 5 11)

0 1 2 3 4 5 6 7

0 1 3 12 32 22 3 1

0 1 3 11 29 15 3 1

0 0 0 1 3 7 0 0

Abbreviation: TIA, transient ischemic attack.

APO (age, paresis, and old cerebral infarction) score gives 1 point each for age of 65 years or older, presence of hemiparesis, and presence of old cerebral infarction. As shown in Table 3, this score more reliably predicted future stroke in patients with tissue-defined TIA. As shown in Figure 1, AUC of this scale was .737 (95% CI .608-.865), which demonstrates its increased predictive value over the ABCD2 score. APOL (age, paresis, old cerebral infarction, and large-artery stenosis) score gives 1 point each for age of 65 years or older, presence of hemiparesis, presence of old cerebral infarction, and largeartery stenosis. As shown in Table 4, this score further improved the prediction of future stroke, with AUC of .807 (95% CI .700-.914; Fig 1). Although the difference did not reach statistical significance, our findings suggest that scales such as APO and APOL, rather than ABCD2, are better able to predict future stroke in patients with tissue-defined TIA.

Discussion Even if no neurologic deficits are apparent when a physician examines a patient with TIA, prompt examination and adequate therapy are extremely important. Approximately 5% of the patients with TIA develop ischemic stroke within 48 hours, and 10%-20% patients do so within 3 months.1-3 The occurrence (or recurrence) of ischemic stroke is more frequent in patients with TIA than in patients with ischemic stroke.16 Because the need for quick treatment is now generally accepted, and MRI assessments have become widely available, it is a standard protocol for patients with TIA to undergo MRI examination. MRI has revealed that between 30% and 50% of patients with TIA have a fresh infarction, seen on DWI.3,17,18 The following studies have demonstrated that patients with TIA having high-intensity lesions as seen by DWI more often develop future ischemic stroke. Coutts et al19 reported that the stroke risk within 90 days increased by 2.6 times with the presence of a DWI lesion. Similarly, Giles et al20 reported that 7.1% of the patients having TIA with a DWI lesion experienced ischemic stroke within 7 days, whereas only .4% of the patients without a DWI lesion did. American Heart Association/American Stroke Association, therefore, considered the classical definition of TIA arbitrary and proposed a new definition (tissue defined) for TIA.3 The ABCD2 score has been commonly used to predict future stroke.6 This score is relatively easy to assess and does not require any radiological examination results; thus, it is now widely used. However, this scoring system was based on the results of studies on classically defined (time defined) TIA.6 Furthermore, even in time-defined TIA, the presence of DWI lesions was found to be more indicative of future stroke risk than any score in the ABCD2.21 Radiological findings had better been

Figure 1. The receiver operating characteristics curve for ABCD2, APO, and APOL scores to predict ischemic stroke within 2 years. The areas under the curve with 95% confidence interval were .662 (.500-.824) for ABCD2, .737 (.608-.865) for APO, and .807 (.700-.914) for APOL. Abbreviations: APO, age, paresis, and old cerebral infarction; APOL, age, paresis, old cerebral infarction, and large-artery stenosis.

STROKE AFTER TISSUE-DEFINED TIA

5

Table 3. APO score and occurrence of ischemic stroke within 2 y of TIA

APO score

Total (N 5 74)

No ischemic stroke within 2 y (n 5 63)

Ischemic stroke within 2 y (n 5 11)

0 1 2 3

1 24 31 18

1 24 25 13

0 0 6 5

Abbreviations: APO, age, paresis, and old cerebral infarction; TIA, transient ischemic attack.

incorporated to stroke risk stratification tool, when radiological examination became widely used for the evaluation of TIA.22 In the present study, age and the presence of hemiparesis and/or dysarthria, both of which are scores in ABCD2, tended to increase future stroke risk, but these tendencies did not reach statistical significance. Blood pressure, diabetes mellitus, and the duration of TIA symptoms had no relationship to the occurrence of future stroke (Table 1). This may be not only because of the relatively small sample size but could also be because of the new definition of TIA being employed in this study. In timedefined TIA, the longer the duration of neurologic deficits, the more probable that a DWI lesion presents.3,23 Indeed, approximately 40% of patients with time-defined TIA experienced neurologic deficits for more than 1 hour 3,23; in contrast, our study on patients with tissue-defined TIA showed that only 25% patients experienced deficits for more than 1 hour (Table 1). To restrict our study to patients having TIA without DWI lesions may have negated the weight of symptom duration on future stroke risk. The reason diabetes mellitus and hypertension were not associated with future stroke is not clear. Both these factors are strongly associated with stroke recurrence after ischemic stroke,24-26 and this might have influenced the findings for patients with time-defined TIA. When we Table 4. APOL score and occurrence of ischemic stroke within 2 y of TIA

APOL score

Total (N 5 74)

No ischemic stroke within 2 y (n 5 63)

Ischemic stroke within 2 y (n 5 11)

0 1 2 3 4

1 17 28 20 8

1 17 27 12 6

0 0 1 8 2

Abbreviations: APOL, age, paresis, old cerebral infarction, and large-artery stenosis; TIA, transient ischemic attack.

restricted the patients to tissue-defined TIA, the effect of these factors could have reduced. The new scales proposed in this study are simpler than the ABCD2 scale. The APO score is a 3-point scale, and APOL is a 4-point scale, whereas the ABCD2 score incorporates 7 points.6 Both APO and APOL require MRI examination; however, because the TIA definition has now changed to a tissue-defined one, an MRI is required for TIA diagnoses itself.3 Therefore, use of the APO or APOL scales requires no additional examination. Even for classically defined TIA, the presence of large-artery stenosis increases stroke risk within 90 days by 4.1 times19 and the presence of old cerebral infarction increases stroke risk within 7 days by 4.3 times.19 These findings should help to increase the accuracy of future tools for predicting stroke. There are several limitations in this study. First, this study was carried out in a single institute, and the sample size was small; it is difficult to make a firm conclusion. Second, the follow-up rate was relatively low, which might have brought selection bias; aged patients with physical or economic problems tended not to attend our hospital for more than 2 years. To overcome these problems and make firmer conclusions, prospective studies with larger patient number are required. Notwithstanding, our results suggest that compared with previously used assessment scales, our new scales are more reliable tools for predicting future stroke in tissuedefined TIA. Acknowledgment: We are thankful to Dr Michio Shiibashi, Information Technology Center, Saitama Medical University, for kind advice about statistical analysis.

References 1. Johnston SC, Gress DR, Browner WS, et al. Short-term prognosis after emergency department diagnosis of TIA. JAMA 2000;284:2901-2906. 2. Hill MD, Yiannakoulias N, Jeerakathil T, et al. The high risk of stroke immediately after transient ischemic attack: a population-based study. Neurology 2004;62:2015-2020. 3. Easton JD, Saver JL, Albers GW, et al. Definition and evaluation of transient ischemic attack: a scientific statement for healthcare professionals from the American Heart Association/American Stroke Association Stroke Council; Council on Cardiovascular Surgery and Anesthesia; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular Nursing; and the Interdisciplinary Council on Peripheral Vascular Disease. The American Academy of Neurology affirms the value of this statement as an educational tool for neurologists. Stroke 2009;40:2276-2293. 4. Feinberg WM, Albers GW, Barnett HJ, et al. Guidelines for the management of transient ischemic attacks. From the Ad Hoc Committee on Guidelines for the Management of Transient Ischemic Attacks of the Stroke Council of the American Heart Association. Circulation 1994; 89:2950-2965.

T. HAYASHI ET AL.

6 5. Giles MF, Rothwell PM. Risk of stroke early after transient ischaemic attack: a systematic review and metaanalysis. Lancet Neurol 2007;6:1063-1072. 6. Johnston SC, Rothwell PM, Nguyen-Huynh MN, et al. Validation and refinement of scores to predict very early stroke risk after transient ischaemic attack. Lancet 2007; 369:283-292. 7. Special report from the National Institute of Neurological Disorders and Stroke. Classification of cerebrovascular diseases III. Stroke 1990;21:637-676. 8. Albers GW, Caplan LR, Easton JD, et al. Transient ischemic attack—proposal for a new definition. N Engl J Med 2002;347:1713-1716. 9. Ay H, Koroshetz WJ, Benner T, et al. Transient ischemic attack with infarction: a unique syndrome? Ann Neurol 2005;57:679-686. 10. Kimura K, Kazui S, Minematsu K, et al. Analysis of 16,922 patients with acute ischemic stroke and transient ischemic attack in Japan. A hospital-based prospective registration study. Cerebrovasc Dis 2004;18:47-56. 11. Deguchi I, Ohe Y, Fukuoka T, et al. Relationship of obesity to recanalization after hyperacute recombinant tissueplasminogen activator infusion therapy in patients with middle cerebral artery occlusion. J Stroke Cerebrovasc Dis 2012;21:161-164. 12. Wardlaw JM, Lewis S. Carotid stenosis measurement on colour Doppler ultrasound: agreement of ECST, NASCET and CCA methods applied to ultrasound with intraarterial angiographic stenosis measurement. Eur J Radiol 2005;56:205-211. 13. Alamowitch S, Eliasziw M, Algra A, et al. Risk, causes, and prevention of ischaemic stroke in elderly patients with symptomatic internal-carotid-artery stenosis. Lancet 2001;357:1154-1160. 14. Madden KP, Karanjia PN, Adams HP Jr, et al. Accuracy of initial stroke subtype diagnosis in the TOAST study. Trial of ORG 10172 in Acute Stroke Treatment. Neurology 1995;45:1975-1979. 15. Hankey GJ, Warlow CP. Lacunar transient ischaemic attacks: a clinically useful concept? Lancet 1991;337:335-338. 16. Chen ZM, Sandercock P, Pan HC, et al. Indications for early aspirin use in acute ischemic stroke: a combined

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

analysis of 40000 randomized patients from the Chinese acute stroke trial and the international stroke trial. On behalf of the CAST and IST collaborative groups. Stroke 2000;31:1240-1249. Inatomi Y, Kimura K, Yonehara T, et al. DWI abnormalities and clinical characteristics in TIA patients. Neurology 2004;62:376-380. Purroy F, Montaner J, Rovira A, et al. Higher risk of further vascular events among transient ischemic attack patients with diffusion-weighted imaging acute ischemic lesions. Stroke 2004;35:2313-2319. Coutts SB, Simon JE, Eliasziw M, et al. Triaging transient ischemic attack and minor stroke patients using acute magnetic resonance imaging. Ann Neurol 2005; 57:848-854. Giles MF, Albers GW, Amarenco P, et al. Early stroke risk and ABCD2 score performance in tissue- vs timedefined TIA: a multicenter study. Neurology 2011; 77:1222-1228. Arhami Dolatabadi A, Meisami A, Hatamabadi H, et al. Improving the prediction of stroke or death after transient ischemic attack (TIA) by adding diffusionweighted imaging lesions and TIA etiology to the ABCD2 score. J Stroke Cerebrovasc Dis 2013;22:e24-e30. Asimos AW, Rosamond WD, Johnson AM, et al. Early diffusion weighted MRI as a negative predictor for disabling stroke after ABCD2 score risk categorization in transient ischemic attack patients. Stroke 2009; 40:3252-3257. Shah SH, Saver JL, Kidwell CS, et al. A multicenter pooled, a patient-level data analysis of diffusionweighed MRI in TIA patients. Stroke 2007;38:463. Tanaka H, Ueda Y, Hayashi M, et al. Risk factors for cerebral hemorrhage and cerebral infarction in a Japanese rural community. Stroke 1982;13:62-73. PROGRESS Collaborative Group. Randomised trial of a perindopril-based blood-pressure-lowering regimen among 6,105 individuals with previous stroke or transient ischaemic attack. Lancet 2001;358:1033-1041. Abbott RD, Donahue RP, MacMahon SW, et al. Diabetes and the risk of stroke. The Honolulu Heart Program. JAMA 1987;257:949-952.

Prediction of ischemic stroke in patients with tissue-defined transient ischemic attack.

The risk of future stroke after transient ischemic attack (TIA) has been widely studied, but most findings were obtained for classically defined TIA (...
313KB Sizes 0 Downloads 0 Views