Browse the corpus

Introduction Stroke is the most prevalent cardiovascular disease and the most prevalent neurological disease in Asia.1 Many countries in East Asia and Southeast Asia have higher mortality rates from stroke than from ischemic heart disease, the opposite of Western countries.1 The prevalence of intracerebral hemorrhage (ICH) and intracranial arterial sclerosis is another unique feature of Asian patients.2,3 Among Asian countries, Japan was the first to become an aging society; the others, in particular Korea, have been rapidly approaching one. Thus, the epidemiologic characteristics of stroke in Japan seem to be good examples for other countries. In this review, epidemiological studies and patients' registry studies of stroke in Japan are briefly introduced.

Introduction Stroke is the most prevalent cardiovascular disease and the most prevalent neurological disease in Asia.1 Many countries in East Asia and Southeast Asia have higher mortality rates from stroke than from ischemic heart disease, the opposite of Western countries.1 The prevalence of intracerebral hemorrhage (ICH) and intracranial arterial sclerosis is another unique feature of Asian patients.2,3 Among Asian countries, Japan was the first to become an aging society; the others, in particular Korea, have been rapidly approaching one. Thus, the epidemiologic characteristics of stroke in Japan seem to be good examples for other countries. In this review, epidemiological studies and patients' registry studies of stroke in Japan are briefly introduced. The Hisayama Study The Hisayama Study was begun as a population-based prospective cohort study of cerebrovascular and cardiovascular diseases in 1961 in the town of Hisayama, a suburban community adjacent to the Fukuoka etropolitan area, Kyushu, in western Japan. Four study cohorts were established from Hisayama residents ≥40 years of age in 1961, 1974, 1988, and 2002 after screening examinations. One of the strengths of this study is that most of the deceased subjectsof the study underwent autopsy examinations from the beginning of the study (80% between 1962 and 1994),4 and thus, the morphological features of the brains examined by autopsy or brain imaging are available for most of the stroke cases in each cohort. The study was initiated to respond to the doubts of Western researchers in the pre-CT era that the very high mortality from ICH in Japan might be due to overdiagnosis of ICH. The autopsy results in the consecutive residents proved that the prevalence of ICH was not so high as was believed by Japanese physicians but also showed that ICH was still more common than ischemic stroke as a cause of death in Japan.5

hat the very high mortality from ICH in Japan might be due to overdiagnosis of ICH. The autopsy results in the consecutive residents proved that the prevalence of ICH was not so high as was believed by Japanese physicians but also showed that ICH was still more common than ischemic stroke as a cause of death in Japan.5 Of the many studies on stroke and other neurological diseases including dementia, those on stroke incidence and mortality are briefly introduced here. After 12-year follow-up for each of the first three study cohorts, the age-adjusted incidences of total stroke were 1,210 per 100,000 person-years for men and 598 for women in the first cohort (1961); they declined steeply in both sexes from the first to the second cohort (1974) and then declined relatively moderately in both sexes from the second to the third cohort (1988, Figure 1).6 Changes in the incidence among cohorts differed greatly between ischemic stroke and ICH. The incidence of ischemic stroke declined by 37% for men from the first to the second cohort, while the incidence of ICH declined by 61% for men. In contrast, the age-adjusted incidences of coronary heart disease were 340 per 100,000 person-years for men and 113 per 100,000 person-years for women in the first cohort, and they increased for both sexes in the newer cohorts, although they were much smaller than the stroke incidences in all of the cohorts. The different tendencies in the changes in incidence between stroke and coronary heart disease seem to be partly due to changes in prevalence of cardiovascular risk factors among the three cohorts: severe hypertension and current smoking became significantly less frequent, while glucose intolerance, dyslipidemia, and obesity became more frequent. Stroke mortality declined continuously as a result of changes in stroke incidence and significant improvements in acute stroke management; the age-adjusted stroke mortalities among the three cohorts were 634 (the first cohort: 1961), 232 (the second cohort: 1974), and 138 (the third cohort: 1988) per 100,000 person-years, respectively, for men and 286, 162, and 102 per 100,000 person-years, respectively, for women.

improvements in acute stroke management; the age-adjusted stroke mortalities among the three cohorts were 634 (the first cohort: 1961), 232 (the second cohort: 1974), and 138 (the third cohort: 1988) per 100,000 person-years, respectively, for men and 286, 162, and 102 per 100,000 person-years, respectively, for women. Among the ischemic stroke subtypes, the age-adjusted incidence of lacunar infarction declined significantly from the first to the third cohort for both sexes (5.68 per 100,000 person-years in the first cohort and 1.59 per 100,000 person-years in the third cohort for men during the 13-year follow-up), whereas the incidences of atherothrombotic and cardioembolic infarctions did not change during this period.7 As a result, the proportion of ischemic stroke subtypes differed greatly among the 3 cohorts; two-thirds of the male patients had lacunar infarction in the first cohort, compared to two-fifths in the third cohort. The high incidence in the first cohort and recent decline of lacunar infarction were similar to those for ICH, suggesting that intracranial small artery disease has been prevalent in the Japanese population and that the effect of recent developments in preventive therapy, especially antihypertensive therapy, are protective from development of the small artery disease.

decline of lacunar infarction were similar to those for ICH, suggesting that intracranial small artery disease has been prevalent in the Japanese population and that the effect of recent developments in preventive therapy, especially antihypertensive therapy, are protective from development of the small artery disease. Of the 410 patients in the first cohort who developed first ever stroke during 32-year follow-up, 108 (26%) experienced recurrent stroke within 10 years after the index stroke.8 The cumulative recurrence rates at 1, 5, and 10 years were: 10.0%, 34.1%, and 49.7% after ischemic stroke; 25.6%, 34.9%, and 55.6% after ICH; and 32.5%, 55.0%, and 70.0% after subarachnoid hemorrhage (SAH), respectively. Of the 333 patients in the first cohort who developed first-ever stroke during 26-year follow-up, 268 (80.5%) died within 10 years after the index stroke, of whom 239 (89.2%) underwent autopsy examinations.9 The risk of death was greatest in the first year (men 40.3%; women 43.7%). The 30-day case fatality rate was substantially greater in patients with ICH (63.3%) or SAH (58.6%) than in patients with ischemic stroke (9.0%). The risk of dying after the index stroke was twelve times higher during the first year and two times higher during the overall 26-year period as compared to the risk for stroke-free controls. The most common cause of death was the index stroke in the first year, and the impact of recurrent stroke increased gradually thereafter.

of dying after the index stroke was twelve times higher during the first year and two times higher during the overall 26-year period as compared to the risk for stroke-free controls. The most common cause of death was the index stroke in the first year, and the impact of recurrent stroke increased gradually thereafter. The Hisayama Study is one ofthe first sophisticated epidemiological study and one of the most successful epidemiological study of cerebrovascular and cardiovascular diseases in the world. Several unique characteristics of Asian stroke patients were ascertained by this study. The Hisayama Study is still developing by expanding the target diseases into common nonvascular diseases and by adding genomic information for the analysis. The Suita Study Following the Hisayama Study, several epidemiological projects on cerebrovascular and cardiovascular diseases were started in Japan. Most of the study cohorts involved rural or suburban residents, since they are likely to continue to live in the area. The Suita Study was unique in that urban residents were registered.

Following the Hisayama Study, several epidemiological projects on cerebrovascular and cardiovascular diseases were started in Japan. Most of the study cohorts involved rural or suburban residents, since they are likely to continue to live in the area. The Suita Study was unique in that urban residents were registered. Suita city, which contains the National Cerebral and Cardiovascular Center where the author works, is located adjacent to Osaka city, which is the second largest metropolitan area in Japan. The Suita Study was based on a random sampling of 12,200 Japanese urban residents. At baseline, participants between the ages of 30 and 79 years were randomly selected from the municipality's population registry and stratified into groups by sex and age in 10-year increments in 1989. Of these, 6,485 people underwent regular health checkups between 1989 and 1994. During an average 11.7-year (64,391 person-years) follow-up period, 213 strokes, consisting of 141 ischemic stroke, 32 ICH, 22 SAH, and 18unclassified strokes, and 133 myocardial infarctions were documented.10,11 Thus, the incidence of stroke did not differ much as compared to that of myocardial infarction in contrast to the high stroke incidence in the Hisayama Study (especially in its first cohort; the age-adjusted incidence of total stroke for menbeing 1,210 per 100,000 person-years and that of coronary heart disease being 340 per 100,000 person-years), although adjustments for age and other conditions are needed for accurate comparison between the studies. These findings suggest that the data from the Suita Study were influenced by the Western lifestyle, particularly diet.

er 100,000 person-years and that of coronary heart disease being 340 per 100,000 person-years), although adjustments for age and other conditions are needed for accurate comparison between the studies. These findings suggest that the data from the Suita Study were influenced by the Western lifestyle, particularly diet. Among the many publications from the Suita Study, those on the association between blood pressure (BP) levels and stroke incidence are briefly introduced here. The association between high-normal BP and cerebrovascular and cardiovascular disease had not been well studied in the Asian population. The percent ages of the participants with optimal, normal, and high-normal BP and hypertension Stage 1 and Stage ≥2, according to the ESH-ESC 2007 criteria, were 31%, 20%, 18%, 20%, and 11% for men and 42%, 17%, 16%, 16%, and 9% for women, respectively.9 Compared with the optimal BP group, the multivariate hazard ratios (HRs) (95% confidence intervals [CIs]) of stroke for normal and high-normal BP and hypertension Stage 1 and Stage ≥2 were 2.12 (1.04 to 4.30), 2.43 (1.21 to 4.86), 2.62 (1.35 to 5.09), and 4.38 (2.24 to 8.56) in men and 1.05 (0.49 to 2.24), 1.29 (0.63 to 2.67), 1.21 (0.61 to 2.45), and 2.20 (1.07 to 4.50) in women, respectively; the risk of myocardial infarction for each BP category was similar to that of stroke. Population-attributable fractions of high-normal BP and hypertension for combined stroke and myocardial infarction were 12.2% and 35.3% in men and 7.1% and 23.4% in women, respectively (Figure 2). These findings indicate the significance of pre-hypertension as a vascular risk factor and the necessity for pre-hypertensive patients to attempt to control BP through lifestyle modifications.

for combined stroke and myocardial infarction were 12.2% and 35.3% in men and 7.1% and 23.4% in women, respectively (Figure 2). These findings indicate the significance of pre-hypertension as a vascular risk factor and the necessity for pre-hypertensive patients to attempt to control BP through lifestyle modifications. The combined impacts of BP categories and other risk factors were also thoroughly investigated in the Suita Study. A study on glucose abnormalities and that on chronic kidney disease (CKD) are summarized.11,12 The percentages of subjects with normoglycemia, impaired fasting glucose, and diabetes mellitus, defined according to the 2003 American Diabetes Association recommendations, were 59%, 35%, and 6% for men and 75%, 21%, and 4% for women, respectively.12 Compared with normoglycemic subjects, the multivariate HRs (95% CIs) for stroke were 1.11 (0.81-1.52) in individuals with impaired fasting glucose and 2.08 (1.29-3.35) in individuals with diabetes mellitus. Compared with normoglycemic and optimal BP subjects, increased risks of combined stroke and coronary heart disease were observed in the normoglycemic subjects with high-normal BP or hypertension, in impaired fasting glucose subjects with normal or higher BP, and in diabetic subjects regardless of BP category (P-value for interaction=0.046). The percentages of CKD subjects, defined as an estimated glomerular filtration rate (GFR) <60 mL/min/1.73 m2, using the Modification of Diet in Renal Disease equation modified by the Japanese coefficient (0.881), were 8.9% for men and 11.3% for women.10 Compared with the GFR ≥90 mL/min/1.73 m2 group, the HRs (95% CIs) for stroke were 1.9 (1.3 to 3.0) in the GFR 50 to 59 mL/min/1.73 m2 group and 2.2 (1.2 to 4.1) in the GFR <50 mL/min/1.73 m2 group. Compared with the optimal BP subjects without CKD, the normal BP, high-normal BP, and hypertensive subjects without CKD showed increased risks of stroke.However, the impact of each BP category on stroke (P for interaction: 0.03 in men, 0.90 in women) was more evident in men with CKD. These results show that pre-hypertension can be a stronger vascular risk factor when combined with other traditional and newer risk factors than when it is the sole risk factor.

of stroke.However, the impact of each BP category on stroke (P for interaction: 0.03 in men, 0.90 in women) was more evident in men with CKD. These results show that pre-hypertension can be a stronger vascular risk factor when combined with other traditional and newer risk factors than when it is the sole risk factor. As is known, extracranial carotid atherosclerotic lesions are less frequent in the Asian population than in the Western population. The prevalence of asymptomatic extracranial carotid artery lesions and its relationship to cardiovascular risk factors were determined using ultrasound in the Suita residents.13 Significant sex differences were shown in the prevalence of atherosclerotic lesions in the extracranial carotid artery; 4.4% of all the subjects, 7.9% of the men, and 1.3% of the women aged 50 to 79 years had atherosclerosis accompanied by area stenosis >50%, and these values increased to 6.5%, 11.1%, and 2.1% for the subjects aged 60 to 79 years, respectively (Figure 3). In addition, accumulation of established major coronary risk factors (i.e., hypertension, smoking, and hypercholesterolemia) affected carotid atherogenesis in both sexes.14 Registry studies on stroke In this chapter, major registry studies on ischemic stroke in Japan are introduced (Table 1).

As is known, extracranial carotid atherosclerotic lesions are less frequent in the Asian population than in the Western population. The prevalence of asymptomatic extracranial carotid artery lesions and its relationship to cardiovascular risk factors were determined using ultrasound in the Suita residents.13 Significant sex differences were shown in the prevalence of atherosclerotic lesions in the extracranial carotid artery; 4.4% of all the subjects, 7.9% of the men, and 1.3% of the women aged 50 to 79 years had atherosclerosis accompanied by area stenosis >50%, and these values increased to 6.5%, 11.1%, and 2.1% for the subjects aged 60 to 79 years, respectively (Figure 3). In addition, accumulation of established major coronary risk factors (i.e., hypertension, smoking, and hypercholesterolemia) affected carotid atherogenesis in both sexes.14 Registry studies on stroke In this chapter, major registry studies on ischemic stroke in Japan are introduced (Table 1). The Japan Multicenter Stroke Investigators' Collaboration (J-MUSIC) was a nationwide, multicenter, prospective, hospital-based registration study from May 1999 through April 2000, when intravenous recombinant tissue plasminogen activator (rt-PA) was not yet approved for clinical use. A total of 156 hospitals participated in the study, and 16,922 patients (70.6±11.5 years old) with acute ischemic stroke (94%) and transient ischemic attack (TIA, 6%) who were hospitalized within 7 days of onset were registered. As was common in the Asian population, lacunar stroke was the leading subtype (38.8%), followed by atherothrombotic (33.3%) and cardioembolic stroke (21.8%). The median National Institutes of Health stroke scale (NIHSS) score on admission was 5 (interquartile range 2 to 11), and 60.8% of the patients had a modified Rankin Scale (mRS)score of 0-2 at discharge, while 6.9% died during acute hospitalization.15 In the follow-up study of survivors, the 1-year cumulative mortality was 6.8%, which was relatively low compared to that from Western countries.16 The cause of death included cerebrovascular disease in 24.1%, pneumonia in 22.6%, heart disease in 18.1%, and cancer in 11.0%.

hile 6.9% died during acute hospitalization.15 In the follow-up study of survivors, the 1-year cumulative mortality was 6.8%, which was relatively low compared to that from Western countries.16 The cause of death included cerebrovascular disease in 24.1%, pneumonia in 22.6%, heart disease in 18.1%, and cancer in 11.0%. The Japan Standard Stroke Registry Study (JSSRS) is an ongoing multicenter stroke registration study based on a computerized database from 162 Japanese institutes. From January 2000 through November 2007, a total of 47,782 patients with acute stroke and TIA who were hospitalized within 7 days after onset was registered. Many subanalyses of the registry data have been reported in Japanese books published every two to four years. As the major findings, 75.4% of stroke patients had ischemic stroke, 17.8% had ICH, and the remaining 6.8% had SAH. As subtypes of ischemic stroke, 33.9% had atherothrombotic, 31.9% had lacunar, and 27.0% had cardioembolic stroke. It is interesting that the leading stroke subtype changed from lacunar stroke in J-MUSIC (1999-2000) to atherothrombotic stroke in JSSRS (2000-2007), although the participating hospitals and designs of the two studies were not identical. Effects of sex and age on stroke subtypes, underlying risk factors, initial conditions at onset, and outcomes of ischemic stroke patients were reported in English.17 Briefly, women were older than men at stroke onset (75.0±11.7 years versus 69.3±11.4 years), and women more frequently had cardioembolic events (odds ratio [OR] 1.090, 95% CI 1.036 to 1.146) after age-adjustment. Onset-to-arrival time was longer (β=0.0554, P=0.026), the initial NIHSS score was higher (β=0.1565, P<0.001), and duration of hospitalization was longer (β=0.0355, P=0.010) in women than in men after multivariate adjustment. At hospital discharge, women less commonly had an mRS score of 0-1 (OR 0.802, 95% CI 0.741 to 0.868) and more commonly had an mRS score of 4-6 (OR 1.410, 95% CI 1.293 to 1.537) than men. Thus, women developed more severe strokes than men in Japan.

(β=0.0355, P=0.010) in women than in men after multivariate adjustment. At hospital discharge, women less commonly had an mRS score of 0-1 (OR 0.802, 95% CI 0.741 to 0.868) and more commonly had an mRS score of 4-6 (OR 1.410, 95% CI 1.293 to 1.537) than men. Thus, women developed more severe strokes than men in Japan. The Fukuoka Stroke Registry (FSR) is an ongoing, multicenter, hospital-based registry in which acute stroke patients were enrolled from seven stroke centers in the Fukuoka metropolitan area. The FSR has the strengths that the database extensively collected underlying patients' information, image data principally using MRI/MRA, long-term follow-up of vital and functional conditions for years, and serological and genome genetic analyses for most participants. The associations of several risk factors, including pre-stroke glycemic control18 and admission proteinuria19 with clinical outcomes of ischemic stroke patients were published in the last couple of years. As a unique risk factor of ischemic stroke in Japanese, and probably in Korean people, a windblown sand dust originating from mineral soil in the deserts of China and Mongolia was significantly associated with the incidence of atherothrombotic brain infarction after adjusting for expected confounders, including meteorologic variables and other air pollutants in this cohort.20

d probably in Korean people, a windblown sand dust originating from mineral soil in the deserts of China and Mongolia was significantly associated with the incidence of atherothrombotic brain infarction after adjusting for expected confounders, including meteorologic variables and other air pollutants in this cohort.20 Finally, let us consider the Stroke Acute Management with Urgent Risk-factor Assessment and Improvement (SAMURAI) rt-PA Registry.21 This registry included 600 consecutive patients (377 men, 72±12 years old) with ischemic stroke and TIA who received intravenous rt-PA therapy in ten Japanese stroke centers that were balanced regionally between October 2005 (when intravenous alteplase therapy was approved in Japan) and July 2008. Symptomatic ICH within 36 hours with ≥1-point increase from the baseline NIHSS score developed in 3.8% of patients (95% CI 2.6 to 5.7%). At 3 months, 33.2% (95% CI 29.5 to 37.0%) of patients had an mRS score of 0-1, and the mortality was 7.2% (95% CI 5.4 to 9.5%). Analysis of 399 patients with a premorbid mRS score ≤1 who met the approved European indications(≤80 years old, an initial NIHSS score ≤24, etc.) showed that 40.6% (95% CI 35.9 to 45.5%) had a 3-month mRS score of 0-1. These percentages were similar to those in Western postmarketing surveys using 0.9 mg/kg alteplase. Several published subanalyses clarified the associations of risk factors and initial stroke features with thrombolysis outcomes.

S score ≤24, etc.) showed that 40.6% (95% CI 35.9 to 45.5%) had a 3-month mRS score of 0-1. These percentages were similar to those in Western postmarketing surveys using 0.9 mg/kg alteplase. Several published subanalyses clarified the associations of risk factors and initial stroke features with thrombolysis outcomes. The publications that were discussed in this review dealt with only a small part of each study, and the studies that were introduced represent only a small part of Japanese epidemiologic and registry studies. The author hopes that the readers of this journal will find the similarities (or differences) in stroke epidemiology between Japanese people and those in other countries of great interest. Acknowledgements I would like to thank Drs. Takanari Kitazono (Kyushu University), Yutaka Kiyohara (Kyushu University), Shotai Kobayashi (Shimane University), Yoshihiro Kokubo (National Cerebral and Cardiovascular Center), and Takenori Yamaguchi (National Cerebral and Cardiovascular Center) for the valuable advice. The authors have no financial conflicts of interest. Figure 1 Age-specific incidences of stroke and coronary heart disease among the 3 cohorts of the Hisayama Study, with 12-year follow-up in each cohort.6 Figure 2 The HRs and positive fractions attributable to exposure to each blood pressure category at baseline for cardiovascular disease (including stroke): the Suita Study. The gray area displays the excessive incidence of CVD due to normal and high-normal blood pressures and hypertension stages 1 and ≥2 (From reference 10 with permission).

ositive fractions attributable to exposure to each blood pressure category at baseline for cardiovascular disease (including stroke): the Suita Study. The gray area displays the excessive incidence of CVD due to normal and high-normal blood pressures and hypertension stages 1 and ≥2 (From reference 10 with permission). Figure 3 Percentage of subjects with asymptomatic carotid artery stenosis: the Suita Study.13 Table 1 Registry studies on Japanese stroke patients

Discovering biomarkers for stroke In a narrow sense, biomarkers refer to indicators measured by chemical or biologic tests using blood or urine that predicts physiologic or disease states, or increased disease risk. Biomarkers are also a valuable tool in drug development, providing more accurate and complete information regarding drug performance, disease progression, or response to a specific drug therapy. In the research field of myocardial infarction, the role of biomarkers has been emphasized over a long period of time. Treatment according to the biomarkers has also been investigated in various diseases including diabetes or immunological disorders. On the contrary, there has been a relative dearth of biomarker research in cerebrovascular disease. Herein, we review the role of current and new stroke biomarkers with their strengths and weaknesses, focusing on the importance of comprehensive approaches. Learning from the failure of recent clinical trials Over the recent 10 years, numerous large multicenter randomized clinical trials (RCTs) on stroke patients have been performed in the stroke research field. However, almost all major studies including RCTs regarding secondary prevention of stroke,1-3 MR-based thrombolytic therapy,4,5 and STAIR (Stroke Treatment Academic Industry Roundtable) criteria-guided neuroprotection6 have failed to show meaningful clinical benefits. In this regard, several issues have been suggested to explain and overcome these failures.

ding RCTs regarding secondary prevention of stroke,1-3 MR-based thrombolytic therapy,4,5 and STAIR (Stroke Treatment Academic Industry Roundtable) criteria-guided neuroprotection6 have failed to show meaningful clinical benefits. In this regard, several issues have been suggested to explain and overcome these failures. First, it is warranted for more larger and methodologically sound RCTs which meet the STAIR7 and CONSORT8 (CON-solidated Standards Of Reporting Trials) criteria. These may enhance the success rate and reliability of the study. It is clear that findings derived from large-scale intervention trials have provided the impetus to change guidelines for stroke treatment. Nonetheless, direct application of RCTs results to daily clinical practice is dubious, because patients enrolled in large RCTs may not be representative of patients in our clinical practice.9 Second, there has been increasing interest in new statistical approaches to end-point analysis in RCTs (from dichotomized outcome scales to global statistics, responder, or shift analysis).10 These are mainly used to reduce sample size or enhance trial efficiency.11 Unfortunately, however, researchers cannot draw any new findings from RCTs through the novel statistical methods.

l approaches to end-point analysis in RCTs (from dichotomized outcome scales to global statistics, responder, or shift analysis).10 These are mainly used to reduce sample size or enhance trial efficiency.11 Unfortunately, however, researchers cannot draw any new findings from RCTs through the novel statistical methods. Third, the importance of considering heterogeneity among stroke patients has emerged. Unlike coronary heart disease, stroke has heterogeneous pathophysiologies and mechanisms. Moreover, individual patients with stroke have different features even among subjects with same stroke mechanisms. These aspects enhance the need for development of personalized medicine based on the characteristics of each patient rather than performing large RCTs. Stroke biomarkers may provide the information on the heterogeneity and could be a guiding tool for more effective personalized therapy among patients with ischemic cerebrovascular disease. Role of biomarkers in stroke research Emerging roles of stroke biomarkers are summarized in Table 1. Although the roles of biomarkers are basically diagnosing the disease and predicting the outcome, biomarkers in patients with stroke can also provide a large variety of other information about the risk of future stroke, possible stroke mechanisms for biomarker-guided treatment, or drug response. In addition, they can be used as surrogate endpoints in clinical trials.

agnosing the disease and predicting the outcome, biomarkers in patients with stroke can also provide a large variety of other information about the risk of future stroke, possible stroke mechanisms for biomarker-guided treatment, or drug response. In addition, they can be used as surrogate endpoints in clinical trials. Screening high-risk subjects Although many attempts, including national publicity and various programs for health promotion, have been made to manage stroke risk factors, the prevalence of stroke has not been markedly reduced. This may be partially attributable to hidden risk factors of stroke. Interestingly, certain regions in the United States (Stroke Belt and Buckle) have an unusually high incidence and mortality of stroke and the phenomenon could not be explained by the differences of the conventional risk factors.12,13 The exact causes of the higher incidence and mortality of stroke in the regions have not been recognized. Therefore, many researchers have devoted themselves to find novel risk factors of stroke to explain it, whereupon numerous possible contributing factors have been identified, including obesity/metabolic syndrome, diet, sleep-related breathing disorders, air pollution, and cultural lifestyle.14

have not been recognized. Therefore, many researchers have devoted themselves to find novel risk factors of stroke to explain it, whereupon numerous possible contributing factors have been identified, including obesity/metabolic syndrome, diet, sleep-related breathing disorders, air pollution, and cultural lifestyle.14 In addition to finding these new risk factors, a series of biomarkers reflecting inflammation, hemostasis, thrombosis, endothelial function, or neurohormonal activity have been evaluated as potential tools in an effort to improve risk prediction of future stroke, and thereby avert future events.15-22 For example, a recent investigation using data from the Framingham offspring study found that plasma asymmetrical dimethylarginine (ADMA) which is an inhibitor of endothelial nitric oxide synthase (eNOS) and a marker of endothelial dysfunction was associated with a prevalence of silent brain infarcts which is an important correlate of risk of future stroke.19 More recently, the same group published data on multiple biomarkers and identified that baseline B-type natriuretic peptide (BNP) having diuretic and vasodilatory activities and a urinary albumin/creatinine ratio indicating endothelial function were associated with the risk of incident stroke, and offered modest improvements in the accuracy of risk stratification.22 In the near future, a genome-wide association study may also greatly contribute to building risk stratification models by identifying genetic variants that confer susceptibility to cerebrovascular disease.23

ated with the risk of incident stroke, and offered modest improvements in the accuracy of risk stratification.22 In the near future, a genome-wide association study may also greatly contribute to building risk stratification models by identifying genetic variants that confer susceptibility to cerebrovascular disease.23 Rapid stroke diagnosis Although the diagnosis of acute stroke mostly relies on neuroimaging techniques, the evaluation of biomarkers of tissue injury would be an alternative strategy for rapid stroke assessment. This approach has already been successfully applied in the early management of other diseases including coronary heart disease (troponin, CK-MB), pulmonary embolism (D-dimer), and congestive heart failure (BNP).24-26 A rapid diagnosis of stroke based on biomarkers may be useful especially for pre-hospital screening, facilitating entry into a fast track care pathway, and for ancillary data when contemplating thrombolysis. However, a widely available, rapid, and sensitive diagnostic test for acute cerebral ischemia has not been available until now.

is of stroke based on biomarkers may be useful especially for pre-hospital screening, facilitating entry into a fast track care pathway, and for ancillary data when contemplating thrombolysis. However, a widely available, rapid, and sensitive diagnostic test for acute cerebral ischemia has not been available until now. Recently, a biomarker panel rather than a single marker in isolation has been increasingly used to improve the diagnostic accuracy of suspected stroke. For instance, a diagnostic panel incorporating the levels of matrix metalloproteinase 9 (MMP-9), BNP, D-dimer, and S-100β into a composite score enhanced sensitivity of early noncontrast CT alone for acute stroke, although the diagnostic accuracy was clearly imperfect.27 Furthermore, the approach was feasible as a point-of-care test in the emergency setting.27 As the number of presumed biomarkers for stroke expands at an exponential rate, it would be expected to develop improved biomarker combinations for more accurate diagnosis of stroke.

e diagnostic accuracy was clearly imperfect.27 Furthermore, the approach was feasible as a point-of-care test in the emergency setting.27 As the number of presumed biomarkers for stroke expands at an exponential rate, it would be expected to develop improved biomarker combinations for more accurate diagnosis of stroke. Detection of possible stroke mechanisms Several studies have focused on the use of biomarkers for detecting possible stroke mechanisms. A recent study investigated whether concentrations of von Willebrand factor (vWF) which plays crucial roles in thrombus formation differ depending on the etiologic subtypes of stroke, and found the highest levels in large artery disease and cardioembolic stroke.28 Recent data by our group also demonstrated that inflammatory markers, rather than traditional risk factors were associated with clinical and radiological differences among patients with atherosclerotic stroke.29 In addition, the molecular markers related to neuronal death can provide information about the presence of tissue at risk of infarction.30,31

onstrated that inflammatory markers, rather than traditional risk factors were associated with clinical and radiological differences among patients with atherosclerotic stroke.29 In addition, the molecular markers related to neuronal death can provide information about the presence of tissue at risk of infarction.30,31 Predicting drug response and outcome It has been well known that different patients respond in different manners to the same medication. Among many factors that influence the effects of drugs, it is estimated that genetic factors can account for 20 to 95% of variability in drug disposition and effects.32 For example, previous studies revealed that CYP2C9 and VKORC1 genetic variants are associated with warfarin dose requirement and clinical outcomes.33,34 Besides pharmacogenetics, several biomarkers are also contributing to predicting drug responses in patients with stroke, particularly when thrombolysis is administered. Specifically, elevated S-100β and MMP-9 which were reported as serum markers of blood-brain barrier (BBB) dysfunction before thrombolysis could predict hemorrhagic transformation after thrombolysis,35-37 whereas baseline levels of α2-antiplasmin were predictive of recanalization in patients treated with rt-PA.38

tered. Specifically, elevated S-100β and MMP-9 which were reported as serum markers of blood-brain barrier (BBB) dysfunction before thrombolysis could predict hemorrhagic transformation after thrombolysis,35-37 whereas baseline levels of α2-antiplasmin were predictive of recanalization in patients treated with rt-PA.38 There has been mounting evidence that a number of biomarkers can predict clinical or radiological outcomes from cerebral ischemic events. Inflammatory markers such as C-reactive protein (CRP) or proinflammatory cytokines are reportedly associated with early neurological worsening or poor functional outcome after stroke.39,40 Biomarkers related to coagulation/fibrinolysis system such as D-dimer or vWF may also have links with outcome prediction, especially in patients with cardioembolic stroke.41-44 Very recently, it was reported that genetic polymorphisms of brain-derived neurotrophic factor (BDNF) was associated with functional outcome after subarachnoid hemorrhage, and cortical plasticity.45,46

as D-dimer or vWF may also have links with outcome prediction, especially in patients with cardioembolic stroke.41-44 Very recently, it was reported that genetic polymorphisms of brain-derived neurotrophic factor (BDNF) was associated with functional outcome after subarachnoid hemorrhage, and cortical plasticity.45,46 Surrogate endpoints in clinical trials In cardiovascular diseases, many investigators have used biomarkers that correlate with clinical outcomes as surrogate endpoints, because event-driven clinical trials require much of the cost and time burden.47 In the area of stroke research, several studies has been started to use biomarkers to monitor the efficacy and safety of treatments in phase III clinical trials. However, changes detected in surrogate markers do not always translate into clinical endpoints, and may even be the opposite with clinical outcomes.48 Thus, biomarkers may be useful as a screening tool and secondary outcomes, but not primary outcome in clinical trials at the present time. Type of biomarkers in stroke research Stroke biomarkers include traditional protein biomarkers and novel genetic, microvesicle, and metabolomics-associated biomarkers (Table 2).

Surrogate endpoints in clinical trials In cardiovascular diseases, many investigators have used biomarkers that correlate with clinical outcomes as surrogate endpoints, because event-driven clinical trials require much of the cost and time burden.47 In the area of stroke research, several studies has been started to use biomarkers to monitor the efficacy and safety of treatments in phase III clinical trials. However, changes detected in surrogate markers do not always translate into clinical endpoints, and may even be the opposite with clinical outcomes.48 Thus, biomarkers may be useful as a screening tool and secondary outcomes, but not primary outcome in clinical trials at the present time. Type of biomarkers in stroke research Stroke biomarkers include traditional protein biomarkers and novel genetic, microvesicle, and metabolomics-associated biomarkers (Table 2). Protein biomarkers Research using protein biomarkers in patients with ischemic cerebrovascular disease have mainly focused on pathophysiology, diagnosis, prognostication, and neuronal death in stroke.49 A typical example of protein biomarkers is CRP.17,50,51 However, a recent study raised the possibility that the relation may result from various biases.52 Moreover, it has raised skepticism about the efficacy of biomarkers in predicting stroke risk because they provide only limited additional information compared to the well-known stroke risk factors.53,54 Further studies with more a systematic approach and analysis are needed in this area.

result from various biases.52 Moreover, it has raised skepticism about the efficacy of biomarkers in predicting stroke risk because they provide only limited additional information compared to the well-known stroke risk factors.53,54 Further studies with more a systematic approach and analysis are needed in this area. Genetic biomarkers Many epidemiological studies suggested that stroke has genetic susceptibility, and various genetic factors were investigated.55 Among them, establishing the association of the 9p21.3 locus with risk of cerebral infarction is one of the biggest advances.56 It is estimated that the genetic influence is caused by enhanced platelet reactivity.56 However, many other genome-wide association studies failed to reproduce the positive results obtained from previous studies57 or have shown that the clinical usefulness was very low.58 For example, the hazard ratio and population attributable risk of hypertension to ischemic stroke is 2.0 and 26%, respectively. Conversely, the genetic influence on stroke was only 1.3-1.33 and 11-12%, respectively.58 Recently, studies in the pharmacogenomic area have been actively carried out. Among them, aspirin, clopidogrel, warfarin, statin, and thrombolytics-related genetic polymorphisms are particularly of interest. It is expected that selecting the type or dose of medication, avoiding side effects, or drug resistance may be guided by simple genetic tests in the near future.

e been actively carried out. Among them, aspirin, clopidogrel, warfarin, statin, and thrombolytics-related genetic polymorphisms are particularly of interest. It is expected that selecting the type or dose of medication, avoiding side effects, or drug resistance may be guided by simple genetic tests in the near future. Microvesicle Microvesicles are defined as a heterogeneous population of small vesicles with a diameter of 0.1-1 µm. It was previously believed that microvesicles were artifacts generated by apoptotic cell death. However, this view has changed because the shedding of these small vesicles was recognized to result from an active process.59 Microvesicles may be a window for target cell/organs, and include genetic information as well as protein inside them.60 Moreover, it has been identified that microvesicles have their own function, revealing that microvesicles from ischemic tissue facilitated vasculogenesis in the ischemic limb model.61 In this regard, biomarker research using microvesicles is a prominent field. Nevertheless, biomarker studies using microvesicles in stroke are mostly performed in small cohorts. As the methods for analyzing microvesicles are complicated and not unique mainly due to their very small size, investigations with microvesicles are currently at a rudimentary state of development. When seeing the results from a large-scale clinical study perspective for the prediction of the future risk of myocardial infarction, microvesicles could be a good candidate to compensate for the limitations of existing biomarker researches.62

ions with microvesicles are currently at a rudimentary state of development. When seeing the results from a large-scale clinical study perspective for the prediction of the future risk of myocardial infarction, microvesicles could be a good candidate to compensate for the limitations of existing biomarker researches.62 Metabolomics The assumption of metabolomics is that occurrence of the disease is directly related to the specific change of biochemical composition in the cell or biological fluid. Metabolomics-associated biomarker research analyzes profiles of fatty acids, amino acids, or polyamine in the blood or urine and determines normal or pathologic states. Furthermore, metabolomics-associated biomarkers can be applied to the monitoring recovery after treatment. Unfortunately, studies using metabolomics in the area of stroke is relatively lacking. Limitations of stroke biomarkers Currently, the application of biomarkers in the management of cerebral infarction has some limitations, despite their evolving role. First, unlike myocardial infarction, changes in the brain are not sufficiently reflected by blood biomarkers due to the presence of the blood-brain barrier (low sensitivity and underpowered).

Limitations of stroke biomarkers Currently, the application of biomarkers in the management of cerebral infarction has some limitations, despite their evolving role. First, unlike myocardial infarction, changes in the brain are not sufficiently reflected by blood biomarkers due to the presence of the blood-brain barrier (low sensitivity and underpowered). Second, biomarkers can change by a variety of comorbid conditions or brain damage itself (confounders and lack of specificities). As asymptomatic coronary atherosclerosis is frequently accompanied in patients with ischemic stroke,63 it may confound the levels of biomarkers. Indeed, our group recently found that the burden of asymptomatic coronary atherosclerosis was the most important factor for levels of C-reactive protein and homocysteine, regardless of vascular risk factors and the degree and distribution of cervicocephalic atherosclerosis (Figure 1).64 Furthermore, the direct role of biomarkers in the disease process may be hard to reveal. For example, matrix metalloproteinase-9 (MMP-9) which is known as a marker of hemorrhagic transformation after thrombolysis65 is significantly associated with the size of the cerebral infarction irrespective of hemorrhagic transformation.29 Therefore, it is difficult to establish the causal relationship between biomarkers and ischemic stroke in a real clinical setting.

ch is known as a marker of hemorrhagic transformation after thrombolysis65 is significantly associated with the size of the cerebral infarction irrespective of hemorrhagic transformation.29 Therefore, it is difficult to establish the causal relationship between biomarkers and ischemic stroke in a real clinical setting. Third, there is no sufficiently robust single marker for stroke. As shown in Table 2, ischemic stroke is a complex process that includes various etiologies. In addition, the brain consists of many different cells with vessels having distinct anatomical characteristics. Future direction and summary: need of comprehensive approach Each biomarker has different aspects, and its own advantages and drawbacks (Table 3). Therefore, developing an integrated panel of biomarkers for specific stroke subtypes is needed. A recent study reported that multiple microparticle biomarkers in addition to existing protein biomarkers are valuable for predicting future cardio- and cerebrovascular events.62 Hence, a comprehensive approach to using a variety of biomarkers is warranted to overcome the limitations. In addition, multidisciplinary approaches including neuroimaging biomarkers are needed.

iomarkers in addition to existing protein biomarkers are valuable for predicting future cardio- and cerebrovascular events.62 Hence, a comprehensive approach to using a variety of biomarkers is warranted to overcome the limitations. In addition, multidisciplinary approaches including neuroimaging biomarkers are needed. A number of biomarkers are under investigation in patients with ischemic stroke. Research about biomarkers can be helpful especially in predicting stroke and monitoring therapeutic effects. Currently, however, the application of biomarkers is only recommended for research purposes. Monitoring traditional risk factors or vessel status is more efficacious than measuring biomarkers in clinical practice. Considering the advantages and disadvantages of each biomarker is important for future study, and a comprehensive approach using multiple biomarkers is needed. It is strongly anticipated that the biomarkers can provide us a turning point for investigating pathophysiology and therapeutic mechanisms of ischemic stroke. This study was supported by the Korean Healthcare Technology R&D Project, Ministry of Health & Welfare (A110208) and the Bio & Medical Technology Developmental Program of the National Research Foundation of Korea, Ministry of Education, Science and Technology (2011-0019389). The authors have no financial conflicts of interest.

This study was supported by the Korean Healthcare Technology R&D Project, Ministry of Health & Welfare (A110208) and the Bio & Medical Technology Developmental Program of the National Research Foundation of Korea, Ministry of Education, Science and Technology (2011-0019389). The authors have no financial conflicts of interest. Figure 1 Distribution of CRP according to the (A) severity of cerebral atherosclerosis or (B) asymptomatic burden of coronary atherosclerosis in patients with stroke. CRP values are proportional to the increase of coronary (P=0.010 for trends), but not cerebral atherosclerosis (P=0.131 for trends). Severity of cerebral or coronary atherosclerosis. I: No atherosclerosis or 1 segment with <50% stenosis, II: ≥2 segments with <50% stenosis, III: 1 segment with ≥50% stenosis, IV: ≥2 segments with ≥50% stenosis. Table 1 Emerging roles of stroke biomarkers *Urinary albumin/creatinine ratio, a marker of endothelial dysfunction; †Asymmetrical dimethylarginine, an inhibitor of eNOS, a marker of endothelial dysfunction; ‡Protein S100β, homologue for CK-MB in coronary heart disease; §Neuronal specific enolase, homologue for troponin in coronary heart disease.

Table 1 Emerging roles of stroke biomarkers *Urinary albumin/creatinine ratio, a marker of endothelial dysfunction; †Asymmetrical dimethylarginine, an inhibitor of eNOS, a marker of endothelial dysfunction; ‡Protein S100β, homologue for CK-MB in coronary heart disease; §Neuronal specific enolase, homologue for troponin in coronary heart disease. CRP, C-reactive protein; vWF, von Willebrand factor; BNP, B-type natriuretic peptide; GFAP, Glial fibrillary acidic protein; MBP, Myelin basic protein; MMP, Matrix metalloproteinase; VCAM, Vascular cell adhesion molecule; TNF, Tumor necrosis factor; IL, Interleukin; VEGF, Vascular endothelial growth factor; GABA, Gamma aminobutyric acid; DWI, Diffusion-weighted imaging; BDNF, Brain-derived neurotrophic factor; SAH, Subarachnoid hemorrhage; rTMS, Repetitive transcranial magnetic stimulation; NT-proBNP, N-terminal probrain natriuretic peptide. Table 2 Biomarkers for specific stroke subtypes

CRP, C-reactive protein; vWF, von Willebrand factor; BNP, B-type natriuretic peptide; GFAP, Glial fibrillary acidic protein; MBP, Myelin basic protein; MMP, Matrix metalloproteinase; VCAM, Vascular cell adhesion molecule; TNF, Tumor necrosis factor; IL, Interleukin; VEGF, Vascular endothelial growth factor; GABA, Gamma aminobutyric acid; DWI, Diffusion-weighted imaging; BDNF, Brain-derived neurotrophic factor; SAH, Subarachnoid hemorrhage; rTMS, Repetitive transcranial magnetic stimulation; NT-proBNP, N-terminal probrain natriuretic peptide. Table 2 Biomarkers for specific stroke subtypes TAT, Thrombin-Antithrombin complex; PIC, Plasmin inhibitor complex; sP-selectin, Soluble P-selectin; ABCA1, ATP-binding cassette, subfamily A, member 1; IRAK1, Interleukin-1 receptor.associated kinase-1; ROS1, v-Ros avian UR2 sarcoma virus oncogene homolog-1; SOX6, sex-determining region Y-box 6; SNP, Single nucleotide polymorphism; MI, Myocardial infarction; CAD, Coronary artery disease; CRP, C-reactive protein; Lp-PLA2, Lipoprotein-associated phospholipase A2; PAI-1, Plasminogen activator inhibitor-1; VEGF, Vascular endothelial growth factor; ADMA, Asymmetrical dimethylarginine; FABP2, Fatty acid-binding protein 2; BNP, B-type natriuretic peptide; AT-III, Antithrombin-III; vWF, von Willebrand factor; sCD40L, Soluble CD40 ligand; tPA, Tissue plasminogen activator; ICAM, Intercellular adhesion molecule; PKC, Protein kinase C; PRKCH, Protein kinase C eta; CADASIL, Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; COL4A1, Collagen, type IV, alpha 1; F1.2, Prothrombin fragment 1.2; LIMK1, LIM domain kinase 1; CYP3A4, Cytochrome P450, subfamily IIIA, polypeptide 4.

esion molecule; PKC, Protein kinase C; PRKCH, Protein kinase C eta; CADASIL, Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; COL4A1, Collagen, type IV, alpha 1; F1.2, Prothrombin fragment 1.2; LIMK1, LIM domain kinase 1; CYP3A4, Cytochrome P450, subfamily IIIA, polypeptide 4. Table 3 Strengths and weaknesses of various biomarkers

Introduction Health-related quality of life (HRQoL) is used to quantify patients' burden of disease.1 This measure focuses on the patient rather than the disease and is helpful to compare health status between different types of disease. HRQoL also allows direct comparisons of benefits from a variety of interventions on a variety of diseases.1,2 Recently, the US Food and Drug Administration announced to the pharmaceutical industry that HRQoL could be a possible endpoint in clinical trials.3 Therefore, HRQoL has begun to be used as an outcome measure in clinical trials on various diseases, including stroke and dementia.4,5 HRQoL was worse in people with Alzheimer's disease (AD) compared to those with mild cognitive impairment or normal cognition, but there was no difference between mild cognitive impairment and normal cognition.6 HRQoL of stroke survivors was lower than that of healthy elderly people living in the community.7 However, HRQoL has not been evaluated and reported in stroke victims with cognitive impairment but no dementia. The utility weight of EQ-5D (EQ-5Dindex) in a representative sample of the Korean population was recently reported as a feasible and valid tool for measuring HRQoL in stroke survivors.8 The objectives of the present study were to compare patients with post-stroke cognitive impairment with no dementia (PSCIND) to those with post-stroke dementia (PSD) and healthy people with normal cognition with respect to HRQoL, measured by EQ-5Dindex, and to investigate the individual effect of each cognitive domain on EQ-5Dindex.

of the present study were to compare patients with post-stroke cognitive impairment with no dementia (PSCIND) to those with post-stroke dementia (PSD) and healthy people with normal cognition with respect to HRQoL, measured by EQ-5Dindex, and to investigate the individual effect of each cognitive domain on EQ-5Dindex. Methods A series of 1,528 patients who visited the stroke prevention clinic at Seoul National University Bundang Hospital, underwent brain magnetic resonance imaging (MRI), and underwent the 60-min neuropsychological assessment protocol of the Korean version of Vascular Cognitive Impairment Harmonization Standards (K-VCIHS-NP) and EQ-5D between May 2007 and March 2009, were identified using the clinical data warehousing system embedded in our institution's electronic medical records system. This study protocol was reviewed and approved by your Institutional Review Board, and that informed consent was waived.

monization Standards (K-VCIHS-NP) and EQ-5D between May 2007 and March 2009, were identified using the clinical data warehousing system embedded in our institution's electronic medical records system. This study protocol was reviewed and approved by your Institutional Review Board, and that informed consent was waived. Of those 1,528, the PSD and PSCIND groups were selected from those who 1) had been hospitalized at our institution for acute stroke management and 2) were impaired in at least one of the frontal executive, language, visuospatial, or memory domains on the K-VCIHS-NP, and 3) in whom the K-VCIHS-NP might have been administered at least 90 days after stroke onset. Among them, those who did not meet the criteria of dementia were classified as the PSCIND group and those who met the criteria formed the PSD group. Dementia was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition9 criteria based on clinical interviews, cognitive functions assessed using the K-VCIHS-NP, and social functioning status assessed on the scale of Instrumental Activities of Daily Living. Patients who had depression were excluded. Stroke (ischemic or hemorrhagic) was defined using the World Health Organization's definition of "rapidly developed clinical signs of focal or global disturbance of cerebral function, lasting more than 24 hours or leading to death, with no apparent cause other than a vascular origin."

who had depression were excluded. Stroke (ischemic or hemorrhagic) was defined using the World Health Organization's definition of "rapidly developed clinical signs of focal or global disturbance of cerebral function, lasting more than 24 hours or leading to death, with no apparent cause other than a vascular origin." The control group was chosen from individuals who visited the stroke prevention clinic worrying about a risk of stroke, but who had no history of stroke and normal cognition on the 60-minutes protocol. Those who had moderate to severe white matter hyperintensities on the fluid-attenuated inversion recovery sequences of brain MRI, psychiatric illness, or other neurological disorders were excluded. We also excluded those who were not evaluated with EQ-5D. A flow chart for the identification of study subjects is displayed in Figure 1. We selected cases and controls in the age band between 50 years (the lowest in PSD group) and 82 (the highest in control group). Each PSCIND patient (N=50) was matched to one PSD patient and one control according to age (±3 years) and sex, blinded to other clinical information.

The control group was chosen from individuals who visited the stroke prevention clinic worrying about a risk of stroke, but who had no history of stroke and normal cognition on the 60-minutes protocol. Those who had moderate to severe white matter hyperintensities on the fluid-attenuated inversion recovery sequences of brain MRI, psychiatric illness, or other neurological disorders were excluded. We also excluded those who were not evaluated with EQ-5D. A flow chart for the identification of study subjects is displayed in Figure 1. We selected cases and controls in the age band between 50 years (the lowest in PSD group) and 82 (the highest in control group). Each PSCIND patient (N=50) was matched to one PSD patient and one control according to age (±3 years) and sex, blinded to other clinical information. Instruments and data collection The K-VCIHS-NP was adapted from the 60-minutes neuropsychology protocol of the Vascular Cognitive Impairment Harmonization Standards (VCIHS) proposed by the National Institute of Neurological Disorders and Stroke and the Canadian Stroke Network.10 It comprises 4 cognitive domains and 8 cognitive tests: frontal executive function (animal naming test, phonemic fluency test, Digit Symbol Coding, Korean Trail-Making Test A and B); language/lexical retrieval (Korean-Boston Naming Test-short form); visuospatial (Rey Complex Figure Test: Copy); and memory (Seoul Verbal Learning Test). In addition, Korean Mini-Mental Status Examination for evaluating global cognitive dysfunctions and the Informant Questionnaire of Cognitive Decline in the Elderly for assessing the patients' pre-morbid history of cognitive dysfunctions, and Geriatric Depression Scale were also included. All tests and scales were validated and standardized in Korean subjects.11

ination for evaluating global cognitive dysfunctions and the Informant Questionnaire of Cognitive Decline in the Elderly for assessing the patients' pre-morbid history of cognitive dysfunctions, and Geriatric Depression Scale were also included. All tests and scales were validated and standardized in Korean subjects.11 Trained clinical psychometricians, blinded to the clinical and neuroradiological profiles of each patient, administered the series of tests. A score on each cognitive test was transformed into a standardized z-score (z-score=[individual score-mean score of the sample]/standard deviation of the sample). Cognitive impairments in language, visuospatial function, or memory domains were defined as a score of less than the 7th percentile (≈mean-1.5 standard deviation [SD]) in each domain-specific test. Cognitive impairment in the frontal executive domain was defined as a score of less than the 7th percentile in 2 or more of the 5 frontal domain-specific tests. Instrumental Activities of Daily Living score ≥0.43 was considered a significant impairment in everyday functioning,12 and Geriatric Depression Scale score of 18-30 as being depressive.13,14

ntal executive domain was defined as a score of less than the 7th percentile in 2 or more of the 5 frontal domain-specific tests. Instrumental Activities of Daily Living score ≥0.43 was considered a significant impairment in everyday functioning,12 and Geriatric Depression Scale score of 18-30 as being depressive.13,14 HRQoL was assessed using the EQ-5D. Subjects were asked to rate their current health state across the five dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression). Subjects were asked to choose one of three responses in each dimension. Trained psychometricians administered the EQ-5D questionnaire after the subjects finished the neuropsychological test series. The EQ-5D allows calculation of a preference-based summary index based on time trade-off techniques in which the value "0" represents death and "1" perfect health. We chose the utility measurement EQ-5Dindex score to identify the difference in HRQoL among the three groups because it is designed to compare HRQoL across different diseases as well as different populations. A utility was derived from a representative sample of the Korean population, and the use of the Korean version of EQ-5Dindex had been validated for the Korean general population and for stroke survivors.8,15

three groups because it is designed to compare HRQoL across different diseases as well as different populations. A utility was derived from a representative sample of the Korean population, and the use of the Korean version of EQ-5Dindex had been validated for the Korean general population and for stroke survivors.8,15 Baseline demographics and clinical characteristics of the study subjects were obtained from a prospective registry database of hospitalized stroke patients16 or by reviewing electronic medical records. Information on the following variables was collected: age, sex, years of education (categorized into 0-3, 4-9, 10-12, and ≥13 years of formal education), blood pressure, history of stroke, hypertension (previously diagnosed and treated, or systolic pressure >140 mmHg and/or diastolic pressure >90 mmHg), diabetes mellitus (previously diagnosed and treated, or fasting glucose 7 mmol/L, or postprandial 2 h blood glucose >11.1 mmol/L), hyperlipidemia (previously diagnosed and treated, or total cholesterol >6.0 mmol/L, or low density lipoprotein cholesterol >4.14 mmol/L), history of coronary artery disease and atrial fibrillation according to clinical diagnosis and electrocardiography, smoking (current smoker: smoked at least 1 cigarette during past 1 month), or alcohol intake (heavy alcoholics: >3 drinks on average/day).

6.0 mmol/L, or low density lipoprotein cholesterol >4.14 mmol/L), history of coronary artery disease and atrial fibrillation according to clinical diagnosis and electrocardiography, smoking (current smoker: smoked at least 1 cigarette during past 1 month), or alcohol intake (heavy alcoholics: >3 drinks on average/day). Sample size calculation and statistical analysis We adopted the generalized estimating equation (GEE) to assess the relationship between matched cases and controls. Because a statistical method to calculate sample size in the GEE model has not been developed yet, the sample size of this study was estimated by analysis of variance with a significance level of 0.05 and a power of 0.90. The mean±SD of EQ-5Dindex in the PSD group came from studies on AD and vascular or other kinds of dementias,17,18 and was assumed to be 0.64±0.21. Those of the control group were assumed to be 0.84±0.18, representative of the general population over 60 years in Korea and based on the Korean National Health and Nutrition Examination Survey between 2007 and 2009. The mean of EQ-5Dindex in the PSCIND group was estimated as that (≒0.77) of the upper two-thirds between the PSD and control groups. The calculated sample size was 40 for SD=0.21. The final sample size was determined to be 50 in each group, considering missing data that originated from the study's retrospective nature.

The mean of EQ-5Dindex in the PSCIND group was estimated as that (≒0.77) of the upper two-thirds between the PSD and control groups. The calculated sample size was 40 for SD=0.21. The final sample size was determined to be 50 in each group, considering missing data that originated from the study's retrospective nature. Differences in baseline characteristics among the PSD, PSCIND, and control groups were examined using Pearson's chi-squared test or one-way analysis of variance as appropriate (Table 1). Results of neuropsychological tests are presented as mean±SD in z-scores and percentage of subjects with impairment in each domain (Table 2). Comparisons of z-scores between groups were performed using GEE adjusting for education. A partial correlation analysis adjusting for modified Rankin scale (mRS) measured at 3 months of stroke onset was performed to examine the relative contribution of each cognitive domain on the EQ-5Dindex with adjustment for physical disability. Finally, GEE adjusting for education was used to determine the significance of difference in EQ-5Dindex score among three groups.

Rankin scale (mRS) measured at 3 months of stroke onset was performed to examine the relative contribution of each cognitive domain on the EQ-5Dindex with adjustment for physical disability. Finally, GEE adjusting for education was used to determine the significance of difference in EQ-5Dindex score among three groups. Results Of 1,528 patients who visited the stroke prevention clinic and underwent brain MRI and the K-VCIHS-NP during the study period, 586 had a history of stroke and 942 had not (Figure 1). Among those 586 with stroke, we identified 70 PSD and 126 PSCIND patients, after excluding those who had depression or in whom the EQ-5D was not available. Among 942 without stroke, 222 were suitable for the control group. Lastly, 150 subjects, 50 for each group matched for age (±3 years) and sex, were enrolled in the study. The mean age was 69 years and 72.0% of subjects were male. The clinical characteristics of the study subjects are presented in Table 1. The level of education was significantly different among the three groups. The PSD and PSCIND groups had higher systolic blood pressure, were more likely to be current smokers, and were less likely to have hyperlipidemia compared to the control group. Three-month mRS score was different between the PSD and PSCIND groups and significantly correlated with the EQ-5Dindex at 3 months of stroke onset in the PSD group (-0.38, P=0.007), but not in the PSCIND group (-0.06, P=0.70).

e current smokers, and were less likely to have hyperlipidemia compared to the control group. Three-month mRS score was different between the PSD and PSCIND groups and significantly correlated with the EQ-5Dindex at 3 months of stroke onset in the PSD group (-0.38, P=0.007), but not in the PSCIND group (-0.06, P=0.70). Comparisons between the 3 groups were made with respect to the z-scores of the individual neuropsychological tests, adjusted for education (Table 2). The z-scores of all the cognitive tests of the PSD group were significantly lower than those of the PSCIND and control groups, and those of the PSCIND group were lower than those of the control group. The most frequently impaired cognitive domain was memory, and the most frequently impaired frontal executive function test was the Korean Trail Making Test B (K-TMT-B), in both the PSCIND and PSD groups. With respect to Instrumental Activities of Daily Living scores, the PSCIND group was not different from the control group, while the PSD group had significantly lower scores than the other groups. The frequency of pre-stroke cognitive decline evaluated using Korean-IQCODE (IQCODE ≥3.6) was 35% (137 of 389) among stroke survivors.

t to Instrumental Activities of Daily Living scores, the PSCIND group was not different from the control group, while the PSD group had significantly lower scores than the other groups. The frequency of pre-stroke cognitive decline evaluated using Korean-IQCODE (IQCODE ≥3.6) was 35% (137 of 389) among stroke survivors. Correlations between EQ-5Dindex and neuropsychological tests in the three groups are presented in Table 3, with and without adjustment for 3-month mRS in stroke survivors. In the PSCIND group, EQ-5Dindex correlated only with visuospatial function before adjustment, but with none after adjustment. In the PSD group, EQ-5Dindex correlated with Korean Mini-Mental Status Examination and visuospatial function with and without adjustment, and with Digit Symbol Coding-correct and language, with adjustment. There was a significant difference in the mean EQ-5Dindex among the three groups, and pairwise comparisons between the PSCIND and control groups and between the PSCIND and PSD groups showed similar results even after adjustment for education and depression (Figure 2, Table 4). Discussion This study finds that the HRQoL of PSCIND is intermediate between that of PSD and normal cognition and that no particular cognitive domain is more influential on the HRQoL of PSCIND. To the best of our knowledge, this study is the first to directly compare PSD, PSCIND, and normal cognition with respect to HRQoL.

ussion This study finds that the HRQoL of PSCIND is intermediate between that of PSD and normal cognition and that no particular cognitive domain is more influential on the HRQoL of PSCIND. To the best of our knowledge, this study is the first to directly compare PSD, PSCIND, and normal cognition with respect to HRQoL. The frequency of cognitive impairment in post-stroke (66.4%) was higher than that among previous studies.19-21 A recent report from a multicenter prospective stroke cohort in Korea appeared to be similar with our study in frequency of cognitive impairment after stroke (62.6%) even though it was a bit of difference in study setting such as cutoff value of cognitive impairment (<10 percentile).11 The higher frequency of cognitive impairment in our study might be partially explained by discrepancy with pre-stroke cognitive dysfunction (7.2%) in recent study.11

e impairment after stroke (62.6%) even though it was a bit of difference in study setting such as cutoff value of cognitive impairment (<10 percentile).11 The higher frequency of cognitive impairment in our study might be partially explained by discrepancy with pre-stroke cognitive dysfunction (7.2%) in recent study.11 A previous study reported that the mean EQ-5Dindex was 0.60 in all types of dementia, which was similar to that in the PSD group of the present study, 0.61.17 In another study on AD, the mean EQ-5Dindex score was 0.62 when rated by proxy, but 0.86 when self-rated.22 The HRQoL may be lower in those with PSD than in those with AD; this needs to be confirmed in another study. For determinants of HRQoL in dementia, there is a clear and consistent association with depression (the more severe the depression, the lower the HRQoL), but there have been inconsistent reports on other factors, such as age, gender, education, ethnicity, and activity limitation.23,24 Subjects who had a history of depression (previously diagnosed and treated) was excluded, despite our findings might be partially biased by the inclusion of healthier controls. With high frequency of depression in PSD evaluated by a geriatric depression scale, depression was included as confounder in final model, but the impact of PSCIND on HRQoL was not different from the unadjusted result in statistical significance. Low HRQoL in PSD could be explained by medical comorbidities such as diabetes or coronary heart disease, which are known to be related to HRQoL25,26 and more prevalent in PSD than in AD.

onfounder in final model, but the impact of PSCIND on HRQoL was not different from the unadjusted result in statistical significance. Low HRQoL in PSD could be explained by medical comorbidities such as diabetes or coronary heart disease, which are known to be related to HRQoL25,26 and more prevalent in PSD than in AD. The mean EQ-5Dindex of PSCIND (0.86) was analogous to that of chronic obstructive pulmonary disease (≒0.81) in Korea.27 The reported mean EQ-5Dindex of minor stroke (mRS≤3) was between 0.6 and 0.7, which was similar to that of PSD and lower than that of PSCIND.28 A few previous studies on stroke survivors have shown no influence of vascular cognitive impairment on HRQoL.7,29 From a viewpoint of HRQoL after stroke, the contribution of mRS to HRQoL should be emphasized.7 We found that 3-month mRS was associated with 3-month EQ-5Dindex in PSD, but the difference in EQ-5Dindex between PSD and PSCIND was still significant after adjusting for 3-month mRS. A positive association between cognition and HRQoL has been reported among functionally independent stroke survivors.30 We cautiously suggest that cognitive impairment may further impact the HRQoL in stroke victims.

e difference in EQ-5Dindex between PSD and PSCIND was still significant after adjusting for 3-month mRS. A positive association between cognition and HRQoL has been reported among functionally independent stroke survivors.30 We cautiously suggest that cognitive impairment may further impact the HRQoL in stroke victims. EQ-5Dindex of PSCIND was significantly lower than that of normal cognition in this study. This result is different from previous ones in which there was no significant difference between patients with mild cognitive impairment and controls.4,31 Between the PSCIND and control group, a difference existed not only in cognition but also in stroke history, which had an impact on HRQoL in Koreans,32 and both of them might lead to a difference in EQ-5Dindex. A further study comparing those with PSCIND and stroke survivors with normal cognition can address the question of whether cognitive impairment that is not enough to be diagnosed as dementia, truly impacts HRQoL in stroke survivors.

mpact on HRQoL in Koreans,32 and both of them might lead to a difference in EQ-5Dindex. A further study comparing those with PSCIND and stroke survivors with normal cognition can address the question of whether cognitive impairment that is not enough to be diagnosed as dementia, truly impacts HRQoL in stroke survivors. EQ-5Dindex in the PSCIND group did not correlate with impairment in any specific cognitive domain. A relationship between poor TMT-B performance and poor perceived quality of life has been reported in stroke patients.33 This study showed that impairment in the K-TMT-B, one of the executive function tests, was observed most frequently in both PCIND and PSD groups, but the K-TMT-B scores did not correlate with EQ-5Dindex. However, it should be noted that 48% in the PSD group and 20% in the PSCIND group were unable to complete the K-TMT-B; this was the worst outcome among all cognitive tests. This study has some limitations. First, the eligible stroke patients were not representative of the general stroke population in Korea, or even hospitalized stroke patients in our institution. Inevitably, patients who were unable to complete neuropsychological tests, did not undergo MRI, and were not followed up at the outpatient clinic after discharge were excluded from this study. Second, as mentioned above, this study did not include stroke survivors with normal cognition. A further study including all cognitive spectrums of stroke survivors is therefore warranted.

ogical tests, did not undergo MRI, and were not followed up at the outpatient clinic after discharge were excluded from this study. Second, as mentioned above, this study did not include stroke survivors with normal cognition. A further study including all cognitive spectrums of stroke survivors is therefore warranted. Summary This study shows that among stroke patients surviving more than 3 months, even mild cognitive impairment may interfere with a sense of well-being. We cautiously suggest that cognitive impairment may be an important determinant lowering HRQoL. Interventions aimed at enhancing cognition may improve HRQoL even in PSCIND. This is inferred from a randomized controlled trial in dementia, which proved that improvement in cognition is associated with improvement in HRQoL.5 Acknowledgements H.-J.B is involved as the principal investigator, a member of the steering committee, and/or a site investigator of multicenter clinical trials or clinical studies sponsored by Otsuka Korea, Bayer Korea, Handok Pahrmaceutical Company, SK Chemicals, ESAI-Korea, Daewoong Pharmaceutical Co. Ltd, Daichi Sankyo, Pfizer, Sanofi-Aventis Korea, and Yuhan Corporation; has served as the consultant or on the scientific advisory board for Bayer Korea, Boehringer Ingelheim Korea, YuYu Pharmaceutical Company, and BMS Korea; and has received lecture honoraria from MSD Korea, AstraZeneca Korea, BMS Korea, Novartis Korea, Ostuka Korea, Pfizer Korea, Daichi Sankyo Korea, and Handok Pharmaceutical Company (modest).

ltant or on the scientific advisory board for Bayer Korea, Boehringer Ingelheim Korea, YuYu Pharmaceutical Company, and BMS Korea; and has received lecture honoraria from MSD Korea, AstraZeneca Korea, BMS Korea, Novartis Korea, Ostuka Korea, Pfizer Korea, Daichi Sankyo Korea, and Handok Pharmaceutical Company (modest). This study was partly supported by a grant from the Korea Health 21 R&D project, Ministry of Health and Welfare, Korea (A102065) and by a grant from Esai Korea Inc. The authors have no financial conflicts of interest. Figure 1 Flow chart for identification of study subjects. PSCIND, post-stroke cognitive impairment with no dementia. PSD, post-stroke dementia; SVD, small vessel disease. Figure 2 Distribution of subjects according to the EQ-5Dindex values and mean±SD of EQ-5Dindex in the PSD, PSCIND, and control groups. Table 1 Characteristics of the control, PSCIND, and PSD groups Values are number of patients (%) if not indicated. PSCIND indicates post-stroke cognitive impairment with no dementia. PSD, post-stroke dementia; SD, standard deviation. *P values were calculated by analysis of variance or Pearson's chi-squared test. Table 2 Mean (SD) of z-scores, frequency of impairment (%), and pairwise comparison of cognitive functions PSCIND indicates post-stroke cognitive impairment with no dementia.

Values are number of patients (%) if not indicated. PSCIND indicates post-stroke cognitive impairment with no dementia. PSD, post-stroke dementia; SD, standard deviation. *P values were calculated by analysis of variance or Pearson's chi-squared test. Table 2 Mean (SD) of z-scores, frequency of impairment (%), and pairwise comparison of cognitive functions PSCIND indicates post-stroke cognitive impairment with no dementia. PSD, post-stroke dementia; SD, standard deviation; K-MMSE, Korean-Mini Mental State Examination; K-TMT, Korean Trail-Making Test; COWAT, Controlled Oral Word Association Test; DSC, Digit Symbol Coding; K-BNT, Short version of the Korean Boston Naming Test; SVLT, Seoul Verbal Learning Test; RCFT, Rey Complex Figure Test; IADL, Instrumental Activity of Daily Living. *P values were calculated by Generalized Estimating Equation (GEE) adjusting for education. All the mean difference is significant at the 0.05 level; †The values for the cognitive-domain measures are standardized composite z-scores; ‡The value was calculated by analysis of variance. Table 3 Pearson correlation coefficients and partial correlation coefficients adjusted for 3-month mRS between EQ-5Dindex and neuropsychological tests K-MMSE, Korean-Mini Mental State Examination; K-TMT, Korean Trail-Making Test; COWAT, Controlled Oral Word Association Test; DSC, Digit Symbol Coding; K-BNT, Short version of the Korean Boston Naming Test; SVLT, Seoul Verbal Learning Test; RCFT, Rey Complex Figure Test. *Correlation is significant at the 0.05 level (2-tailed); †Correlation is significant at the 0.01 level (2-tailed).

K-MMSE, Korean-Mini Mental State Examination; K-TMT, Korean Trail-Making Test; COWAT, Controlled Oral Word Association Test; DSC, Digit Symbol Coding; K-BNT, Short version of the Korean Boston Naming Test; SVLT, Seoul Verbal Learning Test; RCFT, Rey Complex Figure Test. *Correlation is significant at the 0.05 level (2-tailed); †Correlation is significant at the 0.01 level (2-tailed). Table 4 Comparisons of EQ-5Dindex among the Control, PSCIND, and PSD groups PSCIND indicates post-stroke cognitive impairment with no dementia. PSD, post-stroke dementia; SD: standard deviation. *P value is for Generalized Estimating Equation (GEE); †Depression is continuous variable.

Introduction Stroke is the leading cause of serious long-term disability and mortality in Korea.1 Organized treatment in stroke centers for patients with an acute stroke may reduce the mortality from this disorder.2,3 The Brain Attack Coalition categorized the types of stroke centers into primary and comprehensive,4-6 and outlined the recommendations for comprehensive stroke centers (CSC).6 Comprehensive stroke centers are designed to have the "necessary personnel, infrastructure, expertise, and programs to diagnose and treat stroke patients who require a high intensity of medical and surgical care, specialized tests, or interventional therapies."6 Within the first 30 days after an ischemic stroke, the case fatality has been variously reported as 3.5-25% and the major causes of death were the index stroke and its sequelae.7-12 Thirty-day stroke case fatality has been used as a key indicator for hospital performance and quality of care to compare hospitals and to implement quality improvement strategies.8,13-15 Despite widespread establishment of CSCs in Korea, there is limited empirical evidence which demonstrates whether more organized stroke care at a stroke unit reduces stroke mortality.16,17 Therefore, our goal was to compare the 30-day mortality of patients with acute ischemic stroke (AIS) that were treated at our hospital during the 3 years before and the 3 years after the establishment of the CSC.

dence which demonstrates whether more organized stroke care at a stroke unit reduces stroke mortality.16,17 Therefore, our goal was to compare the 30-day mortality of patients with acute ischemic stroke (AIS) that were treated at our hospital during the 3 years before and the 3 years after the establishment of the CSC. Methods Patients selection This study was based on our prospective stroke registry. All patients with AIS and transient ischemic attack admitted to our hospital within 7 days after the onset of symptoms, from January 2006 to present, have been enrolled. Stroke severity was assessed according to the National Institutes of Health Stroke Scale (NIHSS).18 Assessment was performed in the emergency room at the time of admission. After admission, all patients were evaluated using a protocol that included demographic data, medical history, and vascular risk factors. We divided the patients into 5 groups, using the TOAST criteria.19 Their clinical data were registered in the stroke registry by two trained study nurses. We divided these enrolled patients into two groups corresponding to each of the three year intervals, before and after the establishment of the CSC. Patients with incomplete or missing clinical data, and transient ischemic attacks were excluded. This study was approved by the Institutional Review Board at Dong-A University Hospital.

We divided these enrolled patients into two groups corresponding to each of the three year intervals, before and after the establishment of the CSC. Patients with incomplete or missing clinical data, and transient ischemic attacks were excluded. This study was approved by the Institutional Review Board at Dong-A University Hospital. Establishment of the comprehensive stroke center Our hospital is a tertiary teaching hospital located in the Busan metropolitan area. Our stroke center was opened in January 2009 through financial support from the Korean government. Before the establishment of CSC, neither a stroke team nor a stroke care unit was present at our hospital. At the same time as the CSC opened, a 6-bed stroke care unit was installed for care of the acute stroke patients, which had a telemetry system for the monitoring of blood pressure, pulse, respiration and oxygenation. With establishment of the CSC, to shorten the processing time to within the time window of IV t-PA use, we applied a central alerting system to recruit all available stroke team members to the emergency room, as soon as possible, after the arrival of suspected stroke patients. A multidisciplinary group developed stroke care pathways to guide the evaluation and treatment of each stroke subtype. Stroke treatment was generally administered according to a well-organized team approach, e.g., intravenous thrombolysis, endovascular treatment, hemicraniectomy and carotid endarterectomy.

With establishment of the CSC, to shorten the processing time to within the time window of IV t-PA use, we applied a central alerting system to recruit all available stroke team members to the emergency room, as soon as possible, after the arrival of suspected stroke patients. A multidisciplinary group developed stroke care pathways to guide the evaluation and treatment of each stroke subtype. Stroke treatment was generally administered according to a well-organized team approach, e.g., intravenous thrombolysis, endovascular treatment, hemicraniectomy and carotid endarterectomy. After the establishment of the stroke center, several performance measures improved as compared with beforehand.20 Care in our stroke center fulfilled the Brain Attack Coalition's standardized criteria (Table 1).6 Clinical outcome measures We investigated the 30-day all-cause mortality of ischemic stroke patients for 3 years before and after the establishment of the CSC. Thirty-day case fatality was defined as the proportion of strokes for which death occurred within 30 days of stroke admission. Data on patients who were able to visit the outpatient department were captured. In all other cases, the survival or death at 1 month after the stroke was checked by a trained study nurse who undertook a standardized telephone interview with the patient or his or her next of kin. The survival or death was identified, regardless of cause of death.

le to visit the outpatient department were captured. In all other cases, the survival or death at 1 month after the stroke was checked by a trained study nurse who undertook a standardized telephone interview with the patient or his or her next of kin. The survival or death was identified, regardless of cause of death. Statistical analysis Nominal variables were expressed as count and percentages, continuous values as the mean±standard deviation (SD). T-test was used to analyze the difference in continuous variables, and the chi-square test for those in proportions. Non-normally distributed variables were compared using a Mann-Whitney U test. We analyzed the influence of risk factors, clinical variables and the establishment of the CSC on 30-day mortality by a univariate analysis. We reanalyzed 30-day case fatality by means of logistic regression to evaluate whether the establishment of CSC is independently associated with reduced 30-day ischemic stroke mortality. The logistic regression model was applied with adjustment for the baseline variables that showed a P-value less than 0.1 on univariate analysis. In addition, propensity score matched analysis was performed to correct for bias. The results are presented as odds ratio (OR) estimates of relative risk with 95% confidence intervals (CI). P-values<0.05 were considered statistically significant.

ine variables that showed a P-value less than 0.1 on univariate analysis. In addition, propensity score matched analysis was performed to correct for bias. The results are presented as odds ratio (OR) estimates of relative risk with 95% confidence intervals (CI). P-values<0.05 were considered statistically significant. Results During the study period, consecutive patients with AIS or transient ischemic attack (n=3,339) were admitted to our hospital within seven days after the onset of the symptoms to our hospital. Patients with a diagnosis of transient ischemic attack (n= 159) and those without 30-day outcome data (n=63) were excluded. The proportion of the men among patients with 30-day follow-up loss was higher than the proportion of men in the included patients; age: 64.9±10.9 vs. 65.5±12.3, P=0.720, male: 74.6% vs. 62.2%, P=0.037, NIHSS, median [ranges] : 5 [1-29] vs. 5 [1-41], P=0.320. Ultimately, we performed a retrospective study using a sample of 3,117 consecutive patients with AIS.

with 30-day follow-up loss was higher than the proportion of men in the included patients; age: 64.9±10.9 vs. 65.5±12.3, P=0.720, male: 74.6% vs. 62.2%, P=0.037, NIHSS, median [ranges] : 5 [1-29] vs. 5 [1-41], P=0.320. Ultimately, we performed a retrospective study using a sample of 3,117 consecutive patients with AIS. After a 6-bed stroke care unit was installed in the CSC, 86% of AIS patients were admitted to the stroke care unit. The mean duration of hospitalization in stroke care unit was 3.5 days. Decompressive hemicraniectomies were performed in 14 patients for 3 years before the establishment of the CSC. After the establishment of the CSC, among 119 patients with malignant middle cerebral artery infarction, 53 patients (44.5%) underwent decompressive hemicraniectomy. Forty patients with hemicraniectomy (75%) survived for up to a month. Carotid endarterectomy was done in 47 patients with severe carotid artery stenosis (Table 1). The baseline characteristics of ischemic stroke patients, before and after the implementation of the CSC, are presented in Table 2. Patients' mean age was 65.5 years, and 62.2% were men.

After a 6-bed stroke care unit was installed in the CSC, 86% of AIS patients were admitted to the stroke care unit. The mean duration of hospitalization in stroke care unit was 3.5 days. Decompressive hemicraniectomies were performed in 14 patients for 3 years before the establishment of the CSC. After the establishment of the CSC, among 119 patients with malignant middle cerebral artery infarction, 53 patients (44.5%) underwent decompressive hemicraniectomy. Forty patients with hemicraniectomy (75%) survived for up to a month. Carotid endarterectomy was done in 47 patients with severe carotid artery stenosis (Table 1). The baseline characteristics of ischemic stroke patients, before and after the implementation of the CSC, are presented in Table 2. Patients' mean age was 65.5 years, and 62.2% were men. Among the 3,117 patients, 61.8% were treated at our hospital during the 3-year period after opening of the CSC. These patients were older, and had lower frequencies of hypertension and current smoking behavior, and had more prior coronary artery disease and higher NIHSS scores on admission than those admitted before the CSC opened. Overall 30-day mortality was 6.8%. The overall 30-day all-cause mortality rate was 8.2% for patients admitted to our hospital before the establishment of CSCs and 5.9% for patients admitted after (unadjusted mortality difference, -2.3%; P=0.012).

gher NIHSS scores on admission than those admitted before the CSC opened. Overall 30-day mortality was 6.8%. The overall 30-day all-cause mortality rate was 8.2% for patients admitted to our hospital before the establishment of CSCs and 5.9% for patients admitted after (unadjusted mortality difference, -2.3%; P=0.012). The characteristics of those patients who died in the first 30 days, compared with those who were alive, are shown in Table 3. On the unadjusted analysis, advanced age, female gender, previous coronary artery disease, no current smoking, stroke subtype, admission on a holiday, referral from other hospitals, higher NIHSS score on admission, and admission before the establishment of CSC were associated with higher 30-day stroke case fatality. Logistic regression analysis was performed to further evaluate the independent predictor for 30-day mortality (Table 4). After a risk adjustment, the establishment of the CSC was independently associated with lower 30-day mortality (OR, 0.57; 95% CI, 0.412-0.795, P=0.001). Then, we performed one-to-one matching based on propensity scores of each patient. The propensity score matching procedure generated a matched cohort in which the baseline prognostic variables were balanced. The 30-day case-fatality in patients admitted to our hospital after opening the CSC was lower than prior to opening (OR 0.64, 95% CI 0.458-0.955, P=0.011). (Supplementary Tables 1-2).

ch patient. The propensity score matching procedure generated a matched cohort in which the baseline prognostic variables were balanced. The 30-day case-fatality in patients admitted to our hospital after opening the CSC was lower than prior to opening (OR 0.64, 95% CI 0.458-0.955, P=0.011). (Supplementary Tables 1-2). Although reductions in the odds of death at 30-day by logistic regression analysis was relatively high comparing with that by propensity score matching analysis, both were all statistically significant. Thus, it was presumed that consistent results from two analytical methods were derived. Discussion This study showed that 30-day mortality rate of patients with AIS treated after the establishment of the CSC was lower than prior. This was true also after the adjustment for the differences in the baseline characteristics. These findings highlight the need for more organized comprehensive care for AIS patients to improve the stroke outcome.

30-day mortality rate of patients with AIS treated after the establishment of the CSC was lower than prior. This was true also after the adjustment for the differences in the baseline characteristics. These findings highlight the need for more organized comprehensive care for AIS patients to improve the stroke outcome. Incidence of case fatality within 30 days of ischemic stroke in our hospital (6.8%) was below the range 10% to 25% reported in previous hospital-based studies.7-11 However, the 30-day mortality of our hospital was higher than that of a nationwide Korean study.12 This variation may be due to age, sex, race, quality of inpatient stroke care, risk factors and stroke severity among study subjects. For example, in our study, patients with cardiac embolism, large artery disease and undetermined cause of stroke are 3.49, 2.92, and 4.85 times more likely, respectively, to be dead 30 days after ischemic stroke than patients with a small vessel stroke. This is in accordance with the previous study reporting that patients with cardioembolic stroke had worse 30-day survival than patients with non-cardioembolic stroke.9,21 The proportion of patients with cardioembolic stroke in our study (23.5%) was higher than that that of a nationwide study in Korea (13.2-20.5%). Also, our cohort includes approximately 50% of patients with referral from other hospitals who would be expected to have more severe neurological and medical complication precipitating the transfer. In our study and others,7,8 initial stroke severity was the most important predictor of mortality after stroke. However, in the administrative records of some of the previous studies, stroke severity was not taken into account.7-12 Thus, we could not compare the neurological severity of stroke symptoms and medical cormorbidity with all other studies.

ial stroke severity was the most important predictor of mortality after stroke. However, in the administrative records of some of the previous studies, stroke severity was not taken into account.7-12 Thus, we could not compare the neurological severity of stroke symptoms and medical cormorbidity with all other studies. So far, thrombolysis, hemicraniectomy for malignant infarction, and stroke unit care are evidence-based therapies for ischemic stroke that achieve a substantial and long-lasting effect on neurological outcome.22 A CSC can provide these treatments to patients with severe stroke, and patients with AIS who required specialized treatments, including endovascular treatment, decompressive surgery and carotid surgery.6 We can not determine which individual components of the Brain Attack Coalition criteria for stroke center recommendation were the most important contributors for the lower mortality observed in this study.

who required specialized treatments, including endovascular treatment, decompressive surgery and carotid surgery.6 We can not determine which individual components of the Brain Attack Coalition criteria for stroke center recommendation were the most important contributors for the lower mortality observed in this study. One of the most important components of organized treatment for patients with AIS is admission to a stroke unit with a continuous telemetry monitoring system and stroke-directed nursing.23 Stroke patients who received organized inpatient care in a stroke unit had better outcomes and reduced stroke fatality than those with stroke care in the general ward.23-25 Randomized trials have shown that stroke unit care prevents 1 death for 33 patients treated.23 According to 25 reports which compared stroke unit care with general wards, stroke unit care showed reductions in the odds of death recorded within 1 year of stroke follow-up (OR: 0.79; 95% CI: 0.73 to 0.86; P=0.00001).26 In addition, one Korean hospital study reported stroke unit care could reduce 3-month fatality rates by up to 80%.16 There was no difference in group-specific mortality reduction according to the severity of stroke symptoms in our study. This is consistent with the findings that stroke units and organized inpatient care reduce 30-day death or institutionalization with the same magnitude of effect across all age groups27 and stroke subtypes.28

no difference in group-specific mortality reduction according to the severity of stroke symptoms in our study. This is consistent with the findings that stroke units and organized inpatient care reduce 30-day death or institutionalization with the same magnitude of effect across all age groups27 and stroke subtypes.28 Fifty-three patients with malignant middle cerebral artery infarction underwent decompressive hemicraniectomy after the establishment of the CSC. Our results showed an 3.8- fold increase in the number of patients who underwent hemicraniectomy surgery after opening the CSC. Decompressive surgery in patients with malignant cerebral infarction increases the probability of survival from 28% to nearly 80%.29 Therefore, it is possible that improved guideline-based treatment, organized care in a stroke unit and decompressive hemicraniectomy may contribute to the lower mortality rates among the stroke patients that were treated in our hospital since the establishment of CSC. Previous studies reported that inpatient stroke center care was associated with a lower mortality compared with general or non-certified hospitals care.2,3,30,31 According to the Registry of the Canadian Stroke Network, stroke unit and organized inpatient care could reduce stroke death at 30 days, indicating a result similar to our study.27 However, not all stroke patients have practical access to optimal care, often because of lack in facilities. In Europe, only few acute stroke patients are treated in appropriate centers.32

Network, stroke unit and organized inpatient care could reduce stroke death at 30 days, indicating a result similar to our study.27 However, not all stroke patients have practical access to optimal care, often because of lack in facilities. In Europe, only few acute stroke patients are treated in appropriate centers.32 Our study also has several limitations. First, our study included patients admitted to a single Korean CSC, and the results can not necessarily be generalized to those treated in other facilities. Second, our CSC had not yet been certified within a professional accreditation scheme, such as the Joint Commission Primary Stroke Center, because there are currently no certification regulations for stroke centers in Korea. However, our stroke center did fulfill the standardized criteria for CSC that the Brain Attack Coalition had published.6 Third, because of the limitation of data collection, we could not obtain acute complication and recurrence rates data, both of which could have influenced some results. Last, our study focused on short-term mortality and could not consider other important patient outcomes, such as 90-day or 1-year functional status or the quality of life. In summary, the present study reveals that more organized stroke care in our stroke center could successfully lower the 30-day mortality rate. Organized stroke care in a CSC might improve the patient outcome after AIS. This study was supported by Dong-A University Research Fund. The authors have no financial conflicts of interest.

In summary, the present study reveals that more organized stroke care in our stroke center could successfully lower the 30-day mortality rate. Organized stroke care in a CSC might improve the patient outcome after AIS. This study was supported by Dong-A University Research Fund. The authors have no financial conflicts of interest. Table 1 Components of CSC for ischemic stroke patient by a consensus statement from the Brain Attack Coalition6 and changes of each specific element according to before and after the establishment of our CSC CSC, comprehensive stroke center; ICU, intensive care unit; MRI, magnetic resonance imaging; MRA, magnetic resonance angiography; DWI, diffusion-weighted imaging; MRV, magnetic resonance venography; TCD, transcranial Doppler; TEE, transesophageal echocardiography; CEA, carotid endarterectomy. Table 2 Characteristics of stroke patients according to before and after the establishment of comprehensive stroke center CSC, comprehensive stroke center; CAD, coronary artery disease; NIHSS, National Institutes of Health Stroke Scale. Table 3 Characteristics and vascular risk factors associated with 30-day stroke fatality CAD, coronary artery disease; NIHSS, National Institutes of Health Stroke Scale; CSC, comprehensive stroke center. Table 4 Multivariate analysis for 30-day stroke case fatality CAD, coronary artery disease; CSC, comprehensive stroke center. Supplementary Table 1 Characteristics of stroke inpatients before the establishment of comprehensive stroke center and propensity-matched controls

CAD, coronary artery disease; NIHSS, National Institutes of Health Stroke Scale; CSC, comprehensive stroke center. Table 4 Multivariate analysis for 30-day stroke case fatality CAD, coronary artery disease; CSC, comprehensive stroke center. Supplementary Table 1 Characteristics of stroke inpatients before the establishment of comprehensive stroke center and propensity-matched controls CSC, comprehensive stroke center; CAD, coronary artery disease; NIHSS, National Institutes of Health Stroke Scale. Supplementary Table 2 Distribution of the number of patients and death at 30 days according to before and after the establishment of comprehensive stroke center CSC, comprehensive stroke center.

Dear sir Intracranial dural carotid cavernous fistulas account for 10 to 15% of all intracranial arteriovenous malformations.1 Although these fistulas tend to regress spontaneously, permanent neurological deficits are observed in as many as 20-30% of untreated patients.2,3 In general, transvenous or transarterial embolization of the fistula is a safe and successful treatment option. Depending on the amount and direction of venous drainage from the cavernous sinus, symptoms may include distended orbit, periorbital veins, and oculomotor, trochlear, and abducens nerve palsy.2,3 In addition, facial palsy and nystagmus are rarely seen as clinical manifestations of dural carotid cavernous fistulas,4 but can arise as complex consequences of embolization.

inage from the cavernous sinus, symptoms may include distended orbit, periorbital veins, and oculomotor, trochlear, and abducens nerve palsy.2,3 In addition, facial palsy and nystagmus are rarely seen as clinical manifestations of dural carotid cavernous fistulas,4 but can arise as complex consequences of embolization. A 41-year-old female visited our clinic with headache and vertical diplopia. The headache was pulsatile, located in the right orbito-temporal region, and had persisted for a month. Other than bilateral myopia, the patient's medical history revealed no trauma, hypertension, diabetes, or other medical problems such as connective tissue disease. On admission, her vital signs were stable. Chest X-ray and laboratory findings were normal. Neurological examination revealed vertical diplopia aggravated by upward gaze. When her head was tilted to the right side, the right eye showed hypotropia and vertical diplopia was aggravated. Thus, we confirmed the right superior rectus palsy; no other abnormalities were observed. Visual acuities and fields, pupil size and reactivity, and ophthalmoscopic findings were all normal. After a full examination for extraocular movement, right partial third cranial nerve palsy was suspected. The brain MRI revealed a carotid cavernous fistula (Figure 1A, B), and digital subtraction angiography confirmed a dural carotid cavernous fistula in the right cavernous sinus. The middle meningeal artery and internal maxillary arteries were the main feeders, and some branches of the meningohypophyseal trunk drained to the cavernous sinus. The main flow from the fistula and the sinus drained to the internal jugular vein through the inferior petrosal sinus. Hence the dural carotid cavernous fistula was classified as type D with branches of the internal carotid artery and the external carotid artery (Figure 1C, D).5

al trunk drained to the cavernous sinus. The main flow from the fistula and the sinus drained to the internal jugular vein through the inferior petrosal sinus. Hence the dural carotid cavernous fistula was classified as type D with branches of the internal carotid artery and the external carotid artery (Figure 1C, D).5 We first performed a transvenous intervention to the right cavernous sinus using coil embolization on the patient's third day in hospital. The transvenous embolization was performed under local anesthesia via the ipsilateral jugular vein; the vein was punctured and a sheath introduced. A catheter was easily navigated through the inferior petrosal sinus to the cavernous sinus. However, the moment the coil was placed in the anterior chamber of the cavernous sinus the patient complained of severe headache and the procedure was stopped. Immediately after the first partial embolization, right infranuclear facial palsy and left side mixed clockwise torsional nystagmus at a neutral position developed. The brain MRI was normal. Despite the partial embolization, no significant change in the drainage of the dural fistula was observed (Figure 1E, F), and no new cortical drainage developed. There were no arterial spasms or dissection.

alsy and left side mixed clockwise torsional nystagmus at a neutral position developed. The brain MRI was normal. Despite the partial embolization, no significant change in the drainage of the dural fistula was observed (Figure 1E, F), and no new cortical drainage developed. There were no arterial spasms or dissection. In the second transarterial intervention on the sixth hospital day, arterial glue (20% n-butyl cyanoacrylate, 0.8 mL) was injected into the middle meningeal artery. Injection into the internal maxillary artery was also attempted, but it was aborted because of arterial spasm. Eventually fistular flow as revealed by external carotid angiography was reduced (Figure 1G, H), and the severity of the patient's headache and nystagmus decreased. The patient's diplopia, nystagmus, and facial palsy gradually declined and resolved completely over two months. Subsequent angiography five months after embolization did not reveal any remnant of thedural carotid cavernous fistula (Figure 1I, J). Symptoms of dural carotid cavernous fistula include proptosis, elevated intraocular pressure, pulsatile bruits, and less commonly, ophthalmoplegia, facial palsy, and facial pain. In the present case, facial palsy and nystagmus developed de novo after transvenousembolization. To the best of our knowledge, this is the first report of facial palsy and nystagmus occurring after transvenous embolization of a carotid cavernous fistula.

nd less commonly, ophthalmoplegia, facial palsy, and facial pain. In the present case, facial palsy and nystagmus developed de novo after transvenousembolization. To the best of our knowledge, this is the first report of facial palsy and nystagmus occurring after transvenous embolization of a carotid cavernous fistula. The classification of dural carotid cavernous fistulas by Barrow et al. involves four angiographic types, one direct (type A) and three dural (types B, C, and D).5 Type D fistulas are shunts between the meningeal branches of both the internal carotid artery and external carotid arteryand the cavernous sinus. In the present case, the carotid cavernous fistula had multiple arteriovenous connections and was classified as type D, prompting consideration of coiling via the transvenous approach. The first intervention was not completed due to severe headache and the patient presented with facial nerve palsy and nystagmus shortly after transvenous embolization. We cannot fully account for the complications, but it is known that pressure palsy can occur because of slight venous engorgement around the internal auditory meatus connected to the inferior petrosal sinus,4 although no new apparent venous reflux was seen in our case. Moreover, because of the close proximity between the facial nerve and vestibulocochlear nerve, elevated venous pressure could affect the vestibulocochlear nerve and may be a cause of concomitant nystagmus. We carefully suggest such a possibility as a responsible cause of facial palsy and nystagmus after transvenous embolization in our case.

he close proximity between the facial nerve and vestibulocochlear nerve, elevated venous pressure could affect the vestibulocochlear nerve and may be a cause of concomitant nystagmus. We carefully suggest such a possibility as a responsible cause of facial palsy and nystagmus after transvenous embolization in our case. The authors have no financial conflicts of interest. Figure 1 T2-weighted MRI (A) shows increased signal voids around right cavernous sinus and MR angiography (B) shows dural fistula flow from internal and external carotid arteries. Pre-embolization angiograms (C, D) shows fistula flow via the internal carotid artery and the external carotid artery. The main flow drains to the internal jugular vein through the inferior petrosal sinus. Angiogram after the first transvenous embolization (E, F) shows insertion of the coil (arrowhead) into the cavernous sinus. After the second transarterial embolization (G, H), fistula flow via the middle meningeal artery (arrow) has decreased, and flow through the internal maxillary artery (arrowhead) has ceased due to vascular spasm. After 5 months, the dural carotid cavernous fistula has regressed (I, J) and there is no remnant aberrant flow.

Introduction Stroke remains to be one of the most devastating neurological diseases. Globally, stroke is the second leading cause of death in persons older than 60 years and the fifth leading cause of death in persons aged 15 to 59 years.1 More than two thirds of stroke deaths worldwide occur are in developing countries. Among the developing countries, China has the largest population with 1.4 billion people. Stroke now has been the leading cause of death in China.2 Stroke carries an enormous financial burden, not only for the families of patients but also for society as a whole in China.3,4 In the past 30 years, China has experienced a rapid economic development. Over time, life expectancies will lengthen, the proportion of elderly people in the population will increase, and the influence of a Westernized lifestyle might shift disease patterns, which may consequently result to an increase in the number of stroke patients. Stroke epidemiological features can help us identify groups of individuals or regions at higher risk of stroke and therefore push the direction of stroke prevention. In this article, we review the current status of the epidemiological features and risk factors of stroke, a recently ongoing stroke epidemiological survey, and the future direction of stroke management in China.

ps of individuals or regions at higher risk of stroke and therefore push the direction of stroke prevention. In this article, we review the current status of the epidemiological features and risk factors of stroke, a recently ongoing stroke epidemiological survey, and the future direction of stroke management in China. Incidence The stroke incidence remains relatively diverse worldwide. Three earlier studies used a similar research design, using a retrospective door-to-door survey, to investigate the stroke incidence in various populations in China since the 1980s.5-7 The total mean age-adjusted incidence of first-ever stroke ranged from 116 to 219 cases per 100,000 population per year (using the 1960 total US population to adjust age). The first multicenter and community-based study on stroke incidence was the 6-city incidence study,5 which was completed in 1983. A sample population of 63,195 was selected from 6 cities, in which a stroke incidence of 219 cases per 100,000 population per year was reported (Table 1).8 The second survey was accomplished in 1985. The scope of the investigation covered 22 rural areas in 21 provinces and a sample population of 246,812. The survey indicated a stroke incidence of 185 cases per 100,000 population per year (Table 1).6 The lowest mean incidence of total stroke was 115.61 cases per 100,000 population per year, which was reported in 1987 by the largest retrospective epidemiological study7 and substantially lower than the stroke incidence reported by the 2 aforementioned studies. The survey included a 5,800,000 population from 29 provinces or cities. The reason for the difference in the reported incidence is unclear but probably related to the difficulties involved in managing and ensuring good-quality data obtained from a huge sample population.

idence reported by the 2 aforementioned studies. The survey included a 5,800,000 population from 29 provinces or cities. The reason for the difference in the reported incidence is unclear but probably related to the difficulties involved in managing and ensuring good-quality data obtained from a huge sample population. For incidence studies, a prospective design is clearly more ideal than a retrospective one. In China, several multicenter and prospective studies have been designed using the same criteria since the 1980s.9-11 The World Health Organization's Monitoring of Trends and Determinants of Cardiovascular Disease project (1985-1990) is the only source of standard data for comparison in different populations of many countries. Beijing was the only Asian area included in the study. The age-adjusted stroke attack incidence rate in Beijing was 175 cases in women and 247 cases in men per 100,000 population per year, the second and sixth highest rates, respectively, in the World Health Organization's Monitoring of Trends and Determinants of Cardiovascular Disease data.9 Another 2 prospective studies indicated that the stroke incidence in their respective monitoring groups varied from 165.4 to 212.2 cases in women and from 182.1 to 314 cases in men per 100,000 population per year.10,11 These 2 studies were well designed and implemented using standard diagnostic criteria, hence their relatively reliable findings (Table 2). All previous Chinese epidemiological studies suggested that the stroke incidence rate increases with age. The highest rates were reported in people older than 75 years, which were 30 times higher than those reported in the age group 35-44 years.12

diagnostic criteria, hence their relatively reliable findings (Table 2). All previous Chinese epidemiological studies suggested that the stroke incidence rate increases with age. The highest rates were reported in people older than 75 years, which were 30 times higher than those reported in the age group 35-44 years.12 Prevalence Prevalence is related to incidence and mortality. High incidence and low mortality can result in high prevalence. Globally, an estimated 16 million people experience a first-ever stroke in the year 2005, with an estimated prevalence of 62 million stroke survivors.13 As a disease of aging, the prevalence of stroke is expected to increase significantly around the world in the years ahead as the global population older than 65 years continues to increase by approximately 9 million people per year.14 Three studies investigated the stroke prevalence in China. As shown in Table 1, the age-adjusted stroke prevalence in 6 cities was 719 cases per 100,000 population for all ages and that in 22 rural areas in 21 provinces was 394 cases per 100,000 population for all ages.5,6 Another retrospective epidemiological study7 of 29 provinces reported that the lowest age-adjusted stroke prevalence rate was 259.86 cases per 100,000 population, substantially lower than those in the 2 aforementioned studies (Table 1). Geographical variations of 1,249-1,285 and 95 cases per 100,000 population were respectively observed in northern cities such as Harbin and Beijing and southern rural areas in the Guangxi province.12 The prevalence of stroke is higher in urban areas than in rural areas (295.30-377.63 vs. 193.56 cases per 100,000 population).7

hical variations of 1,249-1,285 and 95 cases per 100,000 population were respectively observed in northern cities such as Harbin and Beijing and southern rural areas in the Guangxi province.12 The prevalence of stroke is higher in urban areas than in rural areas (295.30-377.63 vs. 193.56 cases per 100,000 population).7 Mortality The World Health Organization's Monitoring of Trends and Determinants of Cardiovascular Disease project indicated that the stroke mortality in Beijing in the 1980s was 58 deaths in women and 66 deaths in men per 100,000 population per year, the second and fifth highest, respectively, in 17 populations.9 Recently, stroke has exceeded heart disease to become the leading cause of death, according to the third survey in China for the period 2004-2005. The stroke mortality recorded was 124.15 deaths in women and 148.57 deaths in men per 100,000 population per year.2 The China Health Statistical Yearbooks8 reveal the trends of stroke mortality in China (Figure 1). Stroke mortality in the urban and rural areas increased gradually from 1990 to 2000 and then began decreasing since the beginning of the 21st century. Some developed countries such as the United States, Canada, France, Switzerland, and Australia have experienced declines in mortality, which may be associated with the increased use of preventative treatment, better control of vascular risk factors, and the advances in acute stroke care. However, the stroke mortality is on a rebound after 2006 (Figure 1). Another notable phenomenon is that the stroke mortality was higher in the urban areas than in the rural areas during the period 1988-2006, whereas an opposite trend was observed after 2006. The reasons for the rebound in mortality and reversal of the trend in the urban and rural areas remain unclear.

r 2006 (Figure 1). Another notable phenomenon is that the stroke mortality was higher in the urban areas than in the rural areas during the period 1988-2006, whereas an opposite trend was observed after 2006. The reasons for the rebound in mortality and reversal of the trend in the urban and rural areas remain unclear. Stroke types and subtypes In some developed countries, up to 67.3-80.5% of stroke cases are attributed to ischemic stroke, whereas only 6.5-19.6% are attributed to intracerebral hemorrhage; approximately 0.8-7.0%, to subarachnoid hemorrhage; and 2.0-4.5%, to undetermined types.15 The Greater Cincinnati/Northern Kentucky Stroke Study and National Institutes of Neurological Disorders and Stroke suggested higher proportions of ischemic stroke cases. Of all stroke cases, 87% were ischemic, 10% were intracerebral hemorrhage, and 3% were subarachnoid hemorrhage.16 Most of the Chinese studies have suggested that the proportion of intracerebral hemorrhage cases is significantly higher in China than in Western countries. Some studies in China reported the proportion of each stroke type (Table 3).8 As shown in Table 3, earlier studies (before 1990) reported that cerebral infarction accounted for 48.6-51.0% of the total number of stroke cases; intracerebral hemorrhage, for 44.0-44.7%; subarachnoid hemorrhage, for 2.0-3.9%; and undetermined types, for 2.8-3.0%.5 In these studies, only a few patients underwent computed tomography (CT); thus, their results might be unreliable. However, the 2 studies in 1992 and 1998 evaluated patients who all underwent CT scan, reporting almost identical proportions of intracerebral hemorrhage (37% vs. 38%), which were 3 times higher than that reported for Western populations. In recent studies,11,17 more than three quarters of patients (75.3-98.0%) underwent CT scan, hence the relatively reliable results. Jiang et al.11 reported that the proportion of each stroke type ranged from 43.7% to 78.9% for ischemic stroke and from 18.8% to 49.0% for hemorrhagic stroke. Another study observed that intracerebral hemorrhage is the most prevalent (55.4% of all stroke cases) in Changsha, a city in South Central China.18 The reasons for the higher incidence of intracerebral hemorrhage in China are unclear, but the high prevalence of hypertension may be the most plausible explanation.

roke. Another study observed that intracerebral hemorrhage is the most prevalent (55.4% of all stroke cases) in Changsha, a city in South Central China.18 The reasons for the higher incidence of intracerebral hemorrhage in China are unclear, but the high prevalence of hypertension may be the most plausible explanation. A review of published studies and data from clinical trials found that hospital admissions for intracerebral hemorrhage have increased by 18% worldwide in the past 10 years, probably because of the increase in the number of elderly people, poor blood pressure control, and the increasing use of anticoagulants, thrombolytics, and antiplatelet agents.19 On the contrary, the China Acute Cerebrovascular Events Registers reported that ischemic stroke cases were increasing and intracerebral hemorrhage cases in China were decreasing.20 In the last 2 decades, the incidence of hemorrhagic and ischemic strokes declined by 1.7% and increased by 8.7% annually, respectively. The decrease in hemorrhagic stroke should be closely related with the improvements in health care that came with the economic development, but the improvement in hypertension control is the most likely explanation because 50% of acute hemorrhagic stroke cases can be attributed to hypertension in Chinese.21

espectively. The decrease in hemorrhagic stroke should be closely related with the improvements in health care that came with the economic development, but the improvement in hypertension control is the most likely explanation because 50% of acute hemorrhagic stroke cases can be attributed to hypertension in Chinese.21 Few community-based studies on ischemic or hemorrhagic stroke subtypes were identified. In the recent multicenter study on the prevalence and outcomes of intracranial large artery atherosclerosis among patients with stroke and transient ischemic attack in China (the Chinese Intracranial Atherosclerosis Study), the prevalence of intracranial stenosis was 46.6% (1,335 patients, including 261 with coexisting extracranial carotid stenosis).22 In addition, geographic and sex differences were noted in the distribution of symptomatic intracranial atherosclerosis cases in China. The proportion of patients with intracranial atherosclerosis was significantly higher in Northern than in Southern China (50.22% vs. 41.88%; P<0.0001). The patients from Northern China were likely to consume more alcohol and cigarettes, and a significantly higher proportion of whom had diabetes mellitus; family histories of stroke, cerebral ischemia, and heart disease; and higher BMI (in press).23

igher in Northern than in Southern China (50.22% vs. 41.88%; P<0.0001). The patients from Northern China were likely to consume more alcohol and cigarettes, and a significantly higher proportion of whom had diabetes mellitus; family histories of stroke, cerebral ischemia, and heart disease; and higher BMI (in press).23 Geographical variation There are substantial geographic disparities in stroke incidence, mortality, and prevalence worldwide. Higher stroke rates in the Southeastern United States (the so-called Stroke Belt), especially along the coasts of Georgia and the Carolinas (socalled Stroke Buckle),24 are well known. In China, all stroke epidemiological studies also suggested a marked geographical variation. Northeastern China has the highest incidence (441-486 cases per 100,000 population per year), whereas Southern China has a significantly lower incidence (81-136 cases per 100,000 population per year).5,8,10 Studies suggested the north-south gradient, with the highest incidence rate in the Heilongjiang province and the lowest incidence rates in the Zhejiang and Guangxi provinces. The Heilongjiang province has 6 times higher stroke incidence rate and 9 times higher mortality rate than the Guangxi province.8 The geographical variation in stroke incidence and mortality may be mainly due to differences in the prevalence of hypertension, the most important risk factor of stroke. The stroke population attributed to hypertension is as great as 40%.

oke incidence rate and 9 times higher mortality rate than the Guangxi province.8 The geographical variation in stroke incidence and mortality may be mainly due to differences in the prevalence of hypertension, the most important risk factor of stroke. The stroke population attributed to hypertension is as great as 40%. Risk factors With the economic growth, the Chinese lifestyle has rapidly changed in the past 3 decades. The risk from major stroke factors, including obesity and hypercholesterolemia, has substantially increased. For example, total fat intake increased from 88.1 g/day in 1983 to 97.4 g/day in 2002. During the same period, total cholesterol intake increased from 124.8 to 350.7 g/day in rural areas and from 334.5 to 488.4 g/day in urban areas.25,26 From 1984 to 1999, mean blood cholesterol level increased by 24%.27 From 1994 to 2002, diabetes prevalence increased by 97%, and overweight or obesity prevalence increased by 85% in the rural areas and 13% in the urban areas.21 At present, the incidence of risk factors for cardiovascular disease and stroke in China has become similar to that in the Western countries, including hypertension, diabetes, hypercholesterolemia, smoking, coronary artery disease, arterial fibrillation, physical inactivity, and obesity. Among the modifiable risk factors, hypertension remains to be the most important risk factor for all types of strokes, with the highest population-attributable risk at 34.6%.

ncluding hypertension, diabetes, hypercholesterolemia, smoking, coronary artery disease, arterial fibrillation, physical inactivity, and obesity. Among the modifiable risk factors, hypertension remains to be the most important risk factor for all types of strokes, with the highest population-attributable risk at 34.6%. The control rate of hypertension differs between the urban and rural areas: In urban areas, it was 3.4% in 1984 and significantly increased to 17.5% in 1999, whereas in rural areas, it did not change much (1.1% in 1984 and 6.9% in 1999).28 Intracranial atherosclerotic disease is more prevalent in the Chinese population than in Western populations. Recent Chinese Intracranial Atherosclerosis Study data suggest that patients with intracranial stenosis had more severe stroke at admission and longer hospital stay than those without intracranial stenosis (median National Institutes of Health Stroke Scale score, 5 vs. 3; median length of hospital stay, 16 vs. 14 days). At 12 months, stroke recurred in 3.24% of all patients with no stenosis, 3.82% of 50-69% of patients with stenosis, 5.16% of 70-99% of patients with stenosis, and 7.40% of 100% of patients with occlusion (in press).22

nal Institutes of Health Stroke Scale score, 5 vs. 3; median length of hospital stay, 16 vs. 14 days). At 12 months, stroke recurred in 3.24% of all patients with no stenosis, 3.82% of 50-69% of patients with stenosis, 5.16% of 70-99% of patients with stenosis, and 7.40% of 100% of patients with occlusion (in press).22 Ongoing epidemiological survey and future direction Epidemiological studies can help identify groups of individuals or regions at higher risk of stroke. They can also help better understand the natural history of certain associated conditions and therefore push the direction of prevention and therapeutic investigations. However, the most widely cited data on the incidence, prevalence, and mortality of stroke are derived from the studies carried out in the 1980s. Stoke epidemiological features in China have changed with the economic development in the last 3 decades. It is important to understand these changes to establish timely strategies for stroke prevention.

ed data on the incidence, prevalence, and mortality of stroke are derived from the studies carried out in the 1980s. Stoke epidemiological features in China have changed with the economic development in the last 3 decades. It is important to understand these changes to establish timely strategies for stroke prevention. To aim at revealing the characteristics of stroke transition, the China Nationwide Epidemiological Survey for cerebrovascular disease will be implemented in 2013. This survey is one of the National Key Technology R&D Programs during the "Twelve-Fifth Plan" period (grant no. 2011BAI08B01). The main purpose of the survey is to obtain the incidence, prevalence, and mortality of stroke in different regions of China; the secondary purpose is to access relevant data on stroke and transient ischemic attack, including risk factors, treatment, and secondary prevention. The epidemiological survey will conducted in 150-160 disease surveillance points, which come from the National Disease Surveillance System in 31 provinces (municipal cities) in China. A multistage stratified cluster sampling method will be applied to obtain the sample population. The total sample size is expected to be more than 600,000 patients. The survey will be completed in cooperation of staffs from the Centers for Disease Control and Prevention and neurologists in each province. The results of this survey are eagerly anticipated.

ing method will be applied to obtain the sample population. The total sample size is expected to be more than 600,000 patients. The survey will be completed in cooperation of staffs from the Centers for Disease Control and Prevention and neurologists in each province. The results of this survey are eagerly anticipated. Future epidemiological studies on stroke in China should use high-quality methods to update our understanding on the incidence (first and recurrent events), prevalence, mortality, cost, and specific characteristics of stroke (ischemic or hemorrhagic stroke subtypes, small vessel disease, etc.). Nationwide hospital-based and especially community-based stroke registers are required to monitor stroke incidence and trends, as well as the quality of stroke care. Feasible and affordable long-term primary and secondary prevention strategies should be developed and strongly recommended for clinical settings. Complementary alternative therapeutics such as traditional Chinese medicine, acupuncture, urokinase, snake venom, and hematoma aspiration for intracerebral hemorrhage should be assessed by means of internationally recognized methods.29

ntion strategies should be developed and strongly recommended for clinical settings. Complementary alternative therapeutics such as traditional Chinese medicine, acupuncture, urokinase, snake venom, and hematoma aspiration for intracerebral hemorrhage should be assessed by means of internationally recognized methods.29 Support from the Chinese government Prevention remains to be the most viable avenue for lessening the burden of stroke on society, considering the high incidence of stroke worldwide, insidious contribution of stroke risk factors, and the paucity of proven acute stroke therapies. Despite the challenges and amount of work, the Chinese people are giving an organized effort to provide better stroke care. A comprehensive plan to address this issue and improve stroke care has been developed. The Ministry of Health has just established the Chinese National Center for Stroke Care Quality Control and Management. A national stroke data bank is being established based on the blueprint of the Chinese National Stroke Registry. Extensive physician training and community education on evidence-based stroke care have been incorporated into the "Twelve-Fifth Year project."30 As the world has witnessed the impressive economic growth in China in the past 3 decades, we expect substantial improvement in stroke care. This survey is supported by the National Key Technology R&D Programs during the "Twelve-Fifth Plan" period (grant no. 2011BAI08B01). The authors have no financial conflicts of interest. Figure 1 The trends of stroke mortality in China.

Support from the Chinese government Prevention remains to be the most viable avenue for lessening the burden of stroke on society, considering the high incidence of stroke worldwide, insidious contribution of stroke risk factors, and the paucity of proven acute stroke therapies. Despite the challenges and amount of work, the Chinese people are giving an organized effort to provide better stroke care. A comprehensive plan to address this issue and improve stroke care has been developed. The Ministry of Health has just established the Chinese National Center for Stroke Care Quality Control and Management. A national stroke data bank is being established based on the blueprint of the Chinese National Stroke Registry. Extensive physician training and community education on evidence-based stroke care have been incorporated into the "Twelve-Fifth Year project."30 As the world has witnessed the impressive economic growth in China in the past 3 decades, we expect substantial improvement in stroke care. This survey is supported by the National Key Technology R&D Programs during the "Twelve-Fifth Plan" period (grant no. 2011BAI08B01). The authors have no financial conflicts of interest. Figure 1 The trends of stroke mortality in China. Table 1 Stroke incidence, mortality, and prevalence (per 100 000 population per year) in different areas in China8 Rates adjusted for age using the 1960 total US population. Table 2 Stroke incidence and mortality (per 100 000 population) in the monitoring groups in 3 prospective studies8 Table 3 Proportion (%) of each stroke type (first event) in the Chinese population

Table 1 Stroke incidence, mortality, and prevalence (per 100 000 population per year) in different areas in China8 Rates adjusted for age using the 1960 total US population. Table 2 Stroke incidence and mortality (per 100 000 population) in the monitoring groups in 3 prospective studies8 Table 3 Proportion (%) of each stroke type (first event) in the Chinese population ICH, intracerebral haemorrhage; CI, cerebral infarction; SAH, subarachnoid haemorrhage; UND, undefined.

Introduction Calcium deposits in the arterial bed may indicate the extent of atherosclerotic lesions and aortic knob calcification (AC) is associated with increased risks of cardiovascular and cerebrovascular events.1-5 AC is also associated with coronary artery calcification or carotid atherosclerosis, and might have predictive and prognostic value for coronary artery disease.1,6,7 In addition, several reports have shown that aortic atherosclerotic disease or AC is related to ischemic stroke.4,8,9 But its clinical significance for ischemic stroke patients with intracranial (IC) stenosis, one of the major mechanisms of ischemic stroke, remains unclear. Although thoracic computed tomography (CT) or digital subtraction angiograms are reliable in detecting aortic calcification,1,10 these imaging modalities are not routinely used. In this study, we evaluated the clinical importance of AC in ischemic stroke patients with intracranial stenosis by using simple, non-invasive routine chest radiography.

tomography (CT) or digital subtraction angiograms are reliable in detecting aortic calcification,1,10 these imaging modalities are not routinely used. In this study, we evaluated the clinical importance of AC in ischemic stroke patients with intracranial stenosis by using simple, non-invasive routine chest radiography. Methods We studied 307 acute ischemic stroke patients within 7 days of onset, who were admitted to our department from May 2009 to April 2010, and who underwent magnetic resonance angiography or distal subtraction angiography. The patients were interviewed to collect information on the risk factors of ischemic stroke based on personal history of smoking, consumption of alcohol, diabetes mellitus, hypertension, family history and past history of stroke before entering the study. We excluded subjects if their chest X-rays were not properly centered, if they had any deviation of the trachea or shift of the mediastinum or if they had any known disease in the aorta, such as aortitis. We also excluded patients who had cerebrovascular events related to trauma, medical instrumentation, severe concomitant kidney (serum creatinine≥2.0 mg/dL) or liver disease, autoimmune disease, cerebral vaculitis or embolism from implants, such as an artificial heart valves or atrial fibrillation. Samples were collected from venous blood after a 12-hour overnight fast, and lipid profiles and standard blood tests were performed on admission.

ey (serum creatinine≥2.0 mg/dL) or liver disease, autoimmune disease, cerebral vaculitis or embolism from implants, such as an artificial heart valves or atrial fibrillation. Samples were collected from venous blood after a 12-hour overnight fast, and lipid profiles and standard blood tests were performed on admission. All patients had chest radiography in the posteroanterior (PA) view. We examined the presence of calcification in the aortic knob (Figure 1A). Based on the findings of the cerebral angiogram, each segment of the carotid and vertebrobasilar arterial systems were classified as normal, stenosis or occlusion. These classifications were based on a neuroradiologist's report and a consensus by stroke specialists at a weekly stroke conference. The existence of any degree of stenotic lesion was interpreted as potential stenosis (Figure 1B). Data are presented as mean±standard deviation. Statistical analyses were done using SPSS® Ver.11.5 for Windows. Discrete variables were analyzed by the independent t-test and χ2 test. Spearman's correlation was used to examine the relationship between the aortic knob width and other risk factors. Logistic regression analysis was conducted to determine the factors related to intracranial atherosclerosis. Statistical significance was considered at P<0.05.

nalyzed by the independent t-test and χ2 test. Spearman's correlation was used to examine the relationship between the aortic knob width and other risk factors. Logistic regression analysis was conducted to determine the factors related to intracranial atherosclerosis. Statistical significance was considered at P<0.05. Results Demographic characteristics of the study patients and risk factor profiles are shown in Table 1. The mean age of patients was 66.7±8.43 years and there were 191 male patients (62.2%). Among the 307 patients, there was stenosis in IC arteries in 151 patients (49.1%), in EC arteries in 68 patients (22.1%) and in both IC and EC arteries in 57 patients (18.5%).

y patients and risk factor profiles are shown in Table 1. The mean age of patients was 66.7±8.43 years and there were 191 male patients (62.2%). Among the 307 patients, there was stenosis in IC arteries in 151 patients (49.1%), in EC arteries in 68 patients (22.1%) and in both IC and EC arteries in 57 patients (18.5%). AC was present in 82 patients (26.7%) of the chest PA view. AC was observed more commonly in elderly patients and in patients with hypertension. There was no significant difference between the patients with or without AC in terms of diabetes, family history, smoking, aortic knob width and other risk factors, such as lipid profile, homocystein and high-sensitivity C-reactive protein levels. AC was observed in 58 (38.4%) patients with IC stenosis, 17 (25.0%) patients with EC stenosis and 17 (29.8%) patients with both stenosis of both the IC and EC arteries. Patients with AC had IC stenosis or IC+EC stenosis more frequently than those without calcification (P<0.01 and P=0.01, respectively). Unlike the patients with IC stenosis or those with IC and EC stenosis, patients with only EC stenosis did not have significantly more frequent AC (25%) than those without stenosis (15.3%) (P=0.08) (Figure 2). For patients of all EC stenosis (EC stenosis and IC+EC stenosis), there was no significant relationship indicated between AC (P=0.671). Logistic regression analysis showed that age (>65 years; Odds ratio (OR), 1.79; 95% confidence interval (CI), 1.01 to 3.19; P=0.04) and AC (OR, 3.54; 95% CI, 1.90 to 6.61, P<0.01) were independent predictors for IC stenosis after adjusting for age, gender, vascular risk factors and AC (Table 2).

d between AC (P=0.671). Logistic regression analysis showed that age (>65 years; Odds ratio (OR), 1.79; 95% confidence interval (CI), 1.01 to 3.19; P=0.04) and AC (OR, 3.54; 95% CI, 1.90 to 6.61, P<0.01) were independent predictors for IC stenosis after adjusting for age, gender, vascular risk factors and AC (Table 2). Discussion The presence of AC in chest X-ray and age were correlated with IC stenosis in ischemic stroke patients. The detection of AC in chest X-ray may allow prediction of the presence of IC stenosis, which is an important mechanism in developing stroke. Chest radiography is a routine, non-invasive screening method and can reliably assess aortic calcification.7,11 Recently, interest has grown in arterial calcification in terms of risk factors and subsequent outcomes.

allow prediction of the presence of IC stenosis, which is an important mechanism in developing stroke. Chest radiography is a routine, non-invasive screening method and can reliably assess aortic calcification.7,11 Recently, interest has grown in arterial calcification in terms of risk factors and subsequent outcomes. Arterial calcification is a complex, regulated process of biomineralization that resembles osteogenesis, which develops in the intima layer within atherosclerotic plaque, and which is a progressive feature of common atherosclerosis.2,12,13 Arterial calcification is associated with conventional atherogenic risk factors such as old age, diabetes, hypertension, cholesterol, or C-reactive protein and with an increased risk of cardiovascular and cerebrovascular diseases.2,14 This study also correlated AC with old age and hypertension, but had indicated a tendency of association with HDL-cholesterol, as the study population was comprised of ischemic stroke patients. Individuals with atherosclerotic disease occurring in one specific vascular bed have a higher risk of a clinical disease caused by atherosclerosis at another site.2,9,15 Thus, patients who suffer ischemic stroke or peripheral artery disease have a higher incidence of concomitant systemic artery disease, which may explain our results involving the simultaneous involvement of aorta and the IC artery, However, when we reanalyzed the association for predictors of EC stenosis after adjusting for age, sex and vascular risk factors, such as hypertension, diabetes mellitus and hyperlipidemia, a significant association was not found (OR, 0.56; 95% CI, 0.23 to 1.32; P=0.18), although we found a tendency of a correlation between AC and EC stenosis (P=0.08).

sociation for predictors of EC stenosis after adjusting for age, sex and vascular risk factors, such as hypertension, diabetes mellitus and hyperlipidemia, a significant association was not found (OR, 0.56; 95% CI, 0.23 to 1.32; P=0.18), although we found a tendency of a correlation between AC and EC stenosis (P=0.08). Our study has several limitations. The first is a selection bias because the study patients did not represent the general population. This is why AC was more common among the participants than in previous studies (2-3%),4 and AC was not correlated with EC athersosclerosis. Also, regardless of symptomatic or asymtomatic stenosis, we only included study patients of large artery atherosclerotic groups and small artery occlusion groups based on the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification,16 in which AC showed no significance between the two groups (P=0.118). Secondly, we detected AC using only chest radiography. We could not confirm the presence of AC using other methods, such as electron-beam CT. We could not also include the other aortic atherosclerotic lesions, such as atheromas, protruding atheromas, or plaque, and thoracic segments of the aorta identified by transesophageal echocardiography. However, the present findings may well indicate the significance of aortic calcification by the use of just a simple chest x-ray, although the characteristics of the study patients and the diagnostic tools used in this study should be interpreted with caution.

f the aorta identified by transesophageal echocardiography. However, the present findings may well indicate the significance of aortic calcification by the use of just a simple chest x-ray, although the characteristics of the study patients and the diagnostic tools used in this study should be interpreted with caution. In conclusion, our results suggest that AC is a reliable predictor for IC stenosis in ischemic stroke patients. An increase in the use of chest radiography as a screening or risk factor assessment tool may be justified. The authors have no financial conflicts of interest. This study was supported by the Wonkwang Institute of Clinical Medicine in 2011. Figure 1 A 67-year-old man presenting with right hemiparesis with dysarthria (NIHSS 3 points). (A) Chest X-ray shows aortic knob calcification. (B and C) Diffusion-weighted image of magnetic resonance (MR) imaging (B) and MR angiography (C) of brain show a high signal intensity in the left MCA area with left MCA stenosis. Figure 2 Frequency of aortic knob calcification in relation to intracranial stenosis (IC) or extracranial stenosis (EC). Table 1 Clinical characteristics and laboratory findings of the study group Data are mean± SD values. All IC stenosis= IC only or IC +EC stenosis; IC, intracranial; BMI, body mass index; hs CRP, high-sensitivity C-reactive protein; HDL, high-density lipoprotein; LDL, low-density lipoprotein. Table 2 Logistic regression analysis of the clinical factors for IC stenosis IC, intracranial; CI, confidence interval; HDL-high density lipoprotein.

Introduction Over the last decade, a big stride was made in research on neurocritical care, which translated into better outcomes for patients treated in the neurointensive care unit (NeuroICU).1 The main purpose of the NeuroICU is to treat patients with severe brain injuries, such as large ischemic or hemorrhagic stroke, traumatic brain injury (TBI), or status epilepticus. Patients with such injuries develop neurologic damage when the initial injury develops (primary brain injury [PBI]). However, a significant portion of patients develop secondary deterioration while being treated in the NeuroICU, which is termed as secondary brain injury (SBI). Even with severe tissue destruction, some viable tissue still exists in the surrounding area of the PBI and may be more vulnerable to additional damage, which is often triggered by the PBI itself as well as by systemic deterioration. Traditionally, PBI was considered to be an irreversible process. On the other hand, SBI is, at least, partially reversible and preventable if identified early and treated appropriately. Therefore, current neurocritical care aims for early detection and minimization of SBI before it becomes irreversible.2,3

terioration. Traditionally, PBI was considered to be an irreversible process. On the other hand, SBI is, at least, partially reversible and preventable if identified early and treated appropriately. Therefore, current neurocritical care aims for early detection and minimization of SBI before it becomes irreversible.2,3 The common SBIs are brain tissue hypoperfusion or ischemia due to intracranial pressure (ICP) surges, brain tissue hypoxia (BTH), brain tissue hypoglycemia, or excitotoxic damage due to recurrent seizures. Even though SBI is frequently encountered in the NeuroICU, neurological examination alone is not sensitive enough for monitoring on-going SBI because such patients are usually comatose.4 Moreover, when the neurological examination shows worsening, it is usually too late to adequately treat, and permanent damage has already occurred.

I is frequently encountered in the NeuroICU, neurological examination alone is not sensitive enough for monitoring on-going SBI because such patients are usually comatose.4 Moreover, when the neurological examination shows worsening, it is usually too late to adequately treat, and permanent damage has already occurred. Considering that patients in the NeuroICU are vulnerable to SBI, more sensitive and accurate methods are required to detect secondary neurophysiologic deterioration as early as possible. Direct monitoring of physiological parameters is technically possible and would expand the monitoring capabilities across patients with various neurologic injuries. A comprehensive understanding of various physiological parameters will allow medical practitioners to pursue a multifaceted approach to limit the occurrence of SBI, which is currently possible with the help of multimodality monitoring (MMM). MMM gathers a variety of information including ICP, cerebral blood flow (CBF), real time brain metabolism of glucose and oxygen, and the electrical status of the brain, all of which allow for a better understanding of any physiological changes in the brain. A list of currently available MMM techniques are described in Table 1. The majority of clinical outcome studies using MMM have focused on TBI and subarachnoid hemorrhage (SAH). Currently, MMM is expanding its coverage to other neurologic conditions such as severe ischemic stroke or intracerebral hemorrhage (ICH).5-9

n. A list of currently available MMM techniques are described in Table 1. The majority of clinical outcome studies using MMM have focused on TBI and subarachnoid hemorrhage (SAH). Currently, MMM is expanding its coverage to other neurologic conditions such as severe ischemic stroke or intracerebral hemorrhage (ICH).5-9 Here, we briefly review the basic concepts of MMM as a neurophysiologic monitoring tool as well as its clinical applicability in patients with severe stroke. Location of monitoring Currently, the most accurate method for neuromonitoring is direct tissue monitoring. Although various forms of non-invasive monitors have been developed, their accuracy is still far from satisfactory. For direct monitoring, MMM probes are inserted into the brain parenchymal tissue via a burr hole and are fixed to the skull using a cranial bolt system (Figure 1). The diameter of the probe is usually less than a few millimeters, regardless of the form of the monitor. Given the fact that the information only represents a small sampling area (from a few mm3 to cm3), proper probe location is very important in interpreting physiologic data.

ll using a cranial bolt system (Figure 1). The diameter of the probe is usually less than a few millimeters, regardless of the form of the monitor. Given the fact that the information only represents a small sampling area (from a few mm3 to cm3), proper probe location is very important in interpreting physiologic data. The ideal location for MMM tissue probes is still unknown. However, the consensus is to monitor those brain tissues that are at the highest risk of secondary damage. In cases of focal brain injury, such as ICH or large cerebral infarction, the ideal location of monitoring is the perihematomal area or ischemic penumbra, respectively. Therefore, probes are inserted into the perilesional frontal white matter.10 In cases with bilateral pathology, such as diffuse TBI or SAH due to anterior communicating artery aneurysm rupture, non-dominant frontal white matter is generally chosen as the probe location. However, when ipsilateral damage is very severe and the patient has already undergone a hemicraniectomy, the probes cannot be affixed to the ipsilateral skull. In this situation, the monitors are inserted into the contralateral areas in close proximity to the PBI. Complications due to probe placement are reported to be as low as 1-2%.10 The majority of complications are procedure-related hemorrhage, proberelated infection, or misplacement of the probe in the core of the PBI. Hemorrhagic complication risk stemming from MMM probe placement is reported to be similar to that stemming from extraventricular drain (EVD) insertion. Most reported cases with probe-related infection had concomitant EVD, and infection is ascribed to be associated with EVD insertion, rather than direct infection via burr-hole procedure. Currently, the MMM probe is inserted visually while targeting the perilesional tissue. However, more studies are needed to clarify the best location for monitoring.

related infection had concomitant EVD, and infection is ascribed to be associated with EVD insertion, rather than direct infection via burr-hole procedure. Currently, the MMM probe is inserted visually while targeting the perilesional tissue. However, more studies are needed to clarify the best location for monitoring. Intracranial pressure Regardless of the fact that the primary mechanisms are cytotoxic or vasogenic, patients with severe strokes usually develop severe brain edema, which leads to an elevation in ICP. If intraventricular hemorrhage and/or hydrocephalus are present in combination, the chances of elevated ICP are even higher. ICP monitoring is the most crucial step in the understanding of cerebral hemodynamics. Because ICP is regarded as a form of resistance in terms of cerebral perfusion, cerebral perfusion pressure (CPP), a difference between mean arterial pressure (MAP) and ICP, is considered as a net driving force for cerebral perfusion. A major adverse consequence of pathologic ICP elevation is cerebral hypoperfusion, leading to secondary brain ischemia. Global tissue perfusion and CBF are associated with CPP, thus, CPP is regarded as an indicator of CBF.

ean arterial pressure (MAP) and ICP, is considered as a net driving force for cerebral perfusion. A major adverse consequence of pathologic ICP elevation is cerebral hypoperfusion, leading to secondary brain ischemia. Global tissue perfusion and CBF are associated with CPP, thus, CPP is regarded as an indicator of CBF. It is possible to reliably measure the ICP via a ventricular catheter or intraparenchymal fiberoptic system; however, the gold standard for measuring the ICP is through EVD. The pressure measured by EVD represents global ICP levels, since EVD measures intraventricular pressure. Any pressure generated by a focal mass may be transmitted to the lateral ventricle until equilibrium is reached. Meanwhile, more direct parenchymal ICP systems became available; a good example is the Camino fiberoptic ICP monitor (Integra Lifesciences). Because ICP can frequently be compartmentalized in cases of focal injury, local ICP surrounding the primary lesion may more accurately reflect the changes of cerebral hemodynamics than ICP at remote areas. Therefore, focal ICP measurement may be more suitable for cases with focal brain mass.

tor (Integra Lifesciences). Because ICP can frequently be compartmentalized in cases of focal injury, local ICP surrounding the primary lesion may more accurately reflect the changes of cerebral hemodynamics than ICP at remote areas. Therefore, focal ICP measurement may be more suitable for cases with focal brain mass. The normal range of ICP is between 5-15 mmHg. According to the Monro-Kellie doctrine, the total intracranial volume is fixed. If an increase in parenchymal volume is not compensated by a reduction in the cerebrospinal fluid volume or cerebral blood volume, an elevation in ICP ultimately ensues.11 With ICP surges, abnormal clinical features develop, such as the Cushing reflex (high blood pressure with decreased heart rate) or pupillary dilatation without light reflex. However, by the time clinical signs of ICP elevation are evident, it is usually too late to intervene and reverse the process. Therefore, for cases in which an elevation in ICP is suspected, the American Stroke Association (ASA) ischemic stroke and ICH guidelines recommend ICP monitoring.12,13 The ICP pressure-volume curve depicts ICP starts skyrocketing when its value goes higher than 20 mmHg, suggesting a fatigue point of compensation (Figure 2). Therefore, traditional ICP-guided therapy aims to keep the ICP below 20 mmHg. However, a recent randomized clinical trial failed to show the superiority of the ICP number-oriented therapy compared to the conventional image-based ICP treatment.14 Although ICP elevation in TBI is somewhat different in pathophysiology from that in stroke, studies suggest that more complex strategies are needed in treating patients with ICP elevation, rather than simply modulating ICP numbers. In the ASA guidelines, most ICP management strategies are based on studies on TBI. Regardless of the underlying physiologic differences in ICP elevation, maintaining proper CPP before initiating ICP lowering treatments is of utmost importance. With respect to CPP, the ASA ICH guidelines suggest that CPP be maintained at least at 60 mmHg.13 Without ICP monitoring, information on ICP and CPP cannot be obtained, limiting the optimal management of patients with large ICH.

, maintaining proper CPP before initiating ICP lowering treatments is of utmost importance. With respect to CPP, the ASA ICH guidelines suggest that CPP be maintained at least at 60 mmHg.13 Without ICP monitoring, information on ICP and CPP cannot be obtained, limiting the optimal management of patients with large ICH. Another benefit of ICP monitoring is that it provides some information on the status of cerebral autoregulation (AR). Under ideal conditions, in order to determine whether AR is intact, we need to have information on CPP as well as CBF. By definition, when CBF is stably maintained within a certain CPP range, the patient is said to have intact AR. However, direct CBF measurement is not always feasible, except in cases for which concomitant Hemedex monitoring is used. With the simple correlation between ICP and MAP, more practically, we can determine whether AR is intact or not. Theoretically, if the AR is intact, cerebral resistance vessels start to constrict in order to maintain constant CBF as CPP increases. Meanwhile, in cases with AR failure, cerebral blood vessels are just passively dependent on CPP or MAP, and the ICP/MAP relationship is linear (Figure 3). However, this static correlation between MAP and ICP is somewhat arbitrary, and a more objective and continuous method is needed for the real time assessment of the AR status. Continuous assessment of AR is possible by the moving correlation coefficient between ICP and MAP over 200-300 seconds using a 5-second interval data set.15 This Pearson's correlation coefficient is called a pressure reactivity index (PRx), which is a real time surrogate marker for pressure AR. A PRx smaller than 0.2 is generally considered as a sign of intact AR, while a PRx greater than 0.2 suggests AR failure.15

over 200-300 seconds using a 5-second interval data set.15 This Pearson's correlation coefficient is called a pressure reactivity index (PRx), which is a real time surrogate marker for pressure AR. A PRx smaller than 0.2 is generally considered as a sign of intact AR, while a PRx greater than 0.2 suggests AR failure.15 Using PRx values, optimal CPP targets can be identified for individual patients, and this aids in creating an individualized goal-directed therapy. Mean PRx-CPP plots show a U-shaped curve in the majority of patients with brain injury. Hence, CPP ranges with the lowest mean PRx values reveal ideal and optimal CPPs where AR is most actively working (Figure 4). This concept was validated in patients with TBI and SAH.16 In a small group of comatose patients with ICH, survivors were maintained at slightly higher ranges of CPP compared to optimal CPP ranges, while the non-survivors had values in the lower ranges of CPP compared to the optimal CPP.8 Moreover, and especially in patients with ICH, the AR status is regarded as an independent predictor of mortality.8

se patients with ICH, survivors were maintained at slightly higher ranges of CPP compared to optimal CPP ranges, while the non-survivors had values in the lower ranges of CPP compared to the optimal CPP.8 Moreover, and especially in patients with ICH, the AR status is regarded as an independent predictor of mortality.8 In patients with malignant middle cerebral artery infarction, the majority of the ipsilateral hemisphere is already infarcted; hence, it is difficult to find a proper location for monitoring. These patients are often candidates for decompressive hemicraniectomy. Therefore, an ICP monitor is inserted in the contralateral hemisphere where the skull is intact, but this provides a falsely low ICP value compared to the perilesional ICP. Herniation and pupillary dilatation may occur without ICP surges in the contralateral hemisphere.17 More studies are needed regarding ICP monitoring in patients with large ischemic stroke. Although there are different ICP monitoring methods, including subdural techniques, they are not commonly used in clinical practice due to accuracy issues. In addition, non-invasive ICP monitoring is still not accurate enough for routine clinical use.

In patients with malignant middle cerebral artery infarction, the majority of the ipsilateral hemisphere is already infarcted; hence, it is difficult to find a proper location for monitoring. These patients are often candidates for decompressive hemicraniectomy. Therefore, an ICP monitor is inserted in the contralateral hemisphere where the skull is intact, but this provides a falsely low ICP value compared to the perilesional ICP. Herniation and pupillary dilatation may occur without ICP surges in the contralateral hemisphere.17 More studies are needed regarding ICP monitoring in patients with large ischemic stroke. Although there are different ICP monitoring methods, including subdural techniques, they are not commonly used in clinical practice due to accuracy issues. In addition, non-invasive ICP monitoring is still not accurate enough for routine clinical use. Brain tissue oxygen Since cerebral ischemia is the most common form of SBI, early detection of BTH is one of the most important purposes of neuromonitoring. Catheter probes, which can sense partial pressure of oxygen at the tissue level, were introduced into the clinical practice more than a decade ago. Two types of oxygen sensors were introduced: one uses a Clark-type electrode (Licox, Integra Lifesciences), and the other uses fluorescent optical sensors (Neurotrend) which is no longer available.18 In vitro studies have shown that the Licox probe has adequate data accuracy and stability for clinical use, and has been used for brain oxygen monitoring in the NeuroICU. Like other tissue monitors, Licox is a focal monitor with a probe diameter of 0.5 mm and a measurement volume of 7-15 mm3. For stable oxygen measurement, Licox needs to run for a few hours after insertion. Therefore, in clinical practice, the low brain tissue oxygen tension (PbtO2) level that occurs right after probe insertion does not necessarily indicate that the probe is in the infarcted area, since some time is needed for the probe to display a valid value. Because gaseous pressure is temperature dependent, the measured oxygen tension should be adjusted for the tissue temperature. The machine automatically performs temperature adjustments if the brain temperature probe is simultaneously inserted; otherwise, manual temperature correction is warranted. Cerebral oxygen tension is generally driven by CBF and the local oxygen extraction fraction. Therefore, measured PbtO2 can be used as a rough indicator of the CBF in metabolically stable conditions when the oxygen extraction fraction is stable.19

aneously inserted; otherwise, manual temperature correction is warranted. Cerebral oxygen tension is generally driven by CBF and the local oxygen extraction fraction. Therefore, measured PbtO2 can be used as a rough indicator of the CBF in metabolically stable conditions when the oxygen extraction fraction is stable.19 PbtO2 levels in the white matter of a healthy individual are around 25-30 mmHg, which is quite lower than expected.20 While searching for clinically meaningful low cut-off values for predicting poor outcomes, several studies have demonstrated that a PbtO2 measurement of less than 10 mmHg is associated with a decrease in oxygen extraction, suggestive of a poor functional status.21 Moreover, the duration of time when PbtO2 is lesser than 20 mmHg is associated with poor outcomes in SAH and TBI.22 Although it is difficult to pinpoint the clinically critical hypoxic point, a PbtO2 level below 15-20 mmHg is generally regarded as a threshold value for BTH. The Brain Trauma Foundation guidelines suggest that PbtO2 be maintained at more than 15 mmHg.23

r than 20 mmHg is associated with poor outcomes in SAH and TBI.22 Although it is difficult to pinpoint the clinically critical hypoxic point, a PbtO2 level below 15-20 mmHg is generally regarded as a threshold value for BTH. The Brain Trauma Foundation guidelines suggest that PbtO2 be maintained at more than 15 mmHg.23 Since PbtO2 level is influenced by many physiologic variables, critically low PbtO2 levels can be improved by adjusting certain physiological factors.19 A recent study showed that the main factors affecting PbtO2 are the CBF and partial pressure of oxygen in the arterial blood (PaO2). In terms of oxygen delivery, the vast majority of oxygen is transferred as a hemoglobin (Hb) bound form. Because the Hb-carried oxygen level is 450 times more than dissolved-oxygen levels in the blood, the unbound form is usually neglected when calculating oxygen transport. However, Hb-bound oxygen cannot directly increase tissue oxygen tension; it needs to be dissociated from Hb at the tissue level. Therefore, in patients with low PbtO2, several therapeutic approaches can be used to increase PbtO2. An increase in fraction of inspired oxygen (FiO2) level is the simplest way to increase PaO2 but it can mask the underlying cause of the low PbtO2 and is generally discouraged. In the NeuroICU, CBF augmentation is often achieved by increasing CPP (by adding more vasopressors) or decreasing ICP (using ICP lowering therapies). When the cardiac output is low, inotropes such as Dobutamine or Milrinone are used for augmenting CBF. In cases with severe ICP crisis due to large hemispheric infarction, decompressive hemicraniectomy immediately improves PbtO2 levels, most likely by decreasing ICP and increasing CPP.24

(using ICP lowering therapies). When the cardiac output is low, inotropes such as Dobutamine or Milrinone are used for augmenting CBF. In cases with severe ICP crisis due to large hemispheric infarction, decompressive hemicraniectomy immediately improves PbtO2 levels, most likely by decreasing ICP and increasing CPP.24 By defining BTH as a PbtO2 level of less than 15 mmHg, we can identify a CPP threshold below which the chances of BTH dramatically increase in comatose patients with ICH.8 This is important for managing such patients if they are in facilities without sophisticated brain tissue oxygen monitoring. In general, blood pressure is usually lowered to limit hematoma expansion in patients with acute ICH based on the result of a randomized clinical trial.25 However, the patients in that trial were mostly non-comatose; therefore, the results cannot be directly translated to patients with ICH who are comatose. Moreover, it is unclear whether keeping the blood pressure low is still beneficial after the first 24 hours when the critical period for hematoma expansion has already passed. A recent study demonstrated that concomitant ischemic infarcts were more frequently found when the blood pressure was abruptly and drastically lowered. Although it is still speculative, uniformly maintaining low blood pressure over several days may not be safe in patients with ICH who are under coma. More studies are needed to confirm this. In Korea, the Licox monitor is currently under safety investigation at Korean Food and Drug Administration as of March 2013.

red. Although it is still speculative, uniformly maintaining low blood pressure over several days may not be safe in patients with ICH who are under coma. More studies are needed to confirm this. In Korea, the Licox monitor is currently under safety investigation at Korean Food and Drug Administration as of March 2013. Global measurement of cerebral oxygenation: Jugular bulb oxygen saturation In addition to Licox, a focal tissue oxygen monitor, there is another type of global brain oxygen monitoring system.26,27 Oxygen levels in the cerebral venous outflow may inversely correlate with global brain oxygen consumption. Therefore, oxygen saturation in the jugular bulb (SjVO2) may be used to indirectly estimate cerebral oxygen consumption. It is still unclear whether the right or ipsilateral jugular vein should be monitored. The majority of patients have right side dominance in their internal jugular venous drainage, thus monitoring oxygen saturation in the dominant draining vein may be reasonable. Most experts choose the dominant right side because they think that this jugular vein more accurately represents global oxygen consumption regardless of the location of the lesion.26 There is another non-invasive oxygen monitoring technique that uses infrared technology (near infrared spectroscopy). However, its clinical implication remains to be elucidated in patients with stroke.28

t this jugular vein more accurately represents global oxygen consumption regardless of the location of the lesion.26 There is another non-invasive oxygen monitoring technique that uses infrared technology (near infrared spectroscopy). However, its clinical implication remains to be elucidated in patients with stroke.28 Cerebral blood flow monitoring CBF measurement provides a better understand of the perfusion status of brain. Transcranial Doppler or laser Doppler flowmetry can measure blood flow velocity, which is just an surrogate of CBF. Computed tomography or magnetic resonance perfusion can measure regional CBF but these yield qualitative data and only represent the time at which the scan was performed. An accurate quantitative measurement is possible using positron emission tomography, single-photon emission computed tomography, or xenon computed tomography. However, these techniques only provide a snapshot of the perfusion status of brain, and a more continuous measurement of CBF is mandatory in the NeuroICU. A prototype of continuous CBF monitoring is the Hemedex Bowman Perfusion Monitor System (Cambridge, MA).29 The probe has 2 thermistors, one of which is heated to 2℃ higher than the measured brain temperature. Because the probe temperature is higher than the body temperature, if blood flow exists near the probe, relatively colder blood comes in and a change in the heat energy field develops.30 The thermal diffusion probe senses the difference between thermal energy fields and back calculates the actual CBF using a thermal energy transfer equation.29,31 The time resolution for the CBF measurement is 1 Hz, which is an ample resolution time for examining autoregulation (Figure 5). Another benefit of this monitor is that it can measure the thermal conductivity of brain tissue as it calibrates at user-defined intervals. Measured thermal conductivity can then be transformed into water content using a simple conversion equation. Therefore, the Hemedex monitor can be utilized to estimate real-time brain water content, which can be used as an indicator of brain edema around the probe.32

n tissue as it calibrates at user-defined intervals. Measured thermal conductivity can then be transformed into water content using a simple conversion equation. Therefore, the Hemedex monitor can be utilized to estimate real-time brain water content, which can be used as an indicator of brain edema around the probe.32 Continuous CBF measurement could be useful in patients with large ischemic stroke. Progression of the infarction is associated with gradual CBF reduction. The role of Hemedex as a trend monitor was validated in patients with SAH and progressing vasospasm.33 More studies to determine whether Hemedex is useful in detecting deteriorating hemodynamic status in patients having large ischemic stroke are needed. In Korea, Hemedex has been approved by the KFDA and is ready to use.

e role of Hemedex as a trend monitor was validated in patients with SAH and progressing vasospasm.33 More studies to determine whether Hemedex is useful in detecting deteriorating hemodynamic status in patients having large ischemic stroke are needed. In Korea, Hemedex has been approved by the KFDA and is ready to use. Microdialysis: Real time metabolic monitoring Continuous monitoring of tissue metabolism is possible using cerebral microdialysis. Microdialysis infuses lactate-free artificial cerebrospinal fluid (molecular concentration of: Na+ 148 mmol/L, Ca2+ 1.2 mmol/L, Mg2+ 0.9 mmol/L, K+ 2.7 mmol/L, Cl- 155 mmol/L) at a rate of 0.3 µL/min through a sterile infusion pump system.34 The molecules in the interstitial fluid then move across the microdialysis membrane and into the infusion fluid, where they reach equilibrium. Theoretically, any molecule can be measured depending on the pore size of the membrane. However, glucose, lactate, pyruvate, glutamate, and glycerol are most frequently measured in clinical settings.34 Glucose, lactate, and pyruvate are 3 key molecules in the glycolysis pathway; therefore, the changes in the concentration in these molecules are used to identify a shift in the glycolysis pathway. In patients under sedation, the average concentrations of glucose, lactate, and pyruvate are reported to be 1.7 mM, 2 mM, and 120 µM, respectively.34,35 Under normal conditions, glucose is converted to pyruvate after several steps, and then the NAD/NADH ratio determines whether it enters into the citric acid cycle or is transformed into lactate. Under anaerobic conditions, more lactate is produced, so the lactate/pyruvate ratio (LPR) surges. Under aerobic conditions, the average LPR value is around 15. However, when the patient is under metabolic distress, the LPR value starts to increase. A LPR value greater than 25 is considered as an early sign of metabolic distress, while a LPR less than 40 is associated with ongoing cell energy dysfunction and cellular metabolic crisis.36,37

, the average LPR value is around 15. However, when the patient is under metabolic distress, the LPR value starts to increase. A LPR value greater than 25 is considered as an early sign of metabolic distress, while a LPR less than 40 is associated with ongoing cell energy dysfunction and cellular metabolic crisis.36,37 A brain glucose level below 0.7 mM is regarded as a sign of brain tissue energy depletion.38 Because brain glucose concentration is associated with systemic glucose levels, low amounts of systemic glucose may lead to critically low levels of brain glucose (Figure 6). Moreover, abrupt drops (more than 25% reduction) in systemic glucose are independently linked to metabolic crisis, regardless of baseline peripheral glucose levels.39 Therefore, more modest control of brain glucose is required in critical care settings. Since brain glucose concentration is dependent on peripheral glucose concentration, if other conditions are stable, the level of brain glucose can be used to find a threshold for metabolically meaningful CPP in inidividual patients (Figure 7). In patients with ICH, a drop in CPP is associated with a gradual increase in the risk of a metabolic crisis. However, the degree is not strictly dependent upon the level of CPP. The reason for this difference in CPP dependency between metabolic crisis and BTH is not clear, but in patients with ICH, a mitochondrial disturbance may develop when the ICH occurs; thus it is thought to be less dependent on the level of CPP.8 More studies are needed to clarify this issue.

ndent upon the level of CPP. The reason for this difference in CPP dependency between metabolic crisis and BTH is not clear, but in patients with ICH, a mitochondrial disturbance may develop when the ICH occurs; thus it is thought to be less dependent on the level of CPP.8 More studies are needed to clarify this issue. Continuous electroencephalography The primary reason for electroencephalography (EEG) in the NeuroICU is to detect non-convulsive status epilepticus (NCSE). Previously, NCSE was regarded as a rare phenomenon; however, a recent study revealed that it is more frequently observed than expected.40 Seizures are detected on continuous EEG (cEEG) in as high as 10% of the patients with ischemic stroke. Patients with ICH have a higher seizure risk compared to patients with ischemia; the seizure mostly develop within the first 48 hours.41 Recurrent seizures may aggravate brain injury; seizures in patients with ICH are associated with ICP surges and midline shifts, which underscore the importance of early detection and management of seizures using cEEG monitoring.42 Continuous surface EEG monitoring is generally enough, but a study that used simultaneous surface and intracortical EEG monitoring found that depth electrodes identified more seizure activities, which were often recorded as rhythmic delta activity on surface EEG.43

on and management of seizures using cEEG monitoring.42 Continuous surface EEG monitoring is generally enough, but a study that used simultaneous surface and intracortical EEG monitoring found that depth electrodes identified more seizure activities, which were often recorded as rhythmic delta activity on surface EEG.43 In addition to the detection of ictal events, cEEG may detect hypoperfusion as hypoperfusion increases slow activity and leads to background attenuation in EEG.44 This concept has been used for detecting vasospasm and ischemic insult during carotid crossclamping in the carotid endarterectomy procedure. More quantitative indices of slowing (alpha/delta ratio) are utilized for detecting hypoperfusionin various NeuroICU settings, including the detection of vasospasm or progression of ischemia with large vessel steno-occlusion. The acute delta change index is reported to be correlated with the degree of perfusion in ischemic stroke.45

ive indices of slowing (alpha/delta ratio) are utilized for detecting hypoperfusionin various NeuroICU settings, including the detection of vasospasm or progression of ischemia with large vessel steno-occlusion. The acute delta change index is reported to be correlated with the degree of perfusion in ischemic stroke.45 Limitation One big limitation of MMM for patients with severe stroke is its invasiveness. Currently, a non-invasive but accurate monitoring technology does not exist. Moreover, the location of the probe itself is still not standardized. Probes are inserted into the brain tissues at high risk of SBI and therefore, mostly targeted at the perilesional or penumbra area. More studies are needed to accurately identify the proper probe location. We do not know whether additional computed tomography perfusion or magnetic resonance multimodal imaging may help better identify tissues at risk. Although ipsilateral and perilesional brain tissue is optimal for monitoring, patients frequently undergo hemicraniectomy or surgical procedures. Sometimes after such surgical procedures, probes are inserted and fixed to the contralateral side of the lesion. Probe insertion in the apparently healthy side may give false information on tissue health. When handling probes in the contralateral side, special precautions and interpretation are needed to better infer from the monitored information.

procedures, probes are inserted and fixed to the contralateral side of the lesion. Probe insertion in the apparently healthy side may give false information on tissue health. When handling probes in the contralateral side, special precautions and interpretation are needed to better infer from the monitored information. As stated above, most of the studies on MMM focused on different types of brain injuries such as TBI and SAH. However, MMM usability is expanding to include cases of large hemorrhagic or ischemic stroke, cardiac arrest, or status epilepticus. These conditions may share similar physiology in terms of an increase in ICP, on-going BTH, or hypoperfusion. However, more direct outcome studies are needed to establish this. Conclusion MMM enables us to better understand brain physiology and may help in patient-specific goal-directed therapy. Since each parameter may reflect only 1 aspect of brain physiology, more systematic integration of information on brain physiology is needed in order to understand the underlying mechanisms in brain damage. This study was supported in part by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A102065) and in part by a grant 04-2012-0950 from the SNUH Research Fund. The authors have no financial conflicts of interest.

Conclusion MMM enables us to better understand brain physiology and may help in patient-specific goal-directed therapy. Since each parameter may reflect only 1 aspect of brain physiology, more systematic integration of information on brain physiology is needed in order to understand the underlying mechanisms in brain damage. This study was supported in part by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A102065) and in part by a grant 04-2012-0950 from the SNUH Research Fund. The authors have no financial conflicts of interest. Figure 1 A representative figure showing the location of multimodality monitoring probes. Using a double or triple lumen cranial bolt system, multiple probes were fixed to the skull (A). Because the majority of the monitoring probes were radio-opaque, probe location was easily identified in scout images of pre-contrast brain computed tomography in the anteroposterior (B) and lateral views (C). Figure 2 Relationship between intracranial pressure and intracranial volume. As intracranial volume increases, a compensatory capacity limits abrupt surges of intracranial pressure; as brain compliance decreases, a small increase in intracranial volume results in a dramatic increase in the intracranial pressure. The fatigue point is usually located around 20 mmHg in the general population (dashed line).

ranial volume increases, a compensatory capacity limits abrupt surges of intracranial pressure; as brain compliance decreases, a small increase in intracranial volume results in a dramatic increase in the intracranial pressure. The fatigue point is usually located around 20 mmHg in the general population (dashed line). Figure 3 Relationship among the cerebral hemodynamic parameters. In cases with intact cerebral autoregulation, constant cerebral blood flow is maintained within autoregulating ranges of blood pressure. As the blood pressure drops, cerebral blood vessels need to dilate in order to maintain constant blood flow, which translates into surges in the intracranial pressure (A). In cases where autoregulation is disrupted, the capacity of the blood vessel is passively dependent on the perfusion pressure. Therefore, the correlation between intracranial pressure and blood pressure is linear (B). (This figure was modified from Crit Care Clin 2007;23:507-538 and Neurocrit Care 2004;1:289).

ure (A). In cases where autoregulation is disrupted, the capacity of the blood vessel is passively dependent on the perfusion pressure. Therefore, the correlation between intracranial pressure and blood pressure is linear (B). (This figure was modified from Crit Care Clin 2007;23:507-538 and Neurocrit Care 2004;1:289). Figure 4 Identifying optimal cerebral perfusion pressure. In patients with a loss of autoregulation, the relationship between intracranial pressure and mean arterial pressure appears to be linear. When the pressure reactivity index was plotted at a specific cerebral perfusion pressure, there was no nadir in the plot, suggesting no better spot for autoregulation (A). In patients with intact autoregulation, the mean pressure reactivity index values at specific cerebral perfusion pressure ranges were lower than those at other cerebral perfusion pressures, suggesting that the nadir existed in terms of the pressure reactivity-cerebral perfusion pressure plot and that the optimal cerebral perfusion pressure range existed (B). Figure 5 Correlation between cerebral blood flow and cerebral perfusion pressure. The scatter plot illustrates that a patient with intact autoregulation has CPP range between 70-100 mmHg. The red line represents locally weighted scatterplot smoothing regression line.

Figure 4 Identifying optimal cerebral perfusion pressure. In patients with a loss of autoregulation, the relationship between intracranial pressure and mean arterial pressure appears to be linear. When the pressure reactivity index was plotted at a specific cerebral perfusion pressure, there was no nadir in the plot, suggesting no better spot for autoregulation (A). In patients with intact autoregulation, the mean pressure reactivity index values at specific cerebral perfusion pressure ranges were lower than those at other cerebral perfusion pressures, suggesting that the nadir existed in terms of the pressure reactivity-cerebral perfusion pressure plot and that the optimal cerebral perfusion pressure range existed (B). Figure 5 Correlation between cerebral blood flow and cerebral perfusion pressure. The scatter plot illustrates that a patient with intact autoregulation has CPP range between 70-100 mmHg. The red line represents locally weighted scatterplot smoothing regression line. Figure 6 Results of microdialysis. Brain glucose and peripheral fingerstick glucose levels show adequate correlation, suggesting that the brain glucose level is passively dependent on peripheral glucose levels. Although intermittent continuous insulin infusion maintains peripheral glucose levels within an acceptable range (between 100-200 mg/dL), brain glucose levels frequently drop below a critical level (0.7 mM) with intensive glucose control. LPR, Lactate/pyruvate ratio; FSG, fingerstick glucose; A.U.,arbitrary unit.

Figure 6 Results of microdialysis. Brain glucose and peripheral fingerstick glucose levels show adequate correlation, suggesting that the brain glucose level is passively dependent on peripheral glucose levels. Although intermittent continuous insulin infusion maintains peripheral glucose levels within an acceptable range (between 100-200 mg/dL), brain glucose levels frequently drop below a critical level (0.7 mM) with intensive glucose control. LPR, Lactate/pyruvate ratio; FSG, fingerstick glucose; A.U.,arbitrary unit. Figure 7 Relationship between microdialysis results and cerebral perfusion pressure. As cerebral perfusion pressure drops, there is a modest surge in the brain lactate/pyruvate ratio, suggesting the brain glucose metabolism is cerebral perfusion pressure dependent. The red line represents locally weighted scatterplot smoothing regression line. Table 1 Variables assessed in multimodality monitoring KFDA, Korea Food and Drug administration; ICP, intracranial pressure; CPP, cerebral perfusion pressure; PRx, pressure reactivity index; CBF, cerebral blood flow; EEG: electroencephalography; CSA, compressed spectral array; LPR, lactate/pyruvate ratio.

Introduction Stroke is becoming an important cause of premature death and disability in low-income and middle-income countries like India, largely driven by demographic changes and enhanced by the increasing prevalence of the key modifiable risk factors. As a result developing countries are exposed to a double burden of both communicable and non-communicable diseases. The poor are increasingly affected by stroke, because of both the changing population exposures to risk factors and, most tragically, not being able to afford the high cost for stroke care. Majority of stroke survivors continue to live with disabilities, and the costs of on-going rehabilitation and long term-care are largely undertaken by family members, which impoverish their families.1,2 Studying the burden of stroke and the availability of health services will help the policy makers to tackle the rising burden of stroke. Recently there has been an increase in the epidemiology data on stroke from India. This review will address the changing burden of stroke and also the available stroke care services in India. Methods Papers were searched in the search engine PUBMED. The search terms used were 'Stroke in India', 'Stroke epidemiology', 'Stroke burden', 'Stroke and India', 'Stroke care in India', etc. Total number of papers listed was 1620. Out of this around fifty papers were short listed and reviewed by the authors.

Studying the burden of stroke and the availability of health services will help the policy makers to tackle the rising burden of stroke. Recently there has been an increase in the epidemiology data on stroke from India. This review will address the changing burden of stroke and also the available stroke care services in India. Methods Papers were searched in the search engine PUBMED. The search terms used were 'Stroke in India', 'Stroke epidemiology', 'Stroke burden', 'Stroke and India', 'Stroke care in India', etc. Total number of papers listed was 1620. Out of this around fifty papers were short listed and reviewed by the authors. Stroke Epidemiology during the pre-CT era Before the CT era, two population-based studies were undertaken in India. The first study was conducted in Vellore, Tamil Nadu, South India. This population-based study covering 258,576 people in and around Vellore was undertaken during the late 1960s and early 1970s.3 In the first phase (1968-1969), the number of hemiplegia cases was detected. In the second phase, this population was kept under surveillance for the next two years to record all cases of hemiplegia.3,4 This study revealed an incidence of 13/100,000 person-years and a point prevalence of 42/100,000. The second study was conducted at Rohtak, Haryana, North India (1971-1974). Eighty-two cases of stroke were recorded yielding an annual incidence of 33/100,000 person-years.5 Unfortunately, no incidence study was reported from India over the next 2 decades.

13/100,000 person-years and a point prevalence of 42/100,000. The second study was conducted at Rohtak, Haryana, North India (1971-1974). Eighty-two cases of stroke were recorded yielding an annual incidence of 33/100,000 person-years.5 Unfortunately, no incidence study was reported from India over the next 2 decades. Stroke Epidemiology between the 1990s and 2010s Prevalence A stroke study conducted in Kolkata6 from 1998 to 1999 showed a crude prevalence rate of 147/100,000 and an annual incidence rate of 36/100,000. When adjusted to the 1996 US population, the age-adjusted prevalence rate was 334/100,000 and the age-adjusted annual incidence rate was 105/100,000. Compared to men, women had substantially higher age-adjusted prevalence rate (564/100,000 for women versus 196/100,000 for men) and incidence rate (204/100,000 for women versus 36/100,000 for men). For all age groups except for people aged 50-69 years, women had a higher prevalence rate than did men. Among stroke patients who underwent neuroimaging study (59.5% of all strokes), 68% proved to be infarct and the remaining 32% to be haemorrhage. Among risk factors for stroke assessed by case-control study, hypertension was the most important risk factor: odds ratio (OR) for hypertension was 5.04 (95% confidence interval [CI] 4.16-5.92) for women and 21.87 (95% CI 18.69-25.05) for men. Smoking in men had the OR 2.91 (95% CI 1.57-4.25). However, the OR (95% CI) for diabetes was not significant for both women (0.99 [0.28-2.26]) and men (1.61 [0.17-3.05]). Since the smoking prevalence in women was very low, the authors suggested that hypertension being less managed in women than in men might account for the higher incidence and prevalence in women.6

5). However, the OR (95% CI) for diabetes was not significant for both women (0.99 [0.28-2.26]) and men (1.61 [0.17-3.05]). Since the smoking prevalence in women was very low, the authors suggested that hypertension being less managed in women than in men might account for the higher incidence and prevalence in women.6 The prevalence of stroke in India shows a huge variation of 147-922/100,000 across diverse community-based studies.4,7 In several studies which used age standardization with the US population as reference, the prevalence of stroke ranged from 244/100,000 to 424/100,000. According to the India stroke factsheet updated in 2012, the estimated age-adjusted prevalence rate for stroke ranges between 84/100,000 and 262/100,000 in rural and between 334/100,000 and 424/100,000 in urban areas.8 Incidence World Health Organization (WHO) steps approach In an effort to assist low-income and middle-income countries to establish surveillance systems for stroke, WHO recommended a stepwise approach (STEPS Stroke) through the use of standardized tools and methods for on-going core, expanded, and optional data collection.1 This system consists of three steps representing the possible outcomes of stroke patients in the hospital and the community.9 Step-1: The first step is gathering data from hospitalised patients such as demographic characteristics, whether it is the first ever or recurrent stroke, vital status at discharge, treatment during stay, risk factor assessment, classification of subtypes and follow up till discharge or death.

Incidence World Health Organization (WHO) steps approach In an effort to assist low-income and middle-income countries to establish surveillance systems for stroke, WHO recommended a stepwise approach (STEPS Stroke) through the use of standardized tools and methods for on-going core, expanded, and optional data collection.1 This system consists of three steps representing the possible outcomes of stroke patients in the hospital and the community.9 Step-1: The first step is gathering data from hospitalised patients such as demographic characteristics, whether it is the first ever or recurrent stroke, vital status at discharge, treatment during stay, risk factor assessment, classification of subtypes and follow up till discharge or death. Step-2: The second level of survey involves identifying and gathering information about the non-hospitalised fatal stroke cases in the community after proper validation from death certificates, verbal autopsy or from direct autopsies. Step-3: The third step represents non-fatal and non-hospitalized in the community and is the most complex level of stroke data collection.

Step-2: The second level of survey involves identifying and gathering information about the non-hospitalised fatal stroke cases in the community after proper validation from death certificates, verbal autopsy or from direct autopsies. Step-3: The third step represents non-fatal and non-hospitalized in the community and is the most complex level of stroke data collection. In this section we have summarised data of four population-based stroke epidemiology studies which were conducted according to the 'WHO-STEPS Stroke protocol' during the first decade of the 21st century in Mumbai,10 Trivandrum,11 Kolkata12 and Bangalore13 areas (Figure 1). In the Mumbai study, which were conducted during a 2-year study period from January 2005 to December 2006, the crude annual incidence rate of first-ever stroke in people aged 25 years or more was 145/100,000 person-years (age-standardized incidence rate, 152/100,000 person-years).10 In the Trivandrum study, which started its preparatory phase on January 2005 and completed its verification phase on August 2006, the crude annual incidence of any stroke was 117/100,000 person-years (age-standardized incidence rate, 135/100,000 person-years).11 In the Kolkata study using a door-to-door survey, the age-standardized incidence rate of first-ever-in-a-lifetime stroke was 145/100,000 person-years.12

verification phase on August 2006, the crude annual incidence of any stroke was 117/100,000 person-years (age-standardized incidence rate, 135/100,000 person-years).11 In the Kolkata study using a door-to-door survey, the age-standardized incidence rate of first-ever-in-a-lifetime stroke was 145/100,000 person-years.12 Age: It is assumed that the average age of patients with stroke in developing countries is usually 15 years younger than those in developed countries. In India, nearly one-fifth of patients with first-ever stroke admitted to hospitals has been estimated to be aged 40 years or less. But the Mumbai10 and Trivandrum11 registries showed that the mean age of patients with stroke was 66 and 67 years respectively. In contrast, in the Bangalore study the mean age was 54.5 years.13 In Trivandrum, stroke occurred at rate of 7.1 per 1000 per year in people aged ≥55 years, and the rate escalated to 13.3 in people aged ≥75 years (age-adjusted).11 The stroke in the young age group defined as 40 years or less comprised 3.8%. In this study, the mean age of stroke onset did not differ between the urban and rural populations.11

ccurred at rate of 7.1 per 1000 per year in people aged ≥55 years, and the rate escalated to 13.3 in people aged ≥75 years (age-adjusted).11 The stroke in the young age group defined as 40 years or less comprised 3.8%. In this study, the mean age of stroke onset did not differ between the urban and rural populations.11 Gender: In the Mumbai registry, men had a higher stroke incidence rate than did women (crude incidence rate, 149/100,000 person-years for men versus 141/100,000 person-years for women; age-standardized incidence rate, 162/100,000 person-years for men versus 141/100,000 person-years for women).10 Women were older (68.9 years) compared to men (63.4 years).10 In the Trivandrum registry, the crude incidence rate was higher in women than in men (115/100,000 person-years for men and 119/100,000 person-years for women), but the age-standardized incidence rate was higher in men than in women (143/100,000 person-years for men and 128/100,000 person-years for women).11 The Bangalore study also showed a greater preponderance among men (67%) with a male to female ratio of 2:1. The observed difference between age and gender and occurrence of stroke was statistically significant (P<0.01).13

in men than in women (143/100,000 person-years for men and 128/100,000 person-years for women).11 The Bangalore study also showed a greater preponderance among men (67%) with a male to female ratio of 2:1. The observed difference between age and gender and occurrence of stroke was statistically significant (P<0.01).13 Stroke subtypes: Of patients with first-ever stroke captured in the Mumbai registry, CT imaging was done in 89.2%, and 80.2% were ischemic strokes and 17.7% hemorrhagic strokes (Figure 2).10 In the Trivandrum registry, 69.7% of patients underwent imaging. Of those, 83.6% were ischemic strokes, 11.6% intracerebral hemorrhages, and 4.8% subarachnoid haemorrhages, respectively.11 There were more strokes of undetermined type in patients enrolled from the rural communities because of a lack of neuroimaging information (31.2%).11 In the Kolkata study, 32% of the patients had hemorrhagic stroke, which is the highest figure reported so far from India.12

.8% subarachnoid haemorrhages, respectively.11 There were more strokes of undetermined type in patients enrolled from the rural communities because of a lack of neuroimaging information (31.2%).11 In the Kolkata study, 32% of the patients had hemorrhagic stroke, which is the highest figure reported so far from India.12 Risk factors: It has been estimated that hypertension causes 54% of stroke in low-income and middle-income countries, followed by hypercholesterolemia (15%) and tobacco smoking (12%).14 In the Mumbai registry, 82.8% of patients had hypertension. However, verifiable data for other risk factors were not available.10 In the Trivandrum registry, nearly 85% had hypertension, half had diabetes mellitus, 26% had dyslipidemia and 26.8% of men smoked tobacco. Compared to urban males, more rural males smoked tobacco (22.8% vs. 39.3%, P=0.013). One risk factor was present in 38.4% patients, two in 42.0%, and three or more in 14.4% patients.11

the Trivandrum registry, nearly 85% had hypertension, half had diabetes mellitus, 26% had dyslipidemia and 26.8% of men smoked tobacco. Compared to urban males, more rural males smoked tobacco (22.8% vs. 39.3%, P=0.013). One risk factor was present in 38.4% patients, two in 42.0%, and three or more in 14.4% patients.11 Urban vs Rural: Of the 541 validated first-ever strokes in Trivandrum, 431 occurred in the urban community and 110 occurred in the rural community. The annual stroke incidence rate was slightly higher in the rural population than in the urban population (crude incidence rate, 116/100,000 person-years for the urban population versus 119/100,000 person-years for the rural population; age-standardized incidence rate, 135/100,000 person-years for the urban population versus 138/100,000 person-years for the rural population) (Table 1).11 It also showed that the number of smokers (men) and presence of multiple risk factors (more than 3) were significantly more in rural population than in urban population. Also the distribution of conventional stroke risk factors was remarkably similar among the urban and rural communities. However number of stroke patients who had imaging was significantly low in rural population.11 Studies from India on cardiovascular risk factors have shown a 2 to 3 time's high prevalence of hypertension, hyperlipidemia, obesity, diabetes mellitus, and smoking (in men) in urban compared to rural communities.15,16

ties. However number of stroke patients who had imaging was significantly low in rural population.11 Studies from India on cardiovascular risk factors have shown a 2 to 3 time's high prevalence of hypertension, hyperlipidemia, obesity, diabetes mellitus, and smoking (in men) in urban compared to rural communities.15,16 Case fatality rate In the Mumbai study, the 28-day case fatality rate was 29.8% (Figure 3).10 The case fatality rate was 24.5% for urban and 37.1% for rural population (overall 27.2%) in Trivandrum.11 Significantly more rural patients compared to urban patients died within 1 month, which probably reflects the disparity in the quality of acute stroke care between the rural and urban areas.11 The case fatality rate was 42% in the Kolkata study.12 Disability In the Trivandrum study, the functional outcome was available in 86.8% among stroke survivors.11 The distributions of functional outcomes at 28 days among stroke survivors were mild disability (Rankin score 2 or less) in 42.4%, moderate disability (Rankin score 3 or 4) in 43%, and bedridden (Rankin score 5) in 14.6%, respectively (Figure 4). Among stroke survivors, there was no significant difference in the functional outcome between men and women (P=0.17), between urban and rural patients (P=0.51), or across different age groups (P=0.52).11

moderate disability (Rankin score 3 or 4) in 43%, and bedridden (Rankin score 5) in 14.6%, respectively (Figure 4). Among stroke survivors, there was no significant difference in the functional outcome between men and women (P=0.17), between urban and rural patients (P=0.51), or across different age groups (P=0.52).11 Stroke in Ludhiana city In India the incidence studies on stroke have been done only in metropolitan cities so far. The Indian Council of Medical Research (ICMR) started a Task Force Project to develop population-based stroke registries across the country. As a pilot project, a population-based registry was conducted in Ludhiana city, Punjab. The crude annual incidence rate was found to be 140/100,000 person-years (95% CI: 132.90 to 147.09) and the age-adjusted incidence rate was 181.67/100,000 person-years (95% CI: 172.44 to 190.89) (Unpublished data). Good outcome (defined as modified Rankin Scale ≤2) was seen in 58% of patients at 28 days. Case fatality rate at 28 days was 21.4%. For stroke subtypes, ischemic stroke accounted for 73%, intracerebral haemorrhage 21.7%, and subarachnoid haemorrhage 4% (Figure 5).

on-years (95% CI: 172.44 to 190.89) (Unpublished data). Good outcome (defined as modified Rankin Scale ≤2) was seen in 58% of patients at 28 days. Case fatality rate at 28 days was 21.4%. For stroke subtypes, ischemic stroke accounted for 73%, intracerebral haemorrhage 21.7%, and subarachnoid haemorrhage 4% (Figure 5). Stroke care services in India Stroke units Stroke unit is a multidisciplinary team comprising of medical, nursing, physiotherapy, occupational therapy, speech therapy, and social-work staff who coordinate their work through regular meetings (Figure 6). These meetings introduce the patients to the team and provide a forum for multidisciplinary assessment, identification of problems, and setting of short-term and long-term recovery goals. Stroke units typically include early active involvement of carers in the rehabilitation process and usually have a programme of on-going education and training.17 Two previous analyses that explored the potential population effect of different interventions for patients with stroke suggest that a basic model of stroke-unit care could provide the most effective population intervention.18

the rehabilitation process and usually have a programme of on-going education and training.17 Two previous analyses that explored the potential population effect of different interventions for patients with stroke suggest that a basic model of stroke-unit care could provide the most effective population intervention.18 Stroke unit implementation remains a big challenge in India. At present in India there are approximately 35 stroke units and they are predominantly in private sector hospitals in the cities. Many of the private hospitals lack CT scan facility and they refer for incentives to other hospitals or centres. This results in crucial time being lost. On the other hand, public hospitals lack a dedicated team and place to manage stroke patients. Unavailability of CT scan is also an issue in smaller public hospitals as in private hospitals. According to the Mumbai registry study, only 306 of 456 (67.2%) patients with first-ever stroke were managed at health-care facility ('in-hospital') and the remaining 150 (32.8%) patients were cared at home or in nursing homes.10 This indicates that one out of every 3 patients with stroke are not accessing appropriate healthcare probably due to non-affordability, usage of alternative medicines, and difficulty in conveyance.

d at health-care facility ('in-hospital') and the remaining 150 (32.8%) patients were cared at home or in nursing homes.10 This indicates that one out of every 3 patients with stroke are not accessing appropriate healthcare probably due to non-affordability, usage of alternative medicines, and difficulty in conveyance. Needs and Barriers We are facing many obstacles for implementing appropriate stroke care service in India. Stroke care units need a geographical base, a dedicated geographically defined hospital area with dedicated beds and nursing staff, though the need for expensive high-dependency facilities is less.18 An adequate number of skilled staff is an essential requirement of stroke-unit care and various skills will be needed to provide comprehensive care. In settings where skilled staff is few, options might include supplementary training for nursing staff, and training for family members and support workers. Constant medical supervision for early detection of post-stroke complication by nurses or trained stroke care physician is necessary for effective functioning of a stroke care unit which again demands a relatively bigger group of personnel to be involved. Last but not least is the requirement of adequate equipment and medication such as adequate bedding and seating, and equipment to assist with monitoring and fluid management. Some basic drugs are needed, particularly aspirin and antihypertensive drugs. Furthermore, generic drugs for lipid lowering and for diabetes management are important. These drugs which are already in the WHO list are yet to be made evenly available in India.

and equipment to assist with monitoring and fluid management. Some basic drugs are needed, particularly aspirin and antihypertensive drugs. Furthermore, generic drugs for lipid lowering and for diabetes management are important. These drugs which are already in the WHO list are yet to be made evenly available in India. Thrombolysis In India thrombolysis for stroke is being widely used across the country both in private and public sector hospitals. Compared to 2007, there has been an increase in the use of intravenous rtPA in the country (Figure 7).19,20 Even in the same center an increasing trend has been observed (Figure 8).21 The proportion of patients receiving rtPA is still low in our country. Among 967 patients enrolled in the on-going Indo-USA Collaborative National Stroke Registry, 134 patients came within 4.5 hours and 104 (11%) patients received rtPA. Intraarterial and mechanical thrombolysis was given in 34 (3.5%) patients. At present in India there are approximately 100 centres which are able to provide intravenous rtPA treatment and 55 centres capable of performing intraarterial or mechanical thrombolysis (Unpublished)

hours and 104 (11%) patients received rtPA. Intraarterial and mechanical thrombolysis was given in 34 (3.5%) patients. At present in India there are approximately 100 centres which are able to provide intravenous rtPA treatment and 55 centres capable of performing intraarterial or mechanical thrombolysis (Unpublished) Rehabilitation In India the rehabilitation is mainly centred on physiotherapists. There are 32800 physiotherapists registered in Indian Association of physiotherapists till 2011, 3000 occupational therapists registered in Indian Association of Occupational therapists till 2011, and currently, 1700 speech therapists registered in Indian Association of speech therapists and 500 speech therapists registered in Rehabilitation Council of India. However, very few centers have an organized in-hospital and outpatient rehabilitation facilities in the country. Moreover use of complementary and alternative medicine treatments by the patients during the post stroke phase hampers the rehabilitation process.22

therapists registered in Rehabilitation Council of India. However, very few centers have an organized in-hospital and outpatient rehabilitation facilities in the country. Moreover use of complementary and alternative medicine treatments by the patients during the post stroke phase hampers the rehabilitation process.22 Secondary prevention drugs Early initiation of treatments for secondary stroke prevention is associated with an 80% reduction in risk of early recurrent stroke.23 A standard secondary stroke prevention treatment will address multiple vascular risk factors and will usually consist of an antiplatelet agent, a lipid lowering drug mainly HMG-CoA reductase inhibitors (statins), and an antihypertensive agent. Monitoring of oral anticoagulant therapy is a major problem for cardioembolic stroke in India. The monitoring facilities are available mainly in the hospitals and the laboratories in the city. Newer oral aniticoagulants are used mainly in rich patients who could afford. Aspirin, clopidogrel and dipyridamole are the most commonly used antiplatelet drugs. Statins are also used with antihypertensive or antidiabetic drugs if indicated. India is the home of a thriving generic drug pharmaceutical industry. It is a thing to ponder on why poor people cannot afford modern therapies even where facilities are available. Even aspirin, although relatively cheap and readily available, is not routinely administered to patients with ischemic stroke or transient ischemic attack in India.

riving generic drug pharmaceutical industry. It is a thing to ponder on why poor people cannot afford modern therapies even where facilities are available. Even aspirin, although relatively cheap and readily available, is not routinely administered to patients with ischemic stroke or transient ischemic attack in India. Conclusion India like other developing countries is in the midst of a stroke epidemic. There is a huge burden of stroke with significant regional variations. Stroke units, thrombolysis, and rehabilitation are predominantly available in urban areas, particularly in private sector hospitals. As a first step, the Government of India has started the National Programme for Prevention and Control of Cancer, Diabetes, Cardiovascular Diseases & Stroke (NPCDCS). The government is focusing on early diagnosis, management, infrastructure, public awareness, and capacity building at different levels of health care for all the non-communicable diseases including stroke. An organised effort from both the government and the private sector is needed to tackle the rising stroke burden in India. We sincerely thank Dr Benedict Joshua for his valuable contributions and Mr George Koshy final year medical student for his valuable help in the preparation of the Figure 1. The authors have no financial conflicts of interest. Figure 1 Map of India showing the different places where stroke incidence study has been done. Figure 2 Distribution of stroke subtypes in the various incidence studies. Figure 3 28-day case fatality rates in hospital and community from the Mumbai and Bangalore registries.

We sincerely thank Dr Benedict Joshua for his valuable contributions and Mr George Koshy final year medical student for his valuable help in the preparation of the Figure 1. The authors have no financial conflicts of interest. Figure 1 Map of India showing the different places where stroke incidence study has been done. Figure 2 Distribution of stroke subtypes in the various incidence studies. Figure 3 28-day case fatality rates in hospital and community from the Mumbai and Bangalore registries. Figure 4 Distribution of post-stroke disability at 28 days from the onset of stroke among stroke survivors. Figure 5 Stroke subtypes in the population-based Ludhiana stroke registry. Figure 6 Components of stroke care unit. Figure 7 Nationwide data of numbers of patients treated with intravenous rtPA during 2009-2011. Figure 8 Secular trend of intravenous rt PA treatment in Ludhiana city over two study periods. Table 1 Data from Trivandrum stroke registry showing urban and rural distribution

Introduction Treatment of stroke requires specialized processes that are expediently provided. A multidisciplinary team approach in hospital and effective communication between the stroke team and the emergency medical system (EMS) are prerequisites to a good outcome for patients with acute stroke.1 Multiple specialized processes in stroke management must be streamlined. Information technology plays a key role in modern, complex healthcare. The heavy demands of stroke care could be mitigated to a degree by using information technology. Information technology can reduce the rate of errors, improve communication, make information more readily accessible, assist in diagnosis and monitoring, provide decision support, and enhance implementation of guidelines and recommendations.2

stroke care could be mitigated to a degree by using information technology. Information technology can reduce the rate of errors, improve communication, make information more readily accessible, assist in diagnosis and monitoring, provide decision support, and enhance implementation of guidelines and recommendations.2 Health-related use of the internet is increasing. As of May 2013, the Pew Internet and American Life Project showed that in the United States, 85% of adults were using the internet, 91% of adults owned a cell phone, 56% of adults owned a smartphone, and 34% owned a tablet device.3 Fifty-two percent of smartphone owners have used their phones to look up health or medical information, and 19% of smartphone owners have downloaded an application specifically to track or manage health. In South Korea, as of July 2012, 78.4% of Koreans were using the internet, 82.3% had a desktop computer, and 63.7% had a smart device (smartphone or tablet computer). Although the internet-use rate has been reported as lower in the elderly population (38.5% of those aged 50-60 years and 9.7% of those aged 60 years or older) than in the younger population, ownership of smart devices is increasing in the former population (46.8% of those 50-60 years old and 23.4% of those aged 60 years or older).4

internet-use rate has been reported as lower in the elderly population (38.5% of those aged 50-60 years and 9.7% of those aged 60 years or older) than in the younger population, ownership of smart devices is increasing in the former population (46.8% of those 50-60 years old and 23.4% of those aged 60 years or older).4 The World Health Organization has recently defined mobile health as "medical and public health practice supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistants, and other wireless devices".5 It is anticipated that about 500 million smartphone users around the world will use mobile health services by 2015.6 Mobile devices are increasingly being used to diagnose and treat various diseases. Current mobile devices are particularly suited for medical purposes because of their rich multi-touch user interfaces, built-in accelerometers, location-sensing frameworks, fast processors, and widely available network connections. Additional benefits of mobile devices are that they can be carried easily and can be used anytime and anywhere.7

re particularly suited for medical purposes because of their rich multi-touch user interfaces, built-in accelerometers, location-sensing frameworks, fast processors, and widely available network connections. Additional benefits of mobile devices are that they can be carried easily and can be used anytime and anywhere.7 Here, we address the contribution of information technology and mobile health in the clinical management pathway of stroke patients, particularly regarding their use in the recognition of stroke, transport and triage of stroke patients, emergent stroke evaluation in the hospital, and rehabilitation. We also briefly review the applications of such technology for healthcare providers and patients. However, our intention in this review is not to address in depth all of the information technology issues relating to electronic health records and wearable, ambient monitoring sensors. Recognition of stroke symptoms Being aware of stroke warning signs and responding to stroke (calling the emergency services promptly) are two important points in educating the public about stroke. Data show that the public's knowledge of stroke warning signs remains poor worldwide.1 Stroke awareness is also suboptimal in South Korea.8 Several studies demonstrated that use of the EMS was associated with earlier arrival and enabled faster imaging studies.9,10 However, a recent nationwide Korean survey showed that only 33% of respondents chose the proper action of calling the EMS if they exhibited a stroke warning sign.8

lso suboptimal in South Korea.8 Several studies demonstrated that use of the EMS was associated with earlier arrival and enabled faster imaging studies.9,10 However, a recent nationwide Korean survey showed that only 33% of respondents chose the proper action of calling the EMS if they exhibited a stroke warning sign.8 The American Heart Association/American Stroke Association guidelines state that activation of the 911 system is strongly recommended (Class I; Level of Evidence B).1 We developed the "stroke 119" (119 being the Korean equivalent to 911) application that provides rapid self-screening for stroke, identifies nearby hospitals that provide thrombolytic treatment, and facilitates calling the EMS.11 This stroke screening tool was adopted from the Cincinnati Prehospital Stroke Scale12 and is displayed in a cartoon format. The multi-touch user interfaces of smartphones enable easy and interactive delivery of information to users. By tapping the cartoon image representing the presence of neurological deficits, the user is informed that he or she may be having a stroke and is given instructions to call the EMS. A global positioning system, which is a built-in function of smartphones, enables users to determine their location and search for nearby hospitals. This helps stroke victims, witnesses, and EMS personnel find nearby hospitals that provide acute stroke care in real time. Thus, the stroke 119 application is useful for reducing hospital arrival times for thrombolytic patients. Similarly, the American Stroke Association recently developed the "Spot a Stroke F.A.S.T." application. The F.A.S.T. provides users with an easy way to remember the sudden signs of stroke and to find a nearby hospital.13 Wide adoption of applications that improve recognition of stroke symptoms might be associated with getting stroke victims to the hospital more quickly.

developed the "Spot a Stroke F.A.S.T." application. The F.A.S.T. provides users with an easy way to remember the sudden signs of stroke and to find a nearby hospital.13 Wide adoption of applications that improve recognition of stroke symptoms might be associated with getting stroke victims to the hospital more quickly. Transport and triage of stroke patients Remote assessment of stroke patients Telemedicine is the use of information and communication technology to provide healthcare services to individuals who are at a distance from the healthcare provider.14 Patients in remote areas who suffer from stroke can gain access to timely expert care by stroke specialists via telemedicine.15 This may be particularly useful in underserved or rural areas. Because telestroke consultations require high-speed, clear, and reliable audio-video data transmission to ensure adequate neurological evaluation, typical telemedicine systems are expensive.15 With recent advances in one-way/two-way video and teleradiology features adapted to smartphones, several investigators have reported the efficacy of ubiquitous and relatively cheap mobile-device-based telemedicine.16,17 Mobile telemedicine systems can be freely used while at work and home as well as while traveling via a wireless internet service and a smartphone.

and teleradiology features adapted to smartphones, several investigators have reported the efficacy of ubiquitous and relatively cheap mobile-device-based telemedicine.16,17 Mobile telemedicine systems can be freely used while at work and home as well as while traveling via a wireless internet service and a smartphone. National Institutes of Health Stroke Scale telestroke examination Because mobile devices can deliver streaming video and audio, they can be used to evaluate acute stroke patients remotely.18 Comparison between the face-to-face method and the mobile telemedicine method of assessment using the National Institutes of Health Stroke Scale (NIHSS) showed high inter-method agreement according to the correlations in total NIHSS scores between the methods (r=0.94 to 0.98, P<0.001) and an acceptable length of remote assessment time (3.38 to 11.4 minutes).15 Therefore, current guidelines state that the NIHSS telestroke examination, when administered by a stroke specialist using high-quality videoconferencing, is recommended when an NIHSS bedside assessment by a stroke specialist is not immediately available for patients with acute stroke (Class I; Level of Evidence A).19

5 Therefore, current guidelines state that the NIHSS telestroke examination, when administered by a stroke specialist using high-quality videoconferencing, is recommended when an NIHSS bedside assessment by a stroke specialist is not immediately available for patients with acute stroke (Class I; Level of Evidence A).19 Teleradiology Teleradiology is a critical aspect of telestroke assessment and is defined as obtaining radiographic images at one location and transmitting them to another for diagnostic and consultative purposes.20 Several investigators have shown that the smartphone-based client-server teleradiology system may have the potential to allow urgent management decisions in acute stroke.21,22 Using a telestroke network, the agreement on identification of contraindications for thrombolytic treatments in noncontrast CT was excellent between vascular neurologists who used a smartphone application and radiologists who used desktops in a hospital.16 Recent guidelines indicate that the teleradiology systems approved by the Food and Drug Administration or equivalent organization are useful in supporting rapid imaging interpretation in time for making decisions on thrombolytic treatment when implemented within a telestroke network (Class I; Level of Evidence B).1

.16 Recent guidelines indicate that the teleradiology systems approved by the Food and Drug Administration or equivalent organization are useful in supporting rapid imaging interpretation in time for making decisions on thrombolytic treatment when implemented within a telestroke network (Class I; Level of Evidence B).1 Prehospital notification system Previous studies showed that only 15-60% of patients with stroke arrived at the hospital within three hours after the onset of symptoms, and only 14-48% arrived within two hours.23 Moreover, the proportion of patients with ischemic stroke who arrived within two hours of symptom onset did not increase significantly from 2003 to 2009.24

showed that only 15-60% of patients with stroke arrived at the hospital within three hours after the onset of symptoms, and only 14-48% arrived within two hours.23 Moreover, the proportion of patients with ischemic stroke who arrived within two hours of symptom onset did not increase significantly from 2003 to 2009.24 Prenotification of a stroke patient's arrival by EMS personnel can shorten the time before the patient is seen for initial evaluation by a physician in the emergency department, shorten the time to brain imaging, and increase the use of IV recombinant tissue plasminogen activator (rt-PA).10,25,26 Thus, prehospital notification is recommended in current guidelines in the United States and Europe.1,27 EMS prenotification, including time of symptom onset or potential contraindications for rt-PA, facilitates earlier activation of a stroke team and the preparation of imaging modalities before the patient arrives.23 The Get With the Guidelines-Stroke program found that, compared with patients who arrived without prenotification, patients who arrived with prenotification had shorter door-to-imaging times, door-to-needle times, symptom-onset-to-needle times, and a greater likelihood of treatment with rt-PA within three hours. Despite these benefits, EMS prenotification occurred in only two-thirds of patients with acute ischemic stroke arriving by EMS.28

who arrived with prenotification had shorter door-to-imaging times, door-to-needle times, symptom-onset-to-needle times, and a greater likelihood of treatment with rt-PA within three hours. Despite these benefits, EMS prenotification occurred in only two-thirds of patients with acute ischemic stroke arriving by EMS.28 In contrast to telemedical communication between two hospitals, telemedical transmission from the ambulance in the field to the stroke center is still a challenge. German researchers tested the feasibility of telemedicine-equipped ambulances. Among 18 of the 30 scenarios the NIHSS assessment could not be performed due to absence or loss of the audio-video signal. Moreover, the remaining 12 completed scenarios did not meet an acceptable level for clinical use.29 In the actual stroke patients, telemedically assisted prehospital care and regular EMS care were compared, but the researchers found that system dropouts occurred in three out of 18 missions. Moreover, there was no decrease in time intervals or improvement in diagnostic accuracy.30 Thus, although prehospital consultation has the potential to improve emergency care, especially when no highly trained personnel are on-scene,31 insufficient reliability of telemedicine technologies resulting from temporary dropouts and local unavailability of mobile networks is still an unresolved problem that limits their use in clinical practice. Therefore, further technical development is needed in prehospital notification systems. As an intermediate step, communication using personal digital assistants or smartphones was tested.32,33 Although many developing countries do not have widespread or even dependable broadband internet access, much of their population can access the internet using cell phones.5 In this respect, prenotification using mobile devices, especially where the internet connection is limited, might be promising.

es was tested.32,33 Although many developing countries do not have widespread or even dependable broadband internet access, much of their population can access the internet using cell phones.5 In this respect, prenotification using mobile devices, especially where the internet connection is limited, might be promising. Mobile stroke unit The mobile stroke unit is equipped with a CT scanner, a point-of-care laboratory, and a telemedicine connection to the hospital. A randomized trial of the mobile stroke unit strategy demonstrated that its prehospital stroke treatment could reduce the median time from alarm to treatment decision from 76 to 35 minutes (median difference of 41 minutes, P<0.0001). The mobile stroke unit strategy seems to bring guideline-adherent stroke treatment directly to the emergency site, thus speeding up acute stroke management considerably.34 Emergent stroke evaluation in the hospital In patients eligible for IV rt-PA, the benefit of treatment is time-dependent, and treatment should be initiated as quickly as possible.35 For every 15-minute reduction in door-to-needle time, the odds of risk-adjusted in-hospital mortality were reduced by 5%.36 The door-to-needle time should be within 60 minutes from hospital arrival (Class I; Level of Evidence A).1 Nevertheless, The Get With The Guidelines-Stroke data showed that only 26.6% of patients had a door-to-needle time within 60 minutes among the 25,504 acute ischemic stroke patients treated at 1,082 hospital sites in the United States.36

should be within 60 minutes from hospital arrival (Class I; Level of Evidence A).1 Nevertheless, The Get With The Guidelines-Stroke data showed that only 26.6% of patients had a door-to-needle time within 60 minutes among the 25,504 acute ischemic stroke patients treated at 1,082 hospital sites in the United States.36 Within hospitals, established protocols are beneficial to optimize medical processes. Incorporation of protocols into a computerized physician-order entry can enhance communication among team members.37 Shortening door-to-needle time requires rapid triage, notification, and action by many medical staff members in different departments, including the emergency, neurology, nursing, neuroimaging, and laboratory departments. Notification by phone to different medical staff members and technicians takes time. Single-call systems to activate all stroke-team members can reduce door-to-needle times in patients with acute ischemic stroke.38 Using information technology rather than a pager can be more effective for communication between stroke-team members because notification can be done simultaneously. It is also possible to use a predetermined order set and monitor time logs by easily acquiring time data for individual evaluation and treatment steps through the use of information technology. A computerized in-hospital alert system, which uses a computerized physician order entry, was developed and shown to be effective in a single medical center (Table 1).39 The effect of this system was also demonstrated in a multicenter study with various hospital settings.40 A computerized in-hospital alert system may be the impetus for healthcare providers to improve multidisciplinary care delivery and patient care quality.41

hown to be effective in a single medical center (Table 1).39 The effect of this system was also demonstrated in a multicenter study with various hospital settings.40 A computerized in-hospital alert system may be the impetus for healthcare providers to improve multidisciplinary care delivery and patient care quality.41 Mobile devices as supplementary tools for neurologic examination and decision making A mobile-device-supported neurological examination Portable mobile devices have been used to easily administer neurological examinations, including measuring visual acuity and color vision.42 Because mobile devices have rich multitouch user interfaces and built-in accelerometers,42 these features might be useful for assessing patients' neurologic deficits. The pronator drift test is a neurologic examination used to evaluate mild arm weakness. The presence or absence of pronation and drift are determined qualitatively. Several studies have demonstrated that accelerometers are reliable tools for quantifying physical activity and walking speed after stroke.43,44 We developed the "iPronator" application to objectify the pronator drift test using a built-in accelerometer on mobile devices.45,46 The iPronator measures changes in the x- and y-axes that correspond to pronation and drift, respectively. A Bluetooth connection between mobile devices on each arm enables data transfer in real time (Figure 1). We found that the parameters of pronation (average and maximum) and drift (average, maximum and oscillation) were significantly different between patients and healthy controls. The iPronator application was also useful for determining degrees of improvement in patients with arm weakness by providing quantitative data for serial tests.

e parameters of pronation (average and maximum) and drift (average, maximum and oscillation) were significantly different between patients and healthy controls. The iPronator application was also useful for determining degrees of improvement in patients with arm weakness by providing quantitative data for serial tests. Decision making support for stroke classification Reports indicate that physicians have difficulty processing complex information.47 The use of information technology can support decision making by providing installation of preprogrammed algorithms and an easily accessible user interface. Classification of stroke subtypes is based on a logical algorithm. Therefore, information technology can be effectively used for accurate classification of stroke subtypes. The Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification is a widely used classification system of ischemic stroke that primarily utilizes the etiology of strokes for classification.48 However, the inter-rater reliability of the TOAST classification is known to be modest. To overcome the limitations of TOAST classification, several computerized classification systems have been developed. Goldstein et al. reported that a computerized algorithm and standardization of procedures for extracting medical records could improve both intraobserver and interobserver agreement.49 The Causative Classification System50 and the Atherosclerosis-Small Vessel Disease-Cardiac Source-Other Cause classification system showed good-to-excellent agreement in comparison with the original TOAST scheme.51

procedures for extracting medical records could improve both intraobserver and interobserver agreement.49 The Causative Classification System50 and the Atherosclerosis-Small Vessel Disease-Cardiac Source-Other Cause classification system showed good-to-excellent agreement in comparison with the original TOAST scheme.51 Although these computerized classification systems showed excellent agreement among raters, they usually run through desktop internet web pages, which may limit their wide use in practice. Thus, we developed a computerized clinical-decision support system running on mobile devices and named it "iTOAST".7,52 A logical algorithm of the TOAST classification system was analyzed and implemented (Figure 2). After answering questions using the touch interface, the results of the stroke subtype are displayed on the screen. We found that the inter-rater agreement was higher when the raters used iTOAST (0.790, 95% CI: 0.707-0.870) than the conventional method (0.692 , 95% CI: 0.600-0.782; P<0.001). Thus, the iTOAST was easier, more accurate, and more reliable than the conventional method, and it has the additional benefit of accessibility because it runs on a mobile device. Decision-making support in stroke management is also promising for emergent situations. A pilot application of "i-Stroke" had successful information transfer, allowing medical staff to discuss stroke patients' diagnosis and management.17

it has the additional benefit of accessibility because it runs on a mobile device. Decision-making support in stroke management is also promising for emergent situations. A pilot application of "i-Stroke" had successful information transfer, allowing medical staff to discuss stroke patients' diagnosis and management.17 Rehabilitation Virtual reality Virtual reality (VR) is a new type of rehabilitation therapy that can be used with handicapped patients with stroke. Typically, VR systems include computer-based programs designed to simulate real-life objects and events.53 VR systems give sensory feedback to patients, in most cases visual or audio feedback, which can improve patients' cognitive and motor functions. Several studies have shown that VR is beneficial to arm function,54,55 walking,56 and neglect.57,58 However, the Cochrane review stated that there was insufficient evidence for any conclusion on the effectiveness of VR on grip strength or gait speed. It is also unclear at present which characteristics of VR are most important, and it is unknown whether the effects can be sustained in the long term.53 However, VR may be advantageous because it offers a number of unique features. For instance, goal-oriented tasks and repetition of said tasks are important in neurological rehabilitation; VR can encourage repetition of tasks by providing simulated real-life functional activities, which are more interesting and enjoyable for both children and adults. Moreover, activities that are unsafe to practice in the real world, such as crossing the street, driving, or preparing food, can be trained with VR. Furthermore, tasks can be graded and immediate feedback can be provided. Many VR programs are designed to be used without supervision.53

able for both children and adults. Moreover, activities that are unsafe to practice in the real world, such as crossing the street, driving, or preparing food, can be trained with VR. Furthermore, tasks can be graded and immediate feedback can be provided. Many VR programs are designed to be used without supervision.53 Telerehabilitation Telerehabilitation can offer prolonged rehabilitation for patients with stroke after they are discharged from the hospital.59 Telerehabilitation allows the patient to maintain contact with medical professionals, doctors, and therapists in a comfortable home environment, while also enabling continuation of long-term rehabilitation to promote functional abilitites. Furthermore, telediagnostics enable transmission of messages and monitoring of patient outcomes and functional progress, as well as changes in the rehabilitation strategy and planning of outpatient visits. Numerous studies clearly show that telerehabilitation improves functional performance and quality of life as well as increases patient satisfaction.60,61 In the near future, telerehabilitation and VR may be among the major advanced technologies in post-stroke rehabilitation for the assessment and management of patients.

umerous studies clearly show that telerehabilitation improves functional performance and quality of life as well as increases patient satisfaction.60,61 In the near future, telerehabilitation and VR may be among the major advanced technologies in post-stroke rehabilitation for the assessment and management of patients. Delivery of medical information to healthcare providers and the general population Applications for healthcare providers Transient symptoms of stroke patients can be easily captured by caregivers or healthcare professionals onsite and then reported to stroke experts using mobile devices. Furthermore, smartphones are used to calculate medical formulas, find drug references, search literature, and access medical education/training. Healthcare providers can also use smartphones in the diagnosis and objectification of neurological symptoms45 and to obtain support for clinical decisions.7

ng mobile devices. Furthermore, smartphones are used to calculate medical formulas, find drug references, search literature, and access medical education/training. Healthcare providers can also use smartphones in the diagnosis and objectification of neurological symptoms45 and to obtain support for clinical decisions.7 As compared with paper guidelines, computer-based guidelines running on a mobile device can improve diagnostic decision making for pulmonary embolisms.62 The Clinical Research Center for Stroke in South Korea provides a smartphone application for stroke guidelines.63 This smartphone application can be useful in daily clinical practice because it is accessible anytime and anywhere. Providing timely updates on guidelines is another advantage of this application; readily available updates to guidelines can help physicians better adhere to evidence-based medicine. In addition, mobile applications can be helpful in clinical trials for pre-screening, assessment of eligibility, and randomization of patients.64 Applications for patients Self-management and remote monitoring of patients are becoming viable solutions for management of diseases with chronic conditions, including stroke. The risk of stroke is as high as 12.8% during the first week after a transient ischemic attack. At least 80% of recurrent events might be prevented with the use of a comprehensive approach.65 To prevent stroke recurrence, drug compliance, risk-factor control, and lifestyle modification are important.

including stroke. The risk of stroke is as high as 12.8% during the first week after a transient ischemic attack. At least 80% of recurrent events might be prevented with the use of a comprehensive approach.65 To prevent stroke recurrence, drug compliance, risk-factor control, and lifestyle modification are important. Cell phone messaging interventions can serve as preventive measures by improving patients' health status and behavior.66 Indeed, the delivery of stroke warning signs through text messages is one way to get people thinking about the dangers of stroke.67 Furthermore, smartphone applications will doubtlessly play an important role in the future in patient education, self-management, and remote monitoring.68 Lifestyle applications can help manage weight control, diet and cooking, and exercise. The Clinical Research Center for Stroke in South Korea has developed the "Stroke STOP" application, which provides risk-factor-control tools, medication reminders, and public information about stroke.69

ment, and remote monitoring.68 Lifestyle applications can help manage weight control, diet and cooking, and exercise. The Clinical Research Center for Stroke in South Korea has developed the "Stroke STOP" application, which provides risk-factor-control tools, medication reminders, and public information about stroke.69 Comments Information technology and smart devices can be useful in many aspects of stroke management, which may improve patient outcomes. Greater automation of a hospital's information system resulted in fewer complications, lower mortality rates, and lower costs. Furthermore, information technology using smart devices, including smartphones and tablet computers, will be a cornerstone of good clinical care in hospital environments. Although there are now hundreds of applications focusing on wellness, fitness, and nutrition for the general population, the development of healthcare applications for stroke patients will provide more opportunities to improve stroke management. However, many applications are developed by programmers without the involvement of doctors or patients. Thus, participation of stroke healthcare professionals in the design of mobile applications would be beneficial to their development and implementation. The authors are grateful to Mr. Dong-Su Jang (Yonsei University College of Medicine) for providing the illustration. This work was supported by a grant from the Korea Healthcare Technology Research and Development Project, Ministry for Health, Welfare, and Family Affairs, Republic of Korea (HI10C2020).

Comments Information technology and smart devices can be useful in many aspects of stroke management, which may improve patient outcomes. Greater automation of a hospital's information system resulted in fewer complications, lower mortality rates, and lower costs. Furthermore, information technology using smart devices, including smartphones and tablet computers, will be a cornerstone of good clinical care in hospital environments. Although there are now hundreds of applications focusing on wellness, fitness, and nutrition for the general population, the development of healthcare applications for stroke patients will provide more opportunities to improve stroke management. However, many applications are developed by programmers without the involvement of doctors or patients. Thus, participation of stroke healthcare professionals in the design of mobile applications would be beneficial to their development and implementation. The authors are grateful to Mr. Dong-Su Jang (Yonsei University College of Medicine) for providing the illustration. This work was supported by a grant from the Korea Healthcare Technology Research and Development Project, Ministry for Health, Welfare, and Family Affairs, Republic of Korea (HI10C2020). The authors have no financial conflicts of interest.

The authors are grateful to Mr. Dong-Su Jang (Yonsei University College of Medicine) for providing the illustration. This work was supported by a grant from the Korea Healthcare Technology Research and Development Project, Ministry for Health, Welfare, and Family Affairs, Republic of Korea (HI10C2020). The authors have no financial conflicts of interest. Figure 1 The iPronator application for the pronator drift test. Two mobile devices were placed on each forearm and held firmly in place with Velcro above the wrists. In patients with mild arm weakness, drift and pronation could be observed (A). A Bluetooth connection is used to transfer data from the mobile device on each arm (B).

iPronator application for the pronator drift test. Two mobile devices were placed on each forearm and held firmly in place with Velcro above the wrists. In patients with mild arm weakness, drift and pronation could be observed (A). A Bluetooth connection is used to transfer data from the mobile device on each arm (B). Figure 2 Logical algorithm of the iTOAST. The logical algorithm for the iTOAST application was based on the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification. There are six decision-making checkpoints: (1) presence of angiographic evaluations, (2) atherosclerotic stenosis >50% or occlusion on the relevant artery, (3) potential cardiac sources of embolism, (4) classic lacunar syndrome, (5) subcortical infarction <1.5 cm, and (6) other rare causes of stroke. The cumulative yes/no responses to each step yield a final stroke classification. RAA, relevant artery atherosclerosis; PCSE, potential cardiac sources of embolism; CLS, classic lacunar syndrome; LAA, large artery atherosclerosis; CE, cardioembolism; LAC, lacune; SOD, stroke of other determined etiology; UT, undetermined etiology owing to more than two causes; UN, undetermined etiology owing to negative evaluation; UI, undetermined etiology owing to incomplete evaluation. Table 1 The process of the "Brain salvage through Emergent Stroke Therapy (BEST)" program using computerized physician order entry (CPOE)

Introduction Cardioembolic stroke, which is an important subtype of ischemic stroke, involves a large infarct volume and multiple vascular territories. The neurologic deficits are grave and develop abruptly.1,2 Because cardioembolic stroke is generally severe and frequently recurs, its long-term mortality is high.3,4 In addition, hemorrhagic transformation occurs frequently because of early or delayed recanalization.5,6 Cardioembolic stroke accounts for 14-30% of all ischemic stroke.7-12 In Korea, it accounts for approximately 17%, and the proportion is increasing (Figure 1).13,14 Cardioembolic stroke is associated with chronological age7 and is thought to be one of the most important subtypes of ischemic stroke in aged or aging populations.

mbolic stroke accounts for 14-30% of all ischemic stroke.7-12 In Korea, it accounts for approximately 17%, and the proportion is increasing (Figure 1).13,14 Cardioembolic stroke is associated with chronological age7 and is thought to be one of the most important subtypes of ischemic stroke in aged or aging populations. Non-valvular atrial fibrillation (NVAF) is the most important cause of cardioembolic stroke, and patients with NVAF can complain of palpitations, chest pain, breathing difficulty, dizziness, or fainting.15 However, many patients do not have any symptoms or complain of vague and non-specific symptoms.16,17 Probability of stroke incidence in patients with atrial fibrillation is 3-4%,18 and the risk of stroke increased by five times in all age groups.19,20 The percentage of strokes attributable to atrial fibrillation increases steeply from 1.5% at age 50 to 59 years to 23.5% at age 80 to 89 years.21 The prevalence of atrial fibrillation is globally increasing over time. In the United States, the number of patients with atrial fibrillation was 2.1 million in 1997, but it increased to 2.3 million in 2001. It is estimated to increase to 5.6 million in 2050.22 The increased prevalence of atrial fibrillation is caused by improved survival rates of patients with heart disease.23 The prevalence of atrial fibrillation significantly increases with an increase in age.24 In Korea, 57% of individuals with atrial fibrillation are older than 65 years, and the prevalence of atrial fibrillation is the highest in those older than 80 years.25 The importance of atrial fibrillation in ischemic stroke will further increase in Korean society, which is rapidly aging.

ses with an increase in age.24 In Korea, 57% of individuals with atrial fibrillation are older than 65 years, and the prevalence of atrial fibrillation is the highest in those older than 80 years.25 The importance of atrial fibrillation in ischemic stroke will further increase in Korean society, which is rapidly aging. Risk assessment of stroke in patients with NVAF Risk factors increasing the incidence of stroke in patients with NVAF are known to be female gender, old age, history of stroke or transient ischemic attack (TIA), hypertension, heart failure, diabetes, and vascular diseases.23,26,27 History of stroke or TIA increases the risk of stroke in patients with NVAF by three times.26 The incidence of stroke in patients with NVAF in their 70s is seven times that of patients in their 40s.28 When patients with NVAF have hypertension, the risk of stroke is thrice as high.19 Since a risk stratification scale for embolic events in patients with NVAF was developed based on integration of these risk factors, it can be used to assess the risk of stroke in patients with NVAF and to select adequate preventive drugs.

en patients with NVAF have hypertension, the risk of stroke is thrice as high.19 Since a risk stratification scale for embolic events in patients with NVAF was developed based on integration of these risk factors, it can be used to assess the risk of stroke in patients with NVAF and to select adequate preventive drugs. To classify the risk of stroke in patients with NVAF, several models are currently utilized. The representative stratification systems currently being used include the CHADS2 and CHA2DS2-VASc scores (Table 1).29,30 The CHADS2 score is widely used as the risk stratification scale, but after considering additional risk factors such as vascular disease, gender, and age of 65-74 years, a more specific evaluation can be made. The scale reflecting these risk factors is the CHA2DS2-VASc score. If a patient is classified into the group of low risk (0 or 1) in the point system by using the CHADS2 score, the CHA2DS2-VASc score can be helpful for a more comprehensive risk assessment.30 In addition, the HAS-BLED score, which is a convenient bleeding risk scale, is prepared from risk factors and a systematic review of bleeding in patients with NVAF (Table 2).31 If the HAS-BLED score is ≥3, a patient is classified into the high-risk group for bleeding.32 When antithrombotic therapy is started, special care and regular observation are required in this group. In addition, microbleeds predict intracranial bleeding associated with warfarin.33,34

in patients with NVAF (Table 2).31 If the HAS-BLED score is ≥3, a patient is classified into the high-risk group for bleeding.32 When antithrombotic therapy is started, special care and regular observation are required in this group. In addition, microbleeds predict intracranial bleeding associated with warfarin.33,34 For the past 60 years, when the need for anticoagulation therapy was determined based on the calculated risk of stroke in patients with NVAF, warfarin has been administered. Previous studies demonstrated that warfarin was superior to other antiplatelet drugs in preventing stroke in patients with NVAF.35,36 In addition, it has been reported that the stroke is less severe when it occurs in patients taking warfarin.37-39 Some patients have a very high tendency of relapse stroke, and stroke may recur despite the use of warfarin after stroke associated with NVAF. Although the primary reason is that warfarin is not properly administered or the international normalized ratio (INR) does not enter the target, other risk factors for recurrence of stroke are old age, female gender, and a history of stroke.40 If large artery atherosclerosis such as carotid or vertebral atherosclerosis is associated with NVAF, the recurrence rate of stroke also increases.41 In addition, a meta-analysis reported that stroke may be more recurrent in Asian individuals taking warfarin compared to other races.42 In this context, the CHADS2 score can be useful in predicting the recurrence of stroke in patients with stroke associated with NVAF.40,42 Furthermore, it was recently reported that the CHADS2 score can be used to predict the prognosis of stroke associated with NVAF (Figure 2).43,44

g warfarin compared to other races.42 In this context, the CHADS2 score can be useful in predicting the recurrence of stroke in patients with stroke associated with NVAF.40,42 Furthermore, it was recently reported that the CHADS2 score can be used to predict the prognosis of stroke associated with NVAF (Figure 2).43,44 Emergence of new anticoagulants Warfarin, a vitamin K antagonist, was the only oral anticoagulant available for a long time. However, its narrow therapeutic window and numerous food and drug interactions affect its safety, efficacy, and patient compliance. In addition, its slow onset of action and variable pharmacologic effects make it difficult to maintain the appropriate antithrombotic effect.45 A meta-analysis reported that 44% of bleeding complications with warfarin were associated with supratherapeutic INRs, and 48% of thromboembolic events occurred in the subtherapeutic range.46 These drawbacks have encouraged the development of new anticoagulants. Several new oral anticoagulants (NOACs) have been developed in the past 10 years.47 Three of them were approved for use in the prevention of stroke and systemic embolism in patients with NVAF: dabigatran, a direct thrombin inhibitor, and rivaroxaban and apixaban, activated factor X inhibitors.48-50 They have predictable anticoagulant effects and fewer food and drug interactions, which allow a fixed dosing regimen without the need for monitoring. However, strict patient compliance is required to maintain the desired anticoagulation levels. When rapid reversal of the anticoagulant effect is needed in the event of a major bleeding or emergency surgery, there are no specific antidotes and standardized tests to monitor the anticoagulant status, which is a weakness of NOACs.51

rict patient compliance is required to maintain the desired anticoagulation levels. When rapid reversal of the anticoagulant effect is needed in the event of a major bleeding or emergency surgery, there are no specific antidotes and standardized tests to monitor the anticoagulant status, which is a weakness of NOACs.51 In terms of efficacy, according to the three randomized clinical trials that compared the efficacy of three NOACs and warfarin, NOACs were not always superior to warfarin. The rates of the primary efficacy outcomes of any type of stroke or systemic embolism were significantly lower in the dabigatran 150 mg and apixaban groups compared to the warfarin group. Those in the dabigatran 110 mg and rivaroxaban groups were not significantly lower than those in the warfarin group. In addition, all three studies had poor or moderately controlled patients (defined as INR in the target range <65% of the time) as a control group.48-50 Follow-up studies in which the control groups were divided based on INR level did not show the superiority of the dabigatran 150 mg dose compared with the well-controlled group (INR in the target range ≥65% of the time).52 The superiority of the primary efficacy outcome in apixaban use was mostly caused by a reduction in hemorrhagic stroke, and there was no statistically significant difference in ischemic stroke rate.49 Therefore, if the patient is well-controlled with warfarin and has minimal complications, the use of NOACs may be controversial (Table 3).

the primary efficacy outcome in apixaban use was mostly caused by a reduction in hemorrhagic stroke, and there was no statistically significant difference in ischemic stroke rate.49 Therefore, if the patient is well-controlled with warfarin and has minimal complications, the use of NOACs may be controversial (Table 3). In terms of safety, the major bleeding complication rate in the dabigatran 110 mg and apixaban groups was significantly lower than that in the warfarin group. However, there was no significant difference between the dabigatran 150 mg and rivaroxaban groups and the warfarin group (Table 3).48-50 Considering subtypes of bleeding, the intracerebral hemorrhage (ICH) rate was significantly lower with both doses of dabigatran, but the major gastrointestinal (GI) bleeding rate was significantly higher in the dabigatran 150 mg and rivaroxaban groups.50,53 The ICH rate was significantly lower and the GI bleeding rate was not significantly different in the apixaban group.49 Hemorrhagic stroke rates related to receiving warfarin were higher in Asian patients than in non-Asian patients, but hemorrhagic stroke rates were significantly reduced by dabigatran in both Asian and non-Asian patients compared with individuals receiving warfarin.54 In non-hemorrhagic adverse events, approximately 10% of the patients in the dabigatran group complained of severe dyspepsia and approximately 21% of them discontinued the drugs. These symptoms were likely caused by the tartaric acid core composed of dabigatran etexilate. Tartaric acid created an acidic environment and increased the absorption of the drug independent of gastric pH.55 Approximately 0.8% of patients taking dabigatran in the RE-LY trial experienced myocardial infarction. It was not statistically significant compared with the group taking warfarin, but it showed greater tendencies.56 A meta-analysis of 7 studies including the RE-LY trial showed that the risk of myocardial infarction and cardiac death or unstable angina was significantly increased in the dabigatran group compared with the warfarin group.57

cally significant compared with the group taking warfarin, but it showed greater tendencies.56 A meta-analysis of 7 studies including the RE-LY trial showed that the risk of myocardial infarction and cardiac death or unstable angina was significantly increased in the dabigatran group compared with the warfarin group.57 The bioavailability of dabigatran is only 6.5% and is not influenced by the coadministration of food. Approximately 80% of the drug is excreted unchanged in the urine, while only 20% is excreted in the feces. Therefore, it is contraindicated in severe renal impairment. It is not metabolized by cytochrome P450 isoenzymes. Thus, it is substantially unaffected by mild to moderate hepatic failure.58 Because of the low plasma protein binding, it may be dialyzable when rapid reversal of its anticoagulation effects is needed.59 The bioavailability of rivaroxaban is dose-dependent and influenced by the coadministration of food. The bioavailability at 10 mg and 20 mg in the fasting state is 80%-100% and 66%, respectively. Taken with meals, its absorption is delayed but increased. Thus, therapeutic doses of rivaroxaban should be taken with meals.60 Approximately 66% is excreted in the urine, and only 28% is excreted in the feces. The safety and efficacy of rivaroxaban were consistent with warfarin in patients with moderate renal impairment, and it was also observed in Japanese patients.61 Two-thirds of the drug is converted into inactive metabolites through different CYP450 isoenzymes (CYP3A4/5 or CYP2J2) and CYP-independent mechanisms.62 In this context, rivaroxaban should be avoided in patients with severe renal failure or moderate to severe hepatic impairment. The bioavailability of apixaban is approximately 50% and is not influenced by the coadministration of food.63 One-third of the drug is metabolized through the cytochrome P-450 isoenzyme system (mainly CYP3A4/5). Approximately 25% is excreted in the urine, and >50% is excreted in the feces.64 It may be suitable for patients with mild to moderate renal impairment because of its low renal excretion.

he coadministration of food.63 One-third of the drug is metabolized through the cytochrome P-450 isoenzyme system (mainly CYP3A4/5). Approximately 25% is excreted in the urine, and >50% is excreted in the feces.64 It may be suitable for patients with mild to moderate renal impairment because of its low renal excretion. Special circumstances for NOACs As previously mentioned, patients stabilized on warfarin may prefer to continue the same. However, the convenience and efficacy of NOACs in the patients with inadequate INR are cause to consider the transition to NOACs. Transition to warfarin also may be needed in patients who are unable to continue on NOACs. In this situation, safe transition between anticoagulants is an important issue in current practice. In previous trials including patients who had warfarin before starting on dabigatran, the highest INR permitted at the time of transition was 2.0 or 2.3,65,66 whereas in the Rivaroxaban Once-Daily, Oral, Direct Factor X Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF), it was 3.0.50 In these trials, the overlap was not associated with an increased risk for bleeding. In transition from NOACs to warfarin, it is necessary to take into account the expected onset of warfarin as well as the half-life of the NOACs. The mean time to achieve therapeutic range (INR 2-3) was approximately 5 days. The half-life of NOACs highly depends on renal function. The half-life of dabigatran varies from 14-17 hours with normal renal function to 18-27 hours in moderate to severe renal impairment.67,68 The recently recommended approaches based on these data are summarized in Table 4.69-71 The longer overlap period of rivaroxaban and warfarin shown in this table is caused by the shorter half-life of rivaroxaban.72 When assessing the INR during transitions between rivaroxaban and warfarin, we need to be cautious because rivaroxaban may increase the prothrombin time.73

data are summarized in Table 4.69-71 The longer overlap period of rivaroxaban and warfarin shown in this table is caused by the shorter half-life of rivaroxaban.72 When assessing the INR during transitions between rivaroxaban and warfarin, we need to be cautious because rivaroxaban may increase the prothrombin time.73 Perioperative management of the NOACs is based on the urgency of the procedure, bleeding risk, and current renal function. Preoperative discontinuation of the drug is based on pharmacokinetic data and considerations regarding the bleeding risk (Table 5). Interruption of dabigatran for 48 hours should be sufficient to ensure adequate hemostasis in patients with normal renal function because of the half-life of 14-17 hours.74 For procedures with a low risk of bleeding such as cardiac catheterization, diagnostic endoscopy, and minor orthopedic surgery, for which an INR of 1.5 for patients on warfarin would be acceptable, it is reasonable to interrupt dabigatran for 24 hours. In patients with decreased renal function, the period of interruption should be longer. This principle is also applied to rivaroxaban. In the previous trial, ROCKET-AF, rivaroxaban was interrupted approximately 2 days before the procedure.60 The short period of interruption does not require bridging therapy with unfractionated heparin or low-molecular-weight heparin. Reinitiation of NOACs after the procedure depends on the postoperative risk of bleeding. For procedures with good hemostasis, the suggestion is to reinitiate NOACs at a minimum of 4-6 hours after the procedure.65 The first dose of dabigatran should be a half-dose (75 mg), and the next scheduled dose should be the usual maintenance dose. A similar strategy using a 10-mg dose for the first dose can be applied to rivaroxaban. Patients with bowel paralysis may require bridging with parenteral anticoagulants because they cannot take oral anticoagulants.

igatran should be a half-dose (75 mg), and the next scheduled dose should be the usual maintenance dose. A similar strategy using a 10-mg dose for the first dose can be applied to rivaroxaban. Patients with bowel paralysis may require bridging with parenteral anticoagulants because they cannot take oral anticoagulants. Novel strategies to prevent stroke in patients with NVAF For strategies to prevent stroke associated with NVAF in addition to antithrombotic therapy, studies on procedures to surgically block the left atrial appendage (LAA), which is the most important position in which thrombus is formed by NVAF, have been conducted.75-77 After a device to block LAA is inserted, warfarin is administered for 45 days, and then both aspirin and clopidogrel are administered for 4.5 months. It is not inferior to the group of conventional warfarin use when aspirin monotherapy is administered.78 In addition, when long-term follow-up continues for up to 2.3 years, it is non-inferior in occurrence of cardiovascular diseases including stroke compared with the group using warfarin.79 Other strategies to prevent stroke associated with NVAF include rhythm control and heart rate control. Previous studies and meta-analyses showed that the beneficial effect between rate control and rhythm control was not different for prevention of stroke.80-85 A post-hoc analysis from a recent study showed that dronedarone, a newly developed rhythm control drug, was beneficial in preventing stroke,86 but another study using dronedarone in addition to standard therapy for rhythm control showed that dronedarone increased the risk of stroke more than two-fold in comparison with placebo.87 Recently, a large-sized population-based observational study to compare rhythm control and rate control in patients with atrial fibrillation was conducted in Canada.88 Risk for the occurrence of stroke was decreased by 20% in patients who underwent rhythm control compared with those who underwent rate control. When patients were classified into low-, moderate-, and high-risk groups based on CHADS2 scores, rhythm control was superior to rate control in the moderate- and high-risk groups. To date, a complete conclusion about the effect of rate control and rhythm control on occurrence of stroke associated with NVAF has not been made.

atients were classified into low-, moderate-, and high-risk groups based on CHADS2 scores, rhythm control was superior to rate control in the moderate- and high-risk groups. To date, a complete conclusion about the effect of rate control and rhythm control on occurrence of stroke associated with NVAF has not been made. Conclusion Knowledge about risk stratification and treatment for stroke associated with NVAF is increasing dramatically. More delicate risk assessment systems make physicians choose adequate preventive strategies to reduce stroke related to NVAF, and NOACs may be additional therapeutic options for these patients in primary and secondary prevention. In this context, there are some clinical practice guidelines for the management of atrial fibrillation patients from the United States and Europe.26,89,90 Recently, a clinical practice guideline for stroke prevention was published and revised in Korea, and it includes these issues about ischemic stroke associated with NVAF.91 Ischemic stroke related to NVAF has been more prevalent, as life expectancy is increasing worldwide, and disease burden is also increasing steeply. Preventive and therapeutic options should be developed and improved in the future. This work was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (HI10C2020). The authors have no financial conflicts of interest. Figure 1 Trend of cardioembolic stroke in Korea.13 Figure 2 CHADS2 score as a prognostic factor after cardioembolic stroke.43 Table 1 CHADS2 and CHA2DS2-VASc scores

This work was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (HI10C2020). The authors have no financial conflicts of interest. Figure 1 Trend of cardioembolic stroke in Korea.13 Figure 2 CHADS2 score as a prognostic factor after cardioembolic stroke.43 Table 1 CHADS2 and CHA2DS2-VASc scores *Hypertension is defined as blood pressure consistently > 140/90 mmHg (or treated hypertension on medication). Table 2 HAS-BLED score Table 3 Summary of 3 clinical trials related to new oral anticoagulants *Any type of stroke and systemic embolism; †Risk ratio; ‡Hazard ratio. CI, confidence interval; bid, twice-daily; qd, once-daily. Table 4 Suggested strategy for conversion to and from new oral anticoagulants VKA, vitamin K antagonist; NOAC, new oral anticoagulant; INR, international normalized ratio; CrCl, creatinine clearance. Table 5 Perioperative management of the new oral anticoagulants *Colonoscopy, uncomplicated laparoscopic procedures, any aspirations not involving the spinal canal; †mL/min per 1.73 m2; ‡Major cardiac surgery, insertion of pacemakers or defibrillators, neurosurgery, major cancer/urologic/vascular surgery, spinal puncture.

Introduction The term cerebral microbleed (CMB) refers to small, round dark-signal lesions detected by T2*-weighted or gradient-echo (GRE) magnetic resonance imaging (MRI).1 CMBs were introduced to stroke physicians in the late 1990s and early 2000s after development of MRI techniques sensitive to paramagnetic effects.2 The clinical significance of CMBs has been actively investigated, especially in the stroke field and more recently in studies on cognitive impairment and aging.3 Histological investigation has shown that CMBs are tiny foci containing hemosiderin-laden macrophages and abnormal microvessels showing fibrohyalinosis.4,5 Clinical cases with frank symptoms caused by CMBs are uncommon. Because CMBs are manifestations of focal extravascular leakage of blood components, however, investigators have suggested that accumulation of CMBs reflects a bleeding-prone status in individuals with an elevated risk of cerebral hemorrhage. Clinical studies have found strong associations between CMBs and chronic hypertension and low cholesterol levels,6,7 and between the proximity and volume of CMBs and those of subsequent intracerebral hemorrhage (ICH).8,9 Longitudinal studies have found that CMBs are linked to subsequent hemorrhagic stroke in stroke survivors,10 and suggested that CMBs are related to antithrombotic-related hemorrhage.11,12 In this review, we discuss fundamental findings of CMBs, and the clinical implications of these observations for the field of cerebrovascular disease.

es have found that CMBs are linked to subsequent hemorrhagic stroke in stroke survivors,10 and suggested that CMBs are related to antithrombotic-related hemorrhage.11,12 In this review, we discuss fundamental findings of CMBs, and the clinical implications of these observations for the field of cerebrovascular disease. Visualization and detection of cerebral microbleeds GRE sequences are more sensitive to susceptibility effects than are classical MRI sequences. Unlike classical T2-weighted imaging or echo-planar sequences, GRE sequences maximize the paramagnetic effects of blood components such as hemosiderin, deoxyhemoglobin, and ferritin.2,13 Due to the dephasing of MRI signals, GRE sequences tend to exaggerate lesion sizes. Therefore, sub-millimeter CMBs appear as signal-loss lesions of several millimeters, a phenomenon referred to as the blooming effect. Susceptibility-weighted imaging (SWI) is a MRI sequence that maximizes sensitivity to magnetic susceptibility effects.14,15 SWI requires more time than does GRE, and also requires post-processing, but because SWI accentuates the magnetic properties of tissues, it enables visualization of areas such as CMBs containing deoxygenated blood substances.16 SWI permits visualization of a greater number of CMBs than can be seen with conventional GRE sequences, but the clinical implications of this increased sensitivity are not yet fully understood (Figure 1).17

properties of tissues, it enables visualization of areas such as CMBs containing deoxygenated blood substances.16 SWI permits visualization of a greater number of CMBs than can be seen with conventional GRE sequences, but the clinical implications of this increased sensitivity are not yet fully understood (Figure 1).17 Since CMBs are primarily a radiologic concept, their sizes and shapes are dependent upon the parameters of the GRE sequence. Visualization of CMBs is thought to be influenced by pulse sequence, sequence parameters, spatial resolution, magnetic field strength, and post-processing of images.1 Thus, lesions due to old parenchymal hemorrhage, cortical or deep-seated mineralizations, or vascular malformations potentially all can be misinterpreted as CMBs. Small cortical vessels, partial volume effects of the cerebellar cortex, and cavernous hemangiomas also should be considered (Figure 2).21 Moreover, multifocal bleeding spots caused by diffuse axonal injury are virtually impossible to distinguish from cortico-subcortical CMBs.

ntially all can be misinterpreted as CMBs. Small cortical vessels, partial volume effects of the cerebellar cortex, and cavernous hemangiomas also should be considered (Figure 2).21 Moreover, multifocal bleeding spots caused by diffuse axonal injury are virtually impossible to distinguish from cortico-subcortical CMBs. Comparison of magnetic field strength in detection of CMBs revealed 0.5 more microbleeds on average could be observed using a 3.0 Tesla MRI compared to a 1.5 Tesla unit (2.1 in 3.0 T and 1.6 in 1.5 T).18 No universal standard MR parameters exist for the detection of CMBs, and in addition, the definition of CMB size also varies. In published reports, the lower limit of CMB size is usually ≤2 mm, and the upper limit is usually between 5 and 10 mm.19 Greenberg et al. reported that the size distribution of microbleeds and macrobleeds is bimodal, and proposed that the most appropriate cut-off point between them is 5.7 mm.20 Those researchers suggested elsewhere that an upper size limitation is impractical, and that instead, a more desirable approach is to carefully exclude CMB-mimicking lesions such as those described above.1 Other investigators have proposed a specified rating scale for CMBs, the Microbleed Anatomical Rating Scale.21 In our view, setting an appropriate upper-size limitation, and determination of intra- and inter-rater agreement values, are necessary for clinical and radiologic research studies.

h as those described above.1 Other investigators have proposed a specified rating scale for CMBs, the Microbleed Anatomical Rating Scale.21 In our view, setting an appropriate upper-size limitation, and determination of intra- and inter-rater agreement values, are necessary for clinical and radiologic research studies. Pathology of cerebral microbleeds Pathologic-radiologic correlation studies have revealed that CMBs are focal accumulations of hemosiderin adjacent to abnormal blood vessels demonstrating fibrolipohyalinosis or amyloid microangiopathy.4,5 CMBs usually develop adjacent to capillaries, small arteries, or arterioles; hemosiderin-laden macrophages are usually present.22 A recent pathological investigation of CMBs in patients with cerebral amyloid angiopathy demonstrated extravasation of blood and hemosiderin through vulnerable vascular walls, with β-amyloid pigmentation and surrounding inflammation.23 These findings suggest that CMBs are the radiologic correlates of extravasation of blood components through injured or fragile vascular walls, or of frank small hemorrhage spots. In the context of cerebral amyloid angiopathy, Greenberg et al. suggested that increased vascular wall thickness was indicative of microbleeds.20 Because blooming artifact on GRE images exaggerates the size of CMBs, it has been proposed that the actual size of hemosiderin deposition is too small to cause apparent neurological deficit.22 Selected cases, however, have suggested that CMBs can cause neurological symptoms and signs.24,25

Pathology of cerebral microbleeds Pathologic-radiologic correlation studies have revealed that CMBs are focal accumulations of hemosiderin adjacent to abnormal blood vessels demonstrating fibrolipohyalinosis or amyloid microangiopathy.4,5 CMBs usually develop adjacent to capillaries, small arteries, or arterioles; hemosiderin-laden macrophages are usually present.22 A recent pathological investigation of CMBs in patients with cerebral amyloid angiopathy demonstrated extravasation of blood and hemosiderin through vulnerable vascular walls, with β-amyloid pigmentation and surrounding inflammation.23 These findings suggest that CMBs are the radiologic correlates of extravasation of blood components through injured or fragile vascular walls, or of frank small hemorrhage spots. In the context of cerebral amyloid angiopathy, Greenberg et al. suggested that increased vascular wall thickness was indicative of microbleeds.20 Because blooming artifact on GRE images exaggerates the size of CMBs, it has been proposed that the actual size of hemosiderin deposition is too small to cause apparent neurological deficit.22 Selected cases, however, have suggested that CMBs can cause neurological symptoms and signs.24,25 Prevalence and associated factors for cerebral microbleeds Cerebral microbleeds in healthy subjects In subjects without a history of cerebrovascular disease, the prevalence of CMBs was reported to be between 3-7%.26-33 Significant associations have been consistently reported between CMBs and advanced age, as well as hypertension.33 In contrast, association between CMBs and diabetes has been inconsistent across published reports.19,34,35 The Rotterdam Scan Study described CMBs in 1,062 older subjects.39 In a group of patients with mean age 69.6 years and a hypertension prevalence of 71.9%, CMBs were detected in 17.8% of patients aged 60-69, in 31.3% of patients aged 70-79, and in 38.3% of patients aged 80-97. The prevalence of multiple CMBs was also found to increase significantly with age.37 A single-hospital-based cross-sectional study performed in Japan found no deep-seated CMBs in subject younger than 40 years old.38 The Rotterdam Scan Study also noted a strong association of very low serum cholesterol levels (<4.42 mmol/L versus higher values) with the presence of strictly lobar microbleeds,39 an observation consistent with our earlier findings.7

performed in Japan found no deep-seated CMBs in subject younger than 40 years old.38 The Rotterdam Scan Study also noted a strong association of very low serum cholesterol levels (<4.42 mmol/L versus higher values) with the presence of strictly lobar microbleeds,39 an observation consistent with our earlier findings.7 Cerebral microbleeds in ischemic stroke patients The reported prevalence of CMBs in ischemic stroke patients varies significantly (35%-71%; Table 1).27,38,40-43 This variability may be due to the heterogeneity of ischemic stroke per se, or to differences in recruited populations, rating strategies, and MRI parameters. Two studies investigating CMB frequency in different ischemic stroke subtypes found that CMBs were less frequent in cardioembolic stroke than in atherothrombotic stroke or lacunar stroke.28,44 Accumulation of sublethal ischemic injuries in brain parenchyma is thought to differ in atherothrombotic stroke and cardioembolic stroke,45 suggesting that the relationship of CMBs and stroke subtype in turn may reflect a different degree of fragility of vascular walls. A recent study found that the absolute number of CMBs, as well as variability indices of blood pressure (including coefficient of variance and successive variation), were elevated in cases of ischemic stroke with CMBs.46

p of CMBs and stroke subtype in turn may reflect a different degree of fragility of vascular walls. A recent study found that the absolute number of CMBs, as well as variability indices of blood pressure (including coefficient of variance and successive variation), were elevated in cases of ischemic stroke with CMBs.46 Another analysis of patients with ischemic stroke reported that serum uric acid level was associated in a dose-dependent manner with the presence of CMBs, but only in hypertensive patients.47 Complex interactions have been observed between chronic medical conditions and individual serologic markers in development of CMBs, such as in ischemic stroke patients with chronic kidney disease. Proteinuria and impaired kidney function has been linked to with small vessel disease in the brain and to CMBs.48-51 We further analyzed this issue, and found that chronic kidney disease is independently associated with cerebral microbleeds in patients without diabetes, but not in patients with diabetes.52

ney disease. Proteinuria and impaired kidney function has been linked to with small vessel disease in the brain and to CMBs.48-51 We further analyzed this issue, and found that chronic kidney disease is independently associated with cerebral microbleeds in patients without diabetes, but not in patients with diabetes.52 Cerebral microbleeds in hemorrhagic stroke patients The frequency of CMBs in hemorrhagic stroke patients has been consistently higher than that that in ischemic stroke patients, reaching 50% to 80%.4,8,53-55 (Table 2; Figure 3) The detection rate of CMBs is higher in Asian populations, which may reflect a higher prevalence of the hemorrhagic stroke subtype in this ethnic group.56 CMBs also have been reported to be associated with hematoma volume, regardless of perihematomal edema volume.9 Patients with cerebral amyloid angiopathy have a higher CMB detection rate, with a preference for a lobar location,57,58 and the presence of the APOE e4 allele also has been reported to favor a lobar location (Figure 4).39,59

een reported to be associated with hematoma volume, regardless of perihematomal edema volume.9 Patients with cerebral amyloid angiopathy have a higher CMB detection rate, with a preference for a lobar location,57,58 and the presence of the APOE e4 allele also has been reported to favor a lobar location (Figure 4).39,59 Longitudinal changes in the number of cerebral microbleeds The appearance and disappearance of CMBs over time have not been highlighted sufficiently. In an analysis of 237 acute ischemic stroke patients who underwent follow-up imaging, 13% had new CMBs (56 microbleeds in total) in the subsequent GRE images taken on average after 4 days.60 Long-term follow-up also revealed that 23% of cases demonstrated new CMBs on GRE imaging performed on average over 5.6 years.61 Development of new CMBs in this study was associated with high blood pressure and the presence of CMBs on baseline imaging. An additional follow-up study published in 2012 reported that the number of CMBs was increased in 54% of cases after 2.5 years.62 The annual change of CMBs significantly correlated with the number of CMBs on the baseline study. Another inter esting finding in that report was that in 15% of patients, CMBs disappeared on the follow-up GREs. Follow-up analyses of CMBs must be interpreted with caution, however, since spatial registration of baseline and follow-up images has not been performed in any published studies. Use of rigorous rating criteria21 is another important factor necessary for detailed evaluation of the natural history of CMBs, including their development and disappearance.

e interpreted with caution, however, since spatial registration of baseline and follow-up images has not been performed in any published studies. Use of rigorous rating criteria21 is another important factor necessary for detailed evaluation of the natural history of CMBs, including their development and disappearance. Clinical implications of cerebral microbleeds Further understanding of the characteristics of CMBs has generated interest in using detection of CMBs to enable hemorrhagic stroke risk stratification. A number of reports have described the hemorrhagic tendency of CMBs.24,63,64 A hemorrhagic transformation after multiple embolic infarctions occurred only in the site of the known CMB.63 A Hong Kong study followed 121 acute ischemic stroke patients, and observed that stroke survivors with CMBs on their initial MRI scans had a higher risk of subsequent hemorrhagic stroke.10 Hemorrhage counts in initial scans also were found to be pro portional to the elevated risk of a future hemorrhagic stroke.54 The increased risk of hemorrhagic stroke conferred by the presence of CMBs was also confirmed in a prospective study of 112 ICH survivors.65 Furthermore, association between CMBs and larger ICH volume has been suggested by two studies,9,66 and the predictive value of CMBs in ICH in patients with advanced white matter lesions also has been documented.67 The spot sign, an enhancing locus of contrast extravasation in a cerebral hematoma, suggesting ongoing bleeding, was reported to be negatively associated with the number of microbleeds; this result is not consistent with previous findings, and bears further investigation.68 Patients with CMBs also have been reported to be 2.8 times more likely to have a subsequent disabling or fatal stroke.69 A systematic review published in 2013 concluded that the presence of CMBs in patients with ischemic stroke was associated with greatly increased odds of a subsequent hemorrhagic stroke, but was only modestly linked to recurrence of ischemic stroke.56 This meta-analysis also noted that the strength of association between CMBs and subsequent ICH risk was modified by ethnic background, with a greater odds ratio in Asian cohorts than in Western cohorts (Figure 5).

odds of a subsequent hemorrhagic stroke, but was only modestly linked to recurrence of ischemic stroke.56 This meta-analysis also noted that the strength of association between CMBs and subsequent ICH risk was modified by ethnic background, with a greater odds ratio in Asian cohorts than in Western cohorts (Figure 5). The increased likelihood of cerebral hemorrhage associated with the presence of CMBs may allow prediction of hemorrhagic transformation after ischemic stroke. An earlier report suggested that hemorrhagic transformation after thrombolysis was associated with the presence of CMBs.70 In a case series of 100 acute ischemic stroke patients with imaging follow-up, the presence of CMBs was indicative of early hemorrhagic transformation.71 Embolic ischemic strokes occurring at the sites of previous CMBs were noted to become hemorrhagic.63 Contrary to these positive associations, however, a retrospective study of 279 acute ischemic stroke patients reported no association between CMB count and subsequent hemorrhagic transformation,72 and in an analysis of 70 stroke patients on thrombolytic treatment, CMBs failed to predict post-thrombolytic hemorrhagic transformation.73 No relationship between CMBs and prediction of hemorrhagic transformation was observed in a group of 1,034 acute ischemic stroke patients recruited in a single hospital.74 Finally, a pooled analysis of 570 acute ischemic stroke patients from 13 centers in Europe, North America, and Asia reported that symptomatic ICH after thrombolytic treatment developed regardless of the initial CMB frequency or extent.75 These studies suggest that for patients in need of thrombolysis, any increased risk of ICH attributable to CMBs is negligible, and unlikely to exceed the benefits from thrombolytic therapy. A subsequent meta-analysis indicated that the published analyses were vulnerable to publication bias and limited power, and identified a trend toward increased odds of symptomatic hemorrhage after thrombolysis (odds ratio, 1.98; 95% confidence interval 0.90-4.35).76 At present, identification of CMBs on baseline GRE images should not be considered a contraindication to thrombolysis treatment, but further investigation is warranted on this important issue.

end toward increased odds of symptomatic hemorrhage after thrombolysis (odds ratio, 1.98; 95% confidence interval 0.90-4.35).76 At present, identification of CMBs on baseline GRE images should not be considered a contraindication to thrombolysis treatment, but further investigation is warranted on this important issue. Different findings regarding CMBs and the development of ICH or hemorrhagic transformation may be explained by the different pathological mechanisms of the two phenomena. Essentially, ICH involves rupture of a fragile microvascular wall affected by lipohyalinosis or microaneurysms under the chronic influence of hypertension.77 As CMBs have histological characteristics similar to those of vasculopathy, a correspondingly similar mechanism may underlie formation of CMBs and ICHs.22 In contrast, hemorrhagic transformation develops after acute lethal injury in relatively healthy microvasculature.

sms under the chronic influence of hypertension.77 As CMBs have histological characteristics similar to those of vasculopathy, a correspondingly similar mechanism may underlie formation of CMBs and ICHs.22 In contrast, hemorrhagic transformation develops after acute lethal injury in relatively healthy microvasculature. Considerable interest also exists in utilizing detection of CMBs to estimate the risks of hemorrhagic complications in patients on antithrombotic treatment. Two patients on warfarin were reported to have developed lobar hemorrhages right at the location of CMBs.64 I CMBs were found to be more frequent and extensive in patients with aspirin-associated ICH.11,78 In a cross-sectional study, CMBs were more common in patients taking antithrombotic agents, and aspirin use was found to be related to a lobar location.79 Results from our group demonstrated that patients with anticoagulation-associated hemorrhagic stroke complications are 3.6 times more likely to have CMBs than are age- and sex-matched controls.12 A recent pooled analysis involving 1,460 hemorrhagic strokes and 3,817 ischemic strokes concluded that the number of CMBs was greater in warfarin users who developed ICH.80 Given the strong association between CMBs and subsequent ICHs in stroke survivors, a prospective study is needed to assess the predictive power of CMBs in stroke patients on antithrombotic treatment.

strokes and 3,817 ischemic strokes concluded that the number of CMBs was greater in warfarin users who developed ICH.80 Given the strong association between CMBs and subsequent ICHs in stroke survivors, a prospective study is needed to assess the predictive power of CMBs in stroke patients on antithrombotic treatment. Conclusion CMBs were first identified as tiny, round dark-signal lesions on GRE MRI, and are frequently detected in patients with ischemic or hemorrhagic strokes. Pathological analysis demonstrated that CMBs are extravasations of blood components through fragile microvascular walls, and therefore reflect a bleeding-prone vasculopathy in brain. Several clinical studies have concluded that CMBs are associated with hemorrhagic stroke and hemorrhagic complications following antithrombotic medications. The currently available data do not support the exclusion of thrombolytic treatment based solely on CMB presence or extent. Prospective studies are warranted to confirm the clinical implications of CMBs, and to establish their use for predictive models of hemorrhagic stroke in various situations. The authors have no financial conflicts of interest. Figure 1 Cerebral microbleeds (CMBs) visualized on gradient-echo (GRE) images and susceptibility-weighted images (SWI). A lobar CMB on a SWI image (white arrow) is only faintly visible on the corresponding GRE image. Vessels located in the subarachnoid space could be mistakenly identified as CMBs on SWI sequence (hatched arrows).

rebral microbleeds (CMBs) visualized on gradient-echo (GRE) images and susceptibility-weighted images (SWI). A lobar CMB on a SWI image (white arrow) is only faintly visible on the corresponding GRE image. Vessels located in the subarachnoid space could be mistakenly identified as CMBs on SWI sequence (hatched arrows). Figure 2 Cerebellar microbleeds (CMBs; white arrow). Two vertebral arteries in the subarachnoid space look similar to microbleeds (dotted arrow). One vessel signal located inside of a sulcus could be mistakenly interpreted as a CMB (hatched arrow). Figure 3 Gradient-echo (GRE) images from a case of multiple lobar hemorrhage (white arrow). Multiple lobar cerebellar microbleeds (CMBs) are visible (dotted arrows). Figure 4 Spatial distribution of cerebral microbleeds by location of hemorrhagic stroke. A case of basal ganglia cerebral hemorrhage with thalamic microbleed (A), compared to a case of lobar hemorrhage with multiple lobar microbleeds (B) in a patient with possible cerebral amyloid angiopathy. Figure 5 A basal ganglia intracerebral hemorrhage (white arrow) in a patient with a few lobar cerebral microbleeds (dotted arrow). The patient took an antiplatelet medication for several months and developed a subdural hemorrhage (hatched arrow). Table 1 Prevalence of cerebral microbleeds in ischemic stroke patients AIS, acute ischemic stroke; TIA, transient ischemic attack; MB, microbleeds; IQR, interquartile range; NR, not reported. Table 2 Prevalence of cerebral microbleeds in hemorrhagic stroke patients

Introduction Stroke is a common, serious, and disabling health-care problem throughout the world.1 In particular, in Korea, which is very rapidly changing into an "Aging Society," the incidence of stroke has increased, albeit gradually, during the last few decades.2 On the other hand, the mortality rate from stroke has declined over time,2 resulting in an increased prevalence of stroke in Korea. Unfortunately, one third of stroke survivors achieve only a poor functional outcome five years after the onset of stroke.3 Therefore, stroke-related problems are a serious burden to both patients and their families.4 Although great advances have been made in acute stroke management, the majority of post-stroke care to reduce patients' dependency relies on rehabilitation treatments. Neuroplasticity is the basic mechanism underlying improvement in functional outcome after stroke.5 Therefore, one important goal of rehabilitation of stroke patients is the effective use of neuroplasticity for functional recovery. Other principles of stroke rehabilitation are goal setting, high-intensity practice, multidisciplinary team care, and task-specific training.1 Therefore, high-dose intensive training6 and repetitive practice of specific functional tasks7 are important for recovery after stroke. These requirements make stroke rehabilitation a labor-intensive process.

bilitation are goal setting, high-intensity practice, multidisciplinary team care, and task-specific training.1 Therefore, high-dose intensive training6 and repetitive practice of specific functional tasks7 are important for recovery after stroke. These requirements make stroke rehabilitation a labor-intensive process. Robotic technology has developed remarkably in recent years, with faster and more powerful computers and new computational approaches as well as greater sophistication of electro-mechanical components.8 This advancement in technology has made robotics available for rehabilitation intervention. A robot is defined as a re-programmable, multi-functional manipulator designed to move material, parts, or specialized devices through variable programmed motions to accomplish a task.9 The most important advantage of using robot technology in rehabilitation intervention is the ability to deliver high-dosage and high-intensity training.10 This property makes robotic therapy a promising novel technology for rehabilitation of patients with motor disorders caused by stroke or spinal cord disease. Research into rehabilitation robotics has been growing rapidly, and the number of therapeutic rehabilitation robots has increased dramatically during the last two decades.11

ty makes robotic therapy a promising novel technology for rehabilitation of patients with motor disorders caused by stroke or spinal cord disease. Research into rehabilitation robotics has been growing rapidly, and the number of therapeutic rehabilitation robots has increased dramatically during the last two decades.11 Rehabilitation robots can be divided into therapeutic and assistive robots. The purpose of assistive robots is compensation, whereas therapeutic robots provide task-specific training.12 In this manuscript, the authors will focus on the usefulness of therapeutic robots in patients with stroke. The types of robotic devices used for motor training are end-effector-type devices and exoskeleton-type devices (Figure 1).13 End-effector devices work by applying mechanical forces to the distal segments of limbs. End-effector type robots offer the advantage of easy set-up but suffer from limited control of the proximal joints of the limb, which could result in abnormal movement patterns. In contrast, exoskeleton-type robotic devices have robot axes aligned with the anatomical axes of the wearer. These robots provide direct control of individual joints, which can minimize abnormal posture or movement. Their construction is more complex and more expensive than that of the end-effector type. In this manuscript, the authors will summarize the recent research concerning both the end-effector and exoskeleton types of robot devices. We will also discuss the current status of robot-assisted therapy in stroke rehabilitation.

heir construction is more complex and more expensive than that of the end-effector type. In this manuscript, the authors will summarize the recent research concerning both the end-effector and exoskeleton types of robot devices. We will also discuss the current status of robot-assisted therapy in stroke rehabilitation. Robot-assisted therapy for gait function End-effector-type robotic devices Seven randomized controlled trials that compared robot-assisted therapy that uses end-effector-type devices with conventional therapies for improving gait function after stroke were selected for review (Table 1).14-20 Two studies conducted in patients with chronic stroke reported comparable effects on gait function between the robot-assisted therapy and conventional gait training.14,17 These results indicate that the robot-assisted therapy with end-effector-type devices cannot replace conventional therapy in patients with chronic stroke. However, the other five trials, which enrolled patients with subacute stroke, demonstrated that robot-assisted therapy in combination with conventional physiotherapy produced greater improvement in gait function than conventional gait training alone.15,16,18-20 This means that the addition of robot-assisted therapy with end-effector-type devices to conventional physiotherapy can be recommended for use in patients with subacute stroke.

in combination with conventional physiotherapy produced greater improvement in gait function than conventional gait training alone.15,16,18-20 This means that the addition of robot-assisted therapy with end-effector-type devices to conventional physiotherapy can be recommended for use in patients with subacute stroke. Exoskeleton-type robot devices Eight randomized controlled trials that investigated the use of robot-assisted therapy with exoskeleton devices for improvement of gait function in patients with stroke were selected for review (Table 1).21-28 Two studies from 2007 reported superior results from robot-assisted therapy with exoskeleton devices in comparison with conventional physiotherapy.24,26 Both trials recruited relatively small numbers of patients. The first was a pilot study in patients with subacute stroke.24 Then, in 2008, Hornby et al.23 performed a randomized controlled study comparing the effects of robot-assisted gait training that uses exoskeleton devices and manual facilitation that uses an assist-as-needed paradigm on gait function in patients with chronic stroke. Their results demonstrated that therapist-assisted training yields greater improvements in walking ability in ambulatory stroke survivors than does a similar dosage of robot-assisted training. Hidler et al.22 also investigated the usefulness of robot-assisted therapy in patients with subacute stroke in a multicenter randomized trial. They concluded that the diversity of conventional gait training interventions appeared to be more effective than robot-assisted gait training for improving walking ability. Therefore, these two reports agreed that at similar training intensities, conventional therapy is more effective than robot-assisted therapy with exoskeleton devices for recovery of gait function after stroke.

g interventions appeared to be more effective than robot-assisted gait training for improving walking ability. Therefore, these two reports agreed that at similar training intensities, conventional therapy is more effective than robot-assisted therapy with exoskeleton devices for recovery of gait function after stroke. However, other reports documented similar or superior effects of robot-assisted therapy in combination with conventional physiotherapy versus conventional therapy alone on gait recovery, especially in patients with subacute stroke.21,27 In 2009, a study by Schwartz et al.49 with a larger number of participants concluded that locomotor therapy by using robot devices in combination with regular physiotherapy produced promising effects on gait function in patients with subacute stroke in comparison with regular physiotherapy alone.27 Therefore, robot-assisted therapy with exoskeleton devices may not be able to replace conventional physiotherapy for improving gait function in patients with stroke but rather is recommended for use in combination with conventional physiotherapy, preferably in the subacute stage of stroke. However, there is insufficient research on the additional effect of robot-assisted therapy on gait function in the chronic stage of stroke.

py for improving gait function in patients with stroke but rather is recommended for use in combination with conventional physiotherapy, preferably in the subacute stage of stroke. However, there is insufficient research on the additional effect of robot-assisted therapy on gait function in the chronic stage of stroke. Robot-assisted therapy for upper limb and hand motor function End-effector-type robotic devices Fourteen randomized controlled trials comparing robot-assisted therapy that use end-effector-type devices with conventional therapies for improvement of upper limb motor function after stroke were selected for review (Table 2).29-41 The meta-analysis in a 2012 Cochrane review demonstrated that robot-assisted arm training improved upper limb function (standardized mean difference 0.45; 95% confidence interval (CI), 0.20 to 0.69; P=0.0004).42 However, more detailed analysis is needed to develop guidelines for individual stroke rehabilitation. A study by Fasoli et al.32 comprising 56 patients with subacute stroke reported that patients who received conventional therapy alone showed little improvement, whereas patients who received robotic training plus conventional therapy continued to improve in the latter half of the inpatient rehabilitation period. This means that robot-assisted therapy is effective for improving upper limb motor function in patients with subacute stroke. A study by Lo et al.36 that recruited 127 chronic stroke patients reported that robot-assisted therapy and conventional therapy produced similar amounts of improvement after 12 weeks of treatment. However, after 36 weeks of therapy, the robot-assisted therapy achieved greater motor improvement than did conventional therapy. A study in patients with chronic stroke by Hsief et al.34 also found significantly greater improvement in upper limb motor function in the higher-intensity robot-assisted training group than in the control treatment group. In contrast, upper limb motor recovery did not differ significantly between the lower-intensity training group and the control group. These findings suggest that the intensity is the most important parameter of robot-assisted therapy for upper limb motor recovery in patients with chronic stroke.

ontrol treatment group. In contrast, upper limb motor recovery did not differ significantly between the lower-intensity training group and the control group. These findings suggest that the intensity is the most important parameter of robot-assisted therapy for upper limb motor recovery in patients with chronic stroke. Nine of the 14 randomized controlled trials that examined robot-assisted therapy with end-effector-type devices assessed the influence of robot-assisted training on activities of daily living (ADL) in patients with stroke.29,30,32-35,38-41 These nine reports demonstrated that robot-assisted training yielded similar or better effects on ADL in comparison with conventional therapy. The 2012 Cochrane review meta-analysis demonstrated that robot-assisted arm training improved ADL performance (SMD, 0.43; 95% CI, 0.11 to 0.75; P=0.009).42 In addition, studies in patients with subacute stroke suggested that patients who received additional robotic therapy showed greater improvements in ADL.32,40 However, trials in patients with chronic stroke demonstrated no additional improvement in ADL over conventional therapy.34 In summary, robot-assisted therapy for upper limb motor function provides an additional effect on ADL function only in patients with subacute stroke. Further studies may be needed to draw a definite conclusion about the effect of robot-assisted training on ADL in patients with chronic stroke.

er conventional therapy.34 In summary, robot-assisted therapy for upper limb motor function provides an additional effect on ADL function only in patients with subacute stroke. Further studies may be needed to draw a definite conclusion about the effect of robot-assisted training on ADL in patients with chronic stroke. Three randomized controlled trials concerning hand motor function in patients with stroke were selected for review (Table 3).43-45 All three studies showed similar or superior effects of robot-assisted training in comparison with conventional therapy on hand motor function in patients with stroke. Hwang et al.45 demonstrated that robot-assisted therapy provided dose-dependent improvement in hand function. However, all three trials were single-center studies with relatively small numbers of participants, all in the chronic stage of stroke, and there was no randomized controlled trial that included subacute stroke patients as participants Furthermore, there was no assessment of ADL function after robot-assisted therapy for hand motor function. Therefore, these results suggest that robot-assisted therapy with end-effector devices may yield similar or greater improvement in hand motor function in patients with chronic stroke, but there is insufficient research to support an effect in patients with subacute stroke. Therefore, well-designed studies are needed to draw clear conclusions regarding the effect of robot-assisted therapy that use end-effector-type devices on improvement of the hand motor function of patients in both the subacute and chronic stages of stroke.

research to support an effect in patients with subacute stroke. Therefore, well-designed studies are needed to draw clear conclusions regarding the effect of robot-assisted therapy that use end-effector-type devices on improvement of the hand motor function of patients in both the subacute and chronic stages of stroke. Exoskeleton-type robot devices Four randomized controlled trials of robot-assisted therapy with exoskeleton devices for improvement of upper limb motor function after stroke were selected for review (Table 2).46-49 All 4 trials were performed in patients in the chronic stage of stroke. Among them, one study reported a significantly better effect on spasticity in the robot-assisted therapy group than in the conventional therapy group.46 In contrast, ADL function improved more markedly in the conventional therapy group that received the same amount of treatment. The other three reports demonstrated no significant difference between robot-assisted therapy with exoskeleton devices and conventional therapies.47-49 In addition, there was no randomized controlled trial that investigated robot-assisted therapy with exoskeleton devices in patients with subacute stroke. Therefore, at this time there is insufficient evidence to draw a definite conclusion regarding the effect of robot-assisted therapy that uses exoskeleton devices on upper limb function in patients with stroke.

lled trial that investigated robot-assisted therapy with exoskeleton devices in patients with subacute stroke. Therefore, at this time there is insufficient evidence to draw a definite conclusion regarding the effect of robot-assisted therapy that uses exoskeleton devices on upper limb function in patients with stroke. Two randomized controlled trials that examined robot-assisted therapy with exoskeleton devices for improving the hand motor function of patients with stroke were selected (Table 3).50,51 Both studies showed similar or better results on hand motor function in comparison with conventional therapy. However, neither trial recruited patients in the subacute stage of stroke or assessed the effect of robot-assisted therapy on ADL function. In summary, robot-assisted therapy that uses exoskeleton devices may provide similar or additional benefits for hand motor function in comparison with conventional therapy in patients with chronic stroke, but there is insufficient evidence regarding the effect of robot-assisted therapy with exoskeleton devices on the hand motor function of patients in the subacute stage of stroke.

s may provide similar or additional benefits for hand motor function in comparison with conventional therapy in patients with chronic stroke, but there is insufficient evidence regarding the effect of robot-assisted therapy with exoskeleton devices on the hand motor function of patients in the subacute stage of stroke. Conclusions Numerous recent studies have heralded the introduction of robotic devices into the field of stroke rehabilitation. Many reports have described the efficacy of robot-assisted therapy for improving motor and ambulatory function in patients with stroke. However, both ethical and methodological constraints hinder the design of double-blind randomized controlled studies of robot-assisted therapy in patients with stroke. Furthermore, there are only a few well-organized comprehensive reviews of robot-assisted therapy.13,42,52,53 Meta-analysis of robot-assisted therapy is very difficult because of the heterogeneity of the robotic devices and the participants' characteristics as well as the diversity of the study designs in the literature. Therefore, it is important to consider expert opinion as well as research data in order to draw the best conclusions. In this review, we made an effort to analyze the effects of different types of robotic devices on upper limb and hand motor function as well as gait function. In summary, the role of robot-assisted therapy in stroke rehabilitation is currently an adjunct to rather than a replacement for conventional rehabilitation therapy. Well-designed studies with large numbers of participants that demonstrate superior efficacy for motor recovery will be necessary to establish robot-assisted therapy as an integral part of stroke rehabilitation. Analysis of the economic impact as well as the functional benefits of robot-assisted therapy is also needed. Robot-assisted therapy for stroke rehabilitation is in a dynamic phase of development and has achieved remarkable advances. Ongoping improvement of the robotic technology may enhance the efficacy and reduce the cost of such devices. Such advances will elevate robot-assisted therapy to a standard therapeutic modality in stroke rehabilitation.

for stroke rehabilitation is in a dynamic phase of development and has achieved remarkable advances. Ongoping improvement of the robotic technology may enhance the efficacy and reduce the cost of such devices. Such advances will elevate robot-assisted therapy to a standard therapeutic modality in stroke rehabilitation. This article was supported by a KOSEF grant (M10644000022-06N4400-02210) funded by the Korean government. The authors have no financial conflicts of interest. Figure 1 Examples of robotic devices for motor training (A) End-effector type (InMotion 2.0 Interactive Motion Technologies, Watertown, MA, USA), (B) Exoskeleton type (Armeo®, Hocoma, Switzerland). Table 1 Robot-assisted therapy for gait function Table 2 Robot-assisted therapy for upper limb motor function Table 3 Robot-assisted therapy for hand motor function

Introduction The possibility of repeated thrombolysis has increased because of the widespread use of thrombolytic treatments, an improved chance of survival after thrombolytic treatment, and an increase in life expectancy.1,2 In myocardial infarction and pulmonary embolism, repeated thrombolytic therapy using intravenous recombinant tissue plasminogen activator (rt-PA) has been reported to be safe and effective.3-5 However, repeated thrombolytic therapy in patients with acute ischemic stroke is rarely reported, and almost all previous reports were of patients who received intravenous rt-PA.2,6-8 Only 1 case report detailed a successful endovascular treatment for recurrent basilar artery occlusions.9 Recently, multimodal thrombolytic therapy, which includes intravenous and intra-arterial thrombolytic drugs and mechanical thrombectomy, has been introduced and is fre quently used by stroke teams.10-13 The goals of this study are to 1) identify how frequently repeated thrombolytic treatments are performed since the existence of multimodal thrombolytic treatments, and 2) characterize the safety and outcome of repeated thrombolytic therapy in patients with acute ischemic stroke.

s fre quently used by stroke teams.10-13 The goals of this study are to 1) identify how frequently repeated thrombolytic treatments are performed since the existence of multimodal thrombolytic treatments, and 2) characterize the safety and outcome of repeated thrombolytic therapy in patients with acute ischemic stroke. Methods Patients and enrollment We drew subjects for this study from the Yonsei Stroke Registry.14 We selected patients with acute ischemic stroke who had received thrombolytic treatments within a 10-year period (from August 2001 to July 2011). Frequency of thrombolysis was determined in each patient, and initial stroke severity was assessed by National Institutes of Health Stroke Scale (NIHSS) scores. Potential cardiac sources of embolism were defined according to the Trial of ORG 10172 in the Acute Stroke Treatment classification.15 This study was approved by the Severance Hospital Institutional Review Board of Yonsei University Health System (4-2012-0553).

National Institutes of Health Stroke Scale (NIHSS) scores. Potential cardiac sources of embolism were defined according to the Trial of ORG 10172 in the Acute Stroke Treatment classification.15 This study was approved by the Severance Hospital Institutional Review Board of Yonsei University Health System (4-2012-0553). Thrombolytic therapy The detailed protocol for thrombolytic treatment has been previously reported.16-19 Thrombolytic treatment was performed using intravenous rt-PA (Actilyse, Boehringer Ingelheim, Germany), intra-arterial urokinase (Urokinase, Yuhan, Seoul, Korea), or intra-arterial mechanical devices (microwire, Agility 10, Cordis, Miami, Fla., USA; Penumbra, Alameda, CA, USA; Solitaire, ev3 Inc, Irvine, CA, USA). Inclusion and exclusion criteria for thrombolytic treatments were based on previous trials.20,21 Patients who could be treated within 3 hours after the onset of symptoms received intravenous rt-PA (0.9 mg/kg with 10% bolus injection, followed by continuous infusion of the remainder over 60 minutes). After intravenous rt-PA infusion, further treatment with intra-arterial urokinase or mechanical thrombectomy was permitted for patients who showed an unsatisfactory clinical response (improvement on the NIHSS score <50%).19,22 Patients who could be treated within 3-6 hours after symptom onset were considered for intra-arterial urokinase (up to 1 million units) or mechanical thrombectomy. Abciximab was also allowed in patients with reocclusion.16

tients who showed an unsatisfactory clinical response (improvement on the NIHSS score <50%).19,22 Patients who could be treated within 3-6 hours after symptom onset were considered for intra-arterial urokinase (up to 1 million units) or mechanical thrombectomy. Abciximab was also allowed in patients with reocclusion.16 Assessment of outcomes Recanalization was evaluated at 24±4 hours after thrombolysis using magnetic resonance angiography or computed tomography angiography. The Thrombolysis in Cerebral Infarction (TICI) grading system (Grade 0: no perfusion; Grade 1: penetration with minimal perfusion; Grade 2a: partial filling 2/3 of the entire vascular territory; Grade 2b: complete filling, but the filling is slower than normal; Grade 3: complete perfusion) was used to assess the success or failure of recanalization.23 Successful recanalization was defined as TICI ≥2a. Hemorrhagic transformations were assessed by computed tomography or magnetic resonance imaging at 24±4 hours after thrombolysis, or whenever clinical deterioration was suspected. Symptomatic hemorrhagic transformation was defined as any increase in the NIHSS score that could be attributed to intracerebral hemorrhage (ICH) on brain imaging studies.20 Independent neurologists determined functional outcomes at 3 months with the modified Rankin scale (mRS). A good outcome was defined as an mRS score ≤2.

omatic hemorrhagic transformation was defined as any increase in the NIHSS score that could be attributed to intracerebral hemorrhage (ICH) on brain imaging studies.20 Independent neurologists determined functional outcomes at 3 months with the modified Rankin scale (mRS). A good outcome was defined as an mRS score ≤2. Statistical analysis Data were analyzed using SPSS 18.0 (SPSS, Chicago, IL, USA) for Windows. The Pearson chi-square test or Fisher's exact test was used to compare frequencies. For continuous variables, data distributions were examined for normality using the Kolmogorov-Smirnov test. Provided that the data did not deviate from normal distribution, the mean and standard deviation were calculated and parametric tests (independent sample t-tests) were applied. For data that were not normally distributed, we report descriptive statistics as the median and interquartile range; these were compared using non-parametric Mann-Whitney U tests. A P-value less than 0.05 was considered significant.

viation were calculated and parametric tests (independent sample t-tests) were applied. For data that were not normally distributed, we report descriptive statistics as the median and interquartile range; these were compared using non-parametric Mann-Whitney U tests. A P-value less than 0.05 was considered significant. Results Characteristics of patients with repeated thrombolysis During the 10-year study period, a total of 437 patients with acute ischemic stroke underwent thrombolytic therapy. Of these, a total of 7 patients (1.6%) received thrombolytic therapy twice in chronologically separate events. Table 1 summarizes the demographic and clinical characteristics of the patients with repeated thrombolysis. The median age for patients who received repeated thrombolytic therapy was 71 years old, and 4 patients were female. The median time interval between the first and second thrombolytic treatment was 16 months (ranging from 6 days to 76 months). The median time from the onset of symptoms to thrombolytic therapy was 120 hours for the first thrombolytic treatment, and 125 hours for the repeated treatment. Initial median NIHSS scores were 12 for the first treatment and 15 for the repeated treatment. The occluded cerebral arteries were the middle cerebral arteries in 6 patients and the internal carotid artery in 1 patient. In the second attack, all of the patients experienced the occlusion in the same cerebral arteries as the first ischemic strokes. However, the right and left sides of the particular arteries were reversed for 4 patients in the second attack. In the first thrombolytic treatment, the treatment modalities were intravenous rt-PA in 4 patients, intra-arterial urokinase in 3 patients, and mechanical treatments in 3 patients. Multimodal treatments were performed in 2 patients. In the second thrombolytic treatment, 5 patients were treated with intravenous rt-PA, 3 with intra-arterial urokinase, and 5 with mechanical treatments. Four patients were treated using multimodal treatments. All patients who received repeated thrombolytic treatments had 1 or more potential sources of cardiac embolism. After the first thrombolysis, recanalization was achieved in all patients (TICI 3 in 2 patients and 2b in 4 patients) with no symptomatic ICHs. Five (71.4%) out of 7 patients showed good outcome after 3 months (mRS 0 in 4 patients, mRS 1 in 1, and mRS 4 in 2).

tments had 1 or more potential sources of cardiac embolism. After the first thrombolysis, recanalization was achieved in all patients (TICI 3 in 2 patients and 2b in 4 patients) with no symptomatic ICHs. Five (71.4%) out of 7 patients showed good outcome after 3 months (mRS 0 in 4 patients, mRS 1 in 1, and mRS 4 in 2). Similar to the first thrombolysis, after the second thrombolysis, recanalization was achieved in all patients and no symptomatic ICH occurred. All patients recovered to the baseline mRS score for the first thrombolytic treatment. Comparison between single and repeated thrombolysis To examine safety and outcome of patients with repeated thrombolysis, the data of 430 patients who were given a single thrombolytic treatment were compared. Age, sex, initial stroke severity, door to admission time, initial systolic blood pressure and glucose level, and risk factors including hypertension, diabetes mellitus, hypercholesterolemia, and smoking, were not different between patients with single and repeated treatments. The frequency of potential sources of cardiac embolism, however, was different (100% in the repeated thrombolytic treatment vs. 59.6% in the single treatment, P=0.045). Among the thrombolytic modalities, mechanical thrombolysis was more commonly performed in patients with repeated treatments (71.4% vs. 38.4%, P=0.027). Outcome variables, including recanalization, symptomatic ICH, and good outcome were similar (Table 2).

rombolytic treatment vs. 59.6% in the single treatment, P=0.045). Among the thrombolytic modalities, mechanical thrombolysis was more commonly performed in patients with repeated treatments (71.4% vs. 38.4%, P=0.027). Outcome variables, including recanalization, symptomatic ICH, and good outcome were similar (Table 2). Discussion This study demonstrates that repeated thrombolytic treatments for recurrent ischemic strokes is safe and feasible. In 10 years, only 1.6% of all thrombolyzed patients underwent repeated thrombolysis. All repeatedly treated patients had 1 or more sources of cardiac embolism and achieved recanalization without symptomatic ICH in the first thrombolytic treatment. On repeated thrombolysis, these patients were recanalized again without symptomatic ICH, and all recovered to the baseline neurological state of the first thrombolytic treatment. Compared to the patients who had a single treatment, the rate of mechanical thrombectomy was twice as high in patients with repeated treatments. The outcome of patients with repeated thrombolysis was similar that of patients with only one thrombolysis.

he baseline neurological state of the first thrombolytic treatment. Compared to the patients who had a single treatment, the rate of mechanical thrombectomy was twice as high in patients with repeated treatments. The outcome of patients with repeated thrombolysis was similar that of patients with only one thrombolysis. In acute myocardial infarction or pulmonary embolism, repeated thrombolysis has been reported to be safe and feasible.3-5 In the case of acute ischemic stroke, however, its effectiveness is largely unknown. The major risk of thrombolytic treatment in ischemic stroke is ICH.24 Because the brain-blood barrier is disrupted in previously infarcted tissue, repeated thrombolytic treatments may pose a greater possibility for ICH. Few investigators have reported multiple administrations of thrombolytic drugs for recurrent ischemic strokes.2,6-8 A case series of repeated thrombolytic therapy from a German stroke database showed that 9 (1.8%) out of 496 thrombolyzed patients were repeatedly treated with intravenous rt-PA for chronologically separate acute ischemic strokes over a 4-year period.2 In that study, the median time span between the first and second thrombolysis was 10 months (range, 3 to 48), 67% of patients had favorable outcomes, and no symptomatic ICH occurred. These rates of repeated treatments and outcomes are similar to the current study.

separate acute ischemic strokes over a 4-year period.2 In that study, the median time span between the first and second thrombolysis was 10 months (range, 3 to 48), 67% of patients had favorable outcomes, and no symptomatic ICH occurred. These rates of repeated treatments and outcomes are similar to the current study. Current guidelines recommend that a stroke in the prior 3 months is considered an exclusionary criterion for intravenous rt-PA.24 This recommendation, however, is derived from clinical trials for acute myocardial infarction, and the information on acute ischemic stroke is limited.25 Repeated use of rt-PA after 1 day theoretically will not lead to toxic plasma levels because the half-life of rt-PA is 6 min, and more than 80% of rt-PA is eliminated via urine within 18 h after administration.26,27 There are only 4 case reports of repeated thrombolytic therapy within 3 months, where the time interval between the first and second thrombolysis was 24 to 90 hours.7-9,28 Three of those patients received intravenous rt-PA, and the remaining patient underwent intra-arterial chemical and mechanical thrombolysis. All patients showed good outcomes without symptomatic ICH. In our data, 1 patient received the second thrombolysis within 6 days (patient 1). Cardioembolic occlusion occurred on the right internal carotid artery in the first stroke, and recurrent occlusion on the left internal carotid artery in the second stroke. Both thrombi were completely removed using the Solitaire devices, and no chemical thrombolytic agents were used. This patient fully recovered. Therefore, further study of patients with prior stroke in the previous 3 months is needed and may increase the number of eligible patients.

carotid artery in the second stroke. Both thrombi were completely removed using the Solitaire devices, and no chemical thrombolytic agents were used. This patient fully recovered. Therefore, further study of patients with prior stroke in the previous 3 months is needed and may increase the number of eligible patients. Immunologic reactions following repeated thrombolysis are another concern. In patients with myocardial infarction, repeat ed administration of streptokinase is not recommended because of its antigenicity.29 Although rt-PA and urokinase are endogenous proteins and are considered non-antigenic, several studies have reported the development of antibodies against rt-PA and allergic reactions after rt-PA administrations.30-32 Furthermore, activation of fibrinolysis may facilitate anaphylactoid reactions.33 Previous reports did not show any immunological reactions in patients who were given repeated treatments.2 Although we did not check titers of antibodies, our 5 patients who received repeated treatments with rt-PA or urokinase did not show any symptoms or signs of an immunological reaction.

e anaphylactoid reactions.33 Previous reports did not show any immunological reactions in patients who were given repeated treatments.2 Although we did not check titers of antibodies, our 5 patients who received repeated treatments with rt-PA or urokinase did not show any symptoms or signs of an immunological reaction. It is noteworthy that all patients who received repeated thrombolytic treatments had 1 or more sources of cardiac embolism, most of which were atrial fibrillation. Recurrent ischemic stroke after intravenous rt-PA in patients with sources of cardiac embolism is common.28,34,35 Higher recurrence of ischemic stroke in patients with cardiac embolic sources may be connected to the increased likelihood of receiving repeated thrombolysis for those patients. Currently, clinical trials of new anticoagulants yield favorable results and have been adopted for patients with non-valvular atrial fibrillation.36 Recent guidelines for thrombolytic treatments also allow intravenous rt-PA use in patients who are currently taking a direct thrombin inhibitor or factor Xa inhibitors with normal coagulation laboratory tests.24 Further study is needed to investigate proper treatments for recurrent stroke in patients with new anticoagulants.

guidelines for thrombolytic treatments also allow intravenous rt-PA use in patients who are currently taking a direct thrombin inhibitor or factor Xa inhibitors with normal coagulation laboratory tests.24 Further study is needed to investigate proper treatments for recurrent stroke in patients with new anticoagulants. Several limitations exist in this study. First, although our data were collected from consecutive patients of a single stroke team over 10 years, the number of patients with repeated thrombolytic treatment were small. Considering the similar frequency of repeated thrombolysis in a different stroke database,2 it can be assumed that repeated thrombolysis is rarely performed. Second, our data do not enable us to draw conclusions about the optimal thrombolytic modality in repeated cases. In this study, mechanical thrombectomy in repeated treatment was twice as likely to be performed. Thrombolytic modality in each case was decided according to predefined protocol; however, conditions including age, stroke severity, occluded cerebral arteries, underlying medical conditions, and medication may influence the selection of thrombolytic modality for repeated events. The advancement of mechanical devices for acute ischemic stroke may allow for more safe and favorable outcomes in repeated thrombolytic treatments.10,12 Third, we examined the safety and outcome of patients with repeated thrombolysis through comparison to those with single thrombolytic treatment. Caution must be taken in interpreting these results. Because the number of cases was small, the number of severe strokes was not large, and the thrombolysis modalities were heterogeneous, the negative results from this comparison cannot be regarded as evidence for the safety and feasibility of repeated thrombolytic treatments. Further prospective study is warranted.

lts. Because the number of cases was small, the number of severe strokes was not large, and the thrombolysis modalities were heterogeneous, the negative results from this comparison cannot be regarded as evidence for the safety and feasibility of repeated thrombolytic treatments. Further prospective study is warranted. Conclusion Our study investigated the safety and outcome of patients with repeated thrombolysis in the era of multimodal thrombolytic treatments. Although it is rarely performed, repeated thrombolysis appears to be safe and feasible for patients with recurrent acute ischemic stroke. Repeated thrombolytic therapy could be considered, even if patients have had previous history of thrombolytic treatments. This study was supported by a Korea Science and Engineering Foundation grant funded by the Korea government (2009-0069165), the Korea Healthcare Technology Research and Development Project, Ministry for Health, Welfare, and Family Affairs, Republic of Korea (HI10C2020), and a faculty research grant from Yonsei University College of Medicine for 2011 (6-2011-0095). The authors have no financial conflicts of interest. Table 1 Demographic characteristics of patients who received repeated thrombolytic treatment

This study was supported by a Korea Science and Engineering Foundation grant funded by the Korea government (2009-0069165), the Korea Healthcare Technology Research and Development Project, Ministry for Health, Welfare, and Family Affairs, Republic of Korea (HI10C2020), and a faculty research grant from Yonsei University College of Medicine for 2011 (6-2011-0095). The authors have no financial conflicts of interest. Table 1 Demographic characteristics of patients who received repeated thrombolytic treatment A-fib, atrial fibrillation; IA UK, intra-arterial urokinase; ICA, internal carotid artery; ICH, intracranial hemorrhage; IV rt-PA, intravenous recombinant tissue plasminogen activator; LAA, left atrial appendage; LA, left atrium; MCA, middle cerebral artery; MS, mitral stenosis; MVP, mitral valve stenosis; mRS, modified Rankin scale; NIHSS, the National Institute of Health Stroke Scale; TICI, Thrombolysis in Cerebral Infarction. Table 2 Comparison between patients with a repeated treatment and patients with single thrombolytic therapy Values are given as median with IQR or number with percentage. IA UK, intra-arterial urokinase; ICH, intracranial hemorrhage; IV rt-PA, intravenous recombinant tissue plasminogen activator; mRS, modified Rankin scale; NIHSS, the National Institute of Health Stroke Scale; PCSE, potent cardiac source of embolism; SBP, systolic blood pressure; TICI, Thrombolysis in Cerebral Infarction.

Introduction Vascular dementia (VaD) is caused by cerebrovascular disease that directly or indirectly damages the brain structures associated with cognitive functioning. Historically, the concept of VaD has been described since the 19th century when arteriosclerotic brain atrophy due to hardening of the arteries was perceived as the major cause of senile dementia.1,2 The importance of VaD was overshadowed in the past century by the emergence of Alzheimer's disease (AD), which has turned out to be the most common cause of dementia. Interest in VaD has been revived in recent years because vascular lesions have been continuously noted to be substantial contributors to the development of dementia by themselves or by their synergetic effects on the pathogenesis associated with AD.3,4 With accumulating evidences suggesting the role of vascular burden on cognitive functions, VaD has now evolved into the concept of vascular cognitive impairment (VCI), which encompasses not only VaD but also AD with cerebrovascular disorder (AD with CVD or mixed dementia) and VCI with no dementia (VCIND). Thus, VCI is a more broad term that comprises all the states of cognitive impairment associated with CVD.3-5 Around one-third of patients with AD demonstrate evidences of vascular pathology,6-8 while up to two-thirds of patients with CVD develop at least some degree of AD pathology in the brain.9,10 Subcortical ischemic vascular dementia (SIVD) is one of the VaDs which can have both vascular and AD associated pathology in the brain. SIVD is pathologically driven by stenosis and occlusion of small vessels, resulting in white matter ischemia and multiple lacunar infarctions in subcortical structures. SIVD is of particular interest as the relatively slow progression of symptoms and clinical manifestations of the disease often make the differentiation of it from AD difficult. In this article, the authors reviewed the emerging concepts of VaD/VCI and clinical manifestations, biomarkers, treatments, and preclinical models of SIVD based on the pathophysiologic mechanisms of the disease.

symptoms and clinical manifestations of the disease often make the differentiation of it from AD difficult. In this article, the authors reviewed the emerging concepts of VaD/VCI and clinical manifestations, biomarkers, treatments, and preclinical models of SIVD based on the pathophysiologic mechanisms of the disease. Emerging concepts in VaD and VCI VaD can be classified into the following: 1) multi-infarction dementia, 2) strategic infraction dementia, 3) hemorrhagic dementia, 4) mixed dementia, 5) SIVD, and 6) other forms of vascular dementia.5,11 The first three types of VaDs develop relatively suddenly because of acute cerebrovascular diseases and they demonstrate specific cortical or subcortical symptoms depending on the stroke-affected regions. The stepwise progression, fluctuation of symptoms, and focal neurological signs are suggestive of acutely developing VaD. However, the cognitive impairments associated with SIVD often show an insidious onset and progressive decline; thus, mimicking the course of AD.12 Motor symptoms are often not obvious and go unnoticed in this setting, further confusing clinicians. It is suggested that the AD pathologies are the leading cause of dementia and this can take precedence over the status of cerebral ischemia.12-15 A recent amyloid imaging study, however, has indicated that there exists a pure SIVD, in which significant subcortical white matter ischemic changes are present with no evidence of amyloid plaque deposition in the brain (Figure 1).16 Moreover, pure SIVD is more prevalent than previously thought. This condition is distinctive from AD or mixed dementia, which demonstrate fibrillar forms of amyloid deposition in the brain.16,17 Finally, other forms of VaD include dementia that is caused by heterogeneous etiologies (e.g., vasculitis, cerebral amyloid angiopathy, and hereditary diseases, such as cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy [CADASIL]).18-20

ms of amyloid deposition in the brain.16,17 Finally, other forms of VaD include dementia that is caused by heterogeneous etiologies (e.g., vasculitis, cerebral amyloid angiopathy, and hereditary diseases, such as cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy [CADASIL]).18-20 There is growing consensus on the concept of mild cognitive impairment (MCI) due to AD or very mild AD. It is now believed to be early phases of the pernicious accumulation of amyloid beta (Aβ) plaques and neurofibrillary tangles in the brain, with diverse ranges of severity in clinical symptoms.21,22 In contrast, the concept of VCI emphasizes the treatable and preventable components of dementia, including the vascular burden in the brain. As such, the concept of VCI has evolved to amplify the potential treatment of vascular factors in order to improve the cognitive impairments because some vascular components are modifiable to prevent additional cerebrovascular deterioration and to ameliorate the current or ongoing cognitive impairments. Even after cognitive impairment sets in, the modulation of vascular risk factors could provide secondary prevention and improvement of the cognitive deficits.5,11,13

ome vascular components are modifiable to prevent additional cerebrovascular deterioration and to ameliorate the current or ongoing cognitive impairments. Even after cognitive impairment sets in, the modulation of vascular risk factors could provide secondary prevention and improvement of the cognitive deficits.5,11,13 In line with the emerging concept of the VCI, cerebral small-vessel disease has drawn lots of interest in the field of AD. The prevalence and influence of cerebral small-vessel disease are considered to be significant in dementia, although underestimated so far.23 Evidence of small vessel disease found on magnetic resonance images (MRI; e.g., white matter hyperintensities or lacunar infarctions) can be observed in 20%-40% of community-dwelling older persons12 and the prevalence of cognitive impairment in association with small vessel disease has been estimated to be as high as 36%-67%.13,24 In addition, recent studies represent that AD also increased the incidence of stroke.14,15 Thus, it is worthwhile to compare SIVD to AD in terms of the pathophysiological mechanisms, clinical manifestations, biomarkers, treatment options, and preclinical models.

disease has been estimated to be as high as 36%-67%.13,24 In addition, recent studies represent that AD also increased the incidence of stroke.14,15 Thus, it is worthwhile to compare SIVD to AD in terms of the pathophysiological mechanisms, clinical manifestations, biomarkers, treatment options, and preclinical models. Pathophysiology of SIVD SIVD is mainly caused by widespread ischemic changes in the white matter or multiple lacunar infarctions in subcortical structures.13,23,25-28 Predominant arteriosclerotic leukoencephalopathy in the white matter is classically referred to as Binswanger's disease. This type of ischemic change in the brain is formed by widespread incomplete infarctions or hypoperfusion of the white matter due to critical stenosis in the cortical medullary branches. Pathologically, the white matter lesions demonstrate degeneration of the myelin sheath and axons and degradation of the oligodendrocytes without profound infarction or cystic necrosis. On the other hand, the lacunar infarctions (état lacunaire) are caused by perforating arteriolar occlusions in subcortical structures, including the thalamus, basal ganglia, and internal and external capsules. The possible mechanisms underlying the steno-occlusion of small vessels include increased resistance to blood flow, decreased autoregulation, disruption of the blood-brain barrier (BBB), dysfunction of the endothelium, and dilatation of the perivascular spaces.12,23,27 The ultimate disruption of the prefronto-subcortical circuits and thalamocortical circuits by the accumulation of white matter ischemic lesions and lacunar infractions is thought to be an underlying mechanism that result in the cognitive impairments in SIVD. In mixed dementia, which has both the vascular and AD pathologies in the brain, these alterations in cerebrovascular circulation have been described to contribute to the development of dementia by themselves or by a synergistic interaction with AD pathology in the brain.29 However, the concomitant pathology of AD, including cholinergic deficits and the pathologic accumulations of Aβ plaques and neurofibrillary tangles, has also been found in patients with alleged SIVD, which is also believed to accelerate cerebral ischemia in the brain.16

tic interaction with AD pathology in the brain.29 However, the concomitant pathology of AD, including cholinergic deficits and the pathologic accumulations of Aβ plaques and neurofibrillary tangles, has also been found in patients with alleged SIVD, which is also believed to accelerate cerebral ischemia in the brain.16 Clinical manifestations of SIVD and its associations with structural and functional brain images Along with cognitive impairments, the stepwise development of neurological deficits is not uncommon in patients with SIVD. These deficits include hemiparesis, dysphagia, dysarthria, pseudobulbar palsy, emotional incontinence, urinary incontinence, and parkinsonian features, such as short-stepped gait. Abnormal behaviors, including disinhibition and akinetic mutism, can also develop during the course of SIVD, if the prefronto-subcortical circuits are interrupted directly or indirectly.30 Because behavioral changes usually develop later in the course of AD, the early appearance of behavioral or psychological symptoms can suggest the existence of SIVD, particularly when they present in stepwise patterns. Neuropsychological tests can reveal impairments in attention, execution, set-shifting performances, and verbal fluency from the early phases of SIVD. However, patients with AD demonstrate more difficulties with memory, naming, and visuospatial functions from the beginning of the symptomatic phases. A recent amyloid positron emission tomography (PET) imaging study has shown that patients with AD having fibrillar forms of Aβ in the brain demonstrate more difficulties in verbal/visual memory-associated tasks, while those with pure SIVD without amyloid pathology in the brain exhibit more trouble during executive functions, such as phonemic fluency of the Controlled Oral Word Association Test and the Stroop color test (Table 1).17,31-33 Associations between the amount of white matter pathology and impairments in daily living activities have also been described.34

loid pathology in the brain exhibit more trouble during executive functions, such as phonemic fluency of the Controlled Oral Word Association Test and the Stroop color test (Table 1).17,31-33 Associations between the amount of white matter pathology and impairments in daily living activities have also been described.34 Beyond the signal changes visible on MRI, recent methodological advances have enabled further image-based analyses to investigate the integrity of normal-appearing white matter tracts and the mechanisms that underlie cognitive dysfunctions associated with white matter changes. Previous studies on patients with CADASIL described an increased mean diffusivity and decreased diffusion anisotropy in the areas of white matter hyperintensities, which suggests the disruption of the white matter tracts in SIVD.35,36 Although the interpretation is limited due to the small numbers of participants, previous studies on patients with SIVD reported perfusion defects in the bilateral pulvinar nuclei of the thalamus,37 caudate nucleus, and diverse cortical areas, including the cingulate, superior temporal, subcallosal gyri,38 and frontal lobes.39 These findings support the notion of frontosubcortical circuit disruption in patients with SIVD. In addition to structural imaging biomarkers, PET images obtained with amyloid tracers can also be very useful for demonstrating the underlying pathology of SIVD. Although recent guidelines from the Amyloid Imaging Task Force convened by the Alzheimer's Association and the Society of Nuclear Medicine and Molecular Imaging do not recommend using amyloid PET imaging as a routine test for the diagnosis of AD,40,41 amyloid PET imaging can be clinically useful for differentiating pure SIVD and mixed dementia, which would sometime be tough to distinguish and would pose diagnostic dilemmas. A new criterion that could help discriminate pure SIVD from mixed dementia is under way using a combination of clinical and MRI findings based on data from amyloid PET imaging.

cally useful for differentiating pure SIVD and mixed dementia, which would sometime be tough to distinguish and would pose diagnostic dilemmas. A new criterion that could help discriminate pure SIVD from mixed dementia is under way using a combination of clinical and MRI findings based on data from amyloid PET imaging. Preclinical models of SIVD Two different models of SIVD have been described in the literature. One is a mechanical model, in which microcoils are applied to the bilateral common carotid arteries to induce continuous hypoperfusion and ischemia in the white matter without apparent gray matter changes.42,43 This mouse model has demonstrated decreased hippocampal volume and impaired working memory deficits that are possibly due to damaged frontosubcortical circuits. Subsequent demyelination of the white matter tracks is typical in this model, which can mimic the demyelinating pathology of the SIVD. However, the lack of methods that can affect the actual small vessels is a fundamental limitation of this model, even though the manipulation of the carotid artery does mimic the pathological findings of SIVD. The other model of SIVD is a transgenic mouse model of CADASIL that expresses Notch3 mutations at R90C.44 As CADASIL shares many pathophysiological and clinical manifestations with SIVD, research on SIVD and CADASIL is closely associated. Mouse models of CADASIL also develop similar cerebrovascular changes in the brain as in humans.45,46 However, the limited volume of white matter in rodents is a drawback and mouse models of CADASIL do not demonstrate widespread white matter changes in the brain. Recent advances in technology will make the modulation of the small vessels easier and more reliable. If new mouse models of SIVD can be introduced, it will allow better understanding of the direct pathological burden of the CVD on cognition, which can be separated from other cognitive impairments driven by neurodegenerative etiologies, such as accumulation of Aβ plaques or neurofibrillary tangles.

r and more reliable. If new mouse models of SIVD can be introduced, it will allow better understanding of the direct pathological burden of the CVD on cognition, which can be separated from other cognitive impairments driven by neurodegenerative etiologies, such as accumulation of Aβ plaques or neurofibrillary tangles. SIVD Biomarkers Various results have been described in biomarker studies on patients with VaD/VCI. A recent longitudinal study has compared the baseline cerebrospinal fluid (CSF) biomarkers of four different groups: a MCI group that finally progressed to AD (MCI-AD), mixed dementia (MCI-MD), or SIVD (MCI-SIVD), and a MCI group that remained stable as MCI (MCI-MCI).47 The results from CSF Aβ and tau showed that the levels of Aβ were higher and the levels of total tau and the phosphorylated forms of tau (p-tau) were lower in the MCI-SIVD group than in the MCI-AD group. Interestingly, the levels of CSF biomarkers obtained in the MCI-MD group were in between the levels of the MCI-AD group and the MCI-SIVD group. The CSF levels obtained from the MCI-SIVD group were closer to the levels of the MCI-MCI group and the normal control group. The intermediate CSF levels of Aβ and tau in the SIVD group compared to the mixed dementia/AD and normal controls indicate that SIVD possibly shares a common pathophysiology with AD. When patients with VaD and AD are compared, the levels of CSF Aβ had a tendency to be lower and CSF tau had a tendency to be higher in patients with AD than in patients with VaD, but substantial overlaps existed between the groups.48,49 On the other hand, another study that compared the levels of CSF Aβ and tau among non-demented elderly persons who had mild, moderate or severe white matter hyperintensities, demonstrated no differences in the levels of CSF Aβand tau among the groups.50

ith VaD, but substantial overlaps existed between the groups.48,49 On the other hand, another study that compared the levels of CSF Aβ and tau among non-demented elderly persons who had mild, moderate or severe white matter hyperintensities, demonstrated no differences in the levels of CSF Aβand tau among the groups.50 In addition, other markers associated with BBB breakdown have been introduced as alternative biomarkers associated with SIVD. A compromised BBB can be a source of albumin leakage, which can increase protein concentration in the CSF as it crosses the BBB with less resistance. With this pathological evidence of BBB disruption in patients with SIVD, the increased levels of albumin have been described as supportive evidence of SIVD.51-53 Metalloproteinases are markers of neuroinflammation as they attack the basal lamina and tight junctions in blood vessels, in addition to the disruption of myelin.54 In comparison with AD, the levels of metalloproteinases-9 in the CSF were increased in patients with VCI.55,56 Another marker of white matter disruption is neurofilament, which functions as a cytoskeleton in large myelinated axons.57 The concentrations of the neurofilament light subunit were reported to be higher in subjects with SIVD.47,58 Similar findings were noted in nondemented subjects with severe white matter lesions.50 As briefly described in the previous section, advanced imaging technologies can also serve as biomarkers, which enable further discrimination of the underlying pathophysiologies of SIVD.

rted to be higher in subjects with SIVD.47,58 Similar findings were noted in nondemented subjects with severe white matter lesions.50 As briefly described in the previous section, advanced imaging technologies can also serve as biomarkers, which enable further discrimination of the underlying pathophysiologies of SIVD. Treatment of SIVD SIVD treatment strategies include both slowing or mitigating underlying small vessel disease progression and the improvement of clinical symptoms. Microinfarction pathology stands out as a significant and independent factor that contributes to brain atrophy and cognitive impairment.26,59,60 Therefore, as a secondary prevention of stroke by modulating small vessel disease, antiplatelet agents can be administered. However, the effectiveness of antiplatelet therapy for the primary prevention of VCI has not yet been established.61 A calcium-channel blocker, nimodipine, has been introduced as a SIVD treatment because of its vasodilation effects. Following potential success in early trials of nimodipine in patients with VaD, an intention-to-treat trial was conducted with a total number of 230 patients. In that study, treatment with nimodipine did not reach the primary endpoint defined by the researchers, although the treatment group demonstrated improved lexical production and better results on the Mini-Mental State Examination and Global Deterioration Scale.62,63 In addition, cholinesterase inhibitors have been introduced to manage cognitive symptoms. Cholinergic deficits in SIVD were described based on findings that the basal forebrain cholinergic nuclei are supplied by perforating arterioles and that the CA1 sector of the hippocampus is vulnerable to ischemic insults.64,65 Patients with CADASIL have also demonstrated deficits in cholinergic fibers. Accordingly, cholinesterase inhibitors have been introduced in the treatment of VaD for the cognitive impairments associated with cholinergic deficits but not so much for other cognitive dysfunctions.66 A recent randomized controlled drug trial on patients with CADASIL has also reported the potential beneficial effects of cholinesterase inhibitors on cognitive functions.67 In another study of CADASIL, treatment with the cholinesterase inhibitor, donepezil, did not improve the primary endpoints measured by the vascular Alzheimer's disease assessment scale-cognitive subscale (VADAS-Cog) scores.

reported the potential beneficial effects of cholinesterase inhibitors on cognitive functions.67 In another study of CADASIL, treatment with the cholinesterase inhibitor, donepezil, did not improve the primary endpoints measured by the vascular Alzheimer's disease assessment scale-cognitive subscale (VADAS-Cog) scores. However, a subgroup analysis indicated improvements in executive functions in the treatment group.67 Antidepressants, such as selective serotonin reuptake inhibitors, can also be used for the management of depression and atypical neuroleptics can be selectively used for behavioral changes, including agitation and aggression. Recently, modification of cardiovascular risk factors has been recommended for the prevention of AD and VCI, including the SIVD. The modifiable and preventive risk factors include hypertension, diabetes, and hypercholesterolemia. In addition, modifications of lifestyle, including education, diet, physical activity, alcohol consumption, smoking, obesity, and social support/networking, can also be considered for the management of VCI (Figure 2).61

The modifiable and preventive risk factors include hypertension, diabetes, and hypercholesterolemia. In addition, modifications of lifestyle, including education, diet, physical activity, alcohol consumption, smoking, obesity, and social support/networking, can also be considered for the management of VCI (Figure 2).61 Conclusions VaD is a heterogeneous disease entity that encompasses various conditions. SIVD is a relatively homogeneous vascular dementia that can mimic AD due to the slow progression of cognitive decline. Detailed medical histories and neuropsychological tests can reveal early impairments in frontal executive symptoms and behavioral or mood changes. Neuroimaging studies and CSF biomarkers can also help in diagnosing and differentiating SIVD from other causes of dementia. Proper and early diagnosis, with careful review of the patients' medical history, neurological findings, structural and/or functional neuroimaging, and information from biomarkers, will enable clinicians to select the proper management techniques for the secondary prevention of stroke and the symptomatic treatment of SIVD. An improved understanding of the pathophysiology of SIVD with diverse experimental models of SIVD as well as the development and validation of new biomarkers will make the diagnosis and treatment of SIVD more reliable and effective. Finally, well-designed longitudinal studies with imaging and biological biomarkers in SIVD and subcortical vascular MCI, such as the Clinical Research Center for Dementia of South Korea (CREDOS),68 the Amyloid PET Imaging for Subcortical Vascular Dementia study (AMPETIS),16 and the Korean Alzheimer's Disease Neuroimaging Initiative (K-ADNI), will shed light on efforts to better understand the pathophysiology and prognosis of SIVD.

r MCI, such as the Clinical Research Center for Dementia of South Korea (CREDOS),68 the Amyloid PET Imaging for Subcortical Vascular Dementia study (AMPETIS),16 and the Korean Alzheimer's Disease Neuroimaging Initiative (K-ADNI), will shed light on efforts to better understand the pathophysiology and prognosis of SIVD. Future directions Data regarding VaD, including SIVD, need to be collected, maintained as a database in a standardized fashion, and shared among researchers to better define and manage dementia. The standardization of vascular lesions in neuropathology and neuroimaging is well under way. More work remains to be done before the concept of SIVD is widely accepted and put actively into clinical practice. This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2013R1A1A1012925), NRF MRC grant funded by the Korea government (MSIP; 2008-0062286), the KIST Institutional Program (2E24242-13-110), a grant (2013-590) from the Asan Institute for Life Sciences (JHR), and a grant from CJ Pharma (2009-0497) (JL), Seoul, Korea. The authors have no financial conflicts of interest.

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2013R1A1A1012925), NRF MRC grant funded by the Korea government (MSIP; 2008-0062286), the KIST Institutional Program (2E24242-13-110), a grant (2013-590) from the Asan Institute for Life Sciences (JHR), and a grant from CJ Pharma (2009-0497) (JL), Seoul, Korea. The authors have no financial conflicts of interest. Figure 1 Representative imaging scans of patients with pure subcortical ischemic vascular dementia (PiB-PET negative) or mixed dementia (PiB-PET positive). The first two axial fluid-attenuated inversion recovery images represent a similar degree of white matter hyperintensities of representative subjects in the PiB-negative (A) and in the PiB-positive (B) group. The T1-coronal images in the third column demonstrate a similar degree of generalized cortical atrophy, including medial temporal atrophy, in both the PiB-negative (A) and PiB-positive (B) groups. The PiB-PET scans in the fourth column demonstrate PiB-negativity and PiB-positivity scans, with the red color representing more intracranial PiB retentions. Abbreviations: PiB, 11C-Pittsburgh compound B; PET, positron emission tomography. Modified from Neurology 77(1): 18-25, 2011.

ive (A) and PiB-positive (B) groups. The PiB-PET scans in the fourth column demonstrate PiB-negativity and PiB-positivity scans, with the red color representing more intracranial PiB retentions. Abbreviations: PiB, 11C-Pittsburgh compound B; PET, positron emission tomography. Modified from Neurology 77(1): 18-25, 2011. Figure 2 Pathophysiology of vascular cognitive impairment. Both vascular and neurodegenerative processes culminate into the development of vascular cognitive impairment. The intervention of modifiable vascular etiologies potentially mitigates and slows the cognitive impairments in patients with vascular cognitive impairments. The progression of Alzheimer's pathologies in the brain is not amenable to current therapeutic interventions. Images modified from J Korean Neurol Assoc 21(5): 445-454, 2003. Table 1 Cognitive functions in patients with Subcortical Ischemic Vascular Dementia (SIVD) and Alzheimer's Disease (AD)

Introduction The importance of intracranial atherosclerotic disease (ICAD) as a cause of stroke is underscored as compared to that of extracranial carotid stenosis and nonvalvular atrial fibrillation. There have been several studies with long-term follow-up data and randomized clinical trials in extracranial carotid stenosis and nonvalvular atrial fibrillation; the risk of stroke and treatment effects were evaluated separately in both asymptomatic (stroke-free) and symptomatic patients. On the contrary, ICAD was not considered or was lumped with extracranial carotid stenosis as an atherosclerotic stroke subtype in most clinical trials1 and current stroke guidelines.2 However, ICAD differs from extracranial atherosclerotic stroke in many aspects, including risk factors and stroke patterns.3-5

patients. On the contrary, ICAD was not considered or was lumped with extracranial carotid stenosis as an atherosclerotic stroke subtype in most clinical trials1 and current stroke guidelines.2 However, ICAD differs from extracranial atherosclerotic stroke in many aspects, including risk factors and stroke patterns.3-5 Large clinical trials of ICAD have recently evaluated the effectiveness of anticoagulation (the Warfarin Aspirin Symptomatic Intracranial Disease [WASID] trial) and stenting (the Stenting and Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis [SAMMPRIS] trial) to prevent thromboembolism and restore hemodynamic compromise, but failed to reduce major vascular events in patients with ICAD.6,7 The failure of these trials may be caused by limitations in the current understanding of ICAD. In this review, the various mechanisms of stroke will be discussed because treatment effects may differ among the different types of ICAD. In addition, recent advances in risk factors for ICAD are summarized because the trials drew attention to the importance of recognition and adequate control of risk factors for ICAD.

n this review, the various mechanisms of stroke will be discussed because treatment effects may differ among the different types of ICAD. In addition, recent advances in risk factors for ICAD are summarized because the trials drew attention to the importance of recognition and adequate control of risk factors for ICAD. Epidemiology and natural course of ICAD Knowledge of the epidemiology of ICAD is limited.8 There have been no data on the prevalence of ICAD in large clinical trials, and data are limited regarding the prevalence of asymptomatic ICAD in the general population. ICAD may be as prevalent as extracranial carotid artery disease. The population-based Rotterdam study has evaluated the prevalence of intracranial internal carotid artery calcification, a marker of intracranial atherosclerosis which was observed in over 80% of older, white subjects.9 In addition, a transcranial Doppler study showed that the prevalence of asymptomatic middle cerebral artery (MCA) stenosis ranged from 7.2%-30% among Asian patients who had vascular risk factors without a history of stroke or TIA.10 ICAD causes 30%-50% of strokes in Asia,11 and 8%-10% of strokes in North America,12 making it one of the most common causes of stroke worldwide.13

revalence of asymptomatic middle cerebral artery (MCA) stenosis ranged from 7.2%-30% among Asian patients who had vascular risk factors without a history of stroke or TIA.10 ICAD causes 30%-50% of strokes in Asia,11 and 8%-10% of strokes in North America,12 making it one of the most common causes of stroke worldwide.13 ICAD is more prevalent in Asians than in Westerners; the reason for this is unknown. Possible explanations include inherited susceptibility of intracranial vessels to atherosclerosis,14 acquired differences in the prevalence of risk factors,15,16 and differential responses to the same risk factors.17,18 Lifestyle may play a role in the racial-ethnic differences: the pattern of ischemic stroke is changing in Asian patients. With the westernized lifestyle, the number of extracranial cervical disease is rising.19,20 Lastly, it is also possible that patients with adult-onset moyamoya disease (MMD) are misclassified as having ICAD, which may partly explain the high prevalence of intracranial atherosclerosis in Asians. Ring finger 213 (RNF213) was recently identified as the strongest susceptibility gene for MMD in East Asian people by a genome-wide linkage analysis and an exome analysis.21,22 The number of patients with MMD was estimated to be more than 53,800 in East Asian populations.23,24 The prevalence of MMD has recently increased with more careful consideration of the disease and better diagnostic techniques;25 many patients may have been misclassified as having ICAD.

lysis and an exome analysis.21,22 The number of patients with MMD was estimated to be more than 53,800 in East Asian populations.23,24 The prevalence of MMD has recently increased with more careful consideration of the disease and better diagnostic techniques;25 many patients may have been misclassified as having ICAD. The natural history of asymptomatic vs. symptomatic ICAD differs from that of extracranial carotid disease. Compared to asymptomatic extracranial carotid disease, asymptomatic ICAD has a low risk of stroke in the stenotic arterial territory, while the risk is very high in patients with symptomatic ICAD, especially in patients with clinically significant hemodynamic stenosis, early after stroke.6,26 Several clinical and laboratory findings reportedly predict stroke in patients with ICAD. The WASID trial showed that a shorter time from stroke (≤17 days), severe stenosis (≥70%), baseline severe neurological deficits, poorly controlled hypertension, and elevated low-density lipoprotein cholesterol (LDL) levels are independent predictors for recurrence after stroke due to ICAD.27 Regardless, progression in stenosis28,29 and DWI lesion pattern (subcortico-cortical lesion and multiple lesions)30 are also associated with recurrent stroke in ICAD.

trolled hypertension, and elevated low-density lipoprotein cholesterol (LDL) levels are independent predictors for recurrence after stroke due to ICAD.27 Regardless, progression in stenosis28,29 and DWI lesion pattern (subcortico-cortical lesion and multiple lesions)30 are also associated with recurrent stroke in ICAD. Phenotypes and high-resolution MRI findings of ICAD From single, small subcortical perforator infarctions to multiple cortical infarctions, various radiologic stroke patterns are associated with ICAD.31,32 Stroke associated with ICAD occurs in association with various stroke mechanisms such as in situ thrombotic occlusion, artery-to-artery embolism, hemodynamic insufficiency, and branch occlusion (Figure 1). Patients with unstable intracranial plaque may show large territorial lesions via sudden thrombotic occlusion. Artery-to-artery embolism, which commonly causes multiple cortico-subcortical infarcts, can be detected by performing transcranial duplex monitoring. Branch occlusive disease (BOD) is one of the main stroke mechanisms of ICAD, which can be characterized by a milder degree of stenosis33 and comma-shaped infarcts extending to the basal surface of the parent artery.34

s multiple cortico-subcortical infarcts, can be detected by performing transcranial duplex monitoring. Branch occlusive disease (BOD) is one of the main stroke mechanisms of ICAD, which can be characterized by a milder degree of stenosis33 and comma-shaped infarcts extending to the basal surface of the parent artery.34 Two major features of intracranial atherosclerosis include: (a) atherosis caused by cholesterol deposition and inflammation, and (b) sclerosis secondary to endothelial dysfunction, leading to arterial stiffness.35 Risk factors and vessel wall pathology may differ between the two. An autopsy study showed that risk factors differed between those with intracranial-plaque vs. plaque-negative stenosis.36 Older age, male gender, and diabetes were commonly associated with the presence of intracranial plaques and stenoses. Interestingly, history of myocardial infarct was an independent risk factor for intracranial plaque but not for plaque-negative stenosis, whereas stroke history was associated with stenosis but not plaque.

er age, male gender, and diabetes were commonly associated with the presence of intracranial plaques and stenoses. Interestingly, history of myocardial infarct was an independent risk factor for intracranial plaque but not for plaque-negative stenosis, whereas stroke history was associated with stenosis but not plaque. Recently, high-resolution MR techniques have been used to evaluate the frequency and role of intracranial artery plaques in living patients with stroke. Patients with symptomatic (vs. asymptomatic) and non-BOD type (vs. BOD) ICAD had characteristic changes in (a) the wall area (larger plaques), (b) plaque signals (eccentric enhancement and heterogeneous signal intensity suggesting unstable plaque), and (c) remodeling patterns (positive remodeling suggesting outward expansion of the vessel wall).37-39 On the contrary, superiorly located MCA plaques (near to the orifices of penetrating arteries) are associated with BOD-type ICAD.40,41 Results of major clinical trials of ICAD There are three therapeutic strategies for ICAD: (a) antithrombotics, (b) intervention to prevent thromboembolism and restore blood flow, and (c) identification and control of risk factors. Most studies have focused on the prevention of thromboembolism, including the WASID trial6 and the FISS-tris trial,42 which compared the benefit of anticoagulants vs. aspirin, and the recent SAMMPRIS wingspan stenting trial.7 However, no studies have been conducted to evaluate the effect of risk factor control in preventing stroke recurrence in ICAD patients.

on of thromboembolism, including the WASID trial6 and the FISS-tris trial,42 which compared the benefit of anticoagulants vs. aspirin, and the recent SAMMPRIS wingspan stenting trial.7 However, no studies have been conducted to evaluate the effect of risk factor control in preventing stroke recurrence in ICAD patients. Both oral and parenteral anticoagulation failed to show beneficial effects in preventing recurrent stroke in patients with ICAD.6,43 The WASID trial has shown that warfarin and aspirin were equally effective for preventing stroke or vascular death.6 In fact, both warfarin and aspirin were ineffective, given the high event rates in both arms. Risk factors were poorly controlled in the WASID trial. The most important findings in the WASID trial were related to the importance of controlling risk factors to reduce major vascular events in ICAD patients. Patients were poorly controlled in terms of mean systolic blood pressure and LDL level. Although this study did not examine the effect of risk factor control in symptomatic ICAD patients, the post hoc analyses suggest that lowering blood pressure and LDL may reduce major vascular events in ICAD patients.44

AD patients. Patients were poorly controlled in terms of mean systolic blood pressure and LDL level. Although this study did not examine the effect of risk factor control in symptomatic ICAD patients, the post hoc analyses suggest that lowering blood pressure and LDL may reduce major vascular events in ICAD patients.44 Thus, in the following SAMMPRIS trial, aggressive risk factor management was performed, targeting LDL below 70 mg/dL, systolic blood pressure below 140 mmHg, and a comprehensive lifestyle modification program.7 In the SAMMPRIS trial, the rate of stroke or death within the first 30 days was 14% in the Wingspan stenting arm vs. 5.8% in the aggressive medical management arm. The SAMMPRIS investigators stopped enrollment prematurely due to futility; the trend was not changed until 1 year. The higher stroke rate in the stenting arm than in aggressive medical management arm in SAMMPRIS was driven by several factors such as (a) inclusion of patients with perforator syndrome, a smaller vessel size, diffuse stenoses (oversizing of devices) in patients with high peri-procedural stroke risk (ischemic or hemorrhagic), (b) procedural considerations of stringent blood pressure control, general anesthesia, operator experience, etc., and (c) most importantly, improved medical treatment including high-dose statin therapy.45,46 It is interesting that the rate of stroke or death in the aggressive medical management arm was substantially lower than in historical (WASID) controls. This highlights the importance of aggressive control of risk factors. As shown in Table 1, the recurrence rate of stroke decreased in the recent clinical trials of ICAD. This is in line with the carotid intervention (endarterectomy) trials, which showed that in patients with asymptomatic carotid stenosis the stroke risk in the control group has lowered during the last 30 years with aggressive medical treatment including risk factor control.47 The annual risk rate of stroke in the control group of the recent trials has lowered even more than that in the carotid intervention group. Just as carotid endarterectomy is indicated only for high risk (annual stroke risk>1%) patients with asymptomatic carotid stenosis, intracranial stenting should be reserved for high risk patients/lesions to prevent disabling stroke refractory to a comprehensive regimen of medical therapy.

he carotid intervention group. Just as carotid endarterectomy is indicated only for high risk (annual stroke risk>1%) patients with asymptomatic carotid stenosis, intracranial stenting should be reserved for high risk patients/lesions to prevent disabling stroke refractory to a comprehensive regimen of medical therapy. Control of risk factors for ICAD Therefore, it is important to control vascular risk factors aggressively, and to find out risk factors more specific to ICAD. Until now, risk factors for ICAD of various conditions have been reported, from risk factors associated with asymptomatic ICAD to risk factors for stroke recurrence (Table 2). Classic risk factors for stroke such as hypertension, diabetes, and lipid disorders, are commonly associated with these conditions. However, many studies showed that these classic risk factors are not major determinants of the location of extracranial or intracranial atherosclerosis.48-50 Because ICAD risk could not be fully explained by conventional risk factors, there have been efforts to find risk factors specific to ICAD.

with these conditions. However, many studies showed that these classic risk factors are not major determinants of the location of extracranial or intracranial atherosclerosis.48-50 Because ICAD risk could not be fully explained by conventional risk factors, there have been efforts to find risk factors specific to ICAD. Controversial results regarding risk factors for ICAD could be caused by the following reasons. First, the current stroke classification system may have limitations in defining ICAD. We have recently reported that BOD-type ICAD patients were frequently misclassified as having SAD or cryptogenic stroke under the TOAST or revised (SSS) TOAST classification.33 Second, as mentioned earlier, various mechanisms of stroke exist related to ICAD (Figure 1), and the risk factors may differ depending on the stroke mechanisms. Lastly, there may be non-conventional risk factors for ICAD. Metabolic syndrome is a cluster of cardiovascular disease risk factors and metabolic alterations associated with excess fat. We have previously reported an association between metabolic syndrome and ICAD.48 Patients with more severe metabolic abnormalities were more likely to have severe intracranial atherosclerosis but not extracranial atherosclerosis, suggesting a dose-dependent relationship.49,51 These results were confirmed by other studies from different cohorts of asymptomatic multi-ethnic populations52,53 and symptomatic ICAD patients in a multi-center trial.54 A recent genetic study also supported this association.55 Risk factors, components of metabolic SD, elevated serum insulin, and adipokines secreted from adipocytes, all cause oxidative stress and endothelial dysfunction.56,51,18 Adults with the metabolic syndrome have suboptimal concentrations of several antioxidants, and intracranial arteries may become susceptible to oxidative stress.

s, components of metabolic SD, elevated serum insulin, and adipokines secreted from adipocytes, all cause oxidative stress and endothelial dysfunction.56,51,18 Adults with the metabolic syndrome have suboptimal concentrations of several antioxidants, and intracranial arteries may become susceptible to oxidative stress. Oxidative stress leads to the attenuation of endothelial function through decreased production of nitric oxide (NO) and increased destruction of NO by superoxide.57 The Asymptomatic Intracranial Atherosclerosis (AsIA) study showed that asymmetric-dimethylarginine (ADMA, an endogenous inhibitor of endothelial NO) was associated with ICAD.53 In addition, a recent report evaluating the circulating endothelial microparticle pattern in stroke patients showed that ICAD and ECAS may have different pathophysiologies.58 It was speculated that the endothelial activation is related to plaque instability in patients with extracranial arterial stenosis, and endothelial apoptosis is related to vascular narrowing in patients with ICAD.58

croparticle pattern in stroke patients showed that ICAD and ECAS may have different pathophysiologies.58 It was speculated that the endothelial activation is related to plaque instability in patients with extracranial arterial stenosis, and endothelial apoptosis is related to vascular narrowing in patients with ICAD.58 Treatment strategies in ICAD: endothelium, plaque, and platelet Several treatments could improve endothelial function. The most consistently reported treatment strategies that have restored endothelial function are statins, angiotensin converting enzyme inhibitors, phosphodiesterase inhibitors to enhance NO signaling (e.g. cilostazol and sildenafil), control of vascular risk factors, and lifestyle modification (exercise and smoking cessation). Several active clinical trials are currently seeking to restore the endothelial dysfunction. In particular, because the use of a high-dose statin reportedly improved arterial elasticity and reduced the plaque burden of carotid and coronary arteries (anti-atherogenic effects),59 several clinical trials of statins are ongoing in patients with ICAD. In addition, cilostazol reportedly protects endothelial function60 and prevents progression of intracranial stenosis.61 A recent multicenter clinical trial (the Trial of Cilostazol in Symptomatic Intracranial Arterial Stenosis [TOSS]-2 trial) found no significant difference in the vascular events between cilostazol and clopidogrel therapies in patients with symptomatic ICAD.62 In this study, the cilostazol group had a trend toward decreased progression of symptomatic intracranial stenosis and more regression of asymptomatic intracranial stenosis with a reduction in the level of apolipoprotein B, whereas the clopidogrel group showed a tendency toward fewer new lesions. A randomized clinical trial showed that combination therapy with clopidogrel and aspirin is more effective than aspirin alone in reducing microembolic signals in patients with acute symptomatic ICAD.63

reduction in the level of apolipoprotein B, whereas the clopidogrel group showed a tendency toward fewer new lesions. A randomized clinical trial showed that combination therapy with clopidogrel and aspirin is more effective than aspirin alone in reducing microembolic signals in patients with acute symptomatic ICAD.63 Therefore, it is possible that therapeutic target differs between hyperacute (platelet) vs. subacute/chronic stages (endothelium and plaque) after stroke in ICAD patients. In addition, treatment effects may differ among the different types of ICAD. For example, the use of aggressive antiplatelet agents and stenting (in selected patients) can be considered in patients with a higher degree of stenosis and non-BOD type infarcts, whereas stenting may be harmful (perforator stroke due to snowplowing effect) in patients with BOD. Strategies targeting anti-atherogenic effects and restoration of endothelial function may be more efficacious in the latter patients. Last but not least, public health measures may be particularly important in this subtype of stroke. The very recently reported Chinese IntraCranial AtheroSclerosis (CICAS) study showed a geographic and sex difference in the distribution of symptomatic ICAD in China, which may be explained by differences in risk factors such as obesity, alcohol/smoking habit, and diabetes.64 In addition, since the Asian population is known to be preferentially affected, focused trials need to be performed to establish treatment modalities that are most effective in this population.65

ICAD in China, which may be explained by differences in risk factors such as obesity, alcohol/smoking habit, and diabetes.64 In addition, since the Asian population is known to be preferentially affected, focused trials need to be performed to establish treatment modalities that are most effective in this population.65 Summary The failure of the major clinical trials of ICAD may be caused by limitations in the current understanding of ICAD. Various mechanisms are associated with stroke in patients with ICAD, as compared to those with extracranial carotid stenosis or atrial fibrillation. It is unlikely that one therapeutic strategy (such as stenting of a new device) will succeed in the near future because endothelium/plaque and platelets both play an important role in the development of stroke in patients with ICAD. Treatment strategies might be selected based on clinical features (including the time after onset) and serologic and neuroimaging biomarkers (including DWI pattern, plaque images, and microembolic signals). Although treatment effects may differ among the different types of ICAD, patients with all of these types of ICAD were included together in clinical trials of ICAD. Selection of target patients based on the ICAD types is needed in clinical trials of ICAD. In addition, recently developed MRI techniques (e.g. plaque image) and results of novel biomarkers (e.g. endothelial dysfunction) may extend our understanding of pathophysiological mechanisms of stroke in patients with ICAD. Further clinical trials using serologic and neuroimaging biomarkers as surrogate biomarkers would save time and money.

eveloped MRI techniques (e.g. plaque image) and results of novel biomarkers (e.g. endothelial dysfunction) may extend our understanding of pathophysiological mechanisms of stroke in patients with ICAD. Further clinical trials using serologic and neuroimaging biomarkers as surrogate biomarkers would save time and money. This study was supported by the Korean Healthcare Technology R&D Project, Ministry of Health & Welfare (A110208). The authors have no financial conflicts of interest.

eveloped MRI techniques (e.g. plaque image) and results of novel biomarkers (e.g. endothelial dysfunction) may extend our understanding of pathophysiological mechanisms of stroke in patients with ICAD. Further clinical trials using serologic and neuroimaging biomarkers as surrogate biomarkers would save time and money. This study was supported by the Korean Healthcare Technology R&D Project, Ministry of Health & Welfare (A110208). The authors have no financial conflicts of interest. Figure 1 Mechanisms of stroke in patients with ICAD. (A) Thrombotic occlusion is a rare phenotype of ICAD. Magnetic resonance angiography (MRA) shows in situ thrombotic occlusion at the site of stenotic plaque. DWI shows territorial infarcts by severe hemodynamic compromise and embolic infarcts on the cortex. High-resolution MRI can show vulnerable plaque on intracranial vessels. (B) Artery-to-artery embolism is one of common phenotypes of ICAD. Artery-to-artery embolism is usually associated with a severe degree of intracranial stenosis, and transcranial Doppler ultrasonography can detect symptomatic or asymptomatic embolism during microembolic signal monitoring. DWI shows small, scattered, cortical embolic infarcts. (C) Hemodynamic impairment is another phenotype of ICAD. This phenotype is usually associated with a severe stenosis and a marked hemodynamic compromise, as seen on a perfusion-weighted image (PWI). DWI typically shows borderzone-type infarcts, and infarct growth is common with clinical deterioration. (D) Branch occlusive disease is a common phenotype of ICAD. This phenotype is often misclassified as small arterial disease due to a mild degree of stenosis on MRA, small deep infarcts on DWI, and relatively small perfusion defects. High-resolution MRI can reveal plaque without stenosis near the orifices of penetrating arteries.

occlusive disease is a common phenotype of ICAD. This phenotype is often misclassified as small arterial disease due to a mild degree of stenosis on MRA, small deep infarcts on DWI, and relatively small perfusion defects. High-resolution MRI can reveal plaque without stenosis near the orifices of penetrating arteries. Table 1 Low rate of stroke in the recent clinical trials of ICAD ASA, aspirin; AMM, aggressive medical management; MRA. Magnetic resonance angiography. Table 2 Risk factors associated with ICAD

Introduction Evidence suggests that symptomatic steno-occlusion (SYSO) in acute ischemic stroke is associated with initial neurologic severity1 and stroke outcome.2-4 For the treatment of acute stroke, SYSO is not only the conceptual therapeutic target of intravenous thrombolysis5 but also the main target of intra-arterial thrombolysis,6 emergency endarterectomy,7 stent placement,8-10 and bypass operation. With an increased focus on SYSO, several studies have attempted to elucidate the impact of SYSO on stroke outcome. These studies detected SYSO in 30%-80% of patients with ischemic stroke,1,11,12 and its location and severity were related to mortality, functional outcome, and stroke recurrence.13-15 However, a widely variable detection rate and inconsistent findings regarding the effect of the location and severity of SYSO has prohibited translation of these important findings to the general stroke population. For example, a variable time interval from stroke onset to study inclusion and the use of different diagnostic modalities may provide inconsistent results. We hypothesized that therapeutic or spontaneous recanalization of the acute occlusion vessel and discrepancies among diagnostic methods in detecting the vascular pathology may also contribute to heterogeneous findings.16-18 In addition, studies that allow enrollment of only selected cases and are conducted at a single center inhibit the generalization of findings.

s recanalization of the acute occlusion vessel and discrepancies among diagnostic methods in detecting the vascular pathology may also contribute to heterogeneous findings.16-18 In addition, studies that allow enrollment of only selected cases and are conducted at a single center inhibit the generalization of findings. In this study, we used a nationwide multicenter stroke registry to investigate the prevalence and distribution of SYSO in patients with acute ischemic stroke within 24 hours of stroke onset.19 To allow for feasible and accurate results, magnetic resonance angiography (MRA) was used for the vascular workup. The distribution of SYSO was characterized by location, number, and severity. In addition, we evaluated the contribution of SYSO and its characteristics to functional outcome.

n 24 hours of stroke onset.19 To allow for feasible and accurate results, magnetic resonance angiography (MRA) was used for the vascular workup. The distribution of SYSO was characterized by location, number, and severity. In addition, we evaluated the contribution of SYSO and its characteristics to functional outcome. Methods Study subjects and data acquisition In this retrospective observational study, we enrolled a consecutive series of 5,393 patients into the Clinical Research Center for Stroke program; these patients were admitted to one of 9 nationwide university or community-based hospitals in the Republic of Korea between September 2008 and March 2010 because of acute ischemic stroke. The Clinical Research Center for Stroke program was designed to prospectively collect clinical data from patients who experienced a stroke and to build a web-based multicenter stroke registry. The Clinical Research Center for Stroke plans to investigate the clinical evidence in the field of stroke prevention, management, and prognosis; to provide practical information; and to eventually aid decision making based on evidence. The registry also includes information on management performance measures as part of a quality improvement program. Each participating hospital received human research approval to enroll cases without individual patient consent under a common rule as well as exemption from subsequent review by their institutional review boards.

try also includes information on management performance measures as part of a quality improvement program. Each participating hospital received human research approval to enroll cases without individual patient consent under a common rule as well as exemption from subsequent review by their institutional review boards. In total, 3,451 (64.0%) of the enrolled patients who visited the hospital within 24 hours after stroke onset were included. We excluded 371 patients (10.8%) who did not undergo diffusion-weighted imaging or cerebrovascular workup by MRA to evaluate the symptomatic lesion. After further excluding 23 patients who did not undergo assessment of functional outcome at discharge, 3,057 patients were included in the study. We extracted the following data from the registry: demographic profile, medical history, smoking, detected risk factors, baseline stroke severity assessed by the National Institutes of Health Stroke Scale (NIHSS) score, stroke subtype by Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification, thrombolysis, imaging findings, and functional outcome at discharge. A trained physician performed routine assessment of functional outcome with the modified Rankin Disability Scale (mRS) at discharge or transfer to a rehabilitation unit. Scores were categorized as good (mRS, 0-2) or poor (mRS, 3-6) outcome.

ification, thrombolysis, imaging findings, and functional outcome at discharge. A trained physician performed routine assessment of functional outcome with the modified Rankin Disability Scale (mRS) at discharge or transfer to a rehabilitation unit. Scores were categorized as good (mRS, 0-2) or poor (mRS, 3-6) outcome. Cerebral vasculature workup All images were obtained using whole body magnetic resonance imaging. Two-dimensional time-of-flight MRA was used for angiography. The imaging study was conducted on the day of admission, and the procedure was performed using the stroke protocol of each participating hospital.

ification, thrombolysis, imaging findings, and functional outcome at discharge. A trained physician performed routine assessment of functional outcome with the modified Rankin Disability Scale (mRS) at discharge or transfer to a rehabilitation unit. Scores were categorized as good (mRS, 0-2) or poor (mRS, 3-6) outcome. Cerebral vasculature workup All images were obtained using whole body magnetic resonance imaging. Two-dimensional time-of-flight MRA was used for angiography. The imaging study was conducted on the day of admission, and the procedure was performed using the stroke protocol of each participating hospital. Steno-occlusion of a cerebral artery was identified by the on-site neuroradiologist and reviewed by a neurologist. The stenosis was defined as a reduction in diameter or transient signal void with distal flow and categorized as either a mild (<50%) or a moderate to severe (≥50%) stenosis with some degree of modification.20 The occlusion denoted the loss of signal without distal flow. The investigated vessels were divided into 7 groups according to the location of the arterial lesion: extracranial internal carotid artery (EICA), intracranial internal carotid artery, middle cerebral artery (MCA), anterior cerebral artery, posterior cerebral artery, basilar artery (BA), and vertebral artery (VA). Ischemic lesions on diffusion-weighted magnetic resonance imaging with compatible clinical findings were classified according to the distribution of arterial territory. SYSO was diagnosed when stenosis or occlusion of the cerebral artery was observed with a relevant ischemic lesion. The number of SYSOs was categorized as the involvement of a single vessel or multiple vessels (≥2). Involvement of multiple vessels was categorized as ipsilateral multiple anterior circulation, multiple posterior circulation, bilateral anterior circulation, and both anterior and posterior circulation. For involvement of a single vessel, the severity of SYSO was categorized as mild, moderate to severe, and occlusion. Location was investigated by 2 methods: involvement of the anterior versus posterior circulation and individual arterial location.

teral anterior circulation, and both anterior and posterior circulation. For involvement of a single vessel, the severity of SYSO was categorized as mild, moderate to severe, and occlusion. Location was investigated by 2 methods: involvement of the anterior versus posterior circulation and individual arterial location. Statistical analysis For univariate statistical analysis, χ2 test and Student t test were used to study demographic characteristics and clinical variables and to study the relationship of variables with functional outcome. We chose the variables that were associated with the presence of SYSO (P<0.2) for covariates of the multivariate model and considered the clinical plausibility. The χ2 test and Student t test were used to analyze the relationship between functional outcome and SYSO characteristics, including existence, number, severity, and location. Multivariate logistic regression models were developed to measure the adjusted odds ratios (ORs) of SYSO and the characteristics of poor functional outcome. For comparison of effects among each group on the basis of the number and location of SYSOs, we applied a contrast method using a combination of estimated coefficients from the logistic model. For evaluation of the severity of SYSO, a likelihood ratio test was applied to compare the dose-response trends with the outcome. Statistical significance was defined as P<0.05. All statistical analyses were performed using SPSS version 17.0 software (Chicago, IL, USA).

f estimated coefficients from the logistic model. For evaluation of the severity of SYSO, a likelihood ratio test was applied to compare the dose-response trends with the outcome. Statistical significance was defined as P<0.05. All statistical analyses were performed using SPSS version 17.0 software (Chicago, IL, USA). Results Baseline characteristics of study subjects The mean age of the 3,057 study subjects was 67.3±12.5 years (male, 58.9%), and their median baseline NIHSS score was 4 (interquartile range, 2-10). In total, 15.1% of these subjects received thrombolytic therapy. At discharge, 1,401 subjects (45.8%) were assessed as having a poor functional outcome. Poor outcome was associated with age, sex, diabetes, high risk of a cardioembolic source, smoking, history of stroke, baseline NIHSS score, and thrombolysis (P<0.2) (Supplementary Table 1). Distribution and clinical properties of SYSO A total of 1,929 subjects (63.1%) had SYSO with more than mild stenosis. The detection rate of SYSO from stroke onset to hospital arrival within 3 hours, 6 hours, 12 hours, and 24 hours was 66%, 65.1%, 60.2%, and 58.0%, respectively. A comparison of the baseline characteristics of the study subjects based on the presence of SYSO is shown in Table 1. Patients with SYSO were older and had significantly higher proportions of diabetes, high risk of a cardioembolic source, history of stroke, and more severe baseline NIHSS score. In total, 50.8% of strokes were attributable to large artery disease and 25.6% were attributable to cardioembolic stroke.

is shown in Table 1. Patients with SYSO were older and had significantly higher proportions of diabetes, high risk of a cardioembolic source, history of stroke, and more severe baseline NIHSS score. In total, 50.8% of strokes were attributable to large artery disease and 25.6% were attributable to cardioembolic stroke. In patients with a single SYSO (n=1,339), MCA was the most affected vessel, followed by EICA, VA, and BA (Figure 1). Of the single SYSOs, 70.6% were located in anterior vessels. The characteristics of old age, female sex, and high risk of a cardioembolic source were significantly greater in those with an anterior vessel SYSO compared with a posterior vessel SYSO. In terms of severity, 25.5% of patients with a single SYSO had mild stenosis, 36.1% had moderate to severe stenosis, and 38.4% had occlusion. As shown in Figure 1, symptomatic arterial occlusion was frequently observed in the MCA and EICA and moderate to severe stenosis was most frequently observed in the BA. A total of 590 subjects (30.6%) with SYSO had involvement of multiple vessels. Compared with patients with a single SYSO, involvement of multiple vessels was related to old age and a more severe baseline NIHSS score but not with high risk of a cardioembolic source. Among the patients with multiple SYSOs, involvement of both the anterior and posterior circulation (44.4%) was the most commonly observed pattern, followed by involvement of the ipsilateral multiple anterior circulation (34.6%), multiple posterior circulation (13.7%), and bilateral anterior circulation (7.3%).

source. Among the patients with multiple SYSOs, involvement of both the anterior and posterior circulation (44.4%) was the most commonly observed pattern, followed by involvement of the ipsilateral multiple anterior circulation (34.6%), multiple posterior circulation (13.7%), and bilateral anterior circulation (7.3%). Impact of SYSO on functional outcome In total, 55.3% of patients with SYSO had a poor functional outcome, which was almost twice as high as that of patients without SYSO (Table 2). The multiple logistic model confirmed that the presence of SYSO was independently associated with a poor outcome at discharge (adjusted OR, 1.77; 95% confidence interval [CI], 1.46-2.15) (Table 2 and Supplementary Table 2). When the involvement of single and multiple vessels was analyzed separately in the multiple logistic model, both single SYSO (adjusted OR, 1.49; 95% CI, 1.22-1.83) and multiple SYSOs (adjusted OR, 2.71; 95% CI, 2.09-3.53) increased the risk of a poor outcome (Table 2). Additional analysis of the involvement of single and multiple vessels showed that the impact on functional outcome differed depending on the number of SYSOs (P for contrast <0.001).

ted OR, 1.49; 95% CI, 1.22-1.83) and multiple SYSOs (adjusted OR, 2.71; 95% CI, 2.09-3.53) increased the risk of a poor outcome (Table 2). Additional analysis of the involvement of single and multiple vessels showed that the impact on functional outcome differed depending on the number of SYSOs (P for contrast <0.001). Analysis of the relationship between outcome and anterior versus posterior location of SYSO revealed that a poor outcome was more prevalent when SYSO was in the anterior circulation than in the posterior circulation (55.2% vs. 37.1%). For individual arterial locations, the EICA, intracranial internal carotid artery, MCA, and BA showed a statistically significant crude OR for functional outcome (Table 2). However, the effect of the location on outcome disappeared in the multivariable logistic model (P for contrast=0.52), and an additional regression model introducing the individual arterial location showed similar results (P for contrast=0.21). Analysis of the relationship between the severity of a single SYSO and outcome showed that a moderate to severe degree of stenosis (OR, 1.41; 95% CI, 1.07-1.85) and occlusion (OR, 2.00; 95% CI, 1.51-2.65) of the cerebral vasculature significantly increased the risk of poor outcome compared with those without SYSO in the multivariable model. In addition, there was a significant dose-response relationship (P for trend <0.001). There was no additional effect of severity and location of a single SYSO (P for interaction effect=0.45).

rebral vasculature significantly increased the risk of poor outcome compared with those without SYSO in the multivariable model. In addition, there was a significant dose-response relationship (P for trend <0.001). There was no additional effect of severity and location of a single SYSO (P for interaction effect=0.45). Discussion In this study, 63% of patients with acute ischemic stroke had an MRA-defined steno-occlusion with a relevant ischemic lesion that independently affected the prognosis. Previous studies have observed vascular lesions in 30%-80% of patients with ischemic stroke.4,11,12,15 To translate this work to the general population, these discrepancies must be addressed. The time interval from stroke onset to hospital arrival is considered first. Two studies conducted during the hyperacute stage (3-6 hours) reported that approximately 80% of patients had vascular occlusions on MRA and digital subtraction arteriography.12,21 Other studies that included patients 1 day after stroke observed SYSO in approximately 30% of patients.11,15 Recanalization of the steno-occluded vessels during the hyperacute stage could explain the discrepancy between these studies, and rapid examination of the cerebral vasculature could be helpful in the detection of steno-occlusion.16

ncluded patients 1 day after stroke observed SYSO in approximately 30% of patients.11,15 Recanalization of the steno-occluded vessels during the hyperacute stage could explain the discrepancy between these studies, and rapid examination of the cerebral vasculature could be helpful in the detection of steno-occlusion.16 The lack of a consensus on the definition of SYSO and differences in diagnostic modalities also affect the detection rate of SYSO. In this work, we included more than a mild degree of symptomatic stenosis and occlusion detected by MRA. However, previous studies using Doppler ultrasonography detected 30% of symptomatic vessels with more than 50% stenosis or occlusion.11 Consistent with previous studies that showed the effect of ethnicity,22 involvement of the intracranial artery was observed in 72% of patients in this study; this is similar to the findings of a Chinese study23 and higher than that seen in Spanish (56.8%)11 and German (37.6%)15 studies. Interestingly, the ratio of stroke subtype in this study was similar to that in the German study (Table 1). After examining the risk factors for stroke, the most frequent cause of stroke in patients with SYSO was large artery disease, followed by cardioembolic stroke.

in Spanish (56.8%)11 and German (37.6%)15 studies. Interestingly, the ratio of stroke subtype in this study was similar to that in the German study (Table 1). After examining the risk factors for stroke, the most frequent cause of stroke in patients with SYSO was large artery disease, followed by cardioembolic stroke. Approximately 30% of patients with SYSO (19.3% of all subjects in the study) had 2 or more symptomatic vascular lesions. Compared with previous findings of multiple vascular lesions in 4%-18% of patients, an MRA-based study may detect a high proportion of cases of multiple SYSOs. In support of this, a previous study reported that 30% of patients with acute ischemic stroke had multiple ischemic lesions on diffusion-weighted magnetic resonance imaging.24 We found no significant difference in demographic profiles or clinical characteristics between patients with single and multiple SYSOs. The most frequent stroke subtype in patients with multiple SYSOs was large artery disease rather than cardioembolic stroke. In this study, SYSO at acute stages of ischemic stroke increased the relative risk of a poor outcome to 77%. This finding suggests that clinicians need to pay more attention to SYSO in addition to previously noted predictors such as lesion volume,25 stroke severity,26 and administration of various agents at acute stages.5,6

is study, SYSO at acute stages of ischemic stroke increased the relative risk of a poor outcome to 77%. This finding suggests that clinicians need to pay more attention to SYSO in addition to previously noted predictors such as lesion volume,25 stroke severity,26 and administration of various agents at acute stages.5,6 We investigated the effects of the characteristics of SYSO, including quantity, location, and severity, on the severity of outcome. The location of SYSO in terms of anterior versus posterior circulation and individual arterial location did not affect the stroke outcome. This finding was consistent with the post-hoc analysis of the Warfarin versus Aspirin for Symptomatic Intracranial Disease trial, which reported that the location of intracranial stenosis was not associated with recurrence of ischemic stroke.13 The individual arterial location also did not have a statistically significant effect on outcome. However, the results of the effect of individual arteries on stroke outcome should be interpreted with caution. Highly correlated relationships between the individual artery site and the baseline severity of stroke as previously described14,21 might have diminished the statistical power.

significant effect on outcome. However, the results of the effect of individual arteries on stroke outcome should be interpreted with caution. Highly correlated relationships between the individual artery site and the baseline severity of stroke as previously described14,21 might have diminished the statistical power. The severity of SYSO varied, depending on the individual artery. In the MCA and EICA, moderate to severe stenosis and occlusion were frequently observed and the proportion of poor outcomes was related to the severity of SYSO. In contrast, the BA and VA did not show this relationship. However, the effect of severity based on the arterial location did not show a significant interaction. We observed a relationship between severity and the risk of a poor outcome when considering a single SYSO. Finally, multiple SYSOs were associated with a worse outcome than a single SYSO even though there were few differences in the clinical characteristics. Further research to explain these results is warranted.

tion. We observed a relationship between severity and the risk of a poor outcome when considering a single SYSO. Finally, multiple SYSOs were associated with a worse outcome than a single SYSO even though there were few differences in the clinical characteristics. Further research to explain these results is warranted. This study had several limitations. First, it is difficult to assess the generalizability of this study because MRA was used exclusively. However, MRA is used in 96.8% of patients with ischemic stroke, which may resolve this limitation to some degree. Second, there is the possibility of a selection bias in this study because of limited hospital participation. However, the large number of patients nationwide from scattered community-based hospitals is believed to provide enough power to detect relatively large effects of SYSO on stroke outcome. Lastly, stroke outcome was observed at discharge or transfer, which would prohibit elucidating its effect on 3-month stroke outcome. In conclusion, we showed that approximately 60% of patients with acute ischemic stroke have SYSO and that many of these patients have multiple vessels involved. SYSO is an independent risk factor for poor clinical outcome, and this risk is affected by the number and severity of SYSOs. This study was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (HI10C2020). The authors have no financial conflicts of interest.

In conclusion, we showed that approximately 60% of patients with acute ischemic stroke have SYSO and that many of these patients have multiple vessels involved. SYSO is an independent risk factor for poor clinical outcome, and this risk is affected by the number and severity of SYSOs. This study was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (HI10C2020). The authors have no financial conflicts of interest. Supplementary Material Supplementary Table 1 Baseline characteristics of study subjects and comparisons by functional outcome at discharge Supplementary Table 2 Multivariable logistic model to evaluate the effect of SYSO to the stroke outcome Figure 1 Distributed pattern of SYSO at the acute stage of ischemic stroke and relationship to functional outcome. The pie graph in the upper panel shows the distribution of single SYSO (n=1,339). MCA was the most affected vessel, followed by EICA, VA, and BA. The bar graphs in the lower panel show the degree of steno-occlusion (x-axis) and functional outcome (y-axis). Outcome was assessed by mRS at discharge and categorized as good (blue; mRS, 0-2) or poor (green; mRS, 3-6). BA, basilar artery; EICA, extracranial internal carotid artery; IICA, intracranial internal carotid artery; MCA, middle cerebral artery; mRS, modified Rankin Disability Scale; SYSO, symptomatic steno-occlusion; VA, vertebral artery. Table 1 Baseline characteristics and acute stage treatment methods of study subjects

Figure 1 Distributed pattern of SYSO at the acute stage of ischemic stroke and relationship to functional outcome. The pie graph in the upper panel shows the distribution of single SYSO (n=1,339). MCA was the most affected vessel, followed by EICA, VA, and BA. The bar graphs in the lower panel show the degree of steno-occlusion (x-axis) and functional outcome (y-axis). Outcome was assessed by mRS at discharge and categorized as good (blue; mRS, 0-2) or poor (green; mRS, 3-6). BA, basilar artery; EICA, extracranial internal carotid artery; IICA, intracranial internal carotid artery; MCA, middle cerebral artery; mRS, modified Rankin Disability Scale; SYSO, symptomatic steno-occlusion; VA, vertebral artery. Table 1 Baseline characteristics and acute stage treatment methods of study subjects Values are presented as number (percent) unless otherwise noted. The relationships between no SYSO or SYSO and single or multiple SYSOs were analyzed separately. † and ‡ indicate P<0.05 by Student t test and χ2 test, respectively. Table 2 Impact of symptomatic steno-occlusion (SYSO) on poor functional outcome at discharge by multivariate analysis

Values are presented as number (percent) unless otherwise noted. The relationships between no SYSO or SYSO and single or multiple SYSOs were analyzed separately. † and ‡ indicate P<0.05 by Student t test and χ2 test, respectively. Table 2 Impact of symptomatic steno-occlusion (SYSO) on poor functional outcome at discharge by multivariate analysis Each characteristic variable of SYSO adopted the multivariable logistic model. The multivariate model used the covariates of age, sex, history of stroke, initial National Institutes of Health Stroke Scale score, diabetes mellitus, high risk of a cardioembolic source, smoking, and thrombolysis. † and ‡ indicate that P values were obtained using a linear combination of estimated coefficients from the logistic model and the likelihood ratio test, respectively. The anterior circulation was composed of the anterior cerebral artery, middle cerebral artery, and intracranial internal carotid artery, and posterior circulation was composed of the posterior cerebral artery, basilar artery, and vertebral artery.

Introduction Stroke is a sudden onset of focal neurological deficit, a major cause of morbidity and mortality and the second leading cause of death worldwide.1 Stroke has a heterogeneous etiology, caused by modifiable and un-modifiable risk factors. Recent studies have strongly suggested an association of deficiency of 25-hydroxyvitamin D with ischemic stroke2,3 and cardiovascular disease.4,5 Vitamin D is a 9,10-seco steroid and the most common forms in humans are vitamin D3 (cholecalciferol) and Vitamin D2 (ergocalciferol).2 Vitamin D is essential for the human body to maintain a balance between calcium and phosphorus. Inadequate vitamin D can cause weakness, reduced bone mineralization, increased bone loss and hip fracture6 and its prevalence is high in both hemispheric populations.7 Serum 25-hydroxyvitamin D is the circulating form of vitamin D with a half life of 2 to 3 weeks and is converted to the active form -1,25-dihydroxy vitamin D3 in the kidneys.8 25-hydroxyvitamin D is a marker of vitamin D status in the human body.9,10 Some population based studies have shown that 40%-45% of Indians have 25-hydroxyvitamin D deficiency in India.11 We aim to investigate the association between serum 25-hydroxyvitamin D deficiency and ischemic stroke and its subtypes in Indian patients. There is no study, so far, on the association of serum 25-hydroxyvitamin D deficiency with increased risk of stroke in Indian patients.

e 25-hydroxyvitamin D deficiency in India.11 We aim to investigate the association between serum 25-hydroxyvitamin D deficiency and ischemic stroke and its subtypes in Indian patients. There is no study, so far, on the association of serum 25-hydroxyvitamin D deficiency with increased risk of stroke in Indian patients. Methods Two hundred fifty patients with ischemic stroke enrolled consecutively in Yashoda hospital, which is a major referral center in the state of Andhra Pradesh. Two hundred fifty age and sex matched controls were recruited from healthy volunteers with no prior history of stroke or transient ischemic attacks. This cohort consisted of patient attendees and people volunteering for blood donation. The study period was between January 2011 and December 2012. This study is a part of ReVDAS (Role of 25-hydroxvitamin D in Acute Ischemic Stroke) study. This study was approved by the Institutional Ethics Committee and informed consent was obtained from controls and patients and if patients were severely ill, consent was taken from their relatives.

anuary 2011 and December 2012. This study is a part of ReVDAS (Role of 25-hydroxvitamin D in Acute Ischemic Stroke) study. This study was approved by the Institutional Ethics Committee and informed consent was obtained from controls and patients and if patients were severely ill, consent was taken from their relatives. Selection of cases Stroke patients met the following criteria: first ischemic stroke, admitted within 7 days of stroke onset. Stroke was defined according to the World Health organization as "rapidly developing clinical signs of focal/global disturbance of cerebral function, with no apparent cause other than of vascular origin."12 Cerebral infarction was diagnosed on the basis of the first Computer tomography (CT) or Magnetic Resonance Imaging (MRI) brain scan. If patients had a normal CT scan brain, then ischemic stroke was diagnosed based on diffusion weighted MRI.

of cerebral function, with no apparent cause other than of vascular origin."12 Cerebral infarction was diagnosed on the basis of the first Computer tomography (CT) or Magnetic Resonance Imaging (MRI) brain scan. If patients had a normal CT scan brain, then ischemic stroke was diagnosed based on diffusion weighted MRI. Stroke subtype analysis All stroke patients underwent CT scan of brain (to rule out hemorrhagic stroke) initially, followed by MRI of brain. Magnetic Resonance Angiography (MRA), Transthoracic echocardiography (TTE) or Transesophageal echocardiography (TEE), non-invasive vascular imaging (Carotid Doppler) were done in all patients. Ischemic stroke subtypes were classified as large artery atherosclerosis, cardioembolic stroke, small vessel disease (lacunar stroke), stroke of other determined etiology and stroke of undetermined etiology.13 If the etiology was not clear, then additional tests-serum fibrinogen, antithrombin III, protein C, protein S, antinuclear and anticardiolipin antibodies were also done. Lipid profile and serum homocysteine estimation were done in all patients. Standardized questions were modified from the behavioral risk factor surveillance system,14 by the Centers for Disease Control and Prevention. Height, weight, and blood pressure were measured, fasting blood samples were collected to estimate blood glucose, serum lipid profiles, calcium, phosphorus, alkaline phosphatase and C-reactive protein (CRP) for cases and controls.

ral risk factor surveillance system,14 by the Centers for Disease Control and Prevention. Height, weight, and blood pressure were measured, fasting blood samples were collected to estimate blood glucose, serum lipid profiles, calcium, phosphorus, alkaline phosphatase and C-reactive protein (CRP) for cases and controls. Definition for hypertension by JNC VII (Joint National Committee)-defined as a systolic blood pressure >140 mmHg and/or a diastolic blood pressure >90 mmHg based on the average of 2 blood pressure measurements at the time of admission, or a patient's self-reported history of hypertension or antihypertensive use, supported by documents.15 Diabetes was diagnosed if fasting plasma glucose was >110 mg/dL or patient was on anti-diabetic medications.16 As per guidelines of National Institute of Health (NIH), patients with serum cholesterol levels >200 mg/100 mL or those on anticholesterol medication were considered as having hypercholesterolemia.17 Smokers were defined as those reporting daily smoking. Ex-smokers and occasional smokers were classified as nonsmokers.18 Alcoholics were defined as those in whom the alcohol consumption was >50 g/day (equivalent to 500 mL [2 drinks] of wine, 1,000 mL of beer, or N5 drinks [units] of spirits).19 Body Mass Index (BMI) values from 25.0-30.0 were taken as overweight and BMI values >30 were taken as obese.20

s were classified as nonsmokers.18 Alcoholics were defined as those in whom the alcohol consumption was >50 g/day (equivalent to 500 mL [2 drinks] of wine, 1,000 mL of beer, or N5 drinks [units] of spirits).19 Body Mass Index (BMI) values from 25.0-30.0 were taken as overweight and BMI values >30 were taken as obese.20 Blood collection Blood collection was done at the time of enrollment of cases and controls; 5 mL blood sample was used for estimation of 25-hydroxyvitamin D. We used chemiluminescent microparticle immunoassay (CMIA) with automated instruments for estimation of 25-hydroxyvitamin D. Values≤20 ng/mL were diagnosed as 25-hydroxyvitamin D deficiency.10,21,22 Values from 11-20 ng/mL were considered mild and <10 ng/mL were diagnosed as severe vitamin D deficiency. Statistical analysis Statistical analysis was performed using SPSS 14.0 software (Chicago, IL, USA). Mean +SD (Stranded Deviation) were calculated. The paired 't' test was applied to test the differences in continuous variables and McNemar test was applied to study the association in proportions. We estimated odds ratio (OR) and the resulting 95% CI for the matched case-control pairs. Multiple logistic regression was performed before and after adjustment for potential confounders -age, gender, hypertension, diabetes, smoking, alcoholism, hyperlipidemia and obesity. All tests were two sided and P value <0.05 were considered statistically significant.

sulting 95% CI for the matched case-control pairs. Multiple logistic regression was performed before and after adjustment for potential confounders -age, gender, hypertension, diabetes, smoking, alcoholism, hyperlipidemia and obesity. All tests were two sided and P value <0.05 were considered statistically significant. Results Out of 250 subjects, men were 190 (76%), mean age was 58.4±11.1 years (age range - 24-89 years). Hypertension was significantly more common in stroke patients (144 [57.6%]) compared to controls (40 [26.6%]; P<0. 0001) as was diabetes (74 [49.3%] of stroke patients vs. 31 [24%] controls; P<0.0001). Positive CRP was noted in 156 (62.4%) stroke patients while seen only in 83 (33.3%) controls (P<0.0001). 25-hydroxyvitamin D deficiency was more prevalent in stroke patients 22 [48.8%]) than controls 79 (31.6%) which was statistically significant (P=0. 0001). Significantly decreased mean serum calcium (8.8±2.6) (mg/dL) and phosphorus (3.6±1.6) (mg/dL) levels were found in stroke patients compared to controls. Mean alkaline phosphatase level was significantly increased in stroke patients (112.1±38.6) (µ/L) compared to controls (85.5±21.5)(µ/L) (P<0.0001) (Table 1). We subdivided the 25-hydroxyvitamin D deficiency group into mild (10.1-20.0 ng/mL) and severe deficiency (below 10 ng/mL) in stroke patients and controls. We found significantly higher proportions in stroke patients, of both mild (65 [26%] vs. 45 controls [19.6%], P=0.002) and severe 25-hydroxyvitamin D deficiency (57 [22.8%] vs. 34 controls [12%], P=0.02).

eficiency group into mild (10.1-20.0 ng/mL) and severe deficiency (below 10 ng/mL) in stroke patients and controls. We found significantly higher proportions in stroke patients, of both mild (65 [26%] vs. 45 controls [19.6%], P=0.002) and severe 25-hydroxyvitamin D deficiency (57 [22.8%] vs. 34 controls [12%], P=0.02). We compared the values of samples taken during summer (samples were collected from March to September) and winter (samples were collected from October to February). There were no significant differences in prevalence of 25-hydroxy vitamin D deficiency between summer and winter samples in cases (P=0.5) and controls (P=0.3) (Table 2). But in both the seasons the prevalence of 25-hydroxyvitamin D deficiency was significantly higher among stroke patients compared to controls (<0.02). Among stroke subtypes, 25-hydroxyvitamin D deficiency was present in 50 patients (54.9%) with large artery atherosclerosis, 20 patients (54%) with cardioembolic stroke, 20 patients (44.4%) with small artery disease, 15 patients (42.8%) with stoke of other determined etiology and 17 patients (40.4%) with stroke of un-determined etiology. In the 20 patients with cardioembolic stroke and 25 hydroxyvitamin D deficiency, the underlying cardiac disease was varied and included history of myocardial infarction (4 patients), atrial fibrillation (3 patients) congestive heart failure (4 patients), akinetic left ventricular segment (2 patients ) ascending aorta stenosis (2 patients), mitral valve stenosis (2 patients), and rheumatic heart disease (3 patients).

diac disease was varied and included history of myocardial infarction (4 patients), atrial fibrillation (3 patients) congestive heart failure (4 patients), akinetic left ventricular segment (2 patients ) ascending aorta stenosis (2 patients), mitral valve stenosis (2 patients), and rheumatic heart disease (3 patients). On comparing the risk factors with stroke, univariate analysis demonstrated maximum risk with hypertension and diabetes followed by 25-hydroxyvitamin D deficiency. 25-hydroxyvitamin D deficiency showed an independent association with ischemic stroke (Table 3). On evaluation of stroke subtypes 25-hydroxyvitamin D deficiency was independently association with large artery atherosclerosis and cardioembolic stroke (Table 4).

On comparing the risk factors with stroke, univariate analysis demonstrated maximum risk with hypertension and diabetes followed by 25-hydroxyvitamin D deficiency. 25-hydroxyvitamin D deficiency showed an independent association with ischemic stroke (Table 3). On evaluation of stroke subtypes 25-hydroxyvitamin D deficiency was independently association with large artery atherosclerosis and cardioembolic stroke (Table 4). Discussion In our study, we found a significant association between 25-hydroxyvitamin D deficiency and ischemic stroke and established an independent association. Similar results have been found from the western part of the world.2,3,23-26 We noted deficiency of 25-hyroxyvitamin D in 54.9% of stroke patients with large artery atherosclerosis. A similar association of hypovitaminosis D with large artery atherosclerosis and small artery disease has been described earlier.25 A recent study showed low 25-hydroxyvitamin D was significantly associated with increasing intimal media thickness and carotid plaques in individuals.27 We also found a significant association of 25-hyroxyvitamin D deficiency with cardioembolic stroke. Several studies have shown a strong association of vitamin D deficiency with cardiovascular disease.28-30 Giovannucci et al.31 demonstrated low levels of 25-hydroxyvitamin D as a high risk factor for myocardial infarction. Lower 25-hydroxyvitamin D concentration was shown to be an independent risk factors for atherosclerosis, coronary calcification32 and cardiovascular death.33 However some studies have found no association between vitamin D and cardiovascular disease.34,35

ydroxyvitamin D as a high risk factor for myocardial infarction. Lower 25-hydroxyvitamin D concentration was shown to be an independent risk factors for atherosclerosis, coronary calcification32 and cardiovascular death.33 However some studies have found no association between vitamin D and cardiovascular disease.34,35 The mechanism of deficiency of vitamin D and atherosclerosis is not fully understood. Li et al.36 observed that vitamin D regulated blood pressure by suppressing the renin angiotensin system. Aihara et al.37 demonstrated vascular effects of vitamin D with inhibition of thormobosis38 and reduction in arterial calcification.39 In addition smooth muscle cells and lymphocytes express receptors for vitamin D and convert circulating 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D. 1,25-hydroxy vitamin D in turn reduce the proliferation of lymphocytes and the production of cytokines.39,40 This anti-inflammatory effect may have a protective role as there is increasing evidence that systemic inflammation leads to atherosclerosis.41

convert circulating 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D. 1,25-hydroxy vitamin D in turn reduce the proliferation of lymphocytes and the production of cytokines.39,40 This anti-inflammatory effect may have a protective role as there is increasing evidence that systemic inflammation leads to atherosclerosis.41 In our study we also observed significantly higher proportion of stroke patients with elevated levels of alkaline phosphatase and decreased phosphate levels compared with control subjects. This is advocated by Nibet et al.42 and Preece et al.43 Decreased serum phosphate and increased alkaline phosphatase are caused by Vitamin D deficiency.44,45 Alkaline phosphatase may contribute to atherosclerosis by promoting vascular calcification. Elevated alkaline phosphatase has also been shown to be an independent predictor of mortality after ischemic stroke.46 CRP an inflammatory marker and an independent risk factor for acute ischemic stroke.47,48 In our study we found significant association deficiency of 25-hydroxyvitamin D with CRP positive in stroke patients compared to controls. Recent studies have advocated that CRP levels are elevated in deficiency of vitamin D.31 This furthers emphasizes the role of vitamin D in reducing inflammation and thus reducing atherogenesis. The strength of this study is that both cases and controls were collected from a single center with same ethnic background and 25-hydroxyvitamin D analysis was done in one lab. We have also analyzed the association of 25-hydroxyvitamin D deficiency with stroke subtypes.

CRP an inflammatory marker and an independent risk factor for acute ischemic stroke.47,48 In our study we found significant association deficiency of 25-hydroxyvitamin D with CRP positive in stroke patients compared to controls. Recent studies have advocated that CRP levels are elevated in deficiency of vitamin D.31 This furthers emphasizes the role of vitamin D in reducing inflammation and thus reducing atherogenesis. The strength of this study is that both cases and controls were collected from a single center with same ethnic background and 25-hydroxyvitamin D analysis was done in one lab. We have also analyzed the association of 25-hydroxyvitamin D deficiency with stroke subtypes. Age may impact vitamin D levels. Gillor et al found that with increasing age 25-hydroxyvitamin D decreases,49 as capability of skin to produce pre-vitamin D after ultraviolet (UV) B irradiation declines with age.50 In our study this effect is nullified as both cases and controls were age and sex matched. We did not perform seasonal adjustment for vitamin D levels, as in our study. we did not find any significant difference between levels during winter and summer. Similar results were found in healthy subject in previous studies from India.51,52 This difference from west may be due to the fact that as a tropical country the temperature fluctuations are ±10 degrees centigrade and may not impact vitamin D levels.

any significant difference between levels during winter and summer. Similar results were found in healthy subject in previous studies from India.51,52 This difference from west may be due to the fact that as a tropical country the temperature fluctuations are ±10 degrees centigrade and may not impact vitamin D levels. Conclusions This study established that deficiency of 25-hydroxyvitamin D had an independent association with ischemic stroke-especially large artery atherosclerosis and cardioembolic stroke. A single serum measurement of this compound could be a useful marker in epidemiologic studies. The role of vitamin D deficiency as a direct causative factor of stroke has to be established for advocating vitamin D usage for stroke prevention. Large scale interventional studies are required to confirm these findings. Acknowledgements Our sincere thanks to the Dr.G.S. Rao, Managing Director, Yashoda group of hospitals and Dr.A.Lingaih, Director of Medical services for their generous support to carry out this study in Yashoda Hospital. Hyderabad. The authors have no financial conflicts of interest. Table 1 Baseline characteristics Table 2 Seasonal variation of 25-hydroxyvitmain D deficiency in patients and controls Table 3 Univariate and multivariate analysis of various risk factors with stroke Table 4 Univariae and multivariate analysis of the relationship between various stroke subtypes and serum 25-hydroxyvitamin D deficiency *Number of patients insufficient for statistical analysis.

Dear Sir, Although patients with acute ischemic stroke achieve recanalization, approximately 37%-65% of them do not show immediate improvement.1 The term "stunned brain syndrome" refers to delayed neurological recovery after an ischemic insult to the brain.1 We describe the case of a patient with delayed neurological recovery who had a small cerebral infarction but relatively rapid compensatory perfusion and high calcification burden in the intracranial arteries.

e term "stunned brain syndrome" refers to delayed neurological recovery after an ischemic insult to the brain.1 We describe the case of a patient with delayed neurological recovery who had a small cerebral infarction but relatively rapid compensatory perfusion and high calcification burden in the intracranial arteries. A 71-year-old woman visited our emergency room 1.5 hours after developing acute right limb weakness. The initial neurological examination revealed right hemiplegia (arm and leg, grade 2 on the Medical Research Council grading system) and global aphasia (National Institutes of Health Stroke Scale score, 22). Non-enhanced computed tomography did not reveal a definite ischemic or hemorrhagic lesion, and intravenous recombinant tissue plasminogen activator was subsequently infused. Computed tomography angiography revealed a moderate degree of stenosis in the left middle cerebral artery (Figure 1A) and a high calcification burden in almost all intracranial arteries (Figure 1A, B). We assumed that an occlusion of the middle cerebral artery might be in the process of recanalization. Immediate magnetic resonance imaging revealed a small acute infarction near the posterior pole of the lateral ventricle on diffusion-weighted imaging (Figure 1C). Perfusion magnetic resonance imaging performed simultaneously revealed luxury perfusion in the left posterior temporo-parieto-occipital area (Figure 1D). Although her neurological symptoms had improved 24 hours after onset, she still had right hemiparesis (arm and leg, grade 4) and dense sensory aphasia (National Institutes of Health Stroke Scale score, 8). The localization of these neurological deficits corresponded to the area of luxury perfusion. On the fourth day after admission, the small lesion persisted on diffusion-weighted imaging (Figure 1E), and the luxury perfusion had normalized (Figure 1F), but the neurological deficits including sensory aphasia remained (National Institutes of Health Stroke Scale score, 7). By the sixth day after admission, her neurological deficits had completely disappeared.

lesion persisted on diffusion-weighted imaging (Figure 1E), and the luxury perfusion had normalized (Figure 1F), but the neurological deficits including sensory aphasia remained (National Institutes of Health Stroke Scale score, 7). By the sixth day after admission, her neurological deficits had completely disappeared. The patient showed some immediate neurological recovery, but sensory aphasia persisted even though the diffusion-restricted lesion was small and the cerebral perfusion pattern was compensatory. We propose that this delayed neurological recovery was associated with vascular aging, which was reflected by the high intracranial arterial calcification burden. Calcium deposition in the vasculature is one of the main components of vascular aging,2 which is associated with poor adaptation following ischemia/reperfusion injuries.3 In addition, neovascularization following reperfusion is hampered by the vascular calcification process.4 Reperfusion may result in increased levels of reactive oxygen species that subsequently impair endothelial progenitor cell function.5 With regard to post-stroke repair mechanisms, endothelial progenitor cells are known to be essential for neovascularization.6,7 The delayed neurological recovery could be explained by these poor adaptations and hampered neovascularization following revascularization.

uently impair endothelial progenitor cell function.5 With regard to post-stroke repair mechanisms, endothelial progenitor cells are known to be essential for neovascularization.6,7 The delayed neurological recovery could be explained by these poor adaptations and hampered neovascularization following revascularization. The delayed neurological recovery in this patient could also have been associated with the luxury perfusion phenomenon; however, a previous report has shown a relationship with perfusion deficit. Hyperperfusion on perfusion magnetic resonance imaging is thought to reflect a high degree of ischemic stress.8 The hyperperfused area was reported to correspond to the area of cerebral infarction in this patient. When compared to patients with normal perfusion, those with hyperperfusion show lower apparent diffusion coefficient values.8 Moreover, delayed clinical recovery has been reported to be associated with perfusion deficits.1 A defect that extends beyond a diffusion-restricted area is referred to as an ischemic penumbra,9 which exhibits deranged cerebral metabolism and can be identified using perfusion imaging. The present case indicates that either hyperperfusion or hypoperfusion may be associated with delayed neurological recovery after cerebral arterial recanalization.8

ffusion-restricted area is referred to as an ischemic penumbra,9 which exhibits deranged cerebral metabolism and can be identified using perfusion imaging. The present case indicates that either hyperperfusion or hypoperfusion may be associated with delayed neurological recovery after cerebral arterial recanalization.8 In conclusion, the delayed neurological recovery observed in our patient might have resulted from either post-ischemic reperfusion syndrome associated with luxury perfusion, or an impaired post-stroke repair mechanism associated with a high burden of intracranial arterial calcification. Computed tomography and magnetic resonance imaging are both useful for estimating the degree of ischemic injury. Acknowledgements This work was supported by the new faculty research fund of the Ajou University School of Medicine. The authors have no financial conflicts of interest.

In conclusion, the delayed neurological recovery observed in our patient might have resulted from either post-ischemic reperfusion syndrome associated with luxury perfusion, or an impaired post-stroke repair mechanism associated with a high burden of intracranial arterial calcification. Computed tomography and magnetic resonance imaging are both useful for estimating the degree of ischemic injury. Acknowledgements This work was supported by the new faculty research fund of the Ajou University School of Medicine. The authors have no financial conflicts of interest. Figure 1 Initial computed tomography (CT) and magnetic resonance (MR) images of the patient in this study. (A) A coronal maximum intensity projection image from CT angiography shows significant stenosis of the distal left middle cerebral artery, which was lined with calcified plaques (arrow). High calcification burden is also seen along the bilateral distal internal carotid arteries (arrowhead). (B) A volume-rendered image from CT angiography shows a high calcification burden along all intracranial arteries (arrowhead). (C) Diffusion-weighted imaging performed on admission demonstrates a small lesion on the left parieto-occipital area (arrow). (D) A time-to-peak map of MR perfusion imaging performed on admission shows extensive luxury perfusion in the posterior temporo-parieto-occipital area. (E) Diffusion-weighted imaging demonstrates the persistence of the small lesion on the fourth day after admission. (F) The abnormal finding on the perfusion MR image had normalized by the fourth day after admission.

Dear Sir: The presence of clinical-diffusion mismatch (CDM) in patients with acute ischemic stroke may represent the ischemic penumbra, which requires emergency reperfusion therapy to improve stroke outcomes. We report here our experience of treating a patient with acute ischemic stroke and CDM. Specifically, the patient had a small infarct in the left temporal cortex and presented with sensory aphasia, but did not require reperfusion therapy. Magnetic resonance angiography and perfusion imaging findings were normal. Further investigation revealed that the aphasia was associated with ictal symptoms. Ictal aphasia is a considerable cause of non-oligemic CDM, perfusion imaging and angiographic studies may help discriminate true ischemic penumbra from non-oligemic CDM.

resonance angiography and perfusion imaging findings were normal. Further investigation revealed that the aphasia was associated with ictal symptoms. Ictal aphasia is a considerable cause of non-oligemic CDM, perfusion imaging and angiographic studies may help discriminate true ischemic penumbra from non-oligemic CDM. A 44-year-old, right-handed male patient presented with sudden onset language disturbance and headache for 24 hours. Approximately 30 years prior, the patient had undergone surgery to removal a right cerebellar tumor, which resulted in a history of mild dysarthria, hearing difficulties, right-sided ataxia, left-sided facial palsy, and right exotropia. There was no history of tumor recurrence or seizure. Neurologic examination revealed Wernicke's aphasia in addition to the preexisting neurological deficits, with an initial score on the National Institutes of Health Stroke Scale of seven. Diffusion-weighted magnetic resonance imaging of the brain revealed a small, focal, and high-signal intensity (volume 1.3 mL) with the apparent diffusion coefficient restriction in the left temporal lobe, compatible with acute infarction (Figure 1). Because there were no lesions that could explain the patient's aphasia, we diagnosed the condition to be significant neurological deficit with CDM and considered emergency revascularization. However, the revascularization procedure was not performed because we did not note any abnormal findings on magnetic resonance angiography or perfusion imaging.

hat could explain the patient's aphasia, we diagnosed the condition to be significant neurological deficit with CDM and considered emergency revascularization. However, the revascularization procedure was not performed because we did not note any abnormal findings on magnetic resonance angiography or perfusion imaging. The results of the Korean version of the Western Aphasia Battery indicated severe Wernicke's aphasia (10/20 fluency, 1.9/10 auditory comprehension, 0.4/10 repetition, 0.5/10 naming, and 45.6/100 aphasia quotient). There were no abnormal laboratory results except a mild elevation in erythrocyte sedimentation rate (23 mm/h). The electrocardiogram showed a normal sinus rhythm, and chest x-ray showed no active lesion in either lung fields. During hospitalization, the severity of aphasia and mild confusion fluctuated. Electroencephalography demonstrated intermittent theta slowing in the left temporal areas (Figure 1). Fluorodeoxyglucose positron emission tomography (FDG-PET) showed hypermetabolism in the left temporal cortex (Figure 1). The aphasia was due to ictal and postictal symptoms following stroke rather than hypoperfusion-related symptoms. The patient was administered 250 mg/day phenytoin on the fourth day of admission. Symptoms remained on the following day, albeit attenuated; therefore, 1,000 mg/day levetiracetam was added to the treatment regimen. Symptoms resolved on day 8 of admission; however, the patient complained of dizziness and depressive mood. We therefore changed his medication at discharge to include 600 mg/day valproic acid, which was maintained for eight months. The patient has been seizure-free for one year (Figure 2), and the temporal hypermetabolism noted on FDG-PET has improved.

ssion; however, the patient complained of dizziness and depressive mood. We therefore changed his medication at discharge to include 600 mg/day valproic acid, which was maintained for eight months. The patient has been seizure-free for one year (Figure 2), and the temporal hypermetabolism noted on FDG-PET has improved. The predominant cause of sudden onset, prolonged aphasia, is stroke affecting the language network. In the present case, the patient developed sudden onset Wernicke's aphasia, with a small cortical high-signal intensity in the territory of the left middle cerebral artery on magnetic resonance imaging. Because the lesion did not explain the patient's symptoms, the aphasia was initially considered a symptom resulting from hypoperfusion in the territory of the left middle cerebral artery, and was therefore considered CDM. However, the cause of aphasia was subsequently thought to be an ictal and postictal symptom caused by acute ischemic stroke, due to the following reasons: First, there were no steno-occlusive lesions on magnetic resonance angiography or perfusion defects, even though the symptoms lasted more than two days. Second, the small cortical lesion itself could not explain the intermittent slowing in the left temporal area found on electroencephalography. Third, FDG-PET performed when the aphasic symptoms presented, showed hypermetabolism in the left temporal lobe. Fourth, aphasic symptoms and hypermetabolism on FDG-PET disappeared after the administration of antiepileptic drugs. Due to its low prevalence, the diagnosis of ictal aphasia is difficult, and only 56% of patients with aphasic seizures demonstrate electrographic ictal discharge on the first routine surface electroencephalography.1 FDG-PET imaging was useful for the diagnosis of aphasic nonconvulsive status epilepticus, although the electroencephalography findings were inconclusive.2

aphasia is difficult, and only 56% of patients with aphasic seizures demonstrate electrographic ictal discharge on the first routine surface electroencephalography.1 FDG-PET imaging was useful for the diagnosis of aphasic nonconvulsive status epilepticus, although the electroencephalography findings were inconclusive.2 According to previous studies, CDM is associated with perfusion-diffusion mismatch, infarction expansion, and early neurological deterioration.3,4 It is therefore important to consider CDM when determining acute stroke management, such as thrombolysis or endovascular treatment.3,4 However, the false-positive diagnosis of ischemic stroke, labeled "stroke mimics", ranges between 1.3-25%, and seizures (38%) are the most common cause of stroke mimics.5,6 Moreover, acute ischemic stroke is important in the etiology of symptomatic seizures. Nonconvulsive seizures can be confused with CDM, as seen in the case reported here. Initially, revascularization was considered as a treatment for CDM. However, CDM can also be due to stroke mimics, indicating that clinicians should consider the possibility of post-stroke seizures. Perfusion imaging, magnetic resonance angiography and FDG-PET can provide useful information for differentiating non-oligemic CDM from true ischemic penumbra. This study was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare Republic of Korea (HI10C2020). The authors have no financial conflicts of interest.

Nonconvulsive seizures can be confused with CDM, as seen in the case reported here. Initially, revascularization was considered as a treatment for CDM. However, CDM can also be due to stroke mimics, indicating that clinicians should consider the possibility of post-stroke seizures. Perfusion imaging, magnetic resonance angiography and FDG-PET can provide useful information for differentiating non-oligemic CDM from true ischemic penumbra. This study was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare Republic of Korea (HI10C2020). The authors have no financial conflicts of interest. Figure 1 Representative images from magnetic resonance imaging and Fluorodeoxyglucose positron emission tomography (FDG-PET) of the brain. (A) Diffusion-weighted image showing a small cortical infarct (yellow arrow) in the territory of the left middle cerebral artery. (B) Perfusion image showing no obvious abnormalities. (C) Electroencephalography showing intermittent theta slowing in the left temporal cortex. (D) Images of FDG-PET, which was performed at the onset of symptoms, showing hypermetabolism (red arrow) in the left temporal cortex (coronal and axial view). (E) Images of the regions shown in D obtained during follow-up FDG-PET, which was performed when the symptoms resolved; they reveal normalization of left temporal hypermetabolism (coronal and axial view). Figure 2 Hospital course of the patient showing the symptoms and prescribed medications.

Patients with atrial fibrillation (AF) have a particular high risk of cardio-embolic stroke. Cardio-embolic strokes have a high mortality (20%) and >50% of patients remain permanently disabled. Oral anticoagulation with warfarin is highly effective in preventing strokes as long as warfarin is taken. Unfortunately many patients with atrial fibrillation refuse to take warfarin because they are afraid of major bleeding complications in particular intracranial bleeds. In addition long term compliance and adherence with warfarin is terrible due to the need of regular international normalized ratio (INR) monitoring and the many interactions with food and drugs in elderly patients with polypharmacy. Anticoagulation is most effective in old patients who have difficulties to get to a physicians' office or an anticoagulation clinic to get the INR monitored. Just imagine the 86 years old lady with unsteady gate and a walker needing to get to a doctor in winter with ice and freezing temperatures outside. Data from Canadian and German stroke units on admission of patients with AF and a new stroke display shocking facts.1,2 One third of patients with AF and without contraindications for anticoagulation are untreated at the time of stroke. Thirty percent are treated with aspirin which is not effective in secondary stroke prevention in AF.3 Another 30% are treated with warfarin but the INR was <2.0 at the time of stroke. Only 10% of patients were on warfarin and the INR was between 2.0 and 3.0.

ions for anticoagulation are untreated at the time of stroke. Thirty percent are treated with aspirin which is not effective in secondary stroke prevention in AF.3 Another 30% are treated with warfarin but the INR was <2.0 at the time of stroke. Only 10% of patients were on warfarin and the INR was between 2.0 and 3.0. Reimbursement bodies, however, will consider only the last 10% of patients, namely patients with AF well controlled on warfarin. This does not reflect clinical reality. The most restrictions for the use of novel anticoagulants (NOACs) instead of warfarin prevent the patients in highest need of stroke prevention from a drug that is superior to warfarin and has fewer intracranial bleeds, namely the elderly.

nts with AF well controlled on warfarin. This does not reflect clinical reality. The most restrictions for the use of novel anticoagulants (NOACs) instead of warfarin prevent the patients in highest need of stroke prevention from a drug that is superior to warfarin and has fewer intracranial bleeds, namely the elderly. In this issue, Bang et al.4 summarized pieces of evidence that NOACs may be particularly helpful for Asian stroke patients. Indeed, Asians are at particular high risk of major bleeds on warfarin. This risk is much lower with novel anticoagulants. Therefore it is not rational to restrict the use of NOACs in Asian patients with AF. My policy would be to offer all patients who had a transient ischemic attack or ischemic stroke and AF a NOAC, because these patients have a particular high risk of a recurrent stroke. In primary prevention I would offer NOACs to untreated patients, patients on aspirin and patients poorly controlled on warfarin. I see no reason to switch well controlled patients on warfarin to NOACs. In summary putting high hurdles on the prescription of NOACs will lead to many strokes with severe disability and tremendous costs for a national health care system.

Introduction Taiwan lies on the Tropic of Cancer, and its climate is marine tropical. The population in Taiwan is about 23.07 million people. Han Chinese made up the vast majority of people (98%) and the remaining minority was composed of Austronesian indigenous groups.1 According to the Statistical Yearbook of Interior in 2012,2 life expectancy of the total population at birth is 79.5 years (76.2 years in males and 83 years in females). The natural increase rate is 0.323%. The proportion of persons aged 65 or over is approximately 11% and aging index is 76.2% greater than the other Asia countries. All of the data shown that Taiwan's population is aging. The common health problems among older people are chronic diseases such as cardiovascular disease, stroke, and diabetes. The current health care system in Taiwan, known as National Health Insurance, was instituted on March 1st, 1995. The system provides equal access to health care for all inhabitants, and the population coverage had reached 99% by the end of 2004.3 A large computerized de-identified database, termed National Health Insurance research databases (NHIRD), was derived from this system by the Bureau of National Health Insurance, Taiwan (BNHI) and maintained by the National Health Research Institutes, Taiwan. The NHIRD containing registration and original claim data for reimbursement is provided to scientists in Taiwan for research purposes. In 2012, there were 502 hospitals and 20,935 clinics in Taiwan. The availability of hospital beds per 10,000 people and physicians per 10,000 persons in 2012 was approximately 69 beds and 20 physicians, respectively. There is about 1 physician for 500 people, on average.

d to scientists in Taiwan for research purposes. In 2012, there were 502 hospitals and 20,935 clinics in Taiwan. The availability of hospital beds per 10,000 people and physicians per 10,000 persons in 2012 was approximately 69 beds and 20 physicians, respectively. There is about 1 physician for 500 people, on average. Stroke is the third leading cause of death in Taiwan. It is also considered as the most common cause of complex disability. Post-stroke disability may leave survivors unable to work, which can lead to serious financial issues for survivors and their families. According to the data from the Ministry of Health and Welfare,4 the proportionate mortality of stroke is 7.2% in 2012. It is about 47 stroke deaths per minute. Using the 2000 WHO standard population, the annual age-standardized mortality rate of stroke is steadily decreasing from 57.8 to 30.8 per 100,000 between 2001 and 2012. Changes in stroke mortality rates are mainly attributable to the improvement of stroke survival, rather than incidence rates. There are 11,061 deaths from stroke (6,424 males and 4,635 females) in 2012. The average years of potential life lost before age 70 for stroke is 13.8 years, ranked the fifth in the cause of death. Stroke places a substantial burden on the national healthcare system, costing an estimated 475 million US dollars. Its national impact is predicted to be greater with aging population.

ales) in 2012. The average years of potential life lost before age 70 for stroke is 13.8 years, ranked the fifth in the cause of death. Stroke places a substantial burden on the national healthcare system, costing an estimated 475 million US dollars. Its national impact is predicted to be greater with aging population. Stroke types and subtypes Briefly speaking, stroke is caused by the interruption of blood supply to the brain or the rupture of blood vessels in the brain. Between 2006 and 2008, the Taiwan Stroke Registry, sponsored by the Ministry of Health and Welfare, engaged 39 academic and community hospitals to collect clinical data from 30,599 stroke admissions.5 As shown in Figure 1, this large scale stroke case surveillance program revealed that the most common type of stroke was ischemic stroke, accounting for almost 74% of all strokes. Other stroke types in order of frequency were intracerebral hemorrhage (ICH, 16.1%), transient ischemic attack (TIA, 6.7%), subarachnoid hemorrhage (SAH, 2.8%), and cerebral venous thrombosis (0.2%). According to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria,6 the majority of ischemic strokes subtype in Taiwan was small vessel occlusion (37.7%) followed by large artery atherosclerosis (27.7%), cardioembolism (10.9%), specific pathogenesis (1.5%), and undetermined pathogenesis (22.2%). Lee et al.7 reported that ischemic stroke subtypes among young patients (18 to 45 years old) were 20.5% of small-vessel occlusion, 7.2% of large-artery atherosclerosis, 17.8% of cardioembolism, 22.3% of specific pathogenesis, and 23.5% of undetermined pathogenesis.

genesis (1.5%), and undetermined pathogenesis (22.2%). Lee et al.7 reported that ischemic stroke subtypes among young patients (18 to 45 years old) were 20.5% of small-vessel occlusion, 7.2% of large-artery atherosclerosis, 17.8% of cardioembolism, 22.3% of specific pathogenesis, and 23.5% of undetermined pathogenesis. Stroke morbidity Incidence A population-base study from Chen et al.8 reported that there were approximately 230 incidence cases of hospitalization daily attributable to stroke during the period from 1 January 2001 to 31 December 2007. Previously, the stroke incidence was estimated by several population-based studies. Hu et al.9 conducted a cohort study of 8,562 stroke-free people with 4-year follow up to observe new stroke occurrence. The average annual incidence rate of first-ever stroke for people aged 36 years or older in this study was 330 per 100,000 population. The other population-based stroke survey carried out in Kinmen between 1993 and 1996 reported that the average annual incidence rate of first-ever stroke for people aged 50 years or older was 527 per 100,000 population.10 The studies of stroke incidence are summarized in Table 1.

s study was 330 per 100,000 population. The other population-based stroke survey carried out in Kinmen between 1993 and 1996 reported that the average annual incidence rate of first-ever stroke for people aged 50 years or older was 527 per 100,000 population.10 The studies of stroke incidence are summarized in Table 1. Prevalence Hu et al.11 reported the stroke prevalence for 8,705 people aged 36 or older was 16.42/1,000 population (95% CI=13.89-19.42/1,000) in 1986. The other studies estimated stroke prevalence of 14.27 per 1,000 in the people aged 35 or older using data from the 1994 National Health Interview Survey (NHIS)12 and 19.3 per 1,000 people from 2001 NHIS.13 In males, the stroke prevalence is 25.5 per 1,000 people, almost two-fold higher than the stroke prevalence in females (12.6 per 1,000 people).12 The level of urbanization, gender, heart disease, hypertension, and diabetes were important contributory factors to the stroke prevalence in Taiwan.12 The studies of stroke prevalence are summarized in Table 1.

5.5 per 1,000 people, almost two-fold higher than the stroke prevalence in females (12.6 per 1,000 people).12 The level of urbanization, gender, heart disease, hypertension, and diabetes were important contributory factors to the stroke prevalence in Taiwan.12 The studies of stroke prevalence are summarized in Table 1. Risk factors Nonmodifiable risk factors The unchangeable risk factors for stroke are age and gender. In a hospital-based study recruiting 1,085 stroke episodes in 1,021 patients between 2003 and 2005, the onset age was greater than 65 years in 66.6% of ischemic stroke patients.14 According to the Taiwan Stroke Registry, the medium age of patients at the onset of ischemic stroke/TIA, ICH, and SAH were 69.9 years, 62.2 years, and 57.6 years, respectively.5 Among ischemic stroke/TIA and ICH patients, the proportions of males were greater than females. However, in SAH, the cases were more common in females. Modifiable risk factors The important changeable risk factors for stroke are hypertension, diabetes, hyperlipidemia, obesity, atrial fibrillation, and smoking.

Risk factors Nonmodifiable risk factors The unchangeable risk factors for stroke are age and gender. In a hospital-based study recruiting 1,085 stroke episodes in 1,021 patients between 2003 and 2005, the onset age was greater than 65 years in 66.6% of ischemic stroke patients.14 According to the Taiwan Stroke Registry, the medium age of patients at the onset of ischemic stroke/TIA, ICH, and SAH were 69.9 years, 62.2 years, and 57.6 years, respectively.5 Among ischemic stroke/TIA and ICH patients, the proportions of males were greater than females. However, in SAH, the cases were more common in females. Modifiable risk factors The important changeable risk factors for stroke are hypertension, diabetes, hyperlipidemia, obesity, atrial fibrillation, and smoking. Hypertension Hypertension is a very common disease observed in the patients with ischemic stroke/TIA (79.2%) or ICH (84.9%). However, in the patients with SAH, the prevalence of hypertension was only 65.3%.5 A prospective cohort study conducted by Lin et al.15 investigated the associations between adherence to Dietary Approaches to Stop Hypertension (DASH) foods or nutrients and blood pressure or stroke risk in 1,420 and 2,061 Taiwanese adults during 1989 to 2002, respectively. They found that adhering to the DASH diet was beneficial for long term blood pressure control and reduction of stroke risk (HR [hazard ratio], 0.63; 95% CI, 0.41-0.98; P=0.037) in Taiwanese population. Another study using pregnant women as study cohort reported that women with hypertensive disorders in pregnancy had an increased risk of subsequent stroke (adjusted HR, 2.04; 95% CI, 1.18-3.51).16 Among predictors for ischemic stroke in patients with atrial fibrillation, hypertension was the most powerful predictor except for prior TIA/stroke, and the magnitude of association (OR, 2.656; 95% CI, 2.140-3.296) was greater than observed in white populations.17

equent stroke (adjusted HR, 2.04; 95% CI, 1.18-3.51).16 Among predictors for ischemic stroke in patients with atrial fibrillation, hypertension was the most powerful predictor except for prior TIA/stroke, and the magnitude of association (OR, 2.656; 95% CI, 2.140-3.296) was greater than observed in white populations.17 Diabetes The prevalence of type 2 diabetes has rapidly increased in the Taiwanese population, reaching 9.78% among adults aged 20-79 years in 2013.18 Increasing levels of obesity in the general population in the past two decades may be one of the principal factors contributing to the elevated incidence of type 2 diabetes.19 The prevalence of diabetes was 45.4% in ischemic stroke/TIA and 37% in hemorrhage stroke.5 A Taiwanese population-based study showed that diabetes were significantly associated with ischemic stroke in patients with atrial fibrillation (OR, 1.341; 95% CI, 1.092-1.648).17 Cheng et al.20 randomly selected 14,856 patients with diabetes from the Taiwan National Health Research Institute database and followed up for 4 years to calculate their ischemic stroke incidence in two groups, one group used oral hypoglycemic agents metformin and the other without. The risk of ischemic stroke was substantially lower in the group with metformin use (9.2%) than in the group without metformin use (17.5%): adjusted HR (95% CI)=0.468 (0.424-0.518).

years to calculate their ischemic stroke incidence in two groups, one group used oral hypoglycemic agents metformin and the other without. The risk of ischemic stroke was substantially lower in the group with metformin use (9.2%) than in the group without metformin use (17.5%): adjusted HR (95% CI)=0.468 (0.424-0.518). Hyperlipidemia Increasing body of evidence has demonstrated that hyperlipidemia is one of the well-defined and modifiable risk factors for ischemic stroke. The prevalence of dyslipidemia is about 49.4% in Taiwanese patients with ischemic stroke/TIA.5 Among young ischemic stroke patients aged 15-45 years, the prevalence of hyperlipidemia was 53.1%.7 It has been reported that every 1.0 mmol/L reduction in low-density lipoprotein cholesterol is associated with a corresponding 21% reduction in stroke.21 Statins are medications used for the control of hypercholesterolemia.22 A nationwide population-based study using the Taiwan NHIRD reported that after the multivariate Cox regression model analysis adjusted for age, gender, comorbidities, and concomitant medication use, use of statins in hemodialysis patients was associated with a lower risk of developing future ischemic stroke (HR, 0.49; 95% CI, 0.39-0.63) as compared to the hemodialysis patients without exposure to statins.23

riate Cox regression model analysis adjusted for age, gender, comorbidities, and concomitant medication use, use of statins in hemodialysis patients was associated with a lower risk of developing future ischemic stroke (HR, 0.49; 95% CI, 0.39-0.63) as compared to the hemodialysis patients without exposure to statins.23 Obesity General obesity for Taiwanese was defined as BMI≥27 kg/m2, corresponding to a similar degree of fat composition to whites with a BMI of 30.8.24 According to a nationwide population-based survey conducted in Taiwan, obesity prevalence (BMI≥27) in 2002 was 19.2% in men and 13.4% in women.25 Comparing the Taiwanese data with other countries, obesity prevalence was greater in Taiwan than in the other Asian countries, but less than in the western countries.26 Data from the Cardiovascular Diseases Risk Factor Two-Township Study showed that the prevalence of obesity in ischemic stroke patients was 25% in males and 35% in females.27 The authors also found that metabolic syndrome subjects with general obesity exhibited a greater risk of ischemic stroke.

the western countries.26 Data from the Cardiovascular Diseases Risk Factor Two-Township Study showed that the prevalence of obesity in ischemic stroke patients was 25% in males and 35% in females.27 The authors also found that metabolic syndrome subjects with general obesity exhibited a greater risk of ischemic stroke. Atrial fibrillation A community-based prospective cohort study among 3,560 Taiwanese participants conducted by Chien et al.28 reported that atrial fibrillation prevalences were 1.4% in men and 0.7% in women. The authors also found that atrial fibrillation patients had nearly 4 times risk of stroke (age, gender-adjusted relative risk [RR], 3.87; 95% CI, 2.12-7.15) compared with those without atrial fibrillation. Another study conducted by Chao et al.29 recruited 829 subjects with atrial fibrillation as study cohort and 8,290 study subjects without atrial fibrillation as control cohort. The control cohort was matched with study cohort by sex and age. The authors found that atrial fibrillation was a significant risk factor for ischemic stroke among females (HR, 7.77; 95% CI, 3.97-15.19). The event rates for female patients with and without atrial fibrillation were 4.4% and 0.7%, respectively. However, the event rates were similar between groups with and without atrial fibrillation for males. A possible interpretation for the gender difference in atrial fibrillation-related thromboembolism risk was that a higher von Willebrand factor level was observed in females with atrial fibrillation, but not in males with atrial fibrillation, as compared with that of non- atrial fibrillation patients.30 In a hospital-based study of 947 patients with first-ever ischemic stroke, the prevalence of atrial fibrillation was significantly higher in elderly patients aged 80 years or older (29%) than in those aged less than 80 years (11%).31

ith atrial fibrillation, as compared with that of non- atrial fibrillation patients.30 In a hospital-based study of 947 patients with first-ever ischemic stroke, the prevalence of atrial fibrillation was significantly higher in elderly patients aged 80 years or older (29%) than in those aged less than 80 years (11%).31 Smoking According to the data from the Adult Smoking Behavior Surveillance System conducted by the Health Promotion Administration, Ministry of Health and Welfare in Taiwan, the smoking rate significantly dropped from 42.9% to 35% in males during 2004 to 2010.32 However, no marked difference in smoking rate was observed in females, which was 4.6% in 2004 and 4.1% in 2010. Data from a prospective hospital-based stroke registry showed that the prevalence of smoking in ischemic stroke patients was 57.7% in males and 6.8% in females.33 In a nationwide survey of health and living status of 2,600 residents aged 65 years or older in Taiwan, heavy smoking (≥20 cigarettes per day) was an independent risk factor for all stroke (OR, 1.71; 95% CI, 1.04-2.80) and cerebral infarction (OR, 1.72; 95% CI, 1.00-2.96).34 In a hospital-based stroke registry among 2,650 acute ischemic stroke patients in Taiwan, it was found that smoking correlated with greater severity of ischemic stroke on admission particularly for the stroke subtype of small-vessel occlusion.35 However, inverse association was observed between smoking and stroke severity of cardioembolism.

roke registry among 2,650 acute ischemic stroke patients in Taiwan, it was found that smoking correlated with greater severity of ischemic stroke on admission particularly for the stroke subtype of small-vessel occlusion.35 However, inverse association was observed between smoking and stroke severity of cardioembolism. Stroke care in Taiwan The standard treatment for acute ischemic stroke recommended by the Stroke Guidelines in Taiwan is providing the intravenous thrombolysis with recombinant tissue plasminogen activator (IV tPA) therapy for ischemic stroke patients within 3 hours of symptom onset. The major exclusion criteria for IV tPA therapy are age >80 years old, anticoagulant user, the initial NIHSS >25, and diabetes patients with prior history of stroke. According to the data from Taiwan Stroke Registry between 2006 and 2008,5 only 8.84% of ischemic stroke patients who presented within 2 hours of symptom onset were given IV tPA treatment. Among patients with ischemic stroke or TIA, 94% were treated with antithrombotics during hospitalization and 85% were on antithrombotics at discharge. The rates of anticoagulation used for atrial fibrillation and lipid-lowering drug used at discharge were 28.28% and 38.69%, respectively, much less than the rates in the western countries. The discharge destination for majority of stroke cases were their homes, only a few of them went to nursing home. In order to improve acute stroke care quality, a Breakthrough Series-Stroke (BTS) activity was conducted by the Institute for Healthcare Improvement from August 2010 to July 2011 through a collaborative learning model focused on 14 performance indicators of stroke care among 24 hospitals in Taiwan.36 One year after BTS activity, it was successful to improve the achievement rate of most performance indicators of stroke care. Since 2001, stroke centers have been increasingly established in medical centers of Taiwan. Most stroke centers have a multidisciplinary team composed of neurology, neurosurgery, emergency department, neurological radiology, rehabilitation, nursing department, outpatient department, department of nutrition, and social welfare department to provide effective, timely care to stroke victims.

l centers of Taiwan. Most stroke centers have a multidisciplinary team composed of neurology, neurosurgery, emergency department, neurological radiology, rehabilitation, nursing department, outpatient department, department of nutrition, and social welfare department to provide effective, timely care to stroke victims. Nationwide program for controlling risk factors Smoking is an important risk factor for stroke. The smoking rate among ischemic stroke patients was nearly 60% in males. Thus, smoking cessation should be an important initiative for stroke prevention. In January 2009, the Taiwan government amended the 1997 Tobacco Hazards Prevention Act by extending smoke-free areas to almost all enclosed work-places and public places, printing graphic health warnings on cigarette packages, totally banning tobacco advertisements, promotion and sponsorship and increasing tobacco taxes. The implementation of the 2009 Act was reported to be associated with the increased quit attempt rate and the increased annual cessation rate.37 The quit attempt rate raised significantly from 39.4% to 42.9% between 2007 and 2010. Furthermore, the annual cessation rate elevated significantly from 7.1% to 8.9%. Obesity is the other important risk factor of stroke. The prevalence of obesity in ischemic stroke patients was about 25% in males and 35% in females. In order to decrease obesity rate, a nationwide weight-loss program was conducted by Health Promotion Administration from 2011 to 2013 in Taiwan. A series of healthy-living campaigns were used to encourage people to exercise more and eat healthily. In this three-year program, more than a million Taiwanese lost 2.2 m kg.38

order to decrease obesity rate, a nationwide weight-loss program was conducted by Health Promotion Administration from 2011 to 2013 in Taiwan. A series of healthy-living campaigns were used to encourage people to exercise more and eat healthily. In this three-year program, more than a million Taiwanese lost 2.2 m kg.38 In summary, stroke was the third leading cause of death and the main cause of disability in Taiwan. Majority of stroke was ischemic stroke. Small vessel occlusion was the most common ischemic stroke subtype. The recommend treatment for acute ischemic stroke patients within 3 hours of symptom onset is IV tPA. However, the rate of IV tPA therapy in Taiwan was still far behind than the rate in the western countries. Therefore, in order to increase the tPA treatment rate, increasing the public awareness of stroke warning signs and act on stroke and improving in-hospital critical pathway for thrombolysis would be the most important and urgent issues in Taiwan. The authors have no financial conflicts of interest. Figure 1 The type of stroke in Taiwan (the figure was created from the data of the Taiwan Stroke Registry5). Table 1 The studies of stroke morbidity in Taiwan

Introduction Moyamoya disease (MMD) is characterized by a progressive stenosis at the terminal portion of the internal carotid artery (ICA) and an abnormal vascular network at the base of the brain.1,2 The etiology of MMD remains unknown, but recent advances have been made in understanding the molecular biology and pathophysiology of this rare entity. Previous studies explored genetic factors and revealed several loci associated with moyamoya disease; 3p24-p26,3 6q25,4 8q23,5 and 17q25.6 More recently, the RNF213 gene (RNF213) in the 17q25-ter region was identified as a novel susceptibility gene for MMD among East Asian population.7,8,9 A polymorphism in c.14576G>A in RNF213 was identified in 95% of familial patients with MMD and 79% of sporadic cases,7 and RNF213 was found to correlate with the early-onset and severe forms of MMD, indicating its value as a good biomarker for predicting prognosis.10 The exact mechanism by which the RNF213 abnormality relates to moyamoya disease remains unknown, but recent reports using genetically engineered mice lacking RNF213 by homologous recombination provide new prospects for the basic research of this rare entity.11,12 In this review article, we focus on the genetics and biomarkers of MMD, and sought to discuss their clinical implication.

yamoya disease remains unknown, but recent reports using genetically engineered mice lacking RNF213 by homologous recombination provide new prospects for the basic research of this rare entity.11,12 In this review article, we focus on the genetics and biomarkers of MMD, and sought to discuss their clinical implication. Basic pathology of MMD Histo-pathological characteristics Intimal hyperplasia and medial thinness are well-known histo-pathological characteristics of MMD.2,13,14 The thickened intima has increased number of smooth muscle cells, which are thought to be synthetic-type smooth muscle cell migrating from medial layer.15 Degradation of the smooth muscle cells in the medial layer and subsequent death of the vascular smooth muscle cells are thought to result in the characteristic finding of medial thinness not only at the terminal ICA but also at the peripheral middle cerebral artery.14 Together with the waviness and duplication of the internal elastic lamina, intracranial arteries of moyamoya disease generally have intrinsic fragility, which should also be mentioned as one of the pitfalls of revascularization surgery while manipulating peripheral middle cerebral artery during extracranial-intracranial (EC-IC) bypass.16

aviness and duplication of the internal elastic lamina, intracranial arteries of moyamoya disease generally have intrinsic fragility, which should also be mentioned as one of the pitfalls of revascularization surgery while manipulating peripheral middle cerebral artery during extracranial-intracranial (EC-IC) bypass.16 Macroscopic pathology of MMD; physiological reorganization system For the better understanding of MMD as dynamic disease, it is particularly important to revisit Suzuki's angiographic staging,1,2 which represent the natural course of the angio-architecture in moyamoya patients by physiological reorganization system. Suzuki's angiographic staging does not reflect the severity of MMD, while it explains how 'moyamoya disease' compensates its ischemic condition by physiological process.16 At the early stage of MMD, progressive stenosis of the terminal ICA occurs while it is subsequently compensated by abnormal vascular networks at the base of the brain at stage III. In substantial number of patients, ischemic brain is further compensated by trans-dural anastomosis from external carotid system without surgical intervention at stage IV to VI, resulting in the disappearance of moyamoya vessels. We named this physiological reorganization process as 'internal carotid (IC)-external carotid (EC) conversion'.16 Concept of revascularization surgery for MMD, either direct or indirect revascularization, is to accomplish successful 'IC-EC conversion' to prevent insufficient reorganization resulting in cerebral infarction. Therefore, the concept of revascularization surgery for MMD should be based on the idea to support the intrinsic compensatory nature of MMD, rather than to eradicate the pathophysiology of this entity.

s to accomplish successful 'IC-EC conversion' to prevent insufficient reorganization resulting in cerebral infarction. Therefore, the concept of revascularization surgery for MMD should be based on the idea to support the intrinsic compensatory nature of MMD, rather than to eradicate the pathophysiology of this entity. Genetics of MMD Background and previous reports Since some patients with MMD show autosomal dominant inheritance pattern, genetic factors have been implicated in the etiology of MMD. In fact, gene loci have been identified in 3q24-p263 and 8q235 in genome-wide analysis, and in 6q254 and 17q256 in a chromosomal search for familial MMD. The RNF213 in the 17q25-ter region was more recently identified as a susceptibility gene for MMD among East Asian population.7,8,9 We found that a polymorphism in c.14576G>A in RNF213 was identified in 95% of familial patients with MMD and 79% of sporadic cases.7 Miyatake et al. reported that patients with the polymorphism of c.14576G>A in the RNF213 gene had significantly earlier onset and a more severe form of MMD, such as the presentation of cerebral infarction and posterior cerebral artery stenosis.10 To investigate the potential role of RNF213 polymorphism, Liu et al.8 investigated the effects of RNF213 suppression on zebra-fish vasculature by generating RNF213 knockdown zebra-fish, demonstrating that zebra-fish lacking RNF213 showed severely abnormal sprouting vessels in the head region, especially from the optic vessels. It was suggested that RNF213 is involved in a novel signaling pathway in intracranial angiogenesis.8

ression on zebra-fish vasculature by generating RNF213 knockdown zebra-fish, demonstrating that zebra-fish lacking RNF213 showed severely abnormal sprouting vessels in the head region, especially from the optic vessels. It was suggested that RNF213 is involved in a novel signaling pathway in intracranial angiogenesis.8 Generation of RNF213-deficient mice (RNF213-/-) Using Cre-lox system, we generated RNF213-deficient mice with C57BL/6 background by deleting exon 32 of RNF213, which is the largest exon in the RNF213 gene (Figure 1).12 Finally, heterozygous male and female mice were bred to produce homozygous offspring (homozygous knockout mice, RNF213-/-). Genotyping was performed by PCR using specific primers to exon 32 (forward; 5'-CTCAGTGGTGGTGTTGGATG-3', reverse; 5'-CTCTTTCTCGTTGGGACTGC-3') and the deleted allele (forward; 5'-ATAACCTCAGGCACCAATCG-3', reverse; 5'-TCCCTCTAGGCAGGAAGGAT-3').12 The complete removal of RNF213 exon 32 from genomic DNA was confirmed in RNF213-/- (Figure 2A). Both homozygous mutant (RNF213-/-) and heterozygous mutant mice (RNF213-/+) were born and grew normally.

CTCGTTGGGACTGC-3') and the deleted allele (forward; 5'-ATAACCTCAGGCACCAATCG-3', reverse; 5'-TCCCTCTAGGCAGGAAGGAT-3').12 The complete removal of RNF213 exon 32 from genomic DNA was confirmed in RNF213-/- (Figure 2A). Both homozygous mutant (RNF213-/-) and heterozygous mutant mice (RNF213-/+) were born and grew normally. Temporal changes of vascular anatomy in RNF213-/- Then we performed magnetic resonance angiography (MRA) using a dedicated small animal scanner with a 50-mm bore operating at 9.4 Tesla field strengths (AV400WV, Bruker BioSpin).12 A high resolution three-dimensional gradient-echo time-of-flight MRA sequence was used to acquire MRA images. As a result, no significant difference was observed in the MRA findings of cervical/intracranial arteries between RNF213-/- and wild-type littermates (Wt.) from 32 to 64 weeks of age.12 The anatomy of the circle of Willis was evaluated by a trans-cardiac injection of carbon black dye in RNF213-/- and Wt. (Figure 2B, C). No significant difference was observed in the structure of the major arteries at the base of the brain between RNF213-/- and Wt. at 16 weeks of age (Figure 2B, C). Vessel diameter was sustained at 40 weeks of age in RNF213-/-, and no steno-occlusive changes were observed around the terminal portion of the internal carotid artery. An abnormal vascular network did not develop at the base of the brain in RNF213-/-.12

brain between RNF213-/- and Wt. at 16 weeks of age (Figure 2B, C). Vessel diameter was sustained at 40 weeks of age in RNF213-/-, and no steno-occlusive changes were observed around the terminal portion of the internal carotid artery. An abnormal vascular network did not develop at the base of the brain in RNF213-/-.12 Modification of vascular remodeling in RNF213-/- after carotid artery ligation We also evaluated the histo-pathological characteristics of the vascular wall structure in RNF213-/-, and showed no apparent abnormality in mutant mice including intimal hyperplasia or medial layer thinness, both of which are the characteristic findings of MMD. Thus we employed common carotid artery (CCA) ligation model, which reproducibly induces arterial wall hyperplasia.17 After CCA ligation, Wt. showed temporary hyperplasia of the intima and medial layers (Figure 3A), which was consistent with previous reports;18 however, RNF213-/- did not exhibit temporary hyperplasia of the intima and medial layers at the same time point (Figure 3B). Both the intima and medial layers were significantly thinner in RNF213-/- than in Wt. 14 days after CCA ligation, while no significant difference was observed in vascular wall thickness 7, 21, and 28 days after CCA ligation.12

bit temporary hyperplasia of the intima and medial layers at the same time point (Figure 3B). Both the intima and medial layers were significantly thinner in RNF213-/- than in Wt. 14 days after CCA ligation, while no significant difference was observed in vascular wall thickness 7, 21, and 28 days after CCA ligation.12 Role of RNF213 polymorphism in MMD Our results demonstrated that RNF213-/- grew normally, and no significant difference was observed in MRA findings or the anatomy of the circle of Willis between RNF213-/- and Wt. under normal conditions. The histo-pathological characteristics of the vascular wall structure in RNF213-/- showed no apparent abnormality, thus it is conceivable that an abnormality in RNF213 does not sufficiently induce MMD. Recently, Kobayashi and colleagues alternatively generated RNF213-/- and suggested that target disruption of RNF213 improved glucose tolerance by protecting inlet β cells when they are mated with Akita diabetic mice, but they also failed to detect any cerebrovascular pathology mimicking MMD,11 which is consistent with our result.

obayashi and colleagues alternatively generated RNF213-/- and suggested that target disruption of RNF213 improved glucose tolerance by protecting inlet β cells when they are mated with Akita diabetic mice, but they also failed to detect any cerebrovascular pathology mimicking MMD,11 which is consistent with our result. Although it is still undetermined whether the polymorphism of c.14576G>A in patients with MMD is a loss of function mutation or a gain of function mutation, the initial report of knockout zebra-fish by Liu et al.8 suggested that it could be a loss of function mutation. In our study, we did not observe any evidence of the characteristic findings of MMD in RNF213-/- under normal physiological conditions until 64 weeks of age. However, following CCA ligation, which reproducibly induced temporary hyperplasia of the adjacent vascular wall,17,18 we showed that the medial layer of CCA was significantly thinner in RNF213-/- than in Wt. 14 days after CCA ligation, which matched one of the histological characteristic findings of MMD.2 Therefore, RNF213 deficiency could lead to vascular fragility including medial thinness, which may make vessels more vulnerable to hemodynamic stress and secondary insults, which facilitates the development of MMD. Furthermore, it is conceivable that secondary insults in addition to RNF213 abnormality, such as an autoimmune response,19,20 infection/inflammation, and radiation2 may be necessary for the development of MMD. In fact, the incidence rate of MMD was shown to be as low as 0.53 per 100,000 population in Japan,21 while 1% of the Japanese population carries the polymorphism of c.14576G>A in the RNF213 gene,8 again suggesting the importance of additional insult to induce MMD. This observation could be illustrated as 'two-hit theory', as is also true in a variety of disorders (Figure 4).

as 0.53 per 100,000 population in Japan,21 while 1% of the Japanese population carries the polymorphism of c.14576G>A in the RNF213 gene,8 again suggesting the importance of additional insult to induce MMD. This observation could be illustrated as 'two-hit theory', as is also true in a variety of disorders (Figure 4). Possibility of medial thinness as an initial histo-patholohical change in MMD Autopsy studies of MMD have shown that the outer diameter, as well as intra-luminal diameter, of the affected arteries of the circle of Willis is smaller in MMD.22 Intraoperative finding also suggested smaller outer diameter of the terminal ICA in MMD.23 More recently, the arterial constrictive remodeling with early narrowing of outer diameter of the ICA was proposed to be the intrinsic pathogenesis of MMD. Based on their finding of three-dimensional constructive interference in steady-state magnetic resonance imaging demonstrating that marked narrowing of the outer vascular diameter may precede internal vascular stenotic progression,24 it could be alternatively possible that the thinness of the entire vascular wall in RNF213-/- might represent the very early finding of the constrictive remodeling as a characteristic of MMD.

imaging demonstrating that marked narrowing of the outer vascular diameter may precede internal vascular stenotic progression,24 it could be alternatively possible that the thinness of the entire vascular wall in RNF213-/- might represent the very early finding of the constrictive remodeling as a characteristic of MMD. Possibility RNF213 polymorphism as a gain of function mutation Alternatively, we should consider another possibility that the RNF213 polymorphism in patients with MMD is a gain of function mutation.25,26 Since RNF213-/- in our study, which represented a loss of function mutation, did not spontaneously develop MMD under normal conditions, it is also conceivable that the polymorphism of c.14576G>A in RNF213 contributes to the gain of function of the RNF213 product in patients with MMD.12 In fact, we obtained conflicting results; the medial layer was significantly thinner in RNF213-/-, which is again the characteristic finding of MMD, but temporary intimal hyperplasia was significantly suppressed in RNF213-/- during vascular remodeling after CCA ligation, which is not a characteristic of MMD. Further studies using knock-in mice carrying the polymorphism of c.14576G>A in the RNF213 gene would address this important issue.

aracteristic finding of MMD, but temporary intimal hyperplasia was significantly suppressed in RNF213-/- during vascular remodeling after CCA ligation, which is not a characteristic of MMD. Further studies using knock-in mice carrying the polymorphism of c.14576G>A in the RNF213 gene would address this important issue. Biomarkers of MMD: clinical implications Expression of angiogenic factors and pro-inflammatory molecules in MMD It has been well known that patients with MMD have increased expression of various pro-inflammatory molecules as well as angiogenic factors in serum and/or cerebrospinal fluid (CSF). Increased expression of basic fibroblast growth factor (bFGF) in CSF from moyamoya patients was considered to be specific and was not thought to be simply related to cerebral ischemia.27 The bFGF level was apparently elevated in the patients in whom neovascularization formed from indirect revascularization, suggesting that the bFGF level is a useful indicator to predict the efficacy of indirect revascularization after surgery.27 Nanba et al.28 alternatively reported the increased expression of hepatocyte growth factor in CSF and intracranial artery in MMD. Regarding serum level of these molecules, Kang and colleagues comprehensively investigated the expression of matrix metalloproteinases, interleukins, and growth factors, and they found that patients with MMD exhibited significantly higher plasma concentrations of matrix metalloproteinase (MMP)-9, monocyte chemoattractant protein-1, interleukin-1β, vascular endothelial growth factor (VEGF), and platelet-derived growth factor-BB.29 In light of the recent report by Park and colleagues demonstrating that moyamoya patients with VEGF polymorphism of the CC genotype of VEGF-634 had better collateral vessel formation after revascularization surgery, increased serum level of VEGF in patients with MMD should participate in the induction of pial synangiosis after indirect revascularization surgery as well as to the development of abnormal vascular networks at the base of the brain.30 Alternatively, increased expression of serum MMP-9 may contribute to the spontaneous intracranial hemorrhage in MMD, as well as to the higher risk for cerebral hyperperfusion syndrome after direct revascularization surgery in moyamoya patients.31,32

development of abnormal vascular networks at the base of the brain.30 Alternatively, increased expression of serum MMP-9 may contribute to the spontaneous intracranial hemorrhage in MMD, as well as to the higher risk for cerebral hyperperfusion syndrome after direct revascularization surgery in moyamoya patients.31,32 Pharmacological approach to the biomarkers of MMD: clinical application EC-IC bypass is an established management of MMD,2 but cerebral hyperperfusion syndrome (CHS) is one of the critical complications of EC-IC bypass for MMD, which could result in focal neurological deterioration and/or delayed intracranial hemorrhage.32,33,34,35,36 Although the mechanism underlying the higher incidence of CHS among patients with MMD has been undetermined, intrinsic vulnerability of vascular structure including blood-brain barrier was considered to lead to vasogenic edema and hemorrhagic conversion following revascularization to the chronic ischemic brain.31 Since MMP-9 is known to participate in edema formation and hemorrhagic conversion after cerebral ischemia-reperfusion injury,37,38 it was conceivable that the increased expression of MMP-9 may contribute in part to CHS in moyamoya patients. Minocycline hydrochloride was known to play a role in blocking MMP-9,39 and is also known to have a neuroprotective effect against ischemic brain injury,40 thus we attempted to use minocycline perioperatively to prevent CHS.41 The result showed that minocycline significantly reduced the incidence of focal neurological deterioration due to CHS after EC-IC bypass for MMD,41 indicating that increased MMP-9 expression in MMD could participate in the pathology of postoperative CHS. This representative 43-year old man with ischemic-onset MMD underwent left EC-IC bypass, and demonstrated significant focal increase in cerebral blood flow at the site of the anastomosis which prolonged over 2 weeks (Figure 5). He was managed by intravenous administration of minocycline hydrochloride (200 mg/day) under strict blood pressure control, and did not suffer any neurological deterioration during the perioperative period. Postoperative magnetic resonance imaging showed no evidence of vasogenic edema (Figure 6), and he was discharged without neurological deficit. We consider that minocycline has potential role for preventing cerebral hyperperfusion to be symptomatic, by blocking MMP-9 in the acute stage after EC-IC bypass.

rative period. Postoperative magnetic resonance imaging showed no evidence of vasogenic edema (Figure 6), and he was discharged without neurological deficit. We consider that minocycline has potential role for preventing cerebral hyperperfusion to be symptomatic, by blocking MMP-9 in the acute stage after EC-IC bypass. Future investigation of various biomarkers related to MMD may further improve the outcome of MMD. Conclusions Although the etiology of MMD remains unknown, recent advances have been made in understanding the molecular biology and pathophysiology of this rare entity. Genome-wide and locus-specific association studies identified RNF213 as an important susceptibility gene of MMD. Ongoing research using genetically engineered mice lacking RNF213 by homologous recombination may provide new prospects for the basic research of this rare entity. Alternatively, variety of biomarkers are known to be involved in MMD, and an increased expression of angiogenic factors and pro-inflammatory molecules such as vascular endothelial growth factors and MMP-9 could be the therapeutic target in the future. This study was supported by JSPS KAKENHI Grant Number 24659642. The authors have no financial conflicts of interest. Figure 1 Gene construct of RNF213-deficient mice. Conventional knockout mice were generated by the Cre-lox system.

Conclusions Although the etiology of MMD remains unknown, recent advances have been made in understanding the molecular biology and pathophysiology of this rare entity. Genome-wide and locus-specific association studies identified RNF213 as an important susceptibility gene of MMD. Ongoing research using genetically engineered mice lacking RNF213 by homologous recombination may provide new prospects for the basic research of this rare entity. Alternatively, variety of biomarkers are known to be involved in MMD, and an increased expression of angiogenic factors and pro-inflammatory molecules such as vascular endothelial growth factors and MMP-9 could be the therapeutic target in the future. This study was supported by JSPS KAKENHI Grant Number 24659642. The authors have no financial conflicts of interest. Figure 1 Gene construct of RNF213-deficient mice. Conventional knockout mice were generated by the Cre-lox system. Figure 2 (A) Polymerase chain reaction (PCR) genotyping of wild-type mice (Wt.), homozygous RNF213- deficient mice (RNF213-/-) and heterozygous RNF213- deficient mice (RNF213-/+). (B, C) Microscopic view of the base of the brain at 16 weeks demonstrated no difference in the vascular structure of the circle of Willis between RNF213-/- and Wt. Figure 3 Photo-microscopic view of the wall of the common carotid artery (CCA) 14 days after CCA ligation (High-power, Elastica-Masson staining). Only Wt. exhibited temporary intimal hyperplasia and the medial layer after CCA ligation (Scale bar: 50 µm).

Figure 2 (A) Polymerase chain reaction (PCR) genotyping of wild-type mice (Wt.), homozygous RNF213- deficient mice (RNF213-/-) and heterozygous RNF213- deficient mice (RNF213-/+). (B, C) Microscopic view of the base of the brain at 16 weeks demonstrated no difference in the vascular structure of the circle of Willis between RNF213-/- and Wt. Figure 3 Photo-microscopic view of the wall of the common carotid artery (CCA) 14 days after CCA ligation (High-power, Elastica-Masson staining). Only Wt. exhibited temporary intimal hyperplasia and the medial layer after CCA ligation (Scale bar: 50 µm). Figure 4 Demographic view of the possible mechanism underlying the development of moyamoya disease. IC-EC conversion: Internal carotid-external carotid conversion as a compensatory physiological reorganization system for moyamoya disease. Figure 5 A 43-year old man with ischemic-onset moyamoya disease undergoing left superficial temporal artery-middle cerebral artery anastomosis. Postoperative single-photon emission computed tomography one day (A) and 14 days (B) after surgery demonstrating focal intense increase in cerebral blood flow at the site of the anastomosis (arrows). Figure 6 Postoperative magnetic resonance (MR) angiography (A) and MR imaging of T2-wighted images (B) indicating patent bypass (arrow in A) and no evidence of vasogenic edema (B).

Introduction Clinical practice guidelines (CPGs) are systematically developed statements aimed at helping practitioners and patients make informed health care decisions in specific clinical circumstances.1 Investigators of the Clinical Research Center for Stroke (CRCS), funded by the Ministry of Health and Welfare of Korea, developed and published the first edition of the Korean CPGs for stroke in 2009, and have since updated the guidelines to reflect and incorporate new evidence pertinent to clinical practice. The Korean CPGs for stroke have been endorsed by relevant and respected academic societies and have been distributed in the forms of monographs, summary manuals, freely accessible PDF files on the websites of the CRCS and the Korean Stroke Society, and through smart phone applications in order to ensure widespread dissemination and implementation. CPGs, which help incorporate scientific advancements into daily practice, are expected to improve the quality of care, lead to better patient outcomes, avoid unnecessary cost, and serve as good educational tools.2 However, physicians' attitudes toward CPGs, as well as their confidence in the instruments, are essential for successful implementation and physician adherence.3 Since the publication of the Korean CPGs for stroke, the attitudes and confidence levels of Korean neurologists toward the guidelines have not been explored.

However, physicians' attitudes toward CPGs, as well as their confidence in the instruments, are essential for successful implementation and physician adherence.3 Since the publication of the Korean CPGs for stroke, the attitudes and confidence levels of Korean neurologists toward the guidelines have not been explored. Methods This study was designed a priori as a sub-study within the Real world of Lipid-Lowering therapy in Korean Stroke patients (ROLLERKOST), which aimed to assess Korean neurologists' knowledge of current dyslipidemia management guidelines and guideline-based discharge prescriptions for statin amongst patients hospitalized with acute ischemic stroke or transient ischemic attack. Details of the ROLLERKOST study have already been published,4 but to summarize, 33 centers that actively enroll their patients in the Korean Stroke Registry were selected from a total of 86 neurology training hospitals in Korea.5 These 33 centers were sent an e-mail describing the purpose of the study and requesting their participation. Consent was received from 27 centers. Between November 2010 and December 2011, we conducted a survey that directly interviewed neurologists (board-certified neurologists and residents) from the 27 centers, using a structured questionnaire composed of four main principles: physician characteristics (5 items), practice patterns of dyslipidemia management (15 items), knowledge of the current dyslipidemia management guidelines (10 items) and attitudes to and confidence in the current guidelines (21 items).

om the 27 centers, using a structured questionnaire composed of four main principles: physician characteristics (5 items), practice patterns of dyslipidemia management (15 items), knowledge of the current dyslipidemia management guidelines (10 items) and attitudes to and confidence in the current guidelines (21 items). The current study analyzed responses to the 21 questions on physicians' attitudes towards and confidence in the current Korean CPGs for stroke (Table 1). We used a 5-point Likert scale to rate physicians' responses to each of the 21 questions: strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree. The respondents' attitudes were then divided into positive views (strongly agree and agree) and negative views (strongly disagree and disagree). Neutral responses (neither agree nor disagree) were disregarded. The internal consistency of the responses of physicians to the 21 questions was examined using Cronbach's alpha. In measuring the consistency, the responses of items 14, 15, 16, 17, 18, 20, and 21 as listed in Table 1 were reversed to ensure that all items had the same direction. Descriptive statistics were used to present the results of physicians' responses to each item. In addition, we investigated which guidelines Korean neurologists usually referred to during their clinical practice of dyslipidemia management (the Korean CPGs, National Cholesterol Education Program-Adult Treatment Panel III guidelines,6 or the reimbursement guidelines provided by the Korean Health Insurance Review and Assessment Service).

gated which guidelines Korean neurologists usually referred to during their clinical practice of dyslipidemia management (the Korean CPGs, National Cholesterol Education Program-Adult Treatment Panel III guidelines,6 or the reimbursement guidelines provided by the Korean Health Insurance Review and Assessment Service). Results Of the 33 centers contacted, 27 centers participated in the study (81.8%). These consisted of 18 university hospitals, 7 affiliated hospitals, and 2 secondary referral hospitals. A total of 174 neurologists who were both actively involved in managing patients and consented to face-to-face interviews responded to the questionnaire. The average age of respondents was 33.6 years±7.1, 49 (28.2%) were female, 73 (42.0%) were board-certified neurologists and 66 (37.9%) indicated that their subspecialty was stroke medicine. Detailed demographic characteristics are presented in Table 2. The internal consistency of responses was relatively high (Cronbach's α=0.8677).

ge of respondents was 33.6 years±7.1, 49 (28.2%) were female, 73 (42.0%) were board-certified neurologists and 66 (37.9%) indicated that their subspecialty was stroke medicine. Detailed demographic characteristics are presented in Table 2. The internal consistency of responses was relatively high (Cronbach's α=0.8677). Most of the respondents reported a positive attitude to the use of CPGs for stroke in their clinical decision-making (Figure 1). The mean value of negative views regarding all questionnaires (excluding the statements that were neither positive nor negative) was 14.9%. More than 70% of the respondents responded that they used the Korean CPGs for stroke and believed that treatments following these guidelines are likely to be effective without infringing on physician autonomy. Only 34% of the respondents, however, were confident that CPGs for stroke are unbiased statements, and a sizable proportion of respondents (48.9%) complained about a perceived lack of education materials, despite a healthy majority of neurologists holding the belief that CPGs for stroke are developed to improve the quality of care afforded to patients and that they are a good educational tool (Figure 1).

statements, and a sizable proportion of respondents (48.9%) complained about a perceived lack of education materials, despite a healthy majority of neurologists holding the belief that CPGs for stroke are developed to improve the quality of care afforded to patients and that they are a good educational tool (Figure 1). In terms of the level of knowledge of current dyslipidemia management guidelines, there was no difference in the attitudes or confidence levels on the distribution of responses between the higher- and lower-knowledge level groups for all questions. The median score for the neurologists' knowledge of the current guidelines was 70 (range, 30-100). A total of 79 (45.4%) neurologists were thus categorized into the higher-level knowledge group, having achieved a score of >70, and 95 (54.6%) were categorized into the lower-level knowledge group. When between-group responses were compared for specialization (stroke neurologists vs. non-stroke neurologists) and certification (board-certified neurologists vs. residents), stroke neurologists and board-certified neurologists showed significantly higher positive responses (3.58±0.41 vs. 3.39±0.40, P=0.0027; 3.57±0.42 vs. 3.38±0.39, P=0.0034, respectively). Respondents' practical use of CPGs for stroke is presented in Figure 2. It was found that over 60% of the health care providers in this survey adhered to domestic CPGs during dyslipidemia management (American Heart Association guidelines, 32%, Korean Health Insurance guidelines, 5%).

In terms of the level of knowledge of current dyslipidemia management guidelines, there was no difference in the attitudes or confidence levels on the distribution of responses between the higher- and lower-knowledge level groups for all questions. The median score for the neurologists' knowledge of the current guidelines was 70 (range, 30-100). A total of 79 (45.4%) neurologists were thus categorized into the higher-level knowledge group, having achieved a score of >70, and 95 (54.6%) were categorized into the lower-level knowledge group. When between-group responses were compared for specialization (stroke neurologists vs. non-stroke neurologists) and certification (board-certified neurologists vs. residents), stroke neurologists and board-certified neurologists showed significantly higher positive responses (3.58±0.41 vs. 3.39±0.40, P=0.0027; 3.57±0.42 vs. 3.38±0.39, P=0.0034, respectively). Respondents' practical use of CPGs for stroke is presented in Figure 2. It was found that over 60% of the health care providers in this survey adhered to domestic CPGs during dyslipidemia management (American Heart Association guidelines, 32%, Korean Health Insurance guidelines, 5%). Discussion Before designing a strategy for the implementation of CPGs, it is important to explore the attitudes and opinions held by health care professionals. Skeptical opinions negatively influence the implementation of guidance, either directly or indirectly, through the creation of an unfavorable environment characterized by lack of support from peers and superiors.7 The positive attitudes and opinions toward CPGs for stroke held by Korean neurologists as reported in this survey, combined with the findings of previous studies among diverse groups of health care professionals,8,9 indicate that the current attitudes of physicians do not form a barrier to the future implementation of CPGs for stroke.

itudes and opinions toward CPGs for stroke held by Korean neurologists as reported in this survey, combined with the findings of previous studies among diverse groups of health care professionals,8,9 indicate that the current attitudes of physicians do not form a barrier to the future implementation of CPGs for stroke. Responses to the questionnaire were reasonably consistent for all respondents (Cronbach's α=0.8677). It was, however, somewhat discouraging that a negative average response to the statement "The specialists' opinions are coordinated without any bias" (item number 16) was obtained, as similar questions such as "They are made by specialists who lack actual experience in medical fields" (item number 5) and "It is hard to agree on the guidelines" (item number 8), yielded positive responses (Figure 1). This discrepancy might be explained by the negative phraseology of the statement, but the possibility that this might be a true response cannot be excluded.

s who lack actual experience in medical fields" (item number 5) and "It is hard to agree on the guidelines" (item number 8), yielded positive responses (Figure 1). This discrepancy might be explained by the negative phraseology of the statement, but the possibility that this might be a true response cannot be excluded. Respondents' work habits also concurred with their attitudes and opinions toward CPGs for stroke. A total of 64% of respondents stated that they follow the treatment advised by the dyslipidemia guidelines, and 66.5% had already read the dyslipidemia guidelines in the management of stroke patients (Figure 2). This finding might be related to professionals' experience in using CPGs for stroke and a culture of evidence-based practice. Furthermore, specific criteria, such as whether the source of the guidelines is a credible and respected body or organization, encourage health care providers to use and adhere to certain CPGs.10 As a result, the fact that our CPGs for stroke have been accredited by the Korean Stroke Society, the Korean Neurological Association, and the Korean Society of Cerebrovascular Surgeons might be one of the more significant reasons for their implementation.

rage health care providers to use and adhere to certain CPGs.10 As a result, the fact that our CPGs for stroke have been accredited by the Korean Stroke Society, the Korean Neurological Association, and the Korean Society of Cerebrovascular Surgeons might be one of the more significant reasons for their implementation. In our previous ROLLERKOST study, Korean neurologists with a higher knowledge level were more likely to adhere to guideline-based discharge prescription of statin.4 In this analysis, however, there was no significant difference in the attitudes or confidence on the distribution of responses between the higher- and lower-knowledge level groups for all questions. It has not been well demonstrated whether knowledge level is associated with adherence to guidelines in clinical practice. Physicians' increased familiarity with the guidelines and their improved knowledge of them may correlate with a higher rate of adherence to prescription, but there was no association between knowledge level and attitudes toward guidelines for stroke.

evel is associated with adherence to guidelines in clinical practice. Physicians' increased familiarity with the guidelines and their improved knowledge of them may correlate with a higher rate of adherence to prescription, but there was no association between knowledge level and attitudes toward guidelines for stroke. To the authors' knowledge, this is the first study in which attitudes towards CPGs for stroke among Korean neurologists have been examined. However, we also admit several methodological limitations. Respondent physicians were affiliated with neurology training hospitals, so this may limit the generalizability of our findings. Moreover, the mean age of physicians was quite low because most respondents were either trainee residents or had only become certified neurologists within the last six years. Further studies should therefore investigate the reasons behind non-adherence following the implementation of specific CPGs among general physicians of a wider age range. Conclusions This study shows that most of the respondents in our survey held positive attitudes and opinions regarding the use of the Korean CPGs for stroke, whereas only a small percentage responded negatively. The positive attitudes and opinions toward the guidelines suggest that physicians' attitudes should not be regarded as a potential barrier to the implementation of Korean CPGs for stroke. This study was sponsored by Pfizer Pharmaceuticals Korea Ltd. The authors have no financial conflicts of interest. Figure 1 Attitudes and opinions of Korean neurologists toward clinical practice guidelines for stroke.

Conclusions This study shows that most of the respondents in our survey held positive attitudes and opinions regarding the use of the Korean CPGs for stroke, whereas only a small percentage responded negatively. The positive attitudes and opinions toward the guidelines suggest that physicians' attitudes should not be regarded as a potential barrier to the implementation of Korean CPGs for stroke. This study was sponsored by Pfizer Pharmaceuticals Korea Ltd. The authors have no financial conflicts of interest. Figure 1 Attitudes and opinions of Korean neurologists toward clinical practice guidelines for stroke. Figure 2 Respondents' work habits in using clinical practice guidelines for dyslipidemia management. Table 1 Items included in the questionnaire Table 2 Demographic characteristics of interviewed neurologists (n=174)

Introduction The ability of smart phones to deliver directed health information to the patients and healthcare providers is path breaking and has opened up an entirely new era of health communication. As of 2013, more than 50 billion and 48 billion apps have been downloaded from the Apple App Store and Android Play Store respectively by millions of smart phone users all over the world.1,2 An increasing number of medical professionals are using smart phone applications for accessing clinical information on the internet and as clinical calculators. A prior study had estimated that the percentage of medical professionals using smart phones will rise to 66-90% by 2012.3 Increasing prevalence and popularity of smart phone apps among health-care professionals complements the medical practice and has gained wide acceptance as a training and information tool.4,5 Smart phone apps are increasingly being used by the patients for information on early identification of stroke symptoms, risk factor awareness and management.6,7

ity of smart phone apps among health-care professionals complements the medical practice and has gained wide acceptance as a training and information tool.4,5 Smart phone apps are increasingly being used by the patients for information on early identification of stroke symptoms, risk factor awareness and management.6,7 Health-related smart phone apps are being actively studied for potential applications in a wide variety of fields of medical importance including but not limited to dermatology,8 ophthalmology9,10 and cancer.11 Many studies have questioned the validity and specificity of the health care information available on these subjects.12,13 The quality of information disseminated in these apps would therefore play an important role in determining the quality of health care provided. The primary aim of our study was to evaluate the stroke related apps available on Apple app store and the Android Play Store and to analyze the content and potential usefulness in the health care delivery system. Methods The Apple iTunes store and Android Google Play Store were searched for stroke-related applications on July 27, 2013. The search terms used to select the apps were: stroke, brain attack, intracranial hemorrhage, subarachnoid hemorrhage, cerebral infarction. The content of the applications was analyzed by two independent investigators (DD, AA). Each app was analyzed and classified on the basis of cost, target audience, type of information, validity, involvement of health-care agencies and usefulness based on audience reviews and ratings.

id hemorrhage, cerebral infarction. The content of the applications was analyzed by two independent investigators (DD, AA). Each app was analyzed and classified on the basis of cost, target audience, type of information, validity, involvement of health-care agencies and usefulness based on audience reviews and ratings. Costs Apps were categorized on the basis of fee/cost of download. Free: Apps which were downloaded free of cost. Paid: Apps which could be downloaded after making a payment. Target audience Apps were categorized based on the people for whom the apps were designed for. Health care workers: these apps included clinical information and recent advances targeted for health care workers to be used in the hospital/clinic/research laboratory. General population: these apps were those which could be utilized by general population for education, primary or secondary prevention and self management of their illnesses. Type of information Apps were divided on the basis of information provided. General information about the disease process: these included the apps providing information about etiology, pathogenesis and categorization of ischemic stroke, intracranial and subarachnoid hemorrhage. Recent research and advances: these apps provided information about latest research and advances in the field of vascular neurology which have impacted the patient management. Tools to be used by health care professionals: these were various apps (e.g. NIHSS, ABCD2, CHA2DS2-VASc etc.) used by health care professionals for evaluation and management.

Recent research and advances: these apps provided information about latest research and advances in the field of vascular neurology which have impacted the patient management. Tools to be used by health care professionals: these were various apps (e.g. NIHSS, ABCD2, CHA2DS2-VASc etc.) used by health care professionals for evaluation and management. Tools to be used by patients: these apps were utilized by the patients for maintaining medication compliance or management of risk factor(s) for stroke prevention. General awareness and support group apps: included information about various awareness programs and support groups for patients and their relatives. Miscellaneous: this included relevant or irrelevant information on stroke which could not be categorized into any of the above categories. Uploading agencies Apps were divided into two groups based on the agency which sold or uploaded the app on Apple iTunes or Google Play Store. Health Care Agencies (HCAs): these included national or international medical/surgical associations, medical universities/medical schools, pharmaceutical companies, hospitals, research associations or medical journals. Non Health Care Agencies (Non HCAs): private or government agencies which do not belong to the above mentioned organizations.

Health Care Agencies (HCAs): these included national or international medical/surgical associations, medical universities/medical schools, pharmaceutical companies, hospitals, research associations or medical journals. Non Health Care Agencies (Non HCAs): private or government agencies which do not belong to the above mentioned organizations. Scientific validity The scientific validity was based on the references provided by uploading agency/individual along with application. The apps which didn't have references, validity was decided based on whether or not the information or tool was compatible with updated stroke literature, expert reviews, national guidelines and recommendations by American Stroke Association. Scientifically validated apps: content of these apps was backed by scientific evidence. Non Validated apps: these apps did not contain scientific evidence based information. Usefulness It was decided on the basis of audience reviews and ratings. Not useful: apps with poor ratings or comments (e.g. less than 2 stars). Moderately useful: apps which were rated as average by the users (e.g. 2-3 stars). Very useful: apps with good rating and reviews by the users (e.g. more than 3 stars). Statistical analysis Values are expressed as simple proportions. Univariable analysis was performed using Pearson's chi-squared test or Fisher's exact test for categorical variables. A 2-sided P value of <0.05 was considered significant. Kappa coefficient of agreement between two investigators was calculated. All statistical analysis was done using IBM SPSS Statistics 19.0 for Windows (SPSS Inc.).

lysis was performed using Pearson's chi-squared test or Fisher's exact test for categorical variables. A 2-sided P value of <0.05 was considered significant. Kappa coefficient of agreement between two investigators was calculated. All statistical analysis was done using IBM SPSS Statistics 19.0 for Windows (SPSS Inc.). Results A total of 107 apps were identified using the search terms. Fourteen apps didn't have information about stroke, therefore excluded from the study. A total of 93 relevant applications were analyzed. Forty three of these apps were uploaded on Google Play Store and fifty on Apple iTunes. Descriptive analysis of information collected in various categories was performed (Table 1). Costs Out of the 93 apps, 47% (44) were available free of cost. More than half of these free apps (24) were considered very useful by the audience. Whereas 27 of 49 paid apps were considered very useful as per user ratings. 61% (27) free apps and 57% (28) paid apps had scientifically valid information. There was no statistical association between cost of app and scientific validity or usefulness. 55% (24) free apps and 49% (24) paid apps were targeted for health care professionals.

aid apps were considered very useful as per user ratings. 61% (27) free apps and 57% (28) paid apps had scientifically valid information. There was no statistical association between cost of app and scientific validity or usefulness. 55% (24) free apps and 49% (24) paid apps were targeted for health care professionals. Target audience Over half of all stroke related apps (48) were aimed towards health care workers and 45 were meant to be used by general public. 75% of apps aimed at health care workers could be utilized as bedside tools for patient care and remainder had information related to recent research advances. More than half (55%) of the apps for health care workers and 46% apps for general public were free of cost. Nearly 90% (43) apps for health care professionals were scientifically valid where as only 27% (12) targeting general population had the backing of scientific evidence. This difference was found to statistically significant (P<0.01).

of the apps for health care workers and 46% apps for general public were free of cost. Nearly 90% (43) apps for health care professionals were scientifically valid where as only 27% (12) targeting general population had the backing of scientific evidence. This difference was found to statistically significant (P<0.01). Type of information Majority of apps provided tools to be used by HCAs (38.3%) and general public (35.5%). 4.3% apps provided general information about stroke etiology, pathogenesis and classification, where as 15.1% and 4.3% had information about recent advances in stroke research and support groups respectively. Majority (72%) tool applications for HCAs were considered very useful by the reviewers. 60% of tools for non-HCAs were considered moderately useful as per the user ratings. 92% of tools for HCAs and only 20% tools for non HCAs tools were backed by scientific literature. There was statistically significant difference between scientific validity of HCAs and non HCAs tool applications (P<0.05). Uploading agency There was a significant participation of healthcare agencies in dissemination of stroke related information with 47.3% (44) apps being uploaded by them. Remaining 52.7% (49) apps were uploaded by Non HCAs. 66% (29) apps uploaded by HCAs were rated very useful and only 11% (5) considered non useful by the users. 70% (31) and 49% (24) apps uploaded by HCAs and non HCAs were considered scientifically valid. The difference in scientific validity between apps uploaded by HCAs and non HCAs was not statistically significant.

s. 66% (29) apps uploaded by HCAs were rated very useful and only 11% (5) considered non useful by the users. 70% (31) and 49% (24) apps uploaded by HCAs and non HCAs were considered scientifically valid. The difference in scientific validity between apps uploaded by HCAs and non HCAs was not statistically significant. Usefulness Majority (55%) of the apps were rated as very useful by users. 38% were moderately useful and remaining (7%) non-useful. 56% (27 very useful and 20 moderately useful) apps were paid apps. There was no significant difference between cost of the apps and usefulness. Scientific validity Majority of apps (59.1%) were scientifically valid. Most of these (78%) were aimed at healthcare professionals (Table 2). The difference in scientific validity between the apps aimed at general population versus healthcare professionals was statistically significant (P<0.01). 96% (53) of the valid apps were considered useful (42 very useful and 11 moderately useful) by the users. Interestingly 87% (33) of non valid apps were also considered useful (9 very useful and 24 moderately useful). Kappa Coefficient of agreement between the two investigators (DD, AA) was 0.97.

Scientific validity Majority of apps (59.1%) were scientifically valid. Most of these (78%) were aimed at healthcare professionals (Table 2). The difference in scientific validity between the apps aimed at general population versus healthcare professionals was statistically significant (P<0.01). 96% (53) of the valid apps were considered useful (42 very useful and 11 moderately useful) by the users. Interestingly 87% (33) of non valid apps were also considered useful (9 very useful and 24 moderately useful). Kappa Coefficient of agreement between the two investigators (DD, AA) was 0.97. Discussion The world is seeing an exponential increase in the use of smart phone applications as a source of valid and updated health-care related information by the general public and health-care professionals.14,15 Smart phone applications have been studied as a source of information for different medical specialties.4,5,8,9,10,11,12,13,16 Paucity of scientific accuracy and clinical relevance of majority of apps directed at general users is a growing concern.7,17,18,19,20 Even the applications available to the HCAs are not screened for the quality and validity of information.18,19,20 Our study addresses the nature and volume of information about stroke available to users, both HCAs and non-HCAs, through iPhone apps and Android Play Store. In our study we found that there was no statistical association between cost of app and scientific validity or usefulness.

ty and validity of information.18,19,20 Our study addresses the nature and volume of information about stroke available to users, both HCAs and non-HCAs, through iPhone apps and Android Play Store. In our study we found that there was no statistical association between cost of app and scientific validity or usefulness. Nearly 90% (43) apps targeting health care professionals were scientifically valid where as only 27% (12) targeting general population had the backing of scientific evidence. There was statistically significant difference between scientific validity of HCAs and non HCAs tool apps further emphasizing the need for a central agency which can screen the sensitive health care related information being made available to the general population.

al population had the backing of scientific evidence. There was statistically significant difference between scientific validity of HCAs and non HCAs tool apps further emphasizing the need for a central agency which can screen the sensitive health care related information being made available to the general population. Although some of the previous studies on utilization of smart phone applications in health care were focused only on Apple iTunes,11 our study did a comprehensive review by including both the android Play store and Apple iTunes as our database. A limitation of our study was that we analyzed only English language apps and the apps available in other languages were not included in the study. Tools formed a considerable proportion of the apps for stroke used by health care professional or patients. Due to limited number of apps certain variables could not be analyzed to assess for statistical significance. Since the number of users rating the apps varied, there is a potential risk for selection bias while assessing usefulness of stroke related apps. There were 9 apps that were common among Apple iTunes and Android Play store, these were included separately since these apps would be downloaded and utilized by different users. But analyzing these as separate apps may have affected the results pertaining of type of information and target audience.

troke related apps. There were 9 apps that were common among Apple iTunes and Android Play store, these were included separately since these apps would be downloaded and utilized by different users. But analyzing these as separate apps may have affected the results pertaining of type of information and target audience. The smart phone applications are fast becoming the peg of information to both the health care providers and the general population. There is a huge demand in the society for health-care information about debilitating and fatal illnesses like stroke and cardiovascular diseases. Information can now be targeted and provided right at the fingertips of the patients, care-takers and health professionals. It is very important that both valid and accurate health-related information is provided in a cost effective manner to those seeking it. Conclusions The lack of specificity and validity of the app content has a risk of potentially endangering patient safety. Therefore, there is a need to set up regulatory guidelines to improve the quality and validity of information disseminated by the smart phone applications. Also, encouraging the involvement of health-care agencies in developing apps aimed at healthcare professionals and general audience would ensure that valid and relevant information reaches the consumers. The authors have no financial conflicts of interest. Table 1 Descriptive analysis of stroke related smart phone applications Table 2 Outcome of descriptive analysis of stroke related smart phone apps

Introduction Cancer and ischemic stroke share many risk factors that lead to increased risk of stroke in cancer patients. A report of autopsies performed on cancer patients indicated that 14.6% patients had evidence of pathological cerebrovascular disease and 7.4% had clinical symptoms of stroke.1 However, only 3.5% of cancer patients sustained a stroke; these findings were similar with that of the general population.2 The causes of ischemic stroke in patients with and without cancer may be different. Cancer increases a patient's hypercoagulable states that increase deep vein thrombosis and nonbacterial thrombotic endocarditis.3 The pathogenesis of stroke, including the presence of a hypercoagulable state and increased D-dimer levels, an important factor, in most cancer patients is different from that of the general population.4,5 Other reports have stated that atherosclerosis and coagulopathy may be the dominant factors in the pathogenesis of stroke in cancer patients.2 Therefore, there has been dispute regarding the risk factors in the pathogenesis of stroke in cancer patients.

ncer patients is different from that of the general population.4,5 Other reports have stated that atherosclerosis and coagulopathy may be the dominant factors in the pathogenesis of stroke in cancer patients.2 Therefore, there has been dispute regarding the risk factors in the pathogenesis of stroke in cancer patients. Stroke biomarkers can be used in understanding pathophysiology, screening high-risk patients, predicting clinical outcomes, and effective treatment of stroke patients.6 The biomarkers D-dimer, brain natriuretic peptide (BNP), fibrinogen, C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR) are significantly increased in patients with ischemic stroke.7,8 Moreover, increased D-dimer and BNP levels are associated with nonbacterial thrombotic endocarditis, and increased ESR is associated with lacunar and atherothrombotic stroke.7,8 The aim of this study was to investigate risk factors, biomarkers, and etiology of ischemic stroke in cancer patients by comparing them with age- and sex-matched noncancer patients with ischemic stroke.

Stroke biomarkers can be used in understanding pathophysiology, screening high-risk patients, predicting clinical outcomes, and effective treatment of stroke patients.6 The biomarkers D-dimer, brain natriuretic peptide (BNP), fibrinogen, C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR) are significantly increased in patients with ischemic stroke.7,8 Moreover, increased D-dimer and BNP levels are associated with nonbacterial thrombotic endocarditis, and increased ESR is associated with lacunar and atherothrombotic stroke.7,8 The aim of this study was to investigate risk factors, biomarkers, and etiology of ischemic stroke in cancer patients by comparing them with age- and sex-matched noncancer patients with ischemic stroke. Methods Patient selection The target group comprised cancer patients with ischemic stroke who were admitted to a general hospital in Busan, Korea, between January 2003 and December 2012. The cancer ischemic stroke group comprised patients with active cancer prior to the onset of ischemic stroke. This study did not consider the duration between cancer diagnosis and the onset of the ischemic event. The control group comprised age- and sex-matched noncancer patients with ischemic stroke who were admitted to the same hospital during the same period. The control group patients were selected at random.

hemic stroke. This study did not consider the duration between cancer diagnosis and the onset of the ischemic event. The control group comprised age- and sex-matched noncancer patients with ischemic stroke who were admitted to the same hospital during the same period. The control group patients were selected at random. Data collection A retrospective review of cancer patients with ischemic stroke was performed to identify the cancer types, metastasis, stroke subtypes, stroke risk factors, and clinical and laboratory results. Hypertension, hyperlipidemia, atrial fibrillation, ischemic heart disease, smoking, alcohol intake, and past individual and family history of stroke were established as risk factors of stroke. ESR, high-sensitivity CRP (hs-CRP), fibrinogen, pro-BNP, D-dimer, and homocysteine were established as biomarkers of stroke.

s. Hypertension, hyperlipidemia, atrial fibrillation, ischemic heart disease, smoking, alcohol intake, and past individual and family history of stroke were established as risk factors of stroke. ESR, high-sensitivity CRP (hs-CRP), fibrinogen, pro-BNP, D-dimer, and homocysteine were established as biomarkers of stroke. Definition of stroke risk factors Hypertension was established when systolic blood pressure was >140 mmHg and diastolic blood pressure was >90 mmHg at rest, or if the patient was taking hypertension medication. Diabetes was established when fasting glucose level was >126 mg/dL, postprandial 2-h glucose level was >200 mg/dL, or if the patient was taking medication for diabetes. Hyperlipidemia was established when one or more of the following levels were elevated: total cholesterol level, >240 mg/dL; low-density lipoprotein cholesterol level, >130 mg/dL; or triglyceride level >200 mg/dL. Atrial fibrillation was confirmed via an electrocardiogram taken at patient admission. Smoking was established when the patient had a past history of ≥5 cigarettes per day. Alcohol intake was established with binge drinking or intake of ≥1 times per week, irrespective of the amount consumed.

glyceride level >200 mg/dL. Atrial fibrillation was confirmed via an electrocardiogram taken at patient admission. Smoking was established when the patient had a past history of ≥5 cigarettes per day. Alcohol intake was established with binge drinking or intake of ≥1 times per week, irrespective of the amount consumed. Stroke biomarkers In order to examine biomarkers, blood was collected from patients with ischemic stroke during the acute ischemic event. Stroke biomarker concentrations were determined by assessing patient serum. Erythrocyte sedimentation rate was measured via photometrical capillary stopped flow kinetic analysis; hs-CRP level was measured via latex-enhanced immunoturbidimetric analysis; fibrinogen level was measured via clotting assay based on viscosity; D-dimer level was measured via immunoturbidimetry; and homocysteine level was measured via direct chemiluminescent immunoassay. Stroke subtypes Brain magnetic resonance imaging, magnetic resonance angiography, transcranial Doppler ultrasonography, carotid ultrasonography, electrocardiography, and echocardiography were performed on most patients admitted with ischemic stroke. Stroke etiology was based on the TOAST classification9 and was classified as large-artery atherosclerosis, cardioembolism, small artery occlusion, and other determined and undetermined etiologies.

d ultrasonography, electrocardiography, and echocardiography were performed on most patients admitted with ischemic stroke. Stroke etiology was based on the TOAST classification9 and was classified as large-artery atherosclerosis, cardioembolism, small artery occlusion, and other determined and undetermined etiologies. Statistical analysis All data were analyzed using SPSS software (version 17.0; SPSS Inc., Chicago, IL). The Chi-square test was used to compare stroke risk factors and subtypes between cancer patients with ischemic stroke and the control group. Student's t-test was used to compare biomarkers between the two groups. Statistical significance was based on a P value of <0.05. Results Clinical characteristics The study included 156 (103 [66%] male; 53 [34%] female) cancer patients with ischemic stroke. Patient age ranged from 40 to 89 years, with an average of 70.3±9.8 years. The control group also comprised 103 male and 53 female noncancer patients, with their age ranging from 40 to 89 years, and a mean age of 70.3±9.8 years. The types of primary cancer in patients with ischemic cancer are listed in Table 1; the incidences (the descending order) of these types are as follows: stomach cancer in 33 (21.2%) patients, lung cancer in 25 (16%), colon cancer in 22 (14.1%), and hepatobiliary cancer in 20 (12.8%). Metastasis was present in 84 (53.8%) patients.

f primary cancer in patients with ischemic cancer are listed in Table 1; the incidences (the descending order) of these types are as follows: stomach cancer in 33 (21.2%) patients, lung cancer in 25 (16%), colon cancer in 22 (14.1%), and hepatobiliary cancer in 20 (12.8%). Metastasis was present in 84 (53.8%) patients. Comparison of stroke risk factors The risk factors of ischemic stroke in cancer patients, in descending order of incidence, are as follows: hypertension, diabetes, smoking, alcohol intake, atrial fibrillation, ischemic heart disease, past history of stroke, and hyperlipidemia. Compared with the noncancer patients with ischemic stroke, the incidences of these risk factors were statistically lower in hypertension (P<0.01), atrial fibrillation (P<0.01), hyperlipidemia (P<0.01), ischemic heart disease (P=0.02), and family history of stroke (P=0.02) (Table 2). Comparison of stroke biomarkers The stroke biomarkers ESR and hs-CRP, fibrinogen, pro-BNP, and D-dimer levels were statistically higher in cancer patients with ischemic stroke (all P<0.01) than in noncancer patients, whereas no difference was found in the homocysteine levels between the groups (Table 3).

Comparison of stroke risk factors The risk factors of ischemic stroke in cancer patients, in descending order of incidence, are as follows: hypertension, diabetes, smoking, alcohol intake, atrial fibrillation, ischemic heart disease, past history of stroke, and hyperlipidemia. Compared with the noncancer patients with ischemic stroke, the incidences of these risk factors were statistically lower in hypertension (P<0.01), atrial fibrillation (P<0.01), hyperlipidemia (P<0.01), ischemic heart disease (P=0.02), and family history of stroke (P=0.02) (Table 2). Comparison of stroke biomarkers The stroke biomarkers ESR and hs-CRP, fibrinogen, pro-BNP, and D-dimer levels were statistically higher in cancer patients with ischemic stroke (all P<0.01) than in noncancer patients, whereas no difference was found in the homocysteine levels between the groups (Table 3). Comparison of stroke etiology The different etiologies according to ischemic stroke subtypes in cancer patients, in descending order of incidence, are as follows: large-artery atherosclerosis (41%), small artery occlusion (28.8%), cardioembolism (17.3%), undetermined etiology (8.3%), and other determined etiology (4.5%). The incidences of large-artery atherosclerosis and undetermined etiology were higher, whereas those of cardioembolism and small artery occlusion were lower in cancer patients than in noncancer patients (P=0.02) (Figure 1).

.8%), cardioembolism (17.3%), undetermined etiology (8.3%), and other determined etiology (4.5%). The incidences of large-artery atherosclerosis and undetermined etiology were higher, whereas those of cardioembolism and small artery occlusion were lower in cancer patients than in noncancer patients (P=0.02) (Figure 1). Discussion The incidences of hypertension, atrial fibrillation, hyperlipidemia, and ischemic heart disease were significantly lower in cancer patients with ischemic stroke than in noncancer patients with ischemic stroke. In contrast, biomarkers such as ESR and hs-CRP, fibrinogen, pro-BNP, and D-dimer levels were significantly increased in cancer patients with ischemic stroke.

tion, hyperlipidemia, and ischemic heart disease were significantly lower in cancer patients with ischemic stroke than in noncancer patients with ischemic stroke. In contrast, biomarkers such as ESR and hs-CRP, fibrinogen, pro-BNP, and D-dimer levels were significantly increased in cancer patients with ischemic stroke. Smoking is a common risk factor in both ischemic stroke and cancer and is also the etiology for most ischemic strokes in cancer patients.2,10 Most reports have found that hypertension and smoking are the most common causes of ischemic stroke in cancer patients. Additional causes include diabetes, ischemic heart disease, hyperlipidemia, alcohol intake, and atrial fibrillation.10,11,12,13,14 This study compared noncancer and cancer patients with ischemic stroke and found that the cancer group had a lower incidence of hypertension, atrial fibrillation, hyperlipidemia, and ischemic heart disease. Previous reports10,11,12,14 have found no differences in common stroke risk factors between cancer and noncancer patients with ischemic stroke. However, other reports have found low incidences of hypertension and hyperlipidemia in cancer patients with ischemic stroke compared with those of the noncancer group.15,16 Moreover, no significant differences in hyperlipidemia and atrial fibrillation incidences have been reported, but a tendency for lower incidence in cancer patients with ischemic stroke.14 This study hypothesized that in addition to the commonly known stroke risk factors, other factors may also be involved in cancer patients with ischemic stroke.

fferences in hyperlipidemia and atrial fibrillation incidences have been reported, but a tendency for lower incidence in cancer patients with ischemic stroke.14 This study hypothesized that in addition to the commonly known stroke risk factors, other factors may also be involved in cancer patients with ischemic stroke. This study found that stroke biomarkers such as ESR and hs-CRP, fibrinogen, pro-BNP, and D-dimer levels were significantly increased in cancer patients compared with noncancer patients with ischemic stroke. D-dimer, a fibrin degradation product that is directly related to coagulation and fibrinolysis, is a marker for hypercoagulability.17 Its levels are increased in acute, subacute, and chronic ischemic stroke.7 Compared with noncancer patients with ischemic stroke, cancer patients with ischemic stroke have statistically significant higher levels of D-dimer,15,16 which coincides with our results. In previous studies, cancer patients with ischemic stroke (without common stroke risk factors) were found to have statistically significant higher levels, which corresponded with an increased incidence of embolic signal in transcranial Doppler monitoring; thus, it was hypothesized that coagulopathy and emboli in cancer patients may be the etiology for ischemic stroke.4,5,18

e (without common stroke risk factors) were found to have statistically significant higher levels, which corresponded with an increased incidence of embolic signal in transcranial Doppler monitoring; thus, it was hypothesized that coagulopathy and emboli in cancer patients may be the etiology for ischemic stroke.4,5,18 Fibrinogen is involved in thrombokinesis, platelet aggregation, blood viscosity, thrombus formation, and inflammation. Increase in fibrinogen concentrations is directly related to an increase in ischemic stroke incidence.19 Cancer patients with ischemic stroke have been found to have a statistically significant increased fibrinogen levels.8,20 No statistically significant difference was found in fibrinogen levels between cancer and noncancer patients with ischemic stroke; however, the cancer group had a tendency to have a higher fibrinogen levels.13 However, in the current study, cancer patients with ischemic stroke had a statistically significant higher level of fibrinogen than noncancer patients. Therefore, we hypothesize that the high fibrinogen levels may be because cancer is the primary disease.

ancer group had a tendency to have a higher fibrinogen levels.13 However, in the current study, cancer patients with ischemic stroke had a statistically significant higher level of fibrinogen than noncancer patients. Therefore, we hypothesize that the high fibrinogen levels may be because cancer is the primary disease. Brain natriuretic peptide is a vasoactive peptide hormone that increases sodium excretion, urination, and vasodilation as well during the acute stage of ischemic stroke, and it is also a predictor of poor prognosis.7,21 It is increased in the cardioembolic subtype of ischemic stroke and is highly suggestive of cardioembolic etiology when serum concentrations are >140 pg/mL.22 This study found that cancer patients with ischemic stroke had a statistically significant higher BNP levels. This study also found that even though noncancer patients with ischemic stroke had higher incidences of atrial fibrillation and ischemic heart disease, BNP concentrations were statistically higher in the cancer group with ischemic stroke. Therefore, we hypothesize that the high levels may be because cancer was the primary disease.

lso found that even though noncancer patients with ischemic stroke had higher incidences of atrial fibrillation and ischemic heart disease, BNP concentrations were statistically higher in the cancer group with ischemic stroke. Therefore, we hypothesize that the high levels may be because cancer was the primary disease. C-reactive protein is related to inflammation and progression of hemangioendotheliomal damage as an acute peptide, which increases in systemic inflammation and artherosclerotic disease.23 Its levels increase in acute ischemic stroke, and CRP is a useful aid in prognosis.7,8 No statistically significant difference was found in CRP levels between cancer and noncancer patients with ischemic stroke.13 On the other hand, we found that cancer patients with ischemic stroke had a statistically significant higher CRP concentrations. Therefore, we hypothesize that the high levels may be because cancer was the primary disease.

difference was found in CRP levels between cancer and noncancer patients with ischemic stroke.13 On the other hand, we found that cancer patients with ischemic stroke had a statistically significant higher CRP concentrations. Therefore, we hypothesize that the high levels may be because cancer was the primary disease. Erythrocyte sedimentation rate is a nonspecific marker of inflammation and infection. Inflammation activates coagulation, and a continued increase in ESR indicates progressive thrombogenesis and fibrinolysis in patients with stroke. In addition to being a predictor of poor prognosis,24 ESR is a surrogate marker for inflammation in small and large vasculopathies. Moreover, increases in ESR are related to artherosclerotic and lacunar stroke.8 This present study found that cancer patients with ischemic stroke had statistically significant higher ESRs. Therefore, we hypothesize that the high levels may be because of the presence of both cancer and ischemic stroke. Homocysteine is an aminoacid byproduct of methionine metabolism, which aggravates artherosclerosis by damaging hemangioendothelial cells, increasing platelet aggregation through changing arachidonic acid metabolism, and decreasing anticoagulation factor activity.25 Increased homocysteine levels correlate with stenosis and occlusion of intracranial vessels in patients with ischemic stroke, and it also increases stroke risk.26,27 In the current study, no differences in homocysteine concentrations were found between cancer and noncancer patients with ischemic stroke.

tor activity.25 Increased homocysteine levels correlate with stenosis and occlusion of intracranial vessels in patients with ischemic stroke, and it also increases stroke risk.26,27 In the current study, no differences in homocysteine concentrations were found between cancer and noncancer patients with ischemic stroke. We found significantly higher levels of biomarkers such as ESR, hs-CRP, fibrinogen, pro-BNP, and D-dimer in the cancer with ischemic stroke group than those in the noncancer ischemic stroke group. These results imply a relationship between cancer patients and inflammation, thrombus formation, and coagulopathy, which could explain the increased frequency of ischemic stroke in the cancer patients compared with the noncancer patients. The etiology in cancer patients with ischemic stroke consists of coagulopathies due to cytokine or microparticle-releasing cancer cells, intravascular coagulation or nonbacterial thrombotic endocarditis, occlusion due to metastasis, direct vascular compression due to metastasis or neurocarcinomas, coagulopathies due to chemotherapy or radiation therapy, vascular embolization (e.g., treatment of cancer vessels), and paraneoplastic causes.11,18,28 In previous studies, of the cancer patients with ischemic stroke, 40% did not have any etiology of general stroke patients, and these patients are thought to have closer relationship to an embolic etiology due to coagulopathies.4,18

ar embolization (e.g., treatment of cancer vessels), and paraneoplastic causes.11,18,28 In previous studies, of the cancer patients with ischemic stroke, 40% did not have any etiology of general stroke patients, and these patients are thought to have closer relationship to an embolic etiology due to coagulopathies.4,18 This study found that etiology according to stroke subtypes in cancer patients with ischemic stroke consisted of (in descending order of incidence): large-artery atherosclerosis, small artery occlusion, cardioembolism, and undetermined etiology. The proportion of large-artery atherosclerosis and stroke of undetermined cause was more frequent in cancer patients than in noncancer patients with ischemic stroke. These results are different from those of past reports. Previous studies have reported that cancer patients with ischemic stroke have a higher incidence of thrombotic strokes (54%), including cardioembolic strokes (15%), than nonthrombotic strokes (46%), including those due to undetermined causes (19%), artherosclerotic causes (10%), small-artery occlusion (12%), and other determined causes (5%), according to the TOAST classification.10 Another report found that cancer patients with ischemic stroke have 39.8% difference in etiology compared with general stroke patients with artherosclerosis, cardioembolic, or lacunar stroke subtypes.4 It is also reported that 25.5% had a statistically significant higher incidence of not having a general stroke etiology in cancer patients with ischemic stroke compared to that of noncancer patients with ischemic stroke.13 However, stroke subtypes were similar in both cancer and noncancer patients with ischemic stroke.18 Stroke types according to the TOAST classification were not statistically significant between cancer and noncancer patients with ischemic stroke.14

ic stroke compared to that of noncancer patients with ischemic stroke.13 However, stroke subtypes were similar in both cancer and noncancer patients with ischemic stroke.18 Stroke types according to the TOAST classification were not statistically significant between cancer and noncancer patients with ischemic stroke.14 The frequency of large-artery atherosclerosis as the etiology for ischemic stroke was significantly higher in cancer patients than in noncancer patients. We hypothesize that increased levels inflammatory biomarkers such as ESR, hs-CRP, and fibrinogen may be the reason for the increased frequency of large-artery atherosclerosis in ischemic stroke in cancer patients. Inflammatory processes play an important role in atherosclerosis development.29,30 The increased inflammation seen in cancer patients may be due to the increased frequency of large-artery atherosclerosis. There are a few limitations to this study. First, the study group comprised patients who were admitted to a single hospital, which can lead to a selection bias. Second, the duration of cancer and that between cancer diagnosis and the onset of ischemic stroke were not considered. Third, cancer type, metastasis, and the effect of cancer treatment were not considered in this study. Further study may be needed to rectify these limitations.

ital, which can lead to a selection bias. Second, the duration of cancer and that between cancer diagnosis and the onset of ischemic stroke were not considered. Third, cancer type, metastasis, and the effect of cancer treatment were not considered in this study. Further study may be needed to rectify these limitations. Conclusions Cancer patients with ischemic stroke demonstrated different risk factors, stroke biomarkers, and stroke etiology compared with noncancer patients with ischemic stroke. Therefore, it is recommended that in order to reduce the incidence of ischemic stroke in cancer patients, we should not only manage conventional risk factors, but also establish a distinct management of reducing inflammation and thrombus formation. The authors have no financial conflicts of interest. Figure 1 Stroke subtype in cancer versus noncancer patients with ischemic stroke. CE, cardioembolism; LAA, large artery atherosclerosis; Other, other determined etiology; SAD, small artery disease; Undermined, undetermined etiology. Table 1 Clinical Characteristics of Cancer Patients with Ischemic Stroke (N=156) Table 2 Stroke Risk Factors in Cancer vs Noncancer Patients with Ischemic Stroke *Statistics were analyzed by Chi-square test. Table 3 Stroke Biomarkers in Cancer vs Noncancer Patients with Ischemic Stroke Values are presented as mean±SD. *Statistics were analyzed using Student's t-test. ESR, erythrocyte sedimentation rate; hs-CRP, high-sensitivity C-reactive protein; pro-BNP, pro-brain natriuretic peptide; SD, standard deviation.

Introduction The human genome The human genome consists of 23 chromosome pairs i.e. in total 46 chromosomes. The DNA of these chromosomes contain about 6 billion base pairs.1 The genome can be divided into the exon part, called the exome. This is the 1.5% of DNA that contains approximately 25,000 genes coding for about 100,000 proteins in humans.2 The intron part is not coding for proteins but has regulatory properties. In addition, there is also mitochondrial DNA containing approximately 16,000 base pairs with 13 protein coding genes.3 Each human cell contains hundreds or thousands of sets of this mitochondrial DNA.3

oding for about 100,000 proteins in humans.2 The intron part is not coding for proteins but has regulatory properties. In addition, there is also mitochondrial DNA containing approximately 16,000 base pairs with 13 protein coding genes.3 Each human cell contains hundreds or thousands of sets of this mitochondrial DNA.3 Genetic variation Genetic variation in humans may have several causes. The DNA in the genome itself can be altered. Epigenetic factors can influence the gene expression through variations in DNA methylation and variations of histones - proteins controlling the DNA string formation. DNA changes range from large alterations of a part of or of a whole chromosome - cytogenetic changes that are possible to detect on microscopic chromosomal examination and sometimes not compatible with life. Alterations of intermediate size are on a submicroscopic level. The smallest molecular changes are confined to variation of one single base pair - a Single Nucleotide Polymorphism (SNP) where a nucleotide has been exchanged for another (Figure 1). Even though there are about 6 billion base pairs, only about 78 million - i.e. a small proportion of these base pairs have variations described as SNPs.4 Today's clinical molecular research is often heavily focused on SNP analyses and there is a risk that other alterations therefore remain unnoticed. It should therefore be kept in mind that other molecular DNA changes such as copy number variants (CNVs), repeats, insertions, deletions and microsatellites may also account for clinical genetic variation. Another question is how several variations may interact with each other and there is a need for advanced mathematical analytical methods utilizing powerful computer techniques to investigate such associations.

ts (CNVs), repeats, insertions, deletions and microsatellites may also account for clinical genetic variation. Another question is how several variations may interact with each other and there is a need for advanced mathematical analytical methods utilizing powerful computer techniques to investigate such associations. It should also be considered how common a genetic variation is and how common the disease studied is. For a common stroke phenotype the genetic variant can be common, rare, or even private i.e. confined to one individual or family. The impact of the genetic variant may differ considerably: with a high penetrance and importance it may cause a monogenic syndrome, with less influence it may still contribute to commonly occurring disorders with a more complex heritability. Molecular genetic studies of stroke risk One to two decades ago, molecular studies were often conducted as linkage studies where markers such as microsatellites were used to identify areas related to risk. As an example, one study using microsatellite markers and SNP analyses reported a possible association between stroke and variations in the PDE4D gene.5

sk One to two decades ago, molecular studies were often conducted as linkage studies where markers such as microsatellites were used to identify areas related to risk. As an example, one study using microsatellite markers and SNP analyses reported a possible association between stroke and variations in the PDE4D gene.5 An era of candidate gene studies followed the linkage studies. The idea is that by a logical and educated guess a candidate gene can be suggested as possibly related to variation in stroke risk. Numerous studies on this have been published, but many studies have been unsuccessful or not been possible to repeat. Genes of interest have included e.g. PDE4D6 and genes associated with cardiovascular disease.7,8 Many of the candidate gene studies have been on specific SNPs in the area of interest in the genome.

isk. Numerous studies on this have been published, but many studies have been unsuccessful or not been possible to repeat. Genes of interest have included e.g. PDE4D6 and genes associated with cardiovascular disease.7,8 Many of the candidate gene studies have been on specific SNPs in the area of interest in the genome. During the last decade many genome wide association studies (GWAS) have been performed. In a GWAS, an agnostic approach is used. A large number of SNPs, often in the range of 500,000 to 5,000,000 are examined throughout the whole chromosomal genome. Because so many SNPs are examined at the same time an adjustment for multiple testing has to be done. Therefore a p value threshold of 5×10-8 - corresponding to a Bonferroni correction for 1,000,000 tests - is often set as a significance level for GWAS investigations. It could be questioned whether this level should always be the same or if it should be adjusted because of e.g. the actual number of SNPs being analyzed. GWAS examinations have now yielded several very interesting results for ischemic stroke risk (Table 1). More studies are ongoing.9 Several overviews on stroke genetics have recently been published.10,11 Molecular genetic variations affecting risk of monogenic stroke syndromes Today, there are several monogenic stroke syndromes that have been related to molecular genetic variation. Examples of monogenetic stroke syndromes are given in Table 2. Comprehensive accounts of have been published.11,22,23 Information on the Internet is available at e.g. Genetics Home Reference at http://ghr.nlm.nih.gov/.

, there are several monogenic stroke syndromes that have been related to molecular genetic variation. Examples of monogenetic stroke syndromes are given in Table 2. Comprehensive accounts of have been published.11,22,23 Information on the Internet is available at e.g. Genetics Home Reference at http://ghr.nlm.nih.gov/. Molecular genetic variations affecting risk of common stroke syndromes, sometimes with specific effects on specific main types of stroke or subtypes of ischemic and hemorrhagic stroke Ischemic stroke The three main ischemic stroke syndromes: large vessel disease, cardioembolic stroke and small vessel disease have been studied separately regarding genetic risk. Several molecular genetic variations have been reported to be related to large artery disease and to cardioembolic stroke (Table 1). Some findings have been reported in several studies e.g. HDAC9 related to large vessel disease13,15 and PITX2 and ZFHX3 related to cardioembolic stroke.13,14 The effect sizes of the identified SNPs have been modest; e.g. for a HDAC9 variant and stroke related to large vessel disease: odds ratio (OR)=1.42 (95% confidence interval [CI]=1.28-1.57); and a PITX2 variant and cardioembolic stroke OR1.32 (95% CI=1.20-1.46).15 One observation is that the GWAS detected variations related to common ischemic stroke caused by small vessel disease have been fewer or absent in GWAS studies of ischemic stroke. The explanation for this is unknown but it has been suggested that small vessel disease may represent several different phenotypes and thus be a more heterogeneous condition24 and also be subject to different definitions,25 whereas large artery disease and cardioembolic stroke may be less heterogeneous. The relation of genetic variants to overall ischemic stroke has also been reported but with less consistency. There is a trade-off between very specific phenotyping e.g. defined types of ischemic stroke and the number of subjects that can be included in the genetic studies. It seems clear today that ischemic stroke is not caused by one common pathogenetic factor, however there seems to be considerable overlap between the different subtypes of ischemic stroke regarding risk factors such as e.g. hypertension. It has been shown that a risk score taking several genetic variations related to ischemic stroke risk into account is associated with ischemic stroke overall.26

etic factor, however there seems to be considerable overlap between the different subtypes of ischemic stroke regarding risk factors such as e.g. hypertension. It has been shown that a risk score taking several genetic variations related to ischemic stroke risk into account is associated with ischemic stroke overall.26 Intracerebral hemorrhage (ICH) Also hemorrhagic stroke risk has been related to genetic variations (Table 3). The risk of lobar intracerebral hemorrhage has been firmly related to variations in the APOE gene, especially the ε2 or ε4 alleles. There are indications that although variations in the COL4A1 region have been related to monogenic related ICH, other variations in the same region may be related to a somewhat increased risk of sporadic ICH.27 A detailed description of genetic risk of sporadic ICH has recently been published.11

y the ε2 or ε4 alleles. There are indications that although variations in the COL4A1 region have been related to monogenic related ICH, other variations in the same region may be related to a somewhat increased risk of sporadic ICH.27 A detailed description of genetic risk of sporadic ICH has recently been published.11 Genetics of conditions associated with stroke risk - intermediate phenotypes - e.g. white matter hyperintensities, atrial fibrillation, and hypertension Several intermediate phenotypes are related to stroke risk and it is therefore of interest if genetic risk for these phenotypes is also associated with increased stroke risk either independently or through the intermediate phenotype. As mentioned above, a genetic risk score considering genetic variations linked to intermediate phenotypes related to stroke has been associated with overall risk of ischemic stroke.26 The same reference contains a comprehensive supplemental table listing genes related to intermediate phenotypes indicating stroke risk. Another study reported that a risk score including genetic variations related to stroke and its risk factors could improve the prediction of future stroke compared with using a risk score based only on clinical information.31 Some intermediate phenotypes are discussed in more detail below.

types indicating stroke risk. Another study reported that a risk score including genetic variations related to stroke and its risk factors could improve the prediction of future stroke compared with using a risk score based only on clinical information.31 Some intermediate phenotypes are discussed in more detail below. White matter hyperintensities (WMH) The presence of white matter lesions has been related to stroke risk.32,33 Therefore it is possible that genetic changes resulting in WMH may also result in increased risk of stroke. The heritability of cerebral white matter hyperintensities is high.34 An association between the 17q25 locus and white matter hyperintensity volume has been reported.35,36 This association has also been reported in subjects with stroke although there was not a clear relation to the presence of small vessel disease manifested as lacunar infarction.37 It is likely that several different genetic variations contribute to the risk of WMH and a recent study of patients with CADASIL reported a polygenic risk score for WMH volume, illustrating that examination of a subgroup of individuals with high likelihood of a condition can be used to detect additional traits contributing to risk of the condition.38 ApoE ε4 carriers have been reported to have higher subcortical white matter lesion volume.39 However, the effect of the APOE ε4 allele on white matter integrity is uncertain.40 Variations in the NOTCH3 gene may also influence the risk of WMH or small vessel disease in individuals without the typical CADASIL syndrome.41

ApoE ε4 carriers have been reported to have higher subcortical white matter lesion volume.39 However, the effect of the APOE ε4 allele on white matter integrity is uncertain.40 Variations in the NOTCH3 gene may also influence the risk of WMH or small vessel disease in individuals without the typical CADASIL syndrome.41 Atrial fibrillation (AF) Several genes have been related to AF.42,43,44 The mechanisms through which these genes contribute to AF risk are largely unknown although a recent study showed a relation between some of these genes and prolonged atrial action potential duration in an animal model.43 It seems as if adding a genetic risk score of gene variations related to AF in patients with AF improves the risk assessment for stroke in addition to the often used CHADS2 score in these patients.45 Hypertension Hypertension is one of the most important risk factors for stroke, both ischemic and hemorrhagic. Blood pressure has heritability estimates of 30%-50%.46 Several genes have been reported to be related to hypertension in GWAS studies of patients without or with stroke. A risk score of 29 such SNPs was related to stroke, not further subtyped, as well as to hypertension and to coronary heart disease.47 It is of interest that an age related effect of genes related to a certain phenotype e.g. blood pressure has been reported48 and it is possible that such age related effects are of importance also for other phenotypes including stroke.

e, not further subtyped, as well as to hypertension and to coronary heart disease.47 It is of interest that an age related effect of genes related to a certain phenotype e.g. blood pressure has been reported48 and it is possible that such age related effects are of importance also for other phenotypes including stroke. In another study, a risk score consisting of 39 SNP associated with blood pressure levels was related to ICH, especially deep ICH rather than lobar ICH, supporting the concept that elevated blood pressure may be more prone to cause deep ICH.49 It was discussed if gene variations influencing blood pressure may be of importance for stroke risk also in individuals with seemingly "normal" blood pressure.49 Ischemic heart disease The situation for ischemic heart disease (IHD) is similar to that for hypertension. Many traditional risk factors are shared between stroke and IHD. Several genetic variations have been associated with ischemic heart disease.50 A variation in the chromosome 9p21 region has been related to ischemic stroke.17 A subsequent follow-up study was not able to find additional variants from the Cardiogram study to be associated with stroke overall or main ischemic stroke subtypes although the number of patients in the individual subtypes were small.7 A much larger GWAS study could detect that several genetic variations were shared between ischemic stroke - especially the large artery disease subtype, and coronary heart disease.51

e associated with stroke overall or main ischemic stroke subtypes although the number of patients in the individual subtypes were small.7 A much larger GWAS study could detect that several genetic variations were shared between ischemic stroke - especially the large artery disease subtype, and coronary heart disease.51 It can be expected that in the near future new reports on stroke risk scores including several intermediate phenotypes for stroke risk, also other than discussed above, will be published using more elaborate statistical methods, larger number of genetic variants as well as larger number of individuals. Another method to consider for future studies is that if patients with a very specific phenotype can be detected this may decrease the number of subjects needed to detect genetic influence on cerebrovascular risk as illustrated in the study on CADASIL and WMH volume mentioned above.38

l as larger number of individuals. Another method to consider for future studies is that if patients with a very specific phenotype can be detected this may decrease the number of subjects needed to detect genetic influence on cerebrovascular risk as illustrated in the study on CADASIL and WMH volume mentioned above.38 Hereditary causes of familial aggregation of stroke Apart from molecular analyses, family studies and twin studies are important tools to study heritability of stroke. Such studies clearly indicate heritability of stroke. One study reported a prevalence of stroke or TIA of 12.3% among first degree relatives of stroke patients compared with 7.5% among first degree relatives of control subjects.52 Another study showed that occurrence of stroke in a parent by 65 years of age was associated with a 3-fold increase in risk of stroke in their offspring.53 Twin studies suggest that a genetic component of stroke risk is present.54,55 However, subtyping of ischemic stroke and other types of stroke in twin studies would be useful to increase the possibility to better understand heritability of ischemic stroke.56 Also GWAS studies have through statistical analyses shown evidence that there is a hitherto unexplained heritability component in the risk of ischemic stroke of about 38%, and that this may vary between different subtypes of ischemic stroke.8 Future family studies of stroke should preferably include stroke subtypes and also focus on not only first-degree relatives but also somewhat more distant relatives of the probands.

heritability component in the risk of ischemic stroke of about 38%, and that this may vary between different subtypes of ischemic stroke.8 Future family studies of stroke should preferably include stroke subtypes and also focus on not only first-degree relatives but also somewhat more distant relatives of the probands. Epigenetic impact on expression of different proteins before, during and after acute brain injury The genetic expression can be influenced by other causes than changes of the DNA content.57 Such influences may be referred to as epigenetic mechanisms (Figure 2). One such mechanism is that gene transcription can be regulated by e.g.: - Methylation of DNA58 - Histone modifications58/Histone deacetylases (HDACs) - Micro-RNA59 - Other It is possible to influence the above mechanisms with pharmacological agents. E.g. valproate is a HDAC inhibitor and has been suggested to perhaps inhibit atherosclerosis. However, much more studies are needed to examine if and how it is possible to treat patients by modifying these mechanisms.

- Histone modifications58/Histone deacetylases (HDACs) - Micro-RNA59 - Other It is possible to influence the above mechanisms with pharmacological agents. E.g. valproate is a HDAC inhibitor and has been suggested to perhaps inhibit atherosclerosis. However, much more studies are needed to examine if and how it is possible to treat patients by modifying these mechanisms. The role of non-protein coding RNA (ncRNA) is intriguing. The DNA coding for ncRNA is in the intronic portion of the genome, which is the vast majority of the DNA. Several classes of non-protein coding RNA exist, among these micro-RNA. Micro-RNA can regulate gene expression and is considered to be an epigenetic regulator.59 Micro-RNA has been suggested to regulate several mechanisms in brain ischemia and therefore be of importance for recovery after stroke.59 These regulatory mechanisms may also be involved in the situation of cerebral ischemia with influence on cell death as well as on regeneration after stroke.60 The gene expression for coding of different proteins during pathological conditions is of importance for the response of the individual subject. Indeed microarray analyses indicate that there is a very dynamic response varying both with time after stroke as well as between the core and the periinfarct areas of the ischemic area of the brain.61 An additional therapeutic epigenetic method suggested is to regulate endogenous or exogenous stem cells to respond to cerebral injury in stroke.62

indicate that there is a very dynamic response varying both with time after stroke as well as between the core and the periinfarct areas of the ischemic area of the brain.61 An additional therapeutic epigenetic method suggested is to regulate endogenous or exogenous stem cells to respond to cerebral injury in stroke.62 Genetic influence on functional outcome and recovery after stroke Recovery after stroke begins immediately after the stroke onset. Many different biological responses are involved after ischemic stroke and these vary in time and between different areas of the affected brain.63 Several of these responses may be of interest from a genetic point of view. These include epigenetic mechanisms - discussed above - that can be targeted for treatment in the varying temporal phases after ischemic stroke onset.60 It is also of interest whether genetic variation may influence the possibility and degree of functional outcome after stroke. Different drug therapies have been tried as treatments after stroke but the responses to these therapies vary between patients. E.g. genetic polymorphisms affect the response to L-dopa treatment.64 The brain derived neurotrophic factor (BDNF) is involved in brain repair and plasticity and a variation of a SNP (Val66Met) in the BDNF gene has been shown to be related to improved recovery although the early response was in the opposite direction.65 The APOE ε4 has been related to poorer outcome.66 Other examples of genes related to functional outcome include IGF1,67 COX-2 and GPIIIa.68

r and plasticity and a variation of a SNP (Val66Met) in the BDNF gene has been shown to be related to improved recovery although the early response was in the opposite direction.65 The APOE ε4 has been related to poorer outcome.66 Other examples of genes related to functional outcome include IGF1,67 COX-2 and GPIIIa.68 After lobar ICH, APOE ε2 has been related to poorer outcome.69 But also other genes are of interest for outcome after ICH: an heritability estimate of 90-day ICH mortality for non-APOE loci using genomewide complex trait analysis has been calculated to about 41%.70 Apart from the study by Devan et al.,70 all the above mentioned studies on functional outcome and recovery have been candidate gene studies. No large GWAS examining the genetic effect on functional outcome after stroke has been published yet. However, such a study is now ongoing - the Genetics of Ischemic Stroke Functional Outcome Study (GISCOME) and results are expected within the coming year.71 Pharmacogenetics Pharmacogenetics is a research area holding large promise to be of importance both for stroke and other diseases in the years to come. As has been discussed above, both epigenetics and SNP variations may be used for therapeutic considerations in stroke. Two additional examples are discussed below: Thrombolytic therapy with tissue-type plasminogen activator (tPA) and anticoagulation therapy with warfarin or dabigatran.

er diseases in the years to come. As has been discussed above, both epigenetics and SNP variations may be used for therapeutic considerations in stroke. Two additional examples are discussed below: Thrombolytic therapy with tissue-type plasminogen activator (tPA) and anticoagulation therapy with warfarin or dabigatran. Thrombolytic therapy A study examined 140 candidate SNPs in 497 tPA-treated ischemic stroke patients and showed that IL1B and vWF variants were associated with early recanalization.72 The vWF variant was also related to FVIII activity in a subsequent functional study.72 The same group has published results showing that a genetic variation rs669 (Val1000Ile) in the alpha-2-macroglobulin gene is related to hemorrhagic transformation after tPA treatment.73 This indicates that genetic information may possibly be used in the future to predict the response of tPA treatment in ischemic stroke. Such a prediction may help in decision-making regarding iv tPA or alternative treatments such as endovascular treatment.

is related to hemorrhagic transformation after tPA treatment.73 This indicates that genetic information may possibly be used in the future to predict the response of tPA treatment in ischemic stroke. Such a prediction may help in decision-making regarding iv tPA or alternative treatments such as endovascular treatment. Anticoagulation therapy with warfarin or dabigatran The treatment with anticoagulants to prevent cardioembolic stroke is highly efficient in a population of individuals with atrial fibrillation and increased risk. However the metabolism of the anticoagulant administered may vary for several reasons where a genetic variation may be one of these. Cytochrome P-450 enzyme CYP2C9 gene variants as well as variants in VKORC1, coding for vitamin K epoxide reductase (VKOR) are related to warfarin metabolism but the usefulness of genetic testing regarding these variants for guidance on initiation of warfarin treatment has been debated.74 The future may lead to other conclusions - in a very recent study association with APOE ε2 and APOE ε4 for lobar warfarin related ICH was reported.75

OR) are related to warfarin metabolism but the usefulness of genetic testing regarding these variants for guidance on initiation of warfarin treatment has been debated.74 The future may lead to other conclusions - in a very recent study association with APOE ε2 and APOE ε4 for lobar warfarin related ICH was reported.75 In the Randomized Evaluation of Long-term Anticoagulation Therapy (RE-LY) study a GWAS was performed in 2944 RE-LY patients and showed that the CES1 rs2244613 minor allele was associated with lower active dabigatran metabolite levels.76 This minor allele was associated with a lower risk of any bleeding in the dabigatran treated patients but there was no reported association with ischemic events. The value of genetic testing and clopidogrel treatment is also of interest. Clopidogrel is metabolized to its active metabolite by CYP2C19 but the clinical utility for genetic testing to detect influence on CYP3C19 activity is still under debate.77

ed patients but there was no reported association with ischemic events. The value of genetic testing and clopidogrel treatment is also of interest. Clopidogrel is metabolized to its active metabolite by CYP2C19 but the clinical utility for genetic testing to detect influence on CYP3C19 activity is still under debate.77 Conclusion Stroke genetics today is involved in many fields, including risk, outcome and pharmacogenetics. The research on stroke genetics is progressing with high pace and is expected to continue to do so during the next decade. Stroke subtyping is very important for all these areas. New methods are emerging including detailed exome content analysis, exome sequencing, whole genome sequencing and advanced statistical analytical methods. Large numbers of subjects are often needed in genetic studies and therefore co-operation in international consortia such as the International Stroke Genetics Consortium (ISGC) - www.strokegenetics.org is necessary. The author has no financial conflicts of interest. Figure 1 Examples of DNA variations in the human genome. Figure 2 Examples of epigenetic mechanisms. Image courtesy of National Institutes of Health (public domain). http://commonfund.nih.gov/sites/default/files/epigeneticmechanisms.pdf. Table 1 Examples of published results from genome-wide studies, showing SNPs related to ischemic stroke risk LVD, large vessel disease; CE, Cardioembolic Embolism; IS, Ischemic stroke. Table 2 Examples of monogenetic stroke syndromes Table 3 Examples of SNPs related to ICH risk SNP, single nucleotide polymorphism; ICH, intracerebral hemorrhage.

Introduction Approximately, 20% of ischemic events occur in the territory of the posterior (vertebrobasilar) circulation, and dizziness/vertigo is one of the most common symptoms of vertebrobasilar diseases.1 With the aids of recent development in neuroimaging, however, inferior cerebellar and small brainstem infarctions are increasingly recognized as a cause of isolated vertigo. Furthermore, transient isolated vertigo is the common manifestation of vertebrobasilar insufficiency.2 It is important to differentiate isolated vertigo of a vascular cause from more benign disorders involving the inner ear since the therapeutic strategy and prognosis differ in these two conditions.3 Early recognition of isolated vertigo of a vascular cause may allow specific management. Misdiagnosis of acute stroke may result in significant morbidity and mortality while overdiagnosis of vascular vertigo would lead to unnecessary costly work-ups and medication.3 This review aims to highlight advances in isolated vertigo of a vascular cause and to address its clinical significance. Recurrent episodes of isolated vertigo of a vascular cause In cerebrovascular disorders, the dizziness/vertigo usually accompanies other neurological symptoms and signs. Indeed, medical adage had taught us that isolated vertigo mostly comes from peripheral vestibular diseases. However, isolated vascular vertigo might have been underestimated. Diagnosis of isolated vertigo from brainstem and cerebellar strokes has been increasing markedly with recent developments in clinical neurotology and neuroimaging.3

ge had taught us that isolated vertigo mostly comes from peripheral vestibular diseases. However, isolated vascular vertigo might have been underestimated. Diagnosis of isolated vertigo from brainstem and cerebellar strokes has been increasing markedly with recent developments in clinical neurotology and neuroimaging.3 Transient isolated vascular vertigo typically occurs abruptly, and usually lasts several minutes. In patients with vertigo due to vertebrobasilar insufficiency, 62% had a history of at least one isolated episode of vertigo, and 19% developed vertigo as the initial symptom.2 Patients with infarction in the territory of anterior inferior cerebellar artery (AICA) may have isolated recurrent vertigo, fluctuating hearing loss, and/or tinnitus (similar to Meniere's disease) as the initial symptoms 1-10 days prior to the permanent infarction.4

, and 19% developed vertigo as the initial symptom.2 Patients with infarction in the territory of anterior inferior cerebellar artery (AICA) may have isolated recurrent vertigo, fluctuating hearing loss, and/or tinnitus (similar to Meniere's disease) as the initial symptoms 1-10 days prior to the permanent infarction.4 A recent study found that the patients who visited the emergency department with dizziness/vertigo had 2-fold (95% CI, 1.35-2.96; P<0.001) higher risk of stroke or cardiovascular events than those without dizziness/vertigo during a follow-up of 3 years.5 The authors also demonstrated that the patients hospitalized with isolated vertigo have a 3.01-times (95% CI, 2.20-4.11; P<0.001) higher risk for stroke than the general population during the 4-year follow-up.5 Particularly, the vertigo patients with 3 or more risk factors have a 5.51-fold higher risk for stroke (95% CI, 3.10-9.79; P<0.001) than those without risk factors.5 Another study adopted the ABCD2 score,6 a clinical prediction tool to assess the risk of stroke after a transient ischemic attack, to predict cerebrovascular events in emergency department patients with dizziness.7 The authors found that only 1.0% of dizzy patients with a score of 3 or less had a cerebrovascular event compared to 8.1% of the patients with a score of 4 or more.7 Especially, 27% of the patients with a score of 6 or 7 suffered from cerebrovascular episodes.7 Thus, the ABCD2 score may predict cerebrovascular attacks in patients with transient vertigo. All of these data suggest that isolated episodic vertigo with or without auditory symptoms may be the only manifestation of transient ischemia within the vertebrobasilar circulation.

7 suffered from cerebrovascular episodes.7 Thus, the ABCD2 score may predict cerebrovascular attacks in patients with transient vertigo. All of these data suggest that isolated episodic vertigo with or without auditory symptoms may be the only manifestation of transient ischemia within the vertebrobasilar circulation. Labyrinthine infarction Because the blood supply to the inner ear originates from the vertebrobasilar system, vertebrobasilar ischemic stroke can present with vertigo and hearing loss due to infarction of the inner ear (i.e., labyrinthine infarction). The internal auditory artery (IAA) is a branch of AICA. The IAA irrigates the cochlea and vestibular labyrinth, and occlusion of the IAA causes loss of auditory and vestibular function. Since the IAA is an end artery with minimal collaterals from the otic capsule, the labyrinth is especially vulnerable to ischemia.2,8 IAA infarction mostly occurs due to thrombotic narrowing of the AICA itself, or in the basilar artery at the orifice of the AICA.9 Because the inner ear is hardly visualized on the routine MRIs, a definite diagnosis of labyrinthine infarction is not possible unless a pathological study is done.10 The apical region of the cochlea is particularly vulnerable to vascular injury, and, thus, low-frequency hearing loss is common with ischemia of the inner ear.11 However, a labyrinthine infarction is usually associated with infarction of the brainstem and/or cerebellum in the territory of the AICA.12

l study is done.10 The apical region of the cochlea is particularly vulnerable to vascular injury, and, thus, low-frequency hearing loss is common with ischemia of the inner ear.11 However, a labyrinthine infarction is usually associated with infarction of the brainstem and/or cerebellum in the territory of the AICA.12 The labyrinthine infarction should be considered in older patients with acute onset of unilateral hearing loss and vertigo, particularly when there is a history of stroke or known vascular risk factors. Because current means of diagnosing labyrinthine infarction are not adequate (including MRIs), clinicians should consider all the clinical evidences when attempting to determine the etiology of acute audio-vestibular syndromes rather than just emphasizing that MRI is the best way to distinguish viral from vascular etiology.13

current means of diagnosing labyrinthine infarction are not adequate (including MRIs), clinicians should consider all the clinical evidences when attempting to determine the etiology of acute audio-vestibular syndromes rather than just emphasizing that MRI is the best way to distinguish viral from vascular etiology.13 Anatomical structures responsible for central isolated vertigo of a vascular cause Vertigo results from imbalance in the tonic discharges of the vestibular system arising from the inner ears on both sides. The origin of vertigo may be peripheral or central. When the vertigo occurs as a symptom of vertebrobasilar ischemic strokes, it is usually associated with other neurological symptoms or signs. Theoretically, a small infarct localized to structures such as the nodulus, root entry zone of the eighth nerve in the pontomedullary junction, and vestibular nucleus can cause vertigo without other accompanying neurological symptoms or signs since all of these structures receive afferent vestibular inputs from the inner ear (Figure 1). Because the vestibular nucleus is more vulnerable to ischemia than other structures in the brainstem and cerebellum from a recent animal study,14 ischemia of the lateral medulla including the vestibular nucleus may be a common mechanism of isolated vascular vertigo. Rarely, lesions involving the flocculus or dorsal insular cortex can also cause isolated vertigo (Figure 1).15,16,17 Vertigo due to a lesion involving the dorsal insular cortex is usually not associated with nystagmus and a flocculus lesion is commonly associated with other central signs such as gaze-evoked nystagmus (GEN) and asymmetrical oculomotor dysfunction.15,16,17

ular cortex can also cause isolated vertigo (Figure 1).15,16,17 Vertigo due to a lesion involving the dorsal insular cortex is usually not associated with nystagmus and a flocculus lesion is commonly associated with other central signs such as gaze-evoked nystagmus (GEN) and asymmetrical oculomotor dysfunction.15,16,17 Bedside diagnosis of stroke in acute isolated vertigo Because central vestibular signs such as vertical nystagmus, direction changing GEN, perverted head shaking nystagmus (HSN), asymmetrical oculomotor dysfunction, or severe postural instability with falling do not appear in all patients with central vertigo, it is not always easy to differentiate isolated vascular vertigo from acute peripheral vestibulopathy at the bedside. Most of these signs are known to have a high specificity, but low sensitivity for detecting a central cause of acute isolated vertigo.

ty with falling do not appear in all patients with central vertigo, it is not always easy to differentiate isolated vascular vertigo from acute peripheral vestibulopathy at the bedside. Most of these signs are known to have a high specificity, but low sensitivity for detecting a central cause of acute isolated vertigo. It is well believed that the bedside head impulse test (HIT) is a useful tool for differentiating acute vascular vertigo from a more benign disorder involving the inner ear. Normal HIT is regarded a reliable sign for an intact peripheral vestibular function, thus suggesting a central lesion in acute spontaneous vertigo. The significance of HIT for differentiating stroke from acute peripheral vestibulopathy has been confirmed in a recent study18 which showed that a negative HIT result (i.e., normal vestibulo-ocular reflex) is strongly suggestive of a central lesion with a pseudo-VN presentation. The study emphasized that a 3-step bedside oculomotor examination for HINTS (normal HIT, direction-changing GEN, and skew deviation) is more sensitive for stroke than early MRI whilst maintaining a high specificity.18 Indeed, initial diffusion-weighted MRIs may be false negative in 12%-20% of the stroke patients during the first 48 hours.18,19 Another recent report also confirmed diagnostic utility of the signs including normal horizontal HIT, skew deviation, abnormal vertical smooth pursuit, and central type nystagmus at the bedside: they found a 100% sensitivity and 90% specificity for stroke if one of those signs was present in acute isolated vertigo.20 Since mild degree of skew deviation usually goes unnoticed during the bedside examination and GEN is also sometimes absent in cerebellar stroke, bedside HIT may be the best tool for differentiating isolated vertigo due to cerebellar stroke (particularly within the territory of the posterior inferior cerebellar artery, PICA) from acute peripheral vestibulopathy. However, bedside HIT has some limitations, and may be positive in patients with AICA territory cerebellar infarct involving the flocculus, or brainstem stroke involving the vestibular nucleus or root entry zone of the vestibular nerve.18,21,22,23

r inferior cerebellar artery, PICA) from acute peripheral vestibulopathy. However, bedside HIT has some limitations, and may be positive in patients with AICA territory cerebellar infarct involving the flocculus, or brainstem stroke involving the vestibular nucleus or root entry zone of the vestibular nerve.18,21,22,23 Overall, the most consistent bedside predictor of central isolated vertigo of a vascular cause appears to be the HIT and normal HIT usually guarantees an absence of peripheral pathology. A three-component bedside oculomotor examination (i.e., HINTS) identifies stroke with high sensitivity and specificity in patients with acute isolated vertigo and diagnoses stroke more effectively than early diffusion-weighted MRIs. Differential diagnostic points for central and peripheral acute vertigo syndromes are summarized in Table 1. Isolated vertigo in cerebellar stroke Dizziness/vertigo is one of the most common symptoms of cerebellar stroke syndrome. Cerebellar ischemic stroke probably ranks first among central vascular vertigo syndromes. A large prospective study showed that about 11% (25/240) of the patients with isolated cerebellar infarctions had isolated vertigo and most (24/25, 96%) of them had an infarct in the territory of the medial branch of the PICA including the nodulus.24

stroke probably ranks first among central vascular vertigo syndromes. A large prospective study showed that about 11% (25/240) of the patients with isolated cerebellar infarctions had isolated vertigo and most (24/25, 96%) of them had an infarct in the territory of the medial branch of the PICA including the nodulus.24 In PICA territory cerebellar infarction, the direction of nystagmus and degree of postural instability are variable. The prominent cerebellar signs, particularly severe axial instability and direction changing GEN (occurring in 71% and 54%, respectively, in the aforementioned series),24 can help in the differential, but these findings are less sensitive. Similarly, perverted head shaking (mostly down beating) and positional down beating nystagmus as important signs of central vestibular dysfunctions are found in only half of the cases with cerebellar infarction.25 Overall, although the severity of imbalance and the appearance of nystagmus in PICA territory cerebellar infarctions can help in differentiating acute vertigo syndrome from vertigo originating from the inner ear, these findings are less sensitive for differentiating two conditions. PICA territory cerebellar infarction should be considered in the differential diagnosis of central vascular vertigo syndrome, even if the nystagmus and imbalance are more typical of acute vertigo originating from the inner ear.

ng from the inner ear, these findings are less sensitive for differentiating two conditions. PICA territory cerebellar infarction should be considered in the differential diagnosis of central vascular vertigo syndrome, even if the nystagmus and imbalance are more typical of acute vertigo originating from the inner ear. GEN is a sensitive sign of central vestibular lesions involving the cerebellum or brainstem. A recent study showed that unidirectional GEN was found in 33% of the patients with acute unilateral cerebellar stroke and the nystagmus was directed either toward or away from the lesion side.26 The structures responsible for unidirectional GEN included the pyramid, uvula, tonsil, and parts of the biventer and inferior semilunar lobules.26 GEN may be a sign indicating damage to the midline and lower cerebellar structures. Of interest, however, none of the patients showed spontaneous nystagmus and none of the patients with GEN had a lesion in the flocculus, which is known to play a major role in the gaze holding mechanism.

nd inferior semilunar lobules.26 GEN may be a sign indicating damage to the midline and lower cerebellar structures. Of interest, however, none of the patients showed spontaneous nystagmus and none of the patients with GEN had a lesion in the flocculus, which is known to play a major role in the gaze holding mechanism. Another study25 found that 51% of patients (37/72) with isolated cerebellar infarction showed HSN and the horizontal component of HSN was constantly ipsilesional. Perverted HSN occurred in 23 (23/37, 62%) patients and was mostly downbeat (22/23, 96%). Lesion subtraction analyses revealed that damage to the uvula, nodulus and inferior tonsil was mostly responsible for generation of HSN in patients with unilateral PICA territory infarction. The authors argued that ipsilesional HSN may be caused by unilateral disruption of uvulonodular inhibition over the velocity storage. Otherwise, the perverted HSN may be ascribed to impaired control over the spatial orientation of the angular vestibulo-ocular reflex due to uvulonodular lesions or a build-up of vertical vestibular asymmetry favoring upward bias due to lesions involving the inferior tonsil.25 In AICA infarction, HSN is also common with both peripheral and central patterns. Careful evaluation of HSN may provide clues for AICA infarction in patients with acute audiovestibular loss.27 Another study28 on long-term outcome of vestibular loss of a vascular cause showed that the caloric responses were normalized in 70% (20/30) of the patients who had canal paresis due to posterior circulation ischemic strokes and had a follow-up for at least 1 year. Moreover, all patients who were followed for >5 years after the onset of vertigo showed normal caloric responses, suggesting that, like other neurological symptoms or signs due to stroke, canal paresis associated with posterior circulation ischemic strokes has a good long-term outcome although the detailed mechanism underlying the recovery of caloric-induced vestibular responses still unclear. Another study29 on the long-term outcome of acute hearing loss reported a similar finding that approximately 65% of the patients who were followed for at least 1 year after the onset of acute hearing loss showed a partial or complete hearing recovery.

the recovery of caloric-induced vestibular responses still unclear. Another study29 on the long-term outcome of acute hearing loss reported a similar finding that approximately 65% of the patients who were followed for at least 1 year after the onset of acute hearing loss showed a partial or complete hearing recovery. Multivariate analysis showed that multiple risk factors for stroke (odds ratio [OR], 10.46; 95% confidence interval [CI], 1.72 to 13.7; P=0.011) and profound hearing loss (OR, 3.92; 95% CI, 1.03 to 14.97; P<0.046) predicted a poor outcome for recovery of hearing loss. The a bove two reports suggest that recovery of audiovestibular loss of a vascular cause is more common than previously thought. To date, at least eight subgroups of AICA infarction have been identified according to the patterns of neurotological presentations, among which the most common pattern of audiovestibular dysfunction is the combined loss of auditory and vestibular functions.30,31 Because audiovestibular loss may occur in isolation before ponto-cerebellar infarction involving AICA distribution, audiovestibular loss may herald more widespread areas of infarction involving the posterior circulation (mainly in the AICA territory),30,31 especially when patients had basilar artery occlusive diseases presumably close to the origin of the AICA on brain MRA, even if other central signs are absent and MRI does not demonstrate acute infarction.30,31

more widespread areas of infarction involving the posterior circulation (mainly in the AICA territory),30,31 especially when patients had basilar artery occlusive diseases presumably close to the origin of the AICA on brain MRA, even if other central signs are absent and MRI does not demonstrate acute infarction.30,31 In a recent study of isolated superior cerebellar artery (SCA) territory cerebellar infarction, approximately half (19/41) of the patients experienced true vertigo and 11 (27%) showed spontaneous nystagmus mainly beating to the lesion side or GEN.32 The authors emphasized that the vertigo and nystagmus in the SCA territory cerebellar infarctions are more common than previously thought. Ipsilesional spontaneous nystgmus may result from damage to the anterior lobe of the cerebellum, which transmits the vestibular output to the fastigial nucleus.32

ion side or GEN.32 The authors emphasized that the vertigo and nystagmus in the SCA territory cerebellar infarctions are more common than previously thought. Ipsilesional spontaneous nystgmus may result from damage to the anterior lobe of the cerebellum, which transmits the vestibular output to the fastigial nucleus.32 Isolated vertigo in brainstem stroke Mono-symptomatic attacks of vertigo and nystagmus without any other brainstem symptoms and signs would be unusual in brainstem ischemia. Selective damage to the vestibular nuclei and root entry zone of the eighth nerve in the pontomedullary junction can cause isolated vertigo.33,34,35,36,37 Because the root entry zone of the eighth cranial nerve has a rich network of anastomotic vessels arising from the neighboring arteries,38 the possibility of focal infarction in that area is extremely low. Although some case reports showed central isolated vertigo due to a demyelinating lesion localized to the root entry zone of the eighth nerve,33,34 isolated vertigo due to focal infarction in the root entry zone of the vestibular nerve has not been reported in the literature. Focal ischemia involving the vestibular nuclei can cause isolated vertigo and nystagmus mimicking acute vestibular neuritis.35,36,37 Several studies have described patients with an isolated vestibular nucleus infarction who presented with isolated prolonged vertigo, spontaneous horizontal nystagmus, a positive HIT, and unilateral canal paresis.35,36,37 All of these findings are consistent with acute peripheral vestibulopathy. These reports emphasize that isolated vestibular nucleus infarction should be considered in the differential diagnosis of central vascular vertigo syndrome, even though when the patients have unilateral canal paresis and positive HIT on the side of the canal paresis, and other neurologic symptoms or signs are absent. Vertigo in the lateral medullary infarction is usually associated with other neurological symptoms or signs, but tiny infarct in the lateral medulla can present with vertigo without other localizing symptoms.39 In this case, the HIT might be positive, if the medial vestibular nucleus is involved. Vestibular-evoked myogenic potentials (VEMPs) have become important diagnostic tools to assess the central otolithic pathways in brainstem and cerebellar lesions. Recent reports on VEMPs in patients with brainstem infarcts showed a various pattern of abnormal VEMPs responses.40,41

medial vestibular nucleus is involved. Vestibular-evoked myogenic potentials (VEMPs) have become important diagnostic tools to assess the central otolithic pathways in brainstem and cerebellar lesions. Recent reports on VEMPs in patients with brainstem infarcts showed a various pattern of abnormal VEMPs responses.40,41 Isolated vertigo and/or hearing loss as a sign of impending AICA territory cerebellar infarction Recent papers have shown that 8%-30% of patients with posterior circulation ischemic strokes (mainly in the AICA territory) had acute audiovestibular loss with vertigo, fluctuating hearing loss and/or tinnitus before more widespread infarction and at this stage, patients may be misdiagnosed as having a peripheral pathology such as Meniere's disease.30,43,44 Selective ischemia to the inner ear can explain the isolated prodromal audiovestibular disturbance because the inner ear requires high-energy metabolism and has little collateral circulation.30,42,43 Although there are as yet no systematic data on which are the high-risk factors suggesting impending stroke or which interventions are beneficial at the stage of isolated audiovestibular loss, patients with a prodromal audiovestibular disturbance are more likely to have a focal or diffuse stenosis of the basilar artery presumably close to the origin of the AICA than the patients without a prodromal audiovestibular disturbance.4,12,30,43,44 This finding highlights that AICA infarction should be considered, particularly in elderly patients with vascular risk factors and acute audiovestibular loss, even when MRI does not demonstrate acute infarction in the brain. At this stage, clinicians should consider a further investigation and a proper management to prevent progression of acute audiovestibular loss into a more widespread posterior circulation stroke, mainly in the territory of AICA.31

e audiovestibular loss, even when MRI does not demonstrate acute infarction in the brain. At this stage, clinicians should consider a further investigation and a proper management to prevent progression of acute audiovestibular loss into a more widespread posterior circulation stroke, mainly in the territory of AICA.31 When does the patient with isolated vertigo need an urgent brain scan and what role does neuroimaging play in diagnosis? For patients with spontaneous prolonged vertigo, in addition to obvious cases of associated neurological symptoms or signs, an urgent brain scan should be considered to rule out central vascular vertigo syndrome in 1) older patients presenting with isolated spontaneous prolonged vertigo, in 2) any patient with vascular risk factors and isolated spontaneous prolonged vertigo who had a normal HIT, in 3) any patient with isolated spontaneous prolonged vertigo who had direction changing GEN or severe gait ataxia with falling at upright posture, in 4) any patient presenting with acute spontaneous vertigo and new onset headache, especially occipital, and in 5) any patient with vascular risk factors and acute onset of vertigo and hearing loss without a history of Meniere's disease.24,42 Since brain CT is known less accurate in detecting an acute ischemic lesion within the posterior fossa,45 brain MRIs with diffusion imaging are considered the golden standard for diagnosis of isolated vertigo due to ischemic strokes. However, diffusion-weighted MRI can be misleading up to at least 48 hours after onset of vertigo due to a stroke. Recent systemic review showed that the aggregate sensitivity of diffusion-weighted MRI of the posterior fossa during the first 24 hours or so after the vertigo onset to be 80% and a negative likelihood ration of 0.21 (95% CI, 0.16-0.26), which makes MRIs less potent for diagnosing stroke than the composite HINTS examination.19

ystemic review showed that the aggregate sensitivity of diffusion-weighted MRI of the posterior fossa during the first 24 hours or so after the vertigo onset to be 80% and a negative likelihood ration of 0.21 (95% CI, 0.16-0.26), which makes MRIs less potent for diagnosing stroke than the composite HINTS examination.19 Conclusion Patients with isolated vertigo are at a higher risk for stroke than the general population. Stroke in the distribution of the posterior circulation may mimic acute peripheral vestibular disorders. Isolated acute audiovestibular loss may herald an impending AICA territory infarction. A focused bedside examination (i.e., HINTS) identifies stroke with high sensitivity and specificity in patients with acute isolated vertigo and is superior to diffusion-weighted MRIs during the acute phase. The author has no financial conflicts of interest. Dr. Lee serves on the editorial boards of the Research in Vestibular Science, Frontiers in Neuro-otology, and Current Medical Imaging Review. Figure 1 Focal infarction selectively involving the structures responsible for isolated vertigo. (A) Isolated nodulus infarction. (B) Anterior inferior cerebellar artery territory infarct involving the lateral caudal pons extending from the root entry zone (*) of the eighth nerve to the most proximal portion of the vestibular fascicle. (C) Isolated vestibular nucleus infarction. (D) Focal infarction selectively involving the dorsal insula. (E) The flocculus is selectively infracted (from Lee [42], with permission).

e lateral caudal pons extending from the root entry zone (*) of the eighth nerve to the most proximal portion of the vestibular fascicle. (C) Isolated vestibular nucleus infarction. (D) Focal infarction selectively involving the dorsal insula. (E) The flocculus is selectively infracted (from Lee [42], with permission). Table 1 Differentiating among common central and peripheral acute vestibular syndromes *Direction-changed bidirectional gaze-evoked nystagmus that the intensity is maximal when gazes towards the lesion side; †Direction-fixed unidirectional gaze-evoked nystagmus beating toward the healthy side (from Lee [42], with permission). PICA, posterior inferior cerebellar artery, AICA, anterior inferior cerebellar artery.

Introduction Ischemic stroke is one of the major causes of death and disability. For the last few decades, many efforts have been made to improve the outcome of acute ischemic stroke treatment. However, thrombolytic therapy is still the only proven treatment for patients following an acute ischemic stroke within 3 or 4.5 hours of symptom onset.1,2 Studies focusing on expansion of the therapeutic time window and indication for thrombolysis are ongoing, but the adverse effects stemming from reperfusion injury, including hemorrhagic transformation (HT) or massive edema are still a concern.3 Therefore, it is important to select appropriate patients based on an assessment of individual risks and benefits for thrombolysis. Multimodal magnetic resonance imaging (MRI) is useful for diagnosing ischemic stroke and for determining treatment strategies in the acute phase.4,5 In the acute stage, early diagnosis of ischemic stroke and its differentiation from stroke-mimics are important.6,7 Various imaging findings from MRI sequences help determine stroke mechanisms, which affect prognosis, and thereby play an important role in treatment decisions. Lesion mismatch profiles on MRI help us to assess potential risks and benefits of thrombolysis by providing information on salvageable tissue or ischemic lesion age.8,9,10

ing findings from MRI sequences help determine stroke mechanisms, which affect prognosis, and thereby play an important role in treatment decisions. Lesion mismatch profiles on MRI help us to assess potential risks and benefits of thrombolysis by providing information on salvageable tissue or ischemic lesion age.8,9,10 Some of these parameters have been used in previous MRI-based thrombolysis trials, but the results were not satisfactory. While simply using a few parameters may be easily applicable, other valuable information for diagnosing stroke, determining the mechanism, and assessing the potential risks and benefits may be overlooked. Therefore, it is critical to understand the clinical implication of various imaging findings, and comprehensively consider them before deciding the treatment for acute stroke. In this review, we discuss the clinical implication of various MRI findings, specifically focusing on 1) MRI for diagnosis of acute stroke and its mechanism, 2) MRI-based patient selection for reperfusion therapy, 3) MRI outcome measures, and 4) the practicality of using MRI for hyperacute stroke.

Some of these parameters have been used in previous MRI-based thrombolysis trials, but the results were not satisfactory. While simply using a few parameters may be easily applicable, other valuable information for diagnosing stroke, determining the mechanism, and assessing the potential risks and benefits may be overlooked. Therefore, it is critical to understand the clinical implication of various imaging findings, and comprehensively consider them before deciding the treatment for acute stroke. In this review, we discuss the clinical implication of various MRI findings, specifically focusing on 1) MRI for diagnosis of acute stroke and its mechanism, 2) MRI-based patient selection for reperfusion therapy, 3) MRI outcome measures, and 4) the practicality of using MRI for hyperacute stroke. Imaging for acute stroke diagnosis Diagnosis of acute stroke Diagnosis of stroke largely depends on clinical presentation. Stroke-mimics account for 19%-30% of suspected stroke presentations, with diverse underlying etiology (Figure 1).11 Physicians need to consider a broad differential diagnosis when evaluating a patient presenting with a focal neurological deficit. With recent advances in MRI technology, ischemic lesions can be identified with high accuracy using diffusion-weighted image (DWI; 88%-100% sensitivity and 95%-100% specificity). The lesions appear as hyperintense areas on DWI and as correlative hypointense areas on apparent diffusion coefficient (ADC) maps, even within 3 minutes of stroke onset.12 Moreover, small cortical or subcortical lesions, especially in the posterior fossa or brain stem, are more easily detected by MRI than computed tomography (CT) at the acute stage. However, the small lesions located at the brain stem that present mild symptoms, especially ataxic hemiparesis or intranuclear ophthalmoplegia, could be invisible on initial DWI.13,14 In addition, other neurological diseases (i.e., Creutzfeldt-Jakob disease or progressive multifocal leukodystrophy) may also show high-intensity lesions on DWI, mimicking ischemic stroke.15,16 Therefore, considering clinical presentation as well as performing follow-up images may be beneficial for diagnosis.

DWI.13,14 In addition, other neurological diseases (i.e., Creutzfeldt-Jakob disease or progressive multifocal leukodystrophy) may also show high-intensity lesions on DWI, mimicking ischemic stroke.15,16 Therefore, considering clinical presentation as well as performing follow-up images may be beneficial for diagnosis. Assessing the mechanism of stroke Many studies have attempted to unravel stroke pathomechanism by ischemic lesion topography on DWI. It has been reported that multiple lesions in the unilateral anterior circulation or small, scattered lesions in one vascular territory are related to large artery atherosclerosis (Figure 2A, B).17 Perforating infarcts, in addition to pial or borderzone infarcts were distinctive pattern for intracranial atherosclerosis (Figure 2A).18,19 DWI also enables the detection of small lacunar infarcts that had previously been undetectable on CT. Fluid-attenuated inversion recovery (FLAIR) image demonstrate subacute or chronic ischemic lesions, which may help in classifying the subtype of index stroke.20 For example, recurrent deep perforating infarcts may develop in patients who have high burden of small vessel pathology.

ad previously been undetectable on CT. Fluid-attenuated inversion recovery (FLAIR) image demonstrate subacute or chronic ischemic lesions, which may help in classifying the subtype of index stroke.20 For example, recurrent deep perforating infarcts may develop in patients who have high burden of small vessel pathology. Cardioembolism can be suspected when patients exhibit acute multiple territorial lesions21 or a single large cortical and subcortical lesion on DWI.17,22 Particularly, early spontaneous recanalization of initially occluded vessels or abrupt occlusion without other atherosclerotic diseases on magnetic resonance angiography (MRA) may suggest cardioembolic stroke (Figure 2C).23,24 Hypointense signals with blooming artifacts in the vascular cistern on gradient echo (GRE) image (susceptible vessel sign; SVS), which may reflect the presence of deoxygenated hemoglobin of red thrombi, are usually associated with cardioembolic stroke (Figure 2C).25 Patent foramen ovale, another source of cardioembolism, usually causes small cortical lesions in vertebro-basilar circulation, which may reflect high blood flow in the posterior circulation provoked by the Valsalva maneuver.26 Currently, aortic arch atherosclerosis is accepted as a possible embolic source in cryptogenic stroke, and the resulting ischemic lesion patterns and clinical outcomes differ from those of cardiogenic embolism. Moreover, aortic arch embolism has a higher propensity of causing left hemispheric stroke and multiple small infarcts (Figure 2D).27,28

erosclerosis is accepted as a possible embolic source in cryptogenic stroke, and the resulting ischemic lesion patterns and clinical outcomes differ from those of cardiogenic embolism. Moreover, aortic arch embolism has a higher propensity of causing left hemispheric stroke and multiple small infarcts (Figure 2D).27,28 Early assessment of the etiology ischemic stroke in the acute stage using MRI is advantageous to guide specific diagnostic work-ups and proper strategy for hyperacute reperfusion therapy and secondary stroke prevention. Patient selection for acute stroke treatment Evaluation of salvageable tissue Concept of penumbra When the cerebral blood vessel is occluded, a complex series of pathophysiological events evolve in time and space.29 First, the core of the area rapidly develops an infarction (ischemic core). However, the surrounding part of the core still exhibits minimum blood flow supplied by collateral circulation, even when neuronal function has been suspended (ischemic penumbra). The neuronal function of some parts of the ischemic penumbra can recover when blood supply is restored, and goes through a dynamic change during the acute period of ischemic stroke.30 Perfusion-weighted image (PWI) can identify ischemic penumbral tissue, whereas DWI-depicted lesions represent the ischemic core. Therefore, areas with PWI-DWI mismatch (i.e., when the perfusion lesion is larger than the diffusion lesion) have been considered representative salvageable tissue that require active treatment (Figure 3A).

ted image (PWI) can identify ischemic penumbral tissue, whereas DWI-depicted lesions represent the ischemic core. Therefore, areas with PWI-DWI mismatch (i.e., when the perfusion lesion is larger than the diffusion lesion) have been considered representative salvageable tissue that require active treatment (Figure 3A). PWI parameters PWI is a semi-quantitative method for evaluating brain perfusion - microcirculation in the capillary network.31 When a contrast agent is administered into a vein, it passes through cerebral vessels and alters the local magnetic field resulting in the rapid decrease of signal intensity in the surrounding brain tissue by the paramagnetic effect of the contrast. The variation of signal intensity is measured during 1 minute, serially with 1- to 2-second intervals by the echo-planar image technique. From this data, time-concentration curves can be obtained at the tissue level, voxel-by-voxel (Figure 3B). After deconvolution with arterial input functions, a deconvolved curve can be obtained (Figure 3C), and various perfusion parameters can be calculated. These parameters include cerebral blood flow (CBF), cerebral blood volume (CBV), mean transit time (MTT), time to peak (TTP), and Tmax (Figure 3D).32

voxel (Figure 3B). After deconvolution with arterial input functions, a deconvolved curve can be obtained (Figure 3C), and various perfusion parameters can be calculated. These parameters include cerebral blood flow (CBF), cerebral blood volume (CBV), mean transit time (MTT), time to peak (TTP), and Tmax (Figure 3D).32 CBF and CBV CBF is a parameter usually taken at the height of deconvolved curve (Figure 3C). It reflects the blood supply to the brain tissue in a given time and is most directly related to the viability of the infarcted tissue.33 CBF is determined by cerebral perfusion pressure, the dilation of blood vessels, and blood viscosity. An area with normal CBF and delayed MTT or TTP demonstrates an area with blood flow maintained by blood vessel dilation (increased CBV), but reaching the particular area late through collaterals. CBV is measured by the whole blood quantity within the target area (area under the deconvolved curve; Figure 3C). Areas of decreased CBV correlate well with the final size of a cerebral infarction.34 This is especially the case when the area of delayed MTT with decreased CBV represents the area of brain tissue not having sufficient collateral circulation; thus, these regions will eventually evolve into cerebral infarctions.33

3C). Areas of decreased CBV correlate well with the final size of a cerebral infarction.34 This is especially the case when the area of delayed MTT with decreased CBV represents the area of brain tissue not having sufficient collateral circulation; thus, these regions will eventually evolve into cerebral infarctions.33 MTT, TTP, and Tmax MTT is the average time required for blood flow to enter the artery and maintain the inside of the cerebral artery. MTT is calculated by CBV/CBF and is used to estimate vulnerable brain tissue which may evolve from the infarction.35 MTT shows the widest range of perfusion deficits compared to other parameters, and is therefore likely to overestimate areas with risk. If an area with MTT delay shows increased CBV, that area may have received sufficient collateral circulation or may have been currently recanalized.33 TTP describes the time it takes CBF to reach the highest value at the target tissue location (Figure 3B). TTP is an indirect measurement of brain perfusion; therefore, it provides minimum information. Since a delay in TTP can occur in a patient with chronic carotid artery stenosis without acute infarction (prolonged arrival time), TTP can also overestimate the hypo-perfused area in an acute infarction.33

ation (Figure 3B). TTP is an indirect measurement of brain perfusion; therefore, it provides minimum information. Since a delay in TTP can occur in a patient with chronic carotid artery stenosis without acute infarction (prolonged arrival time), TTP can also overestimate the hypo-perfused area in an acute infarction.33 Tmax is the time it takes for the tissue residue function to reach its maximum value. Tmax is a sensitive parameter reflecting changes of brain tissue into an infarction and changes in the perfusion state. Tmax has also been used as a predictor of tissue viability in many studies as a non-physiological parameter of the capability of brain tissue to survive. Since this parameter is not influenced by scan duration, Tmax has the merit that sufficient scanning for a long time is possible so that contrast agent can reach all parts.35 Thus, Tmax is the most widely accepted parameter to measure the penumbra. PWI parameter thresholds in discriminating between penumbra and benign oligemic area Recent studies have focused on the threshold for distinguishing a true penumbra from a benign oligemia. Although time-based perfusion variables are widely used methods assessing penumbra, there are still a lot of controversies on interpreting them.

thresholds in discriminating between penumbra and benign oligemic area Recent studies have focused on the threshold for distinguishing a true penumbra from a benign oligemia. Although time-based perfusion variables are widely used methods assessing penumbra, there are still a lot of controversies on interpreting them. Previously, many different thresholds for Tmax representing the true penumbra area have been suggested. For example, a Tmax >6 seconds was defined as the penumbra in the Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution (DEFUSE) trial,36 whereas in the effects of alteplase beyond 3 hours after stroke in the Echoplanar Imaging Thrombolytic Evaluation (EPITHET) trial cohort, a Tmax of 4 to 6 seconds delay was used.37 Another study showed that a Tmax >4 seconds delay provided the most accurate final infarct volume in patients within 3 to 6 hours after stroke onset.38 Compared with the penumbra measured from a positron emission tomography study, a Tmax ≥5.5 seconds delay matched most well.37 Recently, another study suggested that Tmax ≥10 seconds delay best predicts the final infarction.39 Taken together, these studies indicate that there is no established criterion to discriminate between benign oligemic area and the penumbra. Although a number of studies have been conducted on Tmax threshold for that purpose, there are still a lot of controversies on the optimum criterion since the threshold value to determine the final infarction varies from study by study.

stablished criterion to discriminate between benign oligemic area and the penumbra. Although a number of studies have been conducted on Tmax threshold for that purpose, there are still a lot of controversies on the optimum criterion since the threshold value to determine the final infarction varies from study by study. PWI-DWI mismatch in patient selection Previously, PWI-DWI mismatch was used for patient selection in several clinical trials focusing on acute stroke treatment. In the DEFUSE study, patients with PWI-DWI mismatch in the 3 to 6 hours window demonstrated more favorable clinical response after reperfusion compared to patients without a PWI-DWI mismatch.40 However, this previous study had no placebo control group and it was not designed to demonstrate the efficacy of MRI-based intravenous thrombolysis. In the EPITHET study, infarct growth was compared between patients who received tissue plasminogen activator (tPA) and those who received a placebo. The primary outcome measure was the attenuation of infarct growth using a ratio of geometric means.41 However, their results failed to demonstrate beneficial outcomes of using PWI-DWI mismatch in thrombolysis. The Desmoteplase in Acute Ischemic Stroke (DIAS) clinical trial was a phase-II trial that employed a pre-randomization penumbral imaging screening as an indication for patient selection.42 Within 3-9 hours of symptom onset, patients with ischemic stroke and at least 20% PWI-DWI mismatch (as evaluated by visual inspection) were included in the trials. However, the results of the DIAS study have failed to prove the benefit of desmoteplase versus placebo in patients with MRI-identified penumbras.

lection.42 Within 3-9 hours of symptom onset, patients with ischemic stroke and at least 20% PWI-DWI mismatch (as evaluated by visual inspection) were included in the trials. However, the results of the DIAS study have failed to prove the benefit of desmoteplase versus placebo in patients with MRI-identified penumbras. The previous trials of imaging-based thrombolysis were solely or mostly based on PWI-DWI mismatch and failed to prove any advantages. These failures however, have reinforced the benefit of salvaging the penumbral tissue by thrombolysis, and the fact that HT risk should be considered when deciding to treat a patient via thrombolysis. Hemorrhagic risk stratification after thrombolysis Mechanisms and clinical implication HT is a frequent, often asymptomatic event that occurs after acute ischemic stroke. It is thought to negatively influence the early clinical course and outcome of patients, particularly those receiving thrombolytic therapy.43 HT is presumed to occur in the infarcted brain area due to the extravasation of blood components.44 Disruption of the blood-brain barrier has been proposed to precede HT.45,46 In addition, the dose-dependent side effects of tPA treatment have been discussed as risk factors for HT.47 Early blood-brain barrier disruption is associated with HT even in patients who do not receive tPA treatment for stroke.48 This might be due to an increase in vascular permeability in ischemic tissue, resulting in a leakage from small vessels when exposed to restored blood flow.

been discussed as risk factors for HT.47 Early blood-brain barrier disruption is associated with HT even in patients who do not receive tPA treatment for stroke.48 This might be due to an increase in vascular permeability in ischemic tissue, resulting in a leakage from small vessels when exposed to restored blood flow. Definition and classification of HT In the European Cooperative Acute Stroke Study (ECASS) trials,2 the types of HT after acute ischemic stroke were classified into hemorrhagic infarctions (types 1 and 2) and parenchymal hemorrhages (types 1 and 2) on the basis of CT scan characteristics (Figure 4). In the NINDS tPA Stroke Trial,1 the definition of HT required blood to be detected by CT after treatment. Irrespective of other features, detection of any blood by CT in a patient with neurological deterioration qualified the patient for symptomatic HT. In the ECASS-II trial, investigators looked for an increase of more than 4 points on the National Institutes of Health Stroke Scale (NIHSS) score and detection of blood at any site by CT for this qualification.49 Finally, in ECASS-III, an increase of more than 4 points on the NIHSS score had to be "correlated" with blood detection by CT to suggest HT as the probable cause of neurological deterioration.2

he National Institutes of Health Stroke Scale (NIHSS) score and detection of blood at any site by CT for this qualification.49 Finally, in ECASS-III, an increase of more than 4 points on the NIHSS score had to be "correlated" with blood detection by CT to suggest HT as the probable cause of neurological deterioration.2 Factors associated with HT Several factors are known to be associated with HT, and the combination of clinical and imaging data helps identify patients at high risk of symptomatic HT.50 Clinical factors such as a high initial NIHSS,51 delayed treatment time,52 and high blood pressure53 are well known to increase the incidence of HT and have been used as exclusion criteria for thrombolytic therapy. The imaging data indicative of a large infarct54 and early ischemic signs on CT are also widely used as exclusion criteria for thrombolytic therapy.55 In fact, early ischemic changes were defined in the NINDS tPA trial as the presence of one or more of the following characteristics: 1) loss of grey-white matter distinction, 2) hypodensity (hypo-attenuation) of brain parenchyma, and 3) compression of CSF spaces. However, the sensitivity and reproducibility of early ischemic signs on brain CT is poor.56

e defined in the NINDS tPA trial as the presence of one or more of the following characteristics: 1) loss of grey-white matter distinction, 2) hypodensity (hypo-attenuation) of brain parenchyma, and 3) compression of CSF spaces. However, the sensitivity and reproducibility of early ischemic signs on brain CT is poor.56 MRI predictors for HT A low ADC value on baseline MRIs has been suggested as an independent predictor of HT after thrombolysis.54 Results of other studies have shown that large DWI lesion volume, or very low/absent apparent CBV or CBF can increase the risk of HT.57 Focal FLAIR hyperintensity within acute infarcts is also associated with an increased risk of symptomatic HT in several studies.58 The predictive value of silent microbleeds for HT after thrombolysis is unclear. Although their presence on GRE has been suggested to be predictive of increased HT risk after treatment with tPA,59 data from a large cohort study indicated that micro-hemorrhages were not an independent risk factor for early and symptomatic HT, irrespective of the number of microbleeds.60 In addition, a larger PWI lesion volume (a measure of ischemic duration) was associated with hemorrhagic infarction, but not with parenchymal hemorrhage.61

a large cohort study indicated that micro-hemorrhages were not an independent risk factor for early and symptomatic HT, irrespective of the number of microbleeds.60 In addition, a larger PWI lesion volume (a measure of ischemic duration) was associated with hemorrhagic infarction, but not with parenchymal hemorrhage.61 Considering the pathomechanism of HT, radiological markers that directly indicate blood-brain barrier permeability could predict HT after acute ischemic stroke. In fact, it has been shown that contrast-enhancement can predict tPA-induced hemorrhages in rat models.62 These techniques have been assessed in humans, including delayed gadolinium enhancement of the CSF space (Figure 5A)48,63,64,65 or sulcal enlargement on FLAIR images,66 parenchymal enhancement on post-contrast T1-weighted images (Figure 5B),67 and permeability images derived from pretreatment perfusion MRIs.68

odels.62 These techniques have been assessed in humans, including delayed gadolinium enhancement of the CSF space (Figure 5A)48,63,64,65 or sulcal enlargement on FLAIR images,66 parenchymal enhancement on post-contrast T1-weighted images (Figure 5B),67 and permeability images derived from pretreatment perfusion MRIs.68 Assessing the nature and burden of clots Magnetic susceptibility artifacts distort ferromagnetic objects. Technically, shortening the echo-time, increasing the frequency matrix, and decreasing the slice thickness could reduce artifact size. On the other hand, magnetic susceptibility artifacts may also enhance the detection of red thrombi clots, which are ferromagnetic. As the composition, size, and site of clot occluding cerebral arteries are important factors for selecting the treatment strategy, GRE and SVS may be useful for treatment decisions. Initial, long-standing, platelet-rich, and well-organized white thrombi in the cerebral artery under high shear-stress are more resistant to thrombolytic therapy than fresh, fibrin-rich red thrombi formed under static conditions. Therefore, SVS is known as a marker with higher possibility of recanalization.69 Second, the location of a clot is also important. SVS of M1 is a strong predictor of recanalization failure after tPA.70 Therefore, intra-arterial thrombolysis may be more effective than tPA in certain cases. Third, the clot length is typically used to quantify the thrombotic burden.71 This is important since intra-venous tPA has nearly no potential to recanalize an MCA occlusion with a clot length exceeding 8 mm (Figure 6A).72 Finally, irregular and tortuous clot morphologies can decrease the technical and clinical success of thrombectomies in M1 occlusions (Figure 6B).73

fy the thrombotic burden.71 This is important since intra-venous tPA has nearly no potential to recanalize an MCA occlusion with a clot length exceeding 8 mm (Figure 6A).72 Finally, irregular and tortuous clot morphologies can decrease the technical and clinical success of thrombectomies in M1 occlusions (Figure 6B).73 MRI-based thrombolysis in unclear-onset patients Physicians are frequently confronted with patients in whom the exact time of stroke symptom onset is not known,74 and attempts have been made to use signal changes in FLAIR images as a kind of "tissue clock".75 For example, it is known that signal intensity in FLAIR images proportionately increases with a rise in water content inside the infarcted tissue. The water content rises due to vasogenic edema as the blood-brain barrier is disrupted, and occurs within 1 to 4 hours of stroke onset.76 Therefore, DWI-FLAIR mismatch (i.e., lesion visible on DWI but not on FLAIR) has been used as a surrogate marker for estimating the lesion age of unknown stroke onset, and can help determine the use of thrombolytic agent (Figure 7).77,78 Patients with DWI-FLAIR mismatch are likely to be within the time window for thrombolysis, the specificity and predictive value of which have both been shown to be high (93% and 94%, respectively).77 Finally, in the reperfusion therapy in unclear-onset stroke based on MRI evaluation (RESTORE), patients with unclear stroke onset, within 6 hours of symptom detection with PWI-DWI mismatch >20% and negative or subtle FLAIR change, were treated with tPA or endovascular therapy, and MRI-based reperfusion therapy was found to be feasible and safe.79

therapy in unclear-onset stroke based on MRI evaluation (RESTORE), patients with unclear stroke onset, within 6 hours of symptom detection with PWI-DWI mismatch >20% and negative or subtle FLAIR change, were treated with tPA or endovascular therapy, and MRI-based reperfusion therapy was found to be feasible and safe.79 However, potential confounding factors that interfere with the diagnostic accuracy of DWI-FLAIR mismatch also exist. In addition to the time from symptom onset, younger age and large ischemic lesion volume have been reported to be associated with high FLAIR signal intensity.9 As age and lesion size alter permeability in the blood-brain barrier, these biological variations may cause altered FLAIR signal intensity.80 In addition, there is an issue stemming from the low inter-rater reliability of DWI-FLAIR mismatch. For example, in a previous study it was reported that the inter-rater reliability of FLAIR change was only moderate, based on visual rating.9 In other studies, the mean signal intensity of a circular region-of-interest inside the infarcted tissue was quantitatively measured and used to overcome this issue, but even this approach failed to improve inter-rater reliability.81 On the other hand, color-coding of FLAIR signal change has been reported to enhance visual distinguishability, and seems to be effective in decreasing the disagreement between raters.82 Further results of the on-going trials may provide more knowledge on the risk and benefit of MRI-based thrombolysis in this particular group of patients. Table 1 introduces these major on-going clinical trials.

enhance visual distinguishability, and seems to be effective in decreasing the disagreement between raters.82 Further results of the on-going trials may provide more knowledge on the risk and benefit of MRI-based thrombolysis in this particular group of patients. Table 1 introduces these major on-going clinical trials. MRI-based thrombolysis in patients with minor stroke One-third of acute ischemic strokes that have occurred within the therapeutic time window have been excluded because of mild or rapidly improving symptoms.83 However, poor outcome in patients who did not receive tPA because of minor stroke symptoms has been reported with up to 30% of these patients at risk of death and dependence.84 Moreover, a subgroup of patients with minor stroke and proximal vessel occlusion were found to be at higher risk of neurological deterioration.84 Clinical criteria, such as NIHSS scores, are poorly predictive of proximal intracranial occlusion, and underestimate the risk of subsequent neurological deterioration.85 Thus, off-label thrombolytic therapy was not associated with higher complication rates in cases of minor stroke.86

sk of neurological deterioration.84 Clinical criteria, such as NIHSS scores, are poorly predictive of proximal intracranial occlusion, and underestimate the risk of subsequent neurological deterioration.85 Thus, off-label thrombolytic therapy was not associated with higher complication rates in cases of minor stroke.86 MRI has been suggested as a tool to identify patients with minor stroke who may benefit from thrombolysis. In patients that have experienced a minor stroke, a significant mismatch can persist for days and their symptoms may aggravate without recanalization therapy. Recent studies have reported that treatment with tPA based on PWI-DWI mismatch is safe and effective in patients with minor stroke.87 However, the lesion pattern, rather than the PWI-DWI mismatch, has been shown to be an important predictor of neurological decline in patients that have undergone a minor stroke.88 Still, no systematic data are available on the effect of IV-tPA in cases of minor stroke. Thus, only the inclusion of patients with minor stroke in future randomized controlled trials of intravenous thrombolysis will allow us to answer the question of whether thrombolysis is effective and safe in this group of patients.

till, no systematic data are available on the effect of IV-tPA in cases of minor stroke. Thus, only the inclusion of patients with minor stroke in future randomized controlled trials of intravenous thrombolysis will allow us to answer the question of whether thrombolysis is effective and safe in this group of patients. MRI outcome measures Clinical trials evaluating the success of thrombolysis have relied on various outcome parameters to measure how well an ischemic vascular bed responds to treatment. Usually, clinical parameters based on the severity of neurological deficits (NIHSS scores) or functional disability (modified Rankin Scale or Barthel Index) are used. However, as these clinical parameters are influenced by various confounders, including the lesion location or patient age, they do not directly demonstrate the treatment effect of salvaging the penumbral tissue. On the other hand, imaging parameters have an advantage in quantitatively measuring the effect of treatment by demonstrating the initial infarction core, penumbra, and the final infarction volume.89 The three main categories used in outcome measures by MRI are, 1) DWI lesion volume change, 2) PWI lesion volume change, and 3) recanalization of the occluded vessel on MRA.

age in quantitatively measuring the effect of treatment by demonstrating the initial infarction core, penumbra, and the final infarction volume.89 The three main categories used in outcome measures by MRI are, 1) DWI lesion volume change, 2) PWI lesion volume change, and 3) recanalization of the occluded vessel on MRA. Currently, there is no consensus on how to determine infarct growth. This measure has been based on the change in lesion volume between baseline and follow-up MRIs (DWI at 24 hours and FLAIR at 5 days).90 In the EPITHET study, various measures of infarction volume were suggested, such as geometric mean and the difference in cube-root volumes.41 However, measurement tools in the clinical setting should be easy to use, and PWI parameters used in the previous clinical trials differ from study to study. Thus, there are some difficulties in comparing the results of each study. For example, in the DEFUSE study, early reperfusion was defined as a more than 50% reduction in the volume of the PWI lesion (Tmax >6 seconds).91 On the other hand, reperfusion in the EPITHET study was defined as >90% reduction between baseline and day-3 PWI lesion volumes.41 This discrepancy is also seen in the DIAS study as these investigators defined reperfusion as ≥30% reduction of MTT volume abnormality between baseline and follow-up images.42

6 seconds).91 On the other hand, reperfusion in the EPITHET study was defined as >90% reduction between baseline and day-3 PWI lesion volumes.41 This discrepancy is also seen in the DIAS study as these investigators defined reperfusion as ≥30% reduction of MTT volume abnormality between baseline and follow-up images.42 Two distinct measures for assessing recanalization have been used to evaluate the effectiveness of thrombolytic therapy. First, recanalization of the primary arterial occlusive lesion has been evaluated using the Arterial Occlusive Lesion (AOL) grading system. Second, reperfusion of the distal vascular bed has been evaluated using the Thrombolysis in Myocardial Infarction (TIMI) and the Thrombolysis in Cerebral Infarction (TICI) grading system (Table 2). However, the use of multiple scales has resulted in some confusion and limits the comparison among studies.

rading system. Second, reperfusion of the distal vascular bed has been evaluated using the Thrombolysis in Myocardial Infarction (TIMI) and the Thrombolysis in Cerebral Infarction (TICI) grading system (Table 2). However, the use of multiple scales has resulted in some confusion and limits the comparison among studies. In the Interventional Management of Stroke (IMS)-1 trial,92 the AOL grading system was used. However, recanalization of the primary arterial occlusive lesion does not guarantee complete reperfusion of the downstream arterioles. The TIMI definition, which has been used to describe flow in the coronary arteries, was adapted to evaluate the degree of reperfusion.93,94 Therefore, in order to emphasize the use of a standard grading system specific to intracranial circulation, a TICI grading system has been suggested.95 Recently, the TICI grading system is regarded as the most reliable parameter for recanalization. While there are some limitations of angiography based scales, combining the scales with MR perfusion may be a good parameter for treatment outcome.

ic to intracranial circulation, a TICI grading system has been suggested.95 Recently, the TICI grading system is regarded as the most reliable parameter for recanalization. While there are some limitations of angiography based scales, combining the scales with MR perfusion may be a good parameter for treatment outcome. Correlation of MRI parameters and clinical outcome It is important to ask whether MRI parameters are correlated with clinical outcomes. In fact, it has been shown that early infarct growth within the first week can predict long-term clinical outcome after thrombolysis.96 Specifically, when recanalization scores were dichotomized into poor (AOL 0-2, TIMI 0-1, and TICI 0-2a) versus good revascularization (AOL 3, TIMI 2-3, TICI 2b-3), investigators found a significant difference between the two groups (Table 2). Moreover, it was revealed that patients with good clinical outcomes (modified Rankin scale 0-2) had smaller infarct growth, smaller perfusion growth, and good revascularization.97

) versus good revascularization (AOL 3, TIMI 2-3, TICI 2b-3), investigators found a significant difference between the two groups (Table 2). Moreover, it was revealed that patients with good clinical outcomes (modified Rankin scale 0-2) had smaller infarct growth, smaller perfusion growth, and good revascularization.97 Practicality of MRI for use in hyperacute stroke Multimodal MRI reveals various useful parameters for deciding treatment routes of acute stroke (Table 3). This technology is especially useful when stroke diagnosis is uncertain (stroke-mimics), stroke onset is unclear (wake-up or unwitnessed daytime strokes), or clinic-imaging mismatches exist. However, in spite of the potential benefits of MRI, practical issues exist limiting the use of MRI in the acute stage of ischemic stroke.98 One of the major issues is the long scan time when compared to CT. In general, although MRI-based decisions of thrombolysis have prolonged the door-to-needle time, the clinical outcome has been more favorable.99,100 Therefore, the use of MRI in acute stroke treatment is clinically practical and feasible.101,102 On the other hand, considering that "time is brain" in the hyperacute stage of ischemic stroke, minimizing the evaluation time taken for the decision of treatment may expand the benefit of thrombolysis. Interestingly, a recent study revealed a feasible diagnostic quality of 6-minute multimodal MRI using echo-planar FLAIR and GRE in patients with acute ischemic stroke, and this may enhance the usage of MRI in acute stroke treatment.103 The effect and usefulness of this paradigm in acute stroke treatment should be verified by future trials. In addition, various new MRI advancements are actively under research, such as arterial spin labeling, techniques to measure collateral flows, computational flow dynamics, and high-resolution vessel-wall MRI. These novel imaging biomarkers may be helpful in determining the treatment strategy for acute ischemic stroke.

In addition, various new MRI advancements are actively under research, such as arterial spin labeling, techniques to measure collateral flows, computational flow dynamics, and high-resolution vessel-wall MRI. These novel imaging biomarkers may be helpful in determining the treatment strategy for acute ischemic stroke. Conclusion Multimodal imaging provides information that is useful for diagnosing ischemic stroke, selecting appropriate patients for thrombolytic therapy, and predicting the prognosis of ischemic stroke. Only depending on a single or a few parameters may not be sufficient, instead comprehensively combining the information from each MRI sequence (i.e., DWI, FLAIR, GRE, and PWI) and using various mismatch parameters (DWI-FLAIR mismatch and/or PWI-DWI mismatch) may be more helpful in establishing an indication of MRI-based thrombolysis. This study was supported by grants from the Korea Health Technology R&D Project, Ministry for Health & Welfare, Republic of Korea (HI12C1847 and HI11C1531). The authors have no financial conflicts of interest. Figure 1 Common stroke mimics, identified in a systematic review and meta-analysis of case series.11 Figure 2 DWI lesion patterns according to stroke subtypes. (A) Intracranial atherosclerotic stenosis, (B) extracranial atherosclerotic stenosis, (C) cardioembolism, and (D) aortic arch embolism. Figure 3 PWI-DWI mismatch (A), time-concentration curves (B and C) and various PWI parameters (D). Figure 4 Types of hemorrhagic transformation according to ECASS criteria.

Figure 2 DWI lesion patterns according to stroke subtypes. (A) Intracranial atherosclerotic stenosis, (B) extracranial atherosclerotic stenosis, (C) cardioembolism, and (D) aortic arch embolism. Figure 3 PWI-DWI mismatch (A), time-concentration curves (B and C) and various PWI parameters (D). Figure 4 Types of hemorrhagic transformation according to ECASS criteria. Figure 5 MRI markers predicting hemorrhagic transformation (A) Delayed gadolinium enhancement of the CSF space (arrows); (B) Parenchymal enhancement at post contrast T1-weighted image (arrow) and hemorrhagic transformation at the corresponding area at follow-up (arrow). Figure 6 Clot presented on gradient echo image (arrow) with long segment (A) and tortuous vessel (B). Figure 7 DWI-FLAIR mismatch: positive (A) and negative (B). Table 1 On-going clinical trial to expand the time-window of intra-venous thrombolysis LNNT, last known normal time; FFAT, first found abnormal time; DWI, diffusion-weighted image; FLAIR, fluid attenuated inversion recovery; PWI, perfusion-weighted image; MRI, magnetic resonance image; ASPECT, Alberta stroke program early CT; NIHSS, National Institutes of Health Stroke Scale. Table 2 Recanalization grading system TIMI, thrombolysis in myocardial infarction; TICI, thrombolysis in cerebral infarction; AOL, arterial occlusive lesion. Table 3 Basic principles and clinical implications of multimodal MRI DWI, diffusion weighted image; FLAIR, fluid-attenuated inversion recovery image; GRE, gradient echo image; PWI, perfusion weighted image; MRA, magnetic resonance angiography; CT, computerized tomography.

Introduction Malignant cerebral edema following ischemic stroke is life threatening. The pathophysiology of brain edema involves failure of the sodium-potassium adenosine triphosphatase pump and disruption of the blood-brain barrier, leading to cytotoxic edema and cellular death.1 The Monro-Kellie doctrine dictates that since the brain is encased in a finite space, increased intracranial pressure (ICP) due to cerebral edema can result in herniation through the foramen magnum and openings formed by the falx and tentorium.2 Moreover, elevated ICP can cause secondary brain ischemia through decreased cerebral perfusion and blood flow, brain tissue hypoxia, and metabolic crisis.3 Direct cerebrovascular compression caused by brain tissue shifting can lead to secondary infarction, especially in the territories of the anterior and posterior cerebral artery.4 Tissue shifts can also stretch and tear cerebral vessels, causing intracranial hemorrhage such as Duret's hemorrhage of the brainstem.4 Physicians treating ischemic stroke with swelling should utilize every available medical and surgical therapy to minimize secondary brain injuries. Therefore, patients with space-occupying ischemic stroke are advised to be admitted to an intensive care unit (ICU) for constant neurological monitoring.5

f the brainstem.4 Physicians treating ischemic stroke with swelling should utilize every available medical and surgical therapy to minimize secondary brain injuries. Therefore, patients with space-occupying ischemic stroke are advised to be admitted to an intensive care unit (ICU) for constant neurological monitoring.5 Physicians who manage stroke patients should try to understand the body as a whole and not be limited to the brain. It is possible that provision of comprehensive care is the critical factor that determines patient outcomes, rather than focusing on one specific therapy. However, managing the brain and the rest of the body simultaneously presents a challenge to the physician, as focusing on one organ may be at the cost of another. In this review, we provide a comprehensive overview of the management of cerebral and cerebellar infarction with malignant edema. We discuss the measurement of ICP and strategies to lower ICP, including decompressive craniectomy and therapeutic hypothermia. Finally, we discuss ventilatory support and the advantages and limitations of a sedation-off wake-up test.

verview of the management of cerebral and cerebellar infarction with malignant edema. We discuss the measurement of ICP and strategies to lower ICP, including decompressive craniectomy and therapeutic hypothermia. Finally, we discuss ventilatory support and the advantages and limitations of a sedation-off wake-up test. Measurement of ICP ICP is the pressure exerted by the brain, blood, and cerebrospinal fluid (CSF) in the intracranial vault. The skull is a rigid container filled with the brain, blood, and CSF. Expansion of one component occurs at the expense of others, with increases in ICP leading to intracranial hypertension. Normal ICP values for adults are 7-15 mmHg.6 Generally, ICP values greater than 20 mmHg require treatment.7 However, ICP should be interpreted in the context of cerebral perfusion pressure (CPP). CPP is a major factor that affects cerebral blood flow (CBF). It is determined by calculating the difference between mean arterial pressure (MAP) and ICP (CPP = MAP-ICP). CPP can therefore decrease as a result of decreased MAP, increased ICP, or a combination of both. CPP is proportional to CBF as long as cerebrovascular resistance remains constant. To avoid secondary brain ischemia after stroke, extremely low CPP should be avoided. Optimal CPP values have not been clearly established, but 50-60 mmHg is generally accepted as the minimum pressure required to prevent further brain injury.3

s proportional to CBF as long as cerebrovascular resistance remains constant. To avoid secondary brain ischemia after stroke, extremely low CPP should be avoided. Optimal CPP values have not been clearly established, but 50-60 mmHg is generally accepted as the minimum pressure required to prevent further brain injury.3 The gold standard assessment of ICP requires insertion of an extraventricular drain (EVD) into the lateral ventricle with a connection to an external pressure transducer.8 In addition to monitoring absolute pressure values and waveforms, an EVD catheter allows CSF diversion to reduce ICP. Alternatively, ICP can be measured through intraparenchymal, transducer-tipped catheters, including fiber optic and microstrain-gauge sensors.8 A parenchymal catheter can provide continuous measurement of ICP and allows analysis of the interaction between ICP and MAP (pressure reactivity index) as well as real-time values and ICP waveforms.9 Although this device provides a quantitative value for ICP, it can only be calibrated prior to placement of an ICP probe and is vulnerable to drift.

rovide continuous measurement of ICP and allows analysis of the interaction between ICP and MAP (pressure reactivity index) as well as real-time values and ICP waveforms.9 Although this device provides a quantitative value for ICP, it can only be calibrated prior to placement of an ICP probe and is vulnerable to drift. There are controversies surrounding ICP monitoring in patients with stroke. Early studies showed that ICP monitoring was useful for predicting clinical outcomes after acute hemispheric stroke. ICP often correlated with clinical deterioration, final outcome, and computed tomography findings.10 However, a subsequent study of patients with malignant middle cerebral artery (MCA) infarctions showed that pupillary abnormalities and signs of severe brainstem compression were sometimes present despite normal ICP.11 Randomized clinical trials of ICP monitoring have not been performed in patients with stroke. A randomized clinical trial of ICP monitoring in patients with traumatic brain injury failed to show superior efficacy of ICP monitoring over serial neurological examinations and repeated neuroimaging studies.12 Limitations of ICP monitoring in predicting neurological deterioration may be due to the distance between the site of ICP probe insertion and the site of herniation (Figure 1A), because the pressure gradient force is inversely proportional to the distance. The relative positioning of the tentorial aperture and the brainstem may also explain variations in herniation syndromes in patients with similar ICP values and intracranial pathological conditions. Some patients exhibit a narrow tentorial opening and a high degree of contiguity between the brainstem and the tentorial edge, whereas others exhibit a wide tentorial opening and perimesencephalic space. Such morphometric variations in the tentorial aperture and its regional anatomy may play a role in the varying clinical manifestations (Figure 1B).13 Alternatively, since the brain has compartments divided by hard dural structures, i.e., the falx cerebri and tentorium cerebelli, ICP probes placed in the right frontal cortex may not appropriately reflect the pressure in the left hemisphere or infratentorium (compartmentalized ICP).10,14 Malfunction of a mechanical device and probe drift may also contribute to inaccurate ICP values.

ral structures, i.e., the falx cerebri and tentorium cerebelli, ICP probes placed in the right frontal cortex may not appropriately reflect the pressure in the left hemisphere or infratentorium (compartmentalized ICP).10,14 Malfunction of a mechanical device and probe drift may also contribute to inaccurate ICP values. Therefore, routine ICP monitoring without careful interpretation, neurological examination, and a neuroimaging study cannot be recommended in patients with cerebral and cerebellar infarct with swelling.5 However, it may be helpful in comatose patients especially during barbiturate coma therapy and ventilator care under heavy sedation because in these situations neurological examinations are not sensitive enough to detect secondary brain injury, and constant neuroimaging studies are not available. Multimodal neuromonitoring that includes ICP, CPP, CBF, brain oxygen tension, markers of brain metabolism (glucose, lactate, and pyruvate) and electroencephalography may be useful to further understand the physiology of comatose stroke patients.15

ndary brain injury, and constant neuroimaging studies are not available. Multimodal neuromonitoring that includes ICP, CPP, CBF, brain oxygen tension, markers of brain metabolism (glucose, lactate, and pyruvate) and electroencephalography may be useful to further understand the physiology of comatose stroke patients.15 Medical management of intracranial hypertension In general, there are four causes of intracranial hypertension following ischemic stroke: (1) brain swelling due to cytotoxic or vasogenic edema; (2) increased cerebral blood volume due to dilated cerebral arteries; (3) obstruction of venous outflow; and (4) acute hydrocephalus.16 The optimal approach to manage intracranial hypertension differs according to the underlying cause. Brain swelling requires osmotherapy and/or decompressive craniectomy, cerebral arterial dilatation requires a vasoconstriction stimulus such as hyperventilation, obstruction of venous outflow requires a revision of head position and loosening of tube ties (in cases not complicated by venous sinus thrombosis), and acute hydrocephalus can be managed with CSF drainage through an EVD.16 Our stepwise protocol for controlling ICP is outlined in Table 1. It starts with consideration of decompressive craniectomy and CSF drainage through an EVD because these are the most effective measures for ICP control.17 Details on these surgical options are described in subsequent sections.

Medical management of intracranial hypertension In general, there are four causes of intracranial hypertension following ischemic stroke: (1) brain swelling due to cytotoxic or vasogenic edema; (2) increased cerebral blood volume due to dilated cerebral arteries; (3) obstruction of venous outflow; and (4) acute hydrocephalus.16 The optimal approach to manage intracranial hypertension differs according to the underlying cause. Brain swelling requires osmotherapy and/or decompressive craniectomy, cerebral arterial dilatation requires a vasoconstriction stimulus such as hyperventilation, obstruction of venous outflow requires a revision of head position and loosening of tube ties (in cases not complicated by venous sinus thrombosis), and acute hydrocephalus can be managed with CSF drainage through an EVD.16 Our stepwise protocol for controlling ICP is outlined in Table 1. It starts with consideration of decompressive craniectomy and CSF drainage through an EVD because these are the most effective measures for ICP control.17 Details on these surgical options are described in subsequent sections. General management Assessment and management of the airway, breathing, and circulation is the initial step for treating increased ICP in unconscious stroke patients. Early rapid sequence intubation (RSI) should be considered for patients with comatose mental status and intracranial hypertension, and endotracheal intubation should be performed with medications to blunt any increase in ICP. Details on intubation, ventilatory support, and sedation are described in subsequent sections. Hypoxia and hypotension should be avoided during or after the procedure. Head elevation to 30° has been shown to reduce ICP, although this can be accompanied by a decrease in CPP that offsets the beneficial effects.18 The head should be maintained in a midline position and tight ties around the neck should be avoided to improve jugular venous drainage. Other general measures to treat intracranial hypertension include avoiding hypoxia, hypercapnia and hypotonic conditions, controlling fever and hyperglycemia, and treating seizures. Details on ventilatory support and the strategy for sedation and analgesia are described in subsequent sections.

jugular venous drainage. Other general measures to treat intracranial hypertension include avoiding hypoxia, hypercapnia and hypotonic conditions, controlling fever and hyperglycemia, and treating seizures. Details on ventilatory support and the strategy for sedation and analgesia are described in subsequent sections. Hyperventilation Decreasing PaCO2 to 30-35 mmHg is an effective and rapid means to reduce ICP. A decrease in PaCO2 causes vasoconstriction, which lowers cerebral blood volume and thus lowers ICP. The effect is almost immediate but generally lasts only a few hours because the pH of CSF rapidly equilibrates to the new PaCO2 level.3 However, the effect of hyperventilation may be sustained for days in patients with excessive vasodilation and cerebral hyperemia.19 Prolonged, aggressive hyperventilation can cause cerebral ischemia in patients with brain injury,20 and the routine application of extreme hyperventilation is therefore generally considered harmful. Thus, the use of hyperventilation is best reserved for temporary, rescue therapy for sudden increases in ICP. It is not recommended to routinely hyperventilate patients for many hours or days.

atients with brain injury,20 and the routine application of extreme hyperventilation is therefore generally considered harmful. Thus, the use of hyperventilation is best reserved for temporary, rescue therapy for sudden increases in ICP. It is not recommended to routinely hyperventilate patients for many hours or days. CPP optimization The minimal CPP required to prevent brain ischemia is generally accepted as 50-60 mmHg.3 However, there are two different approaches to whether CPP should be maintained at a higher or lower level. The Rosner concept advocates increasing MAP and targeting a higher CPP to maintain adequate CBF,21 whereas the Lund concept advocates decreasing resistance and intravascular hydrostatic pressure, and reducing cerebral blood volume, thereby increasing CBF and accepting a lower CPP.22 Recently, new approaches to individualize optimal CPP using multimodal neuromonitoring measures and statistics from these measures, such as the pressure reactivity index that is calculated using the moving correlation coefficient between ICP and MAP and the oxygen reactivity index that is calculated using the moving correlation coefficient between CPP and brain tissue oxygen, have been introduced.9,23 Details on different conceptual approaches are beyond the scope of this review.

index that is calculated using the moving correlation coefficient between ICP and MAP and the oxygen reactivity index that is calculated using the moving correlation coefficient between CPP and brain tissue oxygen, have been introduced.9,23 Details on different conceptual approaches are beyond the scope of this review. Osmotherapy Mannitol has been the cornerstone of osmotherapy for the treatment of intracranial hypertension. However, optimal usage of mannitol for stroke patients is uncertain. Usually, this drug is administered intravenously at a dose of 0.25-2.0 g/kg per requirement or every 6 hours. Serum osmolality up to 360 mOsm/kg can be tolerated.3 During osmotherapy with mannitol, it is crucial to maintain a hyperosmotic euvolemic state through adequate fluid balance. If not, the patient could suffer various adverse effects including dehydration, hypovolemia, hypotension, increased ICP, and decreased CPP, in addition to acute kidney injury and electrolyte imbalance. When mannitol needs to be discontinued after repeated use for several days, potential rebound increases in ICP should be monitored with consideration of mannitol tapering. Therapeutic drug monitoring for mannitol is unavailable in most hospitals, but physicians can use the osmolar gap, i.e., the difference between measured serum osmolality and calculated serum osmolarity, as an indirect indicator of serum mannitol levels. There are several approaches to calculate serum osmolarity, but the most simple, frequently used formula is as follows: serum osmolarity=(2×Na)+(glucose/18)+(blood urea nitrogen/2.8). The osmolar gap correlates better than serum osmolality with the serum concentration of mannitol.24 When the osmolar gap is higher than 20 mOsm/L, mannitol is generally withheld to avoid accumulation in the brain and circulation.3 However, 55 mOsm/L has also been suggested as a threshold value.25

od urea nitrogen/2.8). The osmolar gap correlates better than serum osmolality with the serum concentration of mannitol.24 When the osmolar gap is higher than 20 mOsm/L, mannitol is generally withheld to avoid accumulation in the brain and circulation.3 However, 55 mOsm/L has also been suggested as a threshold value.25 Hypertonic saline is an alternative to mannitol for osmotherapy for intracranial hypertension. Although hypertonic saline provides an osmotic effect similar to mannitol, it has a better reflection coefficient (1.0) than mannitol (0.9), and is therefore less able to cross the blood-brain barrier and may have a stronger osmotic effect. Additionally, hypertonic saline normalizes resting membrane potential and has an anti-inflammatory effect. In existing reports, the concentration of hypertonic saline used to treat intracranial hypertension was 1.7%-30%. A recent meta-analysis of five clinical trials showed greater ICP control with hypertonic saline than with mannitol in patients with traumatic brain injury (odds ratio, 1.16; 95% confidence interval, 1.00-1.33).17 Another recent meta-analysis of six studies showed that the percent decrease in ICP from baseline to either 60 minutes or the nadir after administration of 23.4% saline was 55.6% (standard error, 5.90; 95% confidence interval, 43.99-67.12).26 Possible adverse effects of hypertonic saline are rebound cerebral edema, hyperchloremic metabolic acidosis, phlebitis, congestive heart failure, transient hypotension, hemolysis, hypokalemia, renal failure, osmotic demyelination, subdural hemorrhage, seizures, and muscle twitching.27 Various methods have been used to administer hypertonic saline. At the Asan Medical Center, a 60-120 mL dose of 11.7% saline, the highest available osmolar agent in Korea, is administered over 10-20 minutes via a peripherally inserted central catheter as required or every 6 hours. Serum sodium levels up to 155-160 mEq/L are typically well tolerated.

to administer hypertonic saline. At the Asan Medical Center, a 60-120 mL dose of 11.7% saline, the highest available osmolar agent in Korea, is administered over 10-20 minutes via a peripherally inserted central catheter as required or every 6 hours. Serum sodium levels up to 155-160 mEq/L are typically well tolerated. Barbiturate coma therapy Barbiturate coma therapy is reserved for patients with refractory intracranial hypertension and can be delivered using pentobarbital or thiopental.28 These drugs are titrated on the basis of ICP measurements and continuous electroencephalogram monitoring. Pentobarbital is administered at a loading dose of 5 mg/kg over 15-30 minutes (up to 50 mg/min) and then continuously infused at 1-5 mg/kg/h. Barbiturate coma therapy requires electroencephalograms to be continuously recorded and the barbiturate titration usually targets a burst-suppression pattern with approximately 6-8 seconds interbursts.19 Side effects of pentobarbital and thiopental include hypotension, myocardial suppression, respiratory suppression, infections, hepatic and renal dysfunction, thrombocytopenia, metabolic acidosis and gastric stasis. However, randomized clinical trials on the effect of barbiturate therapy for uncontrolled ICP have not been performed in stroke patients. Barbiturate coma therapy is the last resort for treatment of refractory intracranial hypertension due to a lack of sufficient evidence and various systemic side effects associated with it, including cardiac depression, arterial hypotension, and increased risk of infection.29

have not been performed in stroke patients. Barbiturate coma therapy is the last resort for treatment of refractory intracranial hypertension due to a lack of sufficient evidence and various systemic side effects associated with it, including cardiac depression, arterial hypotension, and increased risk of infection.29 Surgical management of brain swelling Medical management of brain edema and increased ICP may not be successful when significant swelling occurs in the cerebral or cerebellar hemisphere. Therefore, a decompressive craniectomy should be considered to relieve the mass effect of the swollen cerebral or cerebellar hemisphere on the thalamus, brainstem, and network projections to the cortex, manifested predominantly by decreased levels of arousal.5

t swelling occurs in the cerebral or cerebellar hemisphere. Therefore, a decompressive craniectomy should be considered to relieve the mass effect of the swollen cerebral or cerebellar hemisphere on the thalamus, brainstem, and network projections to the cortex, manifested predominantly by decreased levels of arousal.5 Malignant MCA infarction Malignant MCA infarction was originally defined as an acute infarction in the entire MCA territory evident on a computed tomography scan within the first 48 hours after symptom onset, with or without involvement of other vascular (anterior or posterior cerebral artery) territories.30 Subsequently, the term has been used to refer to large hemispheric infarcts that have occurred as a result of occlusion of the proximal MCA or internal carotid artery, with variable definitions (e.g., National Institutes of Health Stroke Scale score ≥15-20; brain computed tomography ischemic signs involving >50%-66% of the MCA territory; and diffusion-weighted imaging infarct volume >145 cm3).31,32,33,34,35 Malignant MCA infarction accounts for approximately 5% of ischemic strokes and is most commonly caused by cardioembolism.30,36 Despite the best medical treatment, most patients deteriorate between post-ictus days 2 and 5, and malignant MCA infarction is characterized by severe morbidity and high mortality.30,37 Mortality ranges up to 70%-80% with medical treatment.38 For patients who present outside thrombolysis time windows or who already have areas of marked low density on computed tomography scans, treatment should aim to minimize brain swelling and control ICP. Among the variety of treatment modalities available to treat an increase in ICP, the most effective method is decompressive hemicraniectomy.17 In a meta-analysis of studies into the control of increased ICP, decompressive hemicraniectomy was found to cause a 19 mmHg mean reduction in pressure, which was superior to the reduction in ICP seen with the use of hyperventilation (6 mmHg), mannitol (8 mmHg), barbiturates (8 mmHg), hypothermia (10 mmHg), hypertonic saline (15 mmHg), or CSF drainage (15 mmHg).17

l of increased ICP, decompressive hemicraniectomy was found to cause a 19 mmHg mean reduction in pressure, which was superior to the reduction in ICP seen with the use of hyperventilation (6 mmHg), mannitol (8 mmHg), barbiturates (8 mmHg), hypothermia (10 mmHg), hypertonic saline (15 mmHg), or CSF drainage (15 mmHg).17 In decompressive hemicraniectomy, a bone flap over the frontal, temporal, parietal, and occipital lobe at the site of the infarct is removed, allowing the swollen brain tissue to expand extracranially (Figure 2). Decompressive hemicraniectomy with duroplasty can relieve horizontal and vertical tissue shifts, reduce ICP, improve CPP, and CBF, and alleviate vascular compression. During the last decade, several randomized clinical trials have shown that this procedure is most effective when performed within 48 hours post-ictus (Table 2).31,32,33,34 A hemicraniectomy performed too soon after ictus may result in unnecessary surgical procedures and complications, whereas a hemicraniectomy performed after herniation signs are fully developed may be of little value. Multiple randomized clinical trials have investigated if early (<48 hours post-ictus) hemicraniectomy improves clinical outcomes for patients with malignant MCA infarction. Trials included young patients (age ≤60 years) with severe symptoms (National Institutes of Health Stroke Scale score ≥20 for the dominant hemisphere and ≥15 for the nondominant hemisphere) and a decreased level of consciousness (score ≥1 on item 1a of the National Institutes of Health Stroke Scale score or a gradual decrease in consciousness to a score ≤13 on the Glasgow coma scale for right-sided lesions or ≤9 for left-sided lesions). In a pooled analysis of three European trials, the number needed to treat for survival at 1 year was two and the number needed to treat for independent walking (modified Rankin Scale (mRS) score ≤3) was four.33

consciousness to a score ≤13 on the Glasgow coma scale for right-sided lesions or ≤9 for left-sided lesions). In a pooled analysis of three European trials, the number needed to treat for survival at 1 year was two and the number needed to treat for independent walking (modified Rankin Scale (mRS) score ≤3) was four.33 Despite unequivocal effects on survival, there has been criticism that hemicraniectomy can barely avert death to a vegetative or minimally conscious state. However, patients with malignant MCA infarctions who may have been in a vegetative state (mRS score=5) without hemicraniectomy could walk with some help (mRS=4) after hemicraniectomy. A recent meta-analysis showed that the vast majority of patients who underwent hemicraniectomy and their caregivers were satisfied with the outcome of surgery and would provide consent again for the procedure despite the high rate of physical disability and depression.39 Patients with an infarct in the dominant hemisphere may have a lower quality of life than patients with an infarct in the nondominant hemisphere because they have more severe language impairment. However, a meta-analysis found no differences in functional outcome between right- and left-hemispheric infarcts.40

ssion.39 Patients with an infarct in the dominant hemisphere may have a lower quality of life than patients with an infarct in the nondominant hemisphere because they have more severe language impairment. However, a meta-analysis found no differences in functional outcome between right- and left-hemispheric infarcts.40 Until recently, the efficacy of decompressive hemicraniectomy in patients 60 years of age or older has been uncertain. A recent randomized clinical trial showed that malignant MCA infarction patients aged >60 years who were treated with early (<48 hours post-ictus) hemicraniectomy had a higher survival rate and better functional outcome than patients who were managed conservatively (Table 2).35 Thus, old age per se should not be regarded as an exclusion criterion for hemicraniectomy after malignant MCA infarction. Preemptive decompressive hemicraniectomy is an effective treatment for brain swelling in patients with malignant MCA infarction. Cerebellar infarction A space-occupying cerebellar infarct may cause brain death through brainstem compression and obstructive hydrocephalus. Large infarct size on initial magnetic resonance images, mostly in the territories of the posterior inferior cerebellar artery and superior cerebellar artery rather than the anterior inferior cerebellar artery, can suggest poor prognosis.41 Consciousness typically deteriorates between post-ictus days 2 and 4,42 and surgical options should be considered when maximal medical therapy is failing.

itories of the posterior inferior cerebellar artery and superior cerebellar artery rather than the anterior inferior cerebellar artery, can suggest poor prognosis.41 Consciousness typically deteriorates between post-ictus days 2 and 4,42 and surgical options should be considered when maximal medical therapy is failing. One surgical option for treating a space-occupying cerebellar infarct is a suboccipital craniectomy, where the skull is removed and the dura expanded to relieve ICP caused by the swollen brain tissue (Figure 3). Although the efficacy of suboccipital craniectomy has been reported in observational and retrospective studies,42,43,44 no randomized clinical trials have been performed to test the efficacy of this procedure. The value of preemptive surgery (early surgery based on radiological findings such as cerebellar edema and hydrocephalus in a clinically stable patient) and the best neurosurgical approach (removal of necrotic tissue vs. decompression alone vs. decompression plus ventriculostomy) are unknown.5,45,46 In patients with acute hydrocephalus following cerebellar infarction, placement of an EVD (ventriculostomy) should be accompanied by suboccipital craniectomy.5

ble patient) and the best neurosurgical approach (removal of necrotic tissue vs. decompression alone vs. decompression plus ventriculostomy) are unknown.5,45,46 In patients with acute hydrocephalus following cerebellar infarction, placement of an EVD (ventriculostomy) should be accompanied by suboccipital craniectomy.5 Therapeutic hypothermia Clinical studies on therapeutic hypothermia have investigated the effects on two different outcomes: neuroprotection and ICP control. Therapeutic hypothermia is potentially neuroprotective through multiple mechanisms. Hypothermia reduces cerebral metabolism and consumption of oxygen and glucose by the brain, thereby preventing the failure of sodium transport and calcium influx, and decreasing the risk of cell death.47 Additional beneficial effects of hypothermia include the inhibition of excitatory neurotransmitters, free radicals, inflammation, and apoptosis. Reduced ICP is a robust manifestation of hypothermia and is driven by reductions in cerebral blood volume, vasogenic edema and inflammation, and mitigation of blood-brain barrier leakage.47

eficial effects of hypothermia include the inhibition of excitatory neurotransmitters, free radicals, inflammation, and apoptosis. Reduced ICP is a robust manifestation of hypothermia and is driven by reductions in cerebral blood volume, vasogenic edema and inflammation, and mitigation of blood-brain barrier leakage.47 In stroke patients, therapeutic effects of hypothermia are equivocal despite robust benefits in animal models and preclinical trials. Small pilot studies suggest possible benefits of therapeutic hypothermia in stroke (Table 3) and observational studies suggest that therapeutic hypothermia is a potent anti-edema strategy. Induced hypothermia with a target temperature of 32-33℃ helped control ICP elevation due to cerebral edema in patients with malignant MCA infarction,48,49 and induced hypothermia with a target temperature of 34.5℃ reduced cerebral edema and hemorrhagic transformation after successful recanalization for acute ischemic stroke.50 An anti-edema effect of hypothermia at 35℃ for 8-10 days has also been demonstrated in patients with intracerebral hemorrhage.51,52 ICP reduction does not necessarily equal improved clinical outcome. Nevertheless, therapeutic hypothermia is effective for ICP management, at least in selected patients (Figure 3). Currently, two major clinical trials of therapeutic hypothermia in patients with stroke are underway: The Intravenous Cooling in the Treatment of Stroke 2/3 (ICTuS 2/3) is a phase II-III trial of therapeutic hypothermia at 33℃ for 24 hours in 1,600 stroke patients, and EuroHYP-1 is a phase III trial of therapeutic hypothermia at 34-35℃ for 24 hours in 1,500 stroke patients.53 Despite the current lack of evidence from clinical trials on the neuroprotective effect of therapeutic hypothermia in stroke patients, the effect of therapeutic hypothermia on ICP reduction is unequivocal in patients with brain edema irrespective of the cause.54

hermia at 34-35℃ for 24 hours in 1,500 stroke patients.53 Despite the current lack of evidence from clinical trials on the neuroprotective effect of therapeutic hypothermia in stroke patients, the effect of therapeutic hypothermia on ICP reduction is unequivocal in patients with brain edema irrespective of the cause.54 In practice, therapeutic hypothermia can be separated into three phases: induction, maintenance, and rewarming. Induction is typically performed as rapidly as possible to a target temperature of 32-35℃. For this purpose, ice packs and 4℃ cold saline can be combined with a commercialized cooling device. During the maintenance period, only minor fluctuations in patients' body temperature are allowed. The rewarming period is very important because rebound edema and ICP elevation may occur during this period.48 To avoid such detrimental effects, rewarming rates should be slow and are usually targeted at 0.1-0.25℃/h. Rebound hyperthermia after rewarming should also be avoided. Further clinical studies are required to determine the optimal therapeutic time window, target temperature, duration of cooling, and rewarming rate of therapeutic hypothermia for patients with stroke.

es should be slow and are usually targeted at 0.1-0.25℃/h. Rebound hyperthermia after rewarming should also be avoided. Further clinical studies are required to determine the optimal therapeutic time window, target temperature, duration of cooling, and rewarming rate of therapeutic hypothermia for patients with stroke. The prevention and management of various complications of therapeutic hypothermia is challenging. The most frequently encountered and cumbersome complication is shivering, especially when surface cooling is used rather than endovascular cooling.55 Shivering not only counteracts the cooling process but also has a harmful impact on systemic oxygen consumption, brain tissue oxygenation, and ICP.56,57 Our anti-shivering protocol developed according to a literature review and our clinical experience is shown in Table 4.58 The masking of infection-induced fever is also a problem associated with therapeutic hypothermia. We have experienced some patients whose circulating water temperature in the cooling device could be used as a surrogate marker for increased heat production by the patient. Regular follow-ups of chest radiographs and laboratory tests including complete blood cell counts, C-reactive protein, procalcitonin, urinalysis, and surveillance cultures may also help early detection of infection. Other major complications of therapeutic hypothermia include pneumonia, pulmonary edema, hypotension, decreased cardiac output, bradycardia, conduction blocks, ileus, hepatic dysfunction, pancreatitis, hyperglycemia, cold diuresis, hypokalemia, hypomagnesemia, hypophosphatemia, hypocalcemia, alkalosis, hypocarbia, impaired immune function, bleeding diathesis, and alterations in pharmacokinetics and pharmacodynamics. Physicians utilizing therapeutic hypothermia should be familiar with the prevention and management of such complications.

diuresis, hypokalemia, hypomagnesemia, hypophosphatemia, hypocalcemia, alkalosis, hypocarbia, impaired immune function, bleeding diathesis, and alterations in pharmacokinetics and pharmacodynamics. Physicians utilizing therapeutic hypothermia should be familiar with the prevention and management of such complications. Ventilatory support Most stroke patients with mild respiratory difficulties can be managed with supportive care such as insertion of an oral airway, supplemental oxygen, and antibiotics for aspiration pneumonia. However, patients who are stuporous or comatose following massive stroke can develop reduced oropharyngeal muscle tone that leads to posterior displacement of the tongue and airway obstruction. A decrease in or absence of gag and cough reflexes and impaired immune function further increase the risk of aspiration. Therefore, these patients are vulnerable to hypoxia, aspiration pneumonia, and respiratory distress and arrest unless preemptive endotracheal intubation and mechanical ventilation are applied. For safe intubation, RSI is necessary. RSI involves prompt induction of sedation and subsequent endotracheal intubation. Etomidate is the preferred induction agent for stroke patients because of its short-acting characteristics and hemodynamic stability. However, it is not recommended for patients with seizures or adrenal insufficiency. Other induction agents include propofol and midazolam. Succinylcholine is the most commonly administered neuromuscular blocking agent for RSI as a result of its rapid onset and short duration of action. However, a brief increase in ICP has been reported following administration of succinylcholine.59 Other possible side effects of succinylcholine are malignant hyperthermia, hyperkalemia, and aggravation of neuropathy or myopathy.59 For this reason, some experts recommend non-depolarizing neuromuscular blocking agents such as cisatracurium and rocuronium for RSI.59 Fentanyl can reduce the increase in blood pressure and ICP that occurs during intubation and is usually administered to stroke patients undergoing RSI, whereas lidocaine minimizes coughing and the increase in ICP.60

recommend non-depolarizing neuromuscular blocking agents such as cisatracurium and rocuronium for RSI.59 Fentanyl can reduce the increase in blood pressure and ICP that occurs during intubation and is usually administered to stroke patients undergoing RSI, whereas lidocaine minimizes coughing and the increase in ICP.60 Intubated patients usually require mechanical ventilation. Inappropriate management of mechanically ventilated patients with space-occupying stroke can lead to lung damage as well as brain damage. Injuries to the lungs are attributable to a high end-inspiratory and low end-expiratory lung volume which results in repeated collapse and re-expansion. A high shear force is exerted on alveolar walls and small airways during inflation, especially at the interfaces between collapsed and aerated alveoli.61 Application of positive end-expiratory pressure (PEEP) can avoid repeated alveoli collapse. However, the effect of PEEP on cerebral physiology is controversial. PEEP can elevate ICP through increased thoracic pressure and decreased MAP, venous return, and cardiac output. One study increased PEEP up to 12 cmH2O in patients with acute stroke, but ICP remained unchanged or slightly reduced.62 There was marked decrease in CPP caused by the decrease in MAP. In general, increasing PEEP up to 20 cmH2O does not have deleterious effects on ICP as long as the baseline ICP is not high (<20 mmHg).63 The benefits of PEEP may outweigh the risks of hypoxemia in patients with stroke. A strategy of maintaining a low tidal volume and limiting plateau pressure is also necessary to restrict end-inspiratory overexpansion of alveoli. Tidal volume should be limited to 6-8 mL/kg of predicted body weight, especially for patients with acute respiratory distress syndrome (ARDS).64 However, low tidal volume reduces carbon dioxide elimination via the lungs and can cause hypercapnia and respiratory acidosis. Subsequent hyperventilation stimulates the brainstem respiratory center and may cause ventilator dyssynchrony and require heavy sedation or neuromuscular blockade. Hypercapnia causes intracranial hypertension by dilating cerebral vessels and increasing CBF.

the lungs and can cause hypercapnia and respiratory acidosis. Subsequent hyperventilation stimulates the brainstem respiratory center and may cause ventilator dyssynchrony and require heavy sedation or neuromuscular blockade. Hypercapnia causes intracranial hypertension by dilating cerebral vessels and increasing CBF. However, the risk of increased ICP as a consequence of permissive hypercapnia has not been studied in stroke patients.65 Pumpless extracorporeal lung assist combined with lung-protective ventilation may be used as an adjunctive therapy in patients with both ICP crisis and hypercapnia caused by low tidal volume.66 This can be used to address hypercapnia-associated increases in ICP, and the minute volume of the ventilator can be reduced to avoid hyperinflation of alveoli.

st combined with lung-protective ventilation may be used as an adjunctive therapy in patients with both ICP crisis and hypercapnia caused by low tidal volume.66 This can be used to address hypercapnia-associated increases in ICP, and the minute volume of the ventilator can be reduced to avoid hyperinflation of alveoli. Sedation and analgesia Sedation and analgesia are often required in patients with malignant brain edema following stroke. They can be used to manage ICP and CPP as well as attenuate the stress response, increase endotracheal tube tolerance, reduce metabolic energy demands, prevent delirium and decrease patient-ventilator synchrony.67 In the neurological ICU, sedation and analgesia are typically induced by continuous infusion of fentanyl, remifentanil, propofol, midazolam, or dexmedetomidine. Clinical studies that compare the efficacy of these agents are lacking in patients with stroke. In general, the different intravenous opioids are equally effective when titrated to similar pain intensity endpoints.68 Fentanyl, a µ-receptor agonist, is popular in the ICU as a result of its strong analgesic effect, lack of neurotoxicity, rapid onset of action, and short elimination half-life. Remifentanil, an opioid agent that is chemically related to fentanyl, may be preferred to fentanyl as a result of its shorter elimination half-life and its lack of accumulation in patients with hepatic or renal failure. A short context-sensitive half-time-the time required for blood or plasma concentrations of a drug to decrease by 50% after discontinuation of drug administration-is an attractive characteristic of remifentanil when frequent wake-up tests for neurological evaluation and ventilator weaning are planned. Propofol has sedative, hypnotic, anxiolytic, amnestic, antiemetic, and anticonvulsant properties, but no analgesic effects.69 The rapid onset and short duration of action of propofol are advantageous if frequent wake-up tests are required and this agent may have stronger suppressive effects on cerebral metabolism than midazolam.70 However, propofol may lead to dose-dependent respiratory depression and hypotension.

but no analgesic effects.69 The rapid onset and short duration of action of propofol are advantageous if frequent wake-up tests are required and this agent may have stronger suppressive effects on cerebral metabolism than midazolam.70 However, propofol may lead to dose-dependent respiratory depression and hypotension. When administered at a high dose for a prolonged period (>4 mg/kg/h for >48 h), propofol may cause propofol infusion syndrome, which is characterized by various early signs including lactic acidosis, triglyceridemia, bradycardia, and Brugada-like electrocardiogram changes, and later signs including cardiac failure, tachyarrhythmia or heart block, ventricular tachycardia, rhabdomyolysis, hyperkalemia, renal failure, and fatty degeneration of the liver.71 Among the benzodiazepines, midazolam is often used for continuous infusion in the ICU as it has a faster onset, more rapid clearance, and shorter duration of action than lorazepam. However, with prolonged use, it exhibits greater variability and longer time to awakening than lorazepam. It also contains propylene glycol, which can cause metabolic acidosis and acute kidney injury.72 Dexmedetomidine is a selective alpha 2-receptor agonist with sedative, analgesic, and sympatholytic properties, but no anticonvulsive properties.73 Patients sedated with this agent are more easily aroused and more interactive than patients sedated with other drugs.74 Another characteristic of dexmedetomidine that is useful in the neurological ICU is the minimal respiratory depression and anti-shivering effects. In mechanically ventilated adult ICU patients at risk of developing delirium, dexmedetomidine infusion for sedation was associated with a lower prevalence of delirium than benzodiazepine infusion.75 However, dexmedetomidine frequently causes bradycardia, hypotension, and hypertension, especially after the loading dose. For this reason, we prefer starting with a maintenance dose of 0.5 µg/kg/h without a loading dose. Dexmedetomidine is becoming increasingly popular in the neurological ICU. Physicians who manage stroke patients in the ICU should be familiar with the characteristics of each sedative and analgesic agent (Table 5).68