Association between Seizures and Outcomes among Intracerebral Hemorrhage Patients: The China National Stroke Registry Zixiao Li, MD,*† Xingquan Zhao, MD, PhD,*† Yilong Wang, MD, PhD,*† Chunxue Wang, MD, PhD,*‡ Liping Liu, MD, PhD,*‡ Xiaoqiu Shao, MD, PhD,* Wenjuan Wang, MD, PhD,*x Yuesong Pan, MD,*‡ Chunjuan Wang, MD, PhD,*x Ruijun Ji, MD, PhD,* Changqing Zhang, MD,* Jing Jing, MD, PhD,* and Yongjun Wang, MD*†

Background: To determine whether the presence of seizures in patients with spontaneous intracerebral hemorrhage (ICH) was associated with in-hospital complications and measured outcomes. Methods: This prospective cohort study from the China National Stroke Registry included consecutive patients with ICH between August 2007 and September 2008. In-hospital complications, functional outcomes, and mortality at 3, 6, and 12 months were compared between patients with seizures and those without seizures occurring at ICH onset and during hospitalization. Poor functional outcome was defined as a modified Rankin Scale score between 3 and 6. Poor functional outcome and mortality were stratified by stroke severity using Glasgow Coma Scale scores on admission. Results: The study included 3216 patients with ICH and 139 of them (4.3%) experienced seizures. The presence of seizures was associated with high in-hospital complications including atrial fibrillation (P 5 .004), pneumonia (P 5 .001), as well as lower rehabilitation assessment rates (P 5 .033) compared with patients without seizures. ICH patients with seizures had poorer functional outcome at 3-month (P 5 .012), 6-month (P 5 .007), and 12-month (P 5 .001) follow-up. They also had higher mortality at 3 months (P 5 .045), 6 months (P 5 .005), and 12 months (P 5 .002). Patients with mild strokes had poorer functional outcome and higher mortality (P , .005) if seizures occurred.

From the *Department of Neurology, Beijing TianTan Hospital, Capital Medical University, Beijing; †China National Clinical Research Center for Neurological Diseases, Beijing; ‡Center of Stroke, Beijing Institute for Brain Disorders, Beijing; and xBeijing Key Laboratory of Translational Medicine for Cerebrovascular Disease, Beijing, China. Received August 29, 2014; revision received September 4, 2014; accepted September 16, 2014. Z.L. and X.Z. were responsible for literature review, data collection, data analysis, drafting the article, and review of submitted versions; Y.W. was responsible for research design, study conception, and data collection; C.W. and X.S. were responsible for draft revisions and review of submitted versions; L.L. was responsible for study conception; W.W., C.W., C.Z., and J.J. were responsible for data collection; Y.P. conducted statistical consultation and data analysis; R.J. was responsible for literature review; Y.W. was responsible for research design and concept development, critical revisions, and approval of submitted versions.

Supported by grants from the Ministry of Science and Technology and the Ministry of Health of the People’s Republic of China (National S & T Major Project of China: 2006BA101A11 and 2011BAI08B02), the National Key Technology Research and Development Program of the Ministry of Science and Technology of China (2013BAI09B03), and Beijing Institute for Brain Disorders (BIBDPXM2013_014226_07_000084). All authors report no potential conflict of interest relevant to this article. Address correspondence to YongJun Wang, MD, Department of Neurology, Beijing TianTan Hospital, Capital Medical University, No. 6 Tiantan Xili, Dongcheng District, Beijing, China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.09.021

Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 2 (February), 2015: pp 455-464

455

Z. LI ET AL.

456

Conclusions: The presence of seizures in patients with ICH was associated with high in-hospital complications and indicates poor outcomes at 3-, 6-, and 12-month follow-up. Quality improvement strategies targeting ICH patients with seizures especially mild stroke may help improve prognoses. Key Words: Intracerebral hemorrhage—all epilepsy/seizures—outcome research—risk factors. Ó 2015 by National Stroke Association

Clinical seizures after spontaneous intracerebral hemorrhage (ICH) are encountered in clinical practice relatively frequently with incidence varying from 2.7% to 17%.1-6 However, the impact of seizures on stroke outcome remains a matter of debate. In several prospective and population-based studies, clinical seizures did not have a negative impact on neurologic outcomes or mortality.3,5–8 On the contrary, findings from 1 study showed that post-ICH seizures were associated with worsening neurologic function and a trend toward poorer outcomes.9 Interestingly, most of the available data came from cohorts with stroke including ischemic and hemorrhagic stroke, which are 2 very different diseases with different prognoses. Additionally, not many previous studies took into account patients’ and health care providers’ perspectives, particularly on resource utilization, performance measures,10 and influence of seizures on stroke outcome.11 So, we hypothesized the following: (1) the presence of seizures in patients with ICH might be associated with higher in-hospital complications and (2) the presence of seizures might have a negative impact on short- and long-term stroke outcomes. The primary aim of this study was to assess the association between seizure occurrence and outcomes after ICH stroke, including in-hospital complications, short- and long-term functional outcomes, and mortality.

Methods This study was conducted from the largest stroke registry in China, the China National Stroke Registry (CNSR), which is a national hospital-based, multicenter, and prospective registry of consecutive patients with acute cerebrovascular events sponsored by the Ministry of Health between August 2007 and September 2008. A total of 132 hospitals were selected as registry hospitals. The study was approved by the central Institutional Review Board at Beijing Tiantan Hospital, and all patients or their designated relatives gave informed consents. Further details of CNSR have been published elsewhere.12 The current analysis included all patients in CNSR who had been diagnosed with ICH according to World Health Organization criteria.13 Patients were excluded if they met any following criteria: history of seizures, primary intraventricular hemorrhage or prestroke modified Rankin Scale (mRS) of more than 2, neoplastic cause of ICH,

unavailable data of ICH hematoma volume, disagreement to participate in the study, or loss to follow-up.14

Demographic Characteristics and Medical History In the present study, we prospectively collected demographic information: age (stratification: ,60, 60-79, and $80 years of age), gender, and vascular risk factors: history of stroke, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, cardiovascular disease, current smoking, moderate or heavy alcohol consumption (2 standard alcoholic beverages per day).14

Clinical Evaluation at Admission We assessed the severity of neurologic impairment by using the National Institutes of Health Stroke Scale (NIHSS; 3 categories: mild [0-3], moderate [4-14], and severe [15-42]) and Glasgow Coma Scale (GCS) at admission (3 categories: mild [13-15], moderate [9-12], and severe [3-8]).

Image Analysis Initial noncontrast computed tomography (NCCT) scans were employed with a slice thickness of 9 mm supratentorially and 4.5 mm infratentorially. ICH hematoma volume was measured on the initial head NCCT scan using the ABC/2 method.15 Additionally, the locations of ICH were considered as follows: (1) strictly deep locations (putamen, caudate nucleus, internal capsule, thalamus, brain stem, and cerebellar); (2) deep and lobe locations; and (3) strictly lobe locations (frontal, parietal, temporal, and occipital). The presence or absence of intraventricular extension was also noted on the initial head NCCT. All images were prospectively viewed by a neuroradiologist blinded to clinical data at different study centers. The neuroradiologists of the study centers received the same training regarding the computed tomography protocol.

Definition of Seizures The status of seizures was based on clinical diagnosis. According to International League Against Epilepsy criteria, seizures were defined as paroxysmal disorders of the central nervous system, possible subsequent loss of consciousness, and/or with or without motor involvement at any point since stroke onset or during the period

SEIZURES AND OUTCOMES IN HEMORRHAGE STROKE

457

Table 1. Definition of in-hospital medical complications after intracerebral hemorrhage Complications Hematoma expansion Myocardial infarction Atrial fibrillation/flutter Pneumonia

Gastrointestinal bleeding Urinary tract infection Deep vein thrombosis Decubitus ulcer

Definition An increase in volume of 33%-50% or an absolute change in hematoma volume of 12$5-20 mL (on repeat computed tomography). Electrocardiography (ECG) or laboratory evidence of acute myocardial infarction. Intermittent or new persistent atrial fibrillation/flutter on ECG or ECG monitoring not present on admission. Clinical and laboratory indices of respiratory tract infection (fever, cough, auscultatory respiratory crackles, new purulent sputum, or positive sputum culture), and supported by typical chest X-ray findings. Clinical (any episode of fresh blood or coffee ground emesis, hematemesis, melena, or hematochezia), laboratory or radiographic evidence of gastrointestinal bleeding. Clinical symptoms of urinary tract infection combined with a positive urine examination or culture. Clinical diagnosis of deep vein thrombosis and with radiographic evidence. Any skin breaks or necrosis resulting from either pressure or trivial trauma.

of hospitalization.16,17 At admission, physicians actively inquired about seizures occurring at onset of ICH and observed all seizures that occurred during hospitalization.

Measurements (1) Recorded in-hospital medical complications: hematoma expansion, atrial fibrillation (AF), urinary tract infection (UTI), decubitus ulcer, myocardial infarction (MI), deep venous thrombosis (DVT), pneumonia, and gastrointestinal bleeding (GIB) (specific definitions listed in Table 1)18,19; (2) Medical resource utilization: admission department (neurological ward or others, stroke unit, and intensive care unit [ICU]/neurologic ICU [NICU]), length of stay in the hospital or ICU, and supporting interventions (nasal feeding, central venous catheterization, tracheal intubation, and mechanical ventilation). Management by stroke unit, NICU, and ICU was defined as when a patient was admitted to a stroke unit, NICU, or ICU at any point during the hospitalization; (3) Performance measures: antihypertensive therapy for patients whose blood pressure was more than 140/90 mm Hg during hospitalization, dysphagia screening before any oral intake, and rehabilitation evaluation during hospitalization10; (4) Mortality and functional outcomes: at 3, 6, and 12 months after ICH onset, patients or their authorized caregivers were contacted through telephone by trained research personnel who followed standardized scripts for follow-up. Poor functional outcome (dependency or death) was defined as mRS of 3-6 scores.20 We compared mortality and functional outcomes for patients with and without seizures and stratified according to stroke severity based on GCS.

Statistical Analysis Categorical variables were summarized as proportion; continuous variables were summarized with median and interquartile range. In univariate analysis, the c2 or Fisher exact test was used to compare categorical variables, and the Student t test or Mann–Whitney test was used to compare continuous variables. Impact of Seizures on Mortality and Functional Outcomes The mortality and poor functional outcome at 3, 6, and 12 months after ICH onset in patients with and without seizures and stratified according to stroke severity based on GCS were compared using the c2 test or Fisher exact test. A logistic model was employed to identify risk factors associated with mortality and poor functional outcome at 3, 6, and 12 months after ICH onset. Variables were selected using stepwise regression. These variables included the presence of seizures, age at admission, gender, NIHSS and GCS at admission, vascular factors of stroke, diabetes, hypertension, dyslipidemia, AF, cardiovascular disease, current smoking; and moderate or heavy alcohol consumption, the volume and locations of ICH and intraventricular extension, in-hospital complications of hematoma expansion, AF, MI, UTI, decubitus ulcer, DVT, pneumonia, and GIB, performance measures of antihypertensive therapy, dysphagia screening, and rehabilitation evaluation. All variables with a P less than .2 (equivalent to maximizing the information criteria) in the univariate analysis were considered in the multivariable regression model. Factors Associated with the Occurrence of Seizures Multivariable logistic regression was performed to evaluate relative factors of the occurrence of seizures. Variables in this model were as previously mentioned except for performance measures.

458

All tests were 2 tailed, and statistical significance was determined at a level of .05. Statistical analysis was performed using SAS version 9.2 (SAS Institute, Cary, NC).

Results Study Population From September 2007 to August 2008, 22,216 hospitalized patients with acute cerebrovascular events were consecutively recruited in the CNSR. Among them, 5136 patients were finally diagnosed as acute ICH. The following patients were excluded from this study: (1) 138 patients with primary intraventricular hemorrhage; (2) 881 patients with lack of data of hematoma volume; (3) 12 ICH patients with intracranial tumor; (4) 261 patients with a prestroke mRS more than 2; (5) 308 patients who were not willing to participate in this study; (6) 39 patients who had a history of seizures; and (7) 281 patients lost to follow-up. Finally, our cohort consisted of 3216 patients without a history of seizures (1241 [38.6%] female) with median age of 62 years (interquartile range [IQR], 53-72 years). The median length between onset symptoms and discharge was 19 days (IQR, 13-27 days), the median NIHSS score was 9 (IQR, 4-16), the median GCS score was 14 (IQR, 9-15), and the median volume of ICH was 12.5 mL (IQR, 5.4-28.0).

Demographic and Clinical Characteristics of Patients with or without Seizures Among 3216 consecutive patients with ICH, 139 (4.3%) had seizures. Their median age was 62 years (IQR, 52-73), and 57 (41.0%) patients with seizures were female, which had no difference from those without seizures. Patients with seizures had higher NIHSS (16, [IQR, 7-30]; P , .001) and lower GCS (8, [IQR, 6-14]; P , .001). We did not find difference in vascular risk factors (Table 2).

Imaging Features of Patients with or without Seizures ICH patients experiencing seizures had a larger hemorrhage volume (16.9 [IQR, 9-36]; P 5 .005) and significant difference of strictly deep locations, deep and lobe locations, and strictly lobe locations (P , .001), including higher strictly lobar involvement rate (42.4% vs. 15.2%) and lower strictly deep involvement rate (39.6% vs. 67.2%; Table 1).

In-hospital Complications with or without Seizures Patients with seizures had more complications during hospitalization including AF (6.5% vs. 2.1%, P 5 .004) and pneumonia (29.5% vs. 17.4%, P 5 .001) than those without seizures. We did not find a difference regarding hematoma expansion, UTI, MI, DVT, or GIB (Table 3).

Z. LI ET AL.

Medical Resource Utilization and Performance Measures with or without Seizures There was no difference between patients with or without seizures in terms of ICU/NICU admission and length of hospital or ICU/NICU stay. Patients with seizures had higher rates of using supporting interventions, such as nasal feeding (30.9% vs. 14.4%, P , .001) and tracheal intubation (10.8% vs. 5.0%, P 5 .006) than those without seizures. Regarding performance measures, patients with seizures underwent fewer rehabilitation assessments than those without seizures (30.2% vs. 39.3%, P 5 .033). There was a relatively low adherence rate, but no differences between patients with and without seizures in the number of patients who underwent antihypertensive therapy and screening for dysphagia (Table 3).

Univariate and Multivariable Analysis on 3-, 6-, and 12-Month Functional Outcomes and Mortality after ICH Onset In univariate analysis, patients with seizures had higher rates of poor functional outcome at 3 months (odds ratio [OR], 2.51; confidence interval [CI], 1.73-3.63; P , .001), 6 months (OR, 2.63; CI, 1.83-3.8; P , .001) and 12 months (OR, 2.95; CI, 2.04-4.28; P , .001) after ICH onset (Table 4) than those without seizures. Multivariable logistic analysis adjusted for potential confounding variables in the entire cohort revealed that the presence of seizures was associated with the higher rate of poor functional outcome at 3-month (adjusted OR, 1.9; CI, 1.15-3.14; P 5 .012), 6-month (adjusted OR, 1.94; CI, 1.19-3.15; P 5 .007), and 12-month (adjusted OR, 2.27; CI, 1.41-3.65; P 5 .001) mortality after ICH onset (Table 3). In univariate analysis, patients with seizures had higher mortality at 3 months (OR, 2.64; CI, 1.85-3.76; P , .001), 6 months (OR, 2.89; CI, 2.05-4.09; P , .001), and 12 months (OR, 2.9; CI, 2.06-4.08; P , .001) after ICH onset (Table 3) than those without seizures. Multivariable logistic analysis adjusted for potential confounding variables in the entire cohort revealed that the presence of seizures remained an independent predictor of 3month (adjusted OR, 1.6; CI, 1.01-2.53; P 5 .045), 6month (adjusted OR, 1.88; CI, 1.21-2.94; P 5 .005), and 12-month (adjusted OR, 1.97; CI, 1.27-3.05; P 5 .002) mortality after ICH onset (Table 4).

Comparisons of Functional Outcomes and Mortality by ICH Severity For patients with a GCS score of 13-15, those with seizures had poorer functional outcomes at 3 months (53.6% vs. 30.8%, P 5 .001), 6 months (51.8% vs. 28.9%, P , .001), and 12 months (55.4% vs. 29.0%, P , .001) after ICH onset than those without seizures and had higher mortality at 3 months (19.6% vs. 6.8%, P 5 .002), 6 months (21.4% vs. 8.3%, P 5 .002), and 12 months (26.8% vs.

SEIZURES AND OUTCOMES IN HEMORRHAGE STROKE

459

Table 2. Baseline characteristics and imaging features of patients with spontaneous intracerebral hemorrhage with or without seizures Occurrence of seizures Variable

Overall, N (%)

Yes, n (%)

No, n (%)

Patients, n Demographic Age, median (IQR), y Age group, y ,60 60-79 $80 Female Current smoking Current heavy drinking Medical history Stroke Diabetes mellitus Hypertension Dyslipidemia Cardiovascular disease Atrial fibrillation NIHSS at hospital admission, median (IQR) NIHSS at hospital admission 0-4 5-14 $15 GCS at hospital admission, median (IQR) GCS at hospital admission 13-15 9-12 3-8 Hematoma location Strictly deep locations Deep and lobe locations Strictly lobe locations Hematoma volume, mL, median (IQR) Intraventricular extension (extension into ventricles)

3216

139

3077

62 (53-72)

62 (52-73)

62 (53-72)

1391 (43.3) 1535 (47.7) 290 (9.0) 1241 (38.6) 1220 (37.9) 367 (11.4)

63 (45.3) 56 (40.3) 20 (14.4) 57 (41.0) 46 (33.1) 15 (10.8)

1328 (43.2) 1479 (48.1) 270 (8.8) 1184 (38.5) 1174 (38.2) 352 (11.4)

874 (27.2) 285 (8.9) 2190 (68.1) 228 (7.1) 267 (8.3) 52 (1.6) 9 (4-16)

46 (33.1) 18 (12.9) 84 (60.4) 8 (5.8) 11 (7.9) 3 (2.2) 16 (7-30)

828 (26.9) 267 (8.7) 2106 (68.4) 220 (7.1) 256 (8.3) 49 (1.6) 9 (4-16)

1072 (33.3) 1188 (36.9) 956 (29.7) 14 (9-15)

33 (23.7) 35 (25.2) 71 (51.1) 8 (6-14)

1039 (33.8) 1153 (37.5) 885 (28.8) 14 (9-15)

2002 (62.3) 474 (14.7) 740 (23.0)

56 (40.3) 16 (11.5) 67 (48.2)

1946 (63.2) 458 (14.9) 673 (21.9)

P value

.822 .039

.593 .246 .892 .119 .092 .051 .734 1.000 .491 ,.001 ,.001

,.001 ,.001

,.001 2122 (66.0) 568 (17.7) 526 (16.3) 12.5 (5.4-28.0) 946 (29.4)

55 (39.6) 25 (18.0) 59 (42.4) 16.9 (9.0-36) 49 (35.3)

2067 (67.2) 543 (17.6) 467 (15.2) 12 (5.12-27) 897 (29.2)

.005 .128

Abbreviations: GCS, Glasgow Coma Scale; IQR, interquartile range; NIHSS, National Institute of Health Stroke Scale.

10.8%, P 5 .001) after ICH onset than those without seizures (Table 5). For patients with a GCS score of 9-12 or 3-8, the presence of seizures had not significant difference on functional outcomes or mortality compared with the absence of seizures (Table 5).

Multivariable Analysis of Variables Associated with Seizures Multivariable analysis demonstrated that the following variables were associated with seizures: young age (,60 years as reference, 60-79 years, adjusted OR, .64; CI, .43.94), stroke severity at admission (NIHSS $15 [adjusted OR, 2.24; CI, 1.25-4.01; P 5 .007] and GCS #8 [adjusted OR, 2.57; CI, 1.51-4.38; P 5 .001]), lobar involvement (adjusted OR, 6.45; CI, 4.26-9.77; P , .001), in-hospital complications of AF (adjusted OR, 3.18; CI, 1.45-6.98;

P 5 .004) and pneumonia (adjusted OR, 1.63; CI, 1.082.45; P 5 .021; Fig 1).

Discussion The impact of seizures after ICH on clinical outcomes is controversial and is a challenge for clinicians. In this largest cohort study of ICH in China, we found that presence of seizures in patients with ICH at onset or during hospitalization was an independent risk factor of 3-, 6-, and 12-month stroke mortality and poor functional outcomes after adjusting for other factors. Additionally, it also was associated with higher in-hospital complications and lower adherence to rehabilitation assessment. These findings were somewhat different from those of previous studies. A recent population-based observational study found that patients with seizures

Z. LI ET AL.

460

Table 3. Management and in-hospital complications of patients with spontaneous intracerebral hemorrhage with or without seizures Occurrence of seizures P value

Variable

Overall, N (%)

Yes, n (%)

No, n (%)

Patients, n Treated in Stroke unit NICU/ICU Neurology ward/other ward Length NICU/ICU of stay, median (IQR), d Length hospital of stay, median (IQR), d Supporting interventions Nasal feeding Central venous catheterization Tracheal Intubation Mechanical ventilation Performance measures Antihypertensive therapy Swallow function evaluation Rehabilitation assessment In-hospital complications Hematoma expansion Myocardial infarction AF Pneumonia UTI Decubitus ulcer DVT GI hemorrhage

3216

139

3077

549 (17.1) 577 (17.9) 2090 (65.0) 0 (0-6) 18 (11-26)

23 (16.5) 34 (24.5) 82 (59.0) 2 (0-7) 16 (9-26)

526 (17.1) 543 (17.6) 2008 (65.3) 0 (0-5) 18 (12-26)

485 (15.1) 139 (4.3) 168 (5.2) 106 (3.3)

43 (30.9) 10 (7.2) 15 (10.8) 8 (5.8)

442 (14.4) 129 (4.2) 153 (5.0) 98 (3.2)

,.001 .090 .006 .136

1892 (58.8) 837 (26.0) 1251 (38.9)

73 (52.5) 32 (23.0) 42 (30.2)

1819 (59.1) 805 (26.2) 1209 (39.3)

.134 .431 .033

90 (2.8) 25 (.8) 75 (2.3) 577 (17.9) 192 (6.0) 30 (.9) 22 (.7) 176 (5.5)

6 (4.3) 1 (.7) 9 (6.5) 41 (29.5) 11 (7.9) 3 (2.2) 1 (.7) 11 (7.9)

84 (2.7) 24 (.8) 66 (2.1) 536 (17.4) 181 (5.9) 27 (.9) 21 (.7) 165 (5.4)

.283 1.000 .004 .001 .356 .138 1.000 .183

.118

.506 .138

Abbreviations: AF, atrial fibrillation; DVT, deep venous thrombosis; GI, gastrointestinal bleeding; ICU, intensive care unit; IQR, interquartile range; NICU, neurologic intensive care unit; UTI, urinary tract infection.

experienced higher mortality than patients without seizures but seizures were not an independent risk factor of mortality at 30 days after stroke including ischemic or hemorrhagic stroke.7 Another population-based study showed that early seizures occurring within 2 weeks after symptom onset were not independently associated with poor early functional outcomes and mortality at 1 month and 1 year after adjustment for confounding variables.6 The Prognosis of InTra-Cerebral Hemorrhage cohort, a simple-center hospital-based prospective observational study, showed that the occurrence of early seizures within 7 days from initial onset did not influence mortality at 7 days or at 6 months and functional outcomes at 6 months for patients with ICH.5 Another recent study did not find that seizures were related to functional outcomes in young patients with ICH. However, there were a few cases of epilepsy, which significantly limited statistical power.21 A trend toward increased poor outcomes in patients with posthemorrhagic seizures was found, but a relatively small sample size of patients with ICH limited the result findings.9 Several previous studies included both ischemic and hemorrhagic strokes, whose nature and prognosis were different. Early seizures were twice

as frequent in hemorrhagic stroke as in ischemic stroke patients.6 These heterogeneities of ischemic and hemorrhagic strokes may produce a mixed effect in evaluating the impact of seizures on clinical outcomes. Meanwhile, some previous studies involved a single center, small sample size of patients with ICH, and short-term follow-up, which also affected the measured impact of seizures on clinical outcomes. In addition, in-hospital complications and performance measures, which also affected the clinical outcomes, were not taken into consideration sufficiently in several studies. In our study from CNSR, a national hospital-based, multicenter, and prospective registry of patients with acute cerebrovascular events,12 3216 patients with ICH, excluding a history of seizures, were included and analyzed regarding the impact of seizures on 12-month stroke outcomes. Meanwhile, in-hospital complications and performance measures also were recorded and taken into account. Similar to previous studies,17,22 seizure patients with ICH had more severe stroke and in-hospital complications such as pneumonia in our study, which may further worsen clinical outcomes.18,19 This phenomenon could explain the higher utilization of supporting interventions and mortality in

SEIZURES AND OUTCOMES IN HEMORRHAGE STROKE

461

Table 4. Poor functional outcome and mortality in patients with or without seizures after intracerebral hemorrhage Univariate analysis Clinical outcomes At 3 mo Poor outcomez Death At 6 mo Poor outcome Death At 12 mo Poor outcome Death

Multivariable analysis*

Odds ratioy

95% CI

P value

Odds ratio

95% CI

P value

2.51 2.64

1.73-3.63 1.85-3.76

,.001 ,.001

1.9 1.6

1.15-3.14 1.01-2.53

.012 .045

2.63 2.89

1.83-3.8 2.05-4.09

,.001 ,.001

1.94 1.88

1.19-3.15 1.21-2.94

.008 .005

2.95 2.9

2.04-4.28 2.06-4.08

,.001 ,.001

2.27 1.97

1.41-3.65 1.27-3.05

.001 .002

Abbreviation: CI, confidence interval. *Adjusted for age, sex, smoking, heavy drinking, history of stroke, hypertension, diabetes mellitus, dyslipidemia, cardiovascular disease, atrial fibrillation, National Institutes of Health Stroke Scale score and Glasgow Coma Scale score on admission, hematoma volume, hematoma location, intraventricular hemorrhagic extension, in-hospital complications including hematoma expansion, atrial fibrillation, urinary tract infection, decubitus ulcer, myocardial infarction, deep venous thrombosis, pneumonia, and gastrointestinal bleeding and performance measures of antihypertensive therapy, dysphagia screening, and rehabilitation evaluation. yOdds ratio using the group without seizures as the reference. zPoor outcome defined as modified Rankin Scale 3-6.

patients with seizures. Additionally, our findings indicated that several performance measures including antihypertensive therapy, screening for dysphagia, and rehabilitation assessment played a positive role in improving clinical outcomes.14,23 However, compared with patients without seizures, those with seizures had relatively low adherence of these performance measures in our study. This may be because seizure patients had more severe strokes and disturbance of consciousness, making it difficult to give oral antihypertensive therapy and unsuitable for dysphagia screening or rehabilitation assessment. Interestingly, for patients with moderate or severe strokes (GCS of 9-12 or 3-8), we found no difference for seizure patients in mortality and functional outcomes compared with those without seizures, as these patients demonstrated ceiling effects in mortality and poor functional outcome, which was also found in seizure patients with ischemic severe stroke.17 However, among patients with mild strokes (GCS of 13-15), patients with seizures had higher rates of short- or long-term mortality and worsened functional outcomes compared with those without seizures. This indicated that seizures played a vital role in poor functional outcome and mortality, especially in patients with mild strokes. The potential mechanisms by which seizures might worsen the clinical outcomes of patients with ICH might be as follows. First, the combination of sudden development of a space-occupying lesion with mass effect, perihematoma edema, and blood breakdown products has been postulated to account for seizures especially in early ICH.3 Within minutes to hours from the onset of bleeding, damage begins and is primarily the result of mechanical

damage associated with the mass effect.24 Second, secondary damage after ICH is triggered by the presence of intraparenchymal blood and components, which subsequently activates cytotoxic, oxidative and inflammatory pathways, and causing death of surrounding cells.25,26 Third, secondary brain injury after seizures may have a similar molecular pathophysiology with ICH. Seizures can cause neuronal cell death through dynamic processes that might cause excitotoxicity to induce mitochondrial dysfunction, alter cytokine levels and oxidative stress, and change plasticity or activation in some late cell death pathways.27,28 Early epileptiform activity has a deleterious impact on perihematoma areas, possibly by increasing the cerebral blood flow and glucose metabolic demand in a hypoxic tissue and the previously mentioned molecular pathophysiology.2931 Additionally, early seizures occurring in patients with ICH increased the risk of later seizures.32 Symptomatic seizures due to stroke or other disease had a higher early mortality.33 Therefore, it may be reasonably inferred that epileptic seizures in patients with ICH occurring either at onset or whilst in the hospital further worsen clinical outcomes because ICH is coupling with early or later seizures, and they form overlying effects together. Our study looked at a similar frequency of seizures in ICH patients to that of previous studies ranging from 2.7% to 17%.1–6,34,35 Similarly, previous studies have indicated a direct relationship between stroke severity, lobar or cortical involvement, and seizures.5,32,35 Additionally, in our study, other associated factors with occurrence of seizures also included the in-hospital complications of pneumonia.17,19 Interestingly, we also observed that the presence of seizures was associated

Z. LI ET AL.

462

Table 5. Mortality and functional outcomes comparison stratified by GCS at admission Occurrence of seizures Variable

Overall, N (%)

Yes, n (%)

No, n (%)

All patients Poor outcome* within 3 mo Poor outcome within 6 mo Poor outcome within 12 mo Death within 3 mo Death within 6 mo Death within 12 mo GCS 13-15 Poor outcome within 3 mo Poor outcome within 6 mo Poor outcome within 12 mo Death within 3 mo Death within 6 mo Death within 12 mo GCS 9-12 Poor outcome within 3 mo Poor outcome within 6 mo Poor outcome within 12 mo Death within 3 mo Death within 6 mo Death within 12 mo GCS 3-8 Poor outcome within 3 mo Poor outcome within 6 mo Poor outcome within 12 mo Death within 3 mo Death within 6 mo Death within 12 mo

3216 1572 (48.9) 1507 (46.9) 1474 (45.8) 635 (19.7) 716 (22.3) 832 (25.9) 2002 629 (31.4) 592 (29.6) 595 (29.7) 144 (7.2) 173 (8.6) 225 (11.2) 474 324 (68.4) 308 (65.0) 291 (61.4) 103 (21.7) 125 (26.4) 150 (31.6) 740 619 (83.6) 607 (82.0) 588 (79.5) 388 (52.4) 418 (56.5) 457 (61.8)

139 97 (69.8) 96 (69.1) 98 (70.5) 53 (38.1) 61 (43.9) 68 (48.9) 56 30 (53.6) 29 (51.8) 31 (55.4) 11 (19.6) 12 (21.4) 15 (26.8) 16 10 (62.5) 10 (62.5) 10 (62.5) 4 (25.0) 6 (37.5) 7 (43.8) 67 57 (85.1) 57 (85.1) 57 (85.1) 38 (56.7) 43 (64.2) 46 (68.7)

3077 1475 (48.0) 1411 (45.9) 1376 (44.7) 582 (18.9) 655 (21.3) 764 (24.8) 1946 599 (30.8) 563 (28.9) 564 (29.0) 133 (6.8) 161 (8.3) 210 (10.8) 458 314 (68.6) 298 (65.1) 281 (61.4) 99 (21.6) 119 (26.0) 143 (31.2) 673 562 (83.5) 550 (81.7) 531 (78.9) 350 (52.0) 375 (55.7) 411 (61.1)

P value ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 .001 ,.001 ,.001 .002 .002 .001 .594 .796 1.0 .759 .385 .287 .863 .617 .269 .522 .198 .238

Abbreviation: GCS, Glasgow Coma Scale. *Poor outcome defined as modified Rankin Scale 3-6.

with the atrial fibrillation. However, the potential mechanisms are not currently clear. Epileptic seizures are often companied by the changes in cardiac autonomic function.36 Atrial fibrillation or atrial flutter after epileptic seizures has been reported and the possible underlying cause is augmentation in vagal tone and presence of a concealed cardiac dysfunction.37 Among patients with ischemic stroke, the involvement of insular cortex damage can cause cardiac autonomic dysregulation.38,39 Hemorrhagic stroke involving the cortical lobe is prone to seizures.32,35 So, we speculate that the involvement of the insular cortex among ICH patients might take a potential aspect on the relationship between seizures and atrial fibrillation. Our study has some limitations. First, we did not have information on the results of electroencephalogram; some seizures could have been missed or some paroxysmal nonepileptic events could have been called seizures. The true incidence of seizures might have been underestimated because subclinical seizures can be detected by using continuous electroencephalogram monitoring

only.9,40,41 Second, we had no available information about the specific time, types and frequency of seizures, status epilepticus, and the use of antiepileptic drugs during hospitalization and after discharge. Additionally, the sequence of in-hospital seizures and complications was unclear. It indicated that seizures were related factors of some complications but not precipitating factors. Despite some limitations, our study was the large ICH cohort to assess the influence of seizures on outcomes from CNSR including comprehensive clinical administrative data, which may provide some practical implications for clinical management of ICH patients. Our findings suggest that seizures should be considered as a neurologic complication in patients with ICH that have an important impact on short- and long-term poor prognosis and a relationship with complications, as well as medical resource utilization. These results provide a supportive evidence for the guideline recommendation that patients with clinical seizures should be treated with antiepileptic drugs.23 Currently, there is insufficient evidence to support the routine use of antiepileptic drugs for the primary

SEIZURES AND OUTCOMES IN HEMORRHAGE STROKE

Figure 1. Multivariate analysis: variables associated with the occurrence of seizures in patients with ICH. Abbreviations: CI, confidence interval; GCS, Glasgow Coma Scale; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratios.

or secondary prevention of seizures after stroke.42 However, in many hospitals worldwide, the use of antiepileptic drugs as a prophylactic intervention in ICH patients is a common practice. According to the new definition of epilepsy, in the presence of a cortical lesion, 1 seizure is adequate for diagnosis of epilepsy.16 Thus, we should be attentive to seizures, especially in patients with mild hemorrhagic strokes in lobar or cortical regions, and prophylactic or therapeutic antiepileptic drug for these targeted patients may be evaluated more thoroughly in a randomized study. Strategies of targeting patients with seizures in ICH patients may be performed to initiate specialized care, improve the outcomes, and quality of care in these individuals. In conclusions, the presence of seizures in patients with ICH was associated with increased in-hospital complications, poor functional outcome, and mortality at 3, 6, and 12 months and low adherent rate of performance measures. Quality improvement strategies targeting ICH patients with seizures especially mild stroke may help improve outcomes.

References 1. Yang TM, Lin WC, Chang WN, et al. Predictors and outcome of seizures after spontaneous intracerebral hemorrhage. Clinical article. J Neurosurg 2009;111:87-93. 2. Berger AR, Lipton RB, Lesser ML, et al. Early seizures following intracerebral hemorrhage: implications for therapy. Neurology 1988;38:1363-1365.

463 3. Bladin CF, Alexandrov AV, Bellavance A, et al. Seizures after stroke: a prospective multicenter study. Arch Neurol 2000;57:1617-1622. 4. Srinivasan S, Shin H, Chou SH, et al. Seizures and antiepileptic drugs in patients with spontaneous intracerebral hemorrhages. Seizure 2013;22:512-516. 5. De Herdt V, Dumont F, Henon H, et al. Early seizures in intracerebral hemorrhage: incidence, associated factors, and outcome. Neurology 2011;77:1794-1800. 6. Hamidou B, Aboa-Eboule C, Durier J, et al. Prognostic value of early epileptic seizures on mortality and functional disability in acute stroke: the Dijon Stroke Registry (1985-2010). J Neurol 2013;260:1043-1051. 7. Szaflarski JP, Rackley AY, Kleindorfer DO, et al. Incidence of seizures in the acute phase of stroke: a populationbased study. Epilepsia 2008;49:974-981. 8. Andaluz N, Zuccarello M. Recent trends in the treatment of spontaneous intracerebral hemorrhage: analysis of a nationwide inpatient database. J Neurosurg 2009;110: 403-410. 9. Vespa PM, O’Phelan K, Shah M, et al. Acute seizures after intracerebral hemorrhage: a factor in progressive midline shift and outcome. Neurology 2003;60:1441-1446. 10. Fonarow GC, Reeves MJ, Smith EE, et al. Characteristics, performance measures, and in-hospital outcomes of the first one million stroke and transient ischemic attack admissions in get with the guidelines-stroke. Circ Cardiovasc Qual Outcomes 2010;3:291-302. 11. Burneo JG, Fang J, Saposnik G, Investigators of the Registry of the Canadian Stroke Network. Impact of seizures on morbidity and mortality after stroke: a Canadian multi-centre cohort study. Eur J Neurol 2010;17:52-58. 12. Wang Y, Cui L, Ji X, et al. The China National Stroke Registry for patients with acute cerebrovascular events: design, rationale, and baseline patient characteristics. Int J Stroke 2011;6:355-361. 13. Stroke–1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke 1989;20:1407-1431. 14. Wang WJ, Lu JJ, Wang YJ, et al. Clinical characteristics, management, and functional outcomes in Chinese patients within the first year after intracerebral hemorrhage: analysis from China National Stroke Registry. CNS Neurosci Ther 2012;18:773-780. 15. Kothari RU, Brott T, Broderick JP, et al. The ABCs of measuring intracerebral hemorrhage volumes. Stroke 1996;27:1304-1305. 16. Fisher RS, van Emde Boas W, Blume W, et al. Epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia 2005;46:470-472. 17. Huang CW, Saposnik G, Fang J, et al. Influence of seizures on stroke outcomes: a large multicenter study. Neurology 2014;82:768-776. 18. Balami JS, Buchan AM. Complications of intracerebral haemorrhage. Lancet neurology 2012;11:101-118. 19. Ji R, Wang D, Shen H, et al. Interrelationship among common medical complications after acute stroke: pneumonia plays an important role. Stroke 2013;44:3436-3444. 20. Burn JP. Reliability of the modified Rankin Scale. Stroke 1992;23:438. 21. Arntz RM, Maaijwee NA, Rutten-Jacobs LC, et al. Epilepsy after TIA or stroke in young patients impairs long-term functional outcome: the FUTURE Study. Neurology 2013;81:1907-1913.

464 22. Pezzini A, Grassi M, Del Zotto E, et al. Complications of acute stroke and the occurrence of early seizures. Cerebrovasc Dis 2013;35:444-450. 23. Morgenstern LB, Hemphill JC 3rd, Anderson C, et al. Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/ American Stroke Association. Stroke 2010;41:2108-2129. 24. Qureshi AI, Mendelow AD, Hanley DF. Intracerebral haemorrhage. Lancet 2009;373:1632-1644. 25. Aronowski J, Zhao X. Molecular pathophysiology of cerebral hemorrhage: secondary brain injury. Stroke 2011; 42:1781-1786. 26. Xi G, Keep RF, Hoff JT. Mechanisms of brain injury after intracerebral haemorrhage. Lancet Neurol 2006;5:53-63. 27. Nardou R, Ferrari DC, Ben-Ari Y. Mechanisms and effects of seizures in the immature brain. Semin Fetal Neonatal Med 2013;18:175-184. 28. DeLorenzo RJ, Sun DA, Blair RE, et al. An in vitro model of stroke-induced epilepsy: elucidation of the roles of glutamate and calcium in the induction and maintenance of stroke-induced epileptogenesis. Int Rev Neurobiol 2007;81:59-84. 29. La Fougere C, Rominger A, Forster S, et al. PET and SPECT in epilepsy: a critical review. Epilepsy Behav 2009;15:50-55. 30. Goffin K, Dedeurwaerdere S, Van Laere K, et al. Neuronuclear assessment of patients with epilepsy. Semin Nucl Med 2008;38:227-239. 31. Newton MR, Berkovic SF, Austin MC, et al. Postictal switch in blood flow distribution and temporal lobe seizures. J Neurol Neurosurg Psychiatry 1992;55:891-894.

Z. LI ET AL. 32. Haapaniemi E, Strbian D, Rossi C, et al. The CAVE score for predicting late seizures after intracerebral hemorrhage. Stroke 2014;45:1971-1976. 33. Hesdorffer DC, Benn EK, Cascino GD, et al. Is a first acute symptomatic seizure epilepsy? Mortality and risk for recurrent seizure. Epilepsia 2009;50:1102-1108. 34. Sung CY, Chu NS. Epileptic seizures in intracerebral haemorrhage. J Neurol Neurosurg Psychiatry 1989; 52:1273-1276. 35. Passero S, Rocchi R, Rossi S, et al. Seizures after spontaneous supratentorial intracerebral hemorrhage. Epilepsia 2002;43:1175-1180. 36. Devinsky O. Effects of seizures on autonomic and cardiovascular function. Epilepsy Curr 2004;4:43-46. 37. Herskovitz M, Schiller Y. Atrial fibrillation associated with epileptic seizures. Arch Neurol 2012;69:1197-1199. 38. Abboud H, Berroir S, Labreuche J, et al. Insular involvement in brain infarction increases risk for cardiac arrhythmia and death. Ann Neurol 2006;59:691-699. 39. Laowattana S, Zeger SL, Lima JA, et al. Left insular stroke is associated with adverse cardiac outcome. Neurology 2006;66:477-483. 40. Garrett MC, Komotar RJ, Starke RM, et al. Predictors of seizure onset after intracerebral hemorrhage and the role of long-term antiepileptic therapy. J Crit Care 2009; 24:335-339. 41. Claassen J, Jette N, Chum F, et al. Electrographic seizures and periodic discharges after intracerebral hemorrhage. Neurology 2007;69:1356-1365. 42. Sykes L, Wood E, Kwan J. Antiepileptic drugs for the primary and secondary prevention of seizures after stroke. Cochrane Database Syst Rev 2014;1:CD005398.