- Email: [email protected]
Electrocardiograph abnormalities in intracerebral hemorrhage
Journal of Clinical Neuroscience xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Electrocardiograph abnormalities in intracerebral hemorrhage Satoru Takeuchi ⇑, Kimihiro Nagatani, Naoki Otani, Kojiro Wada, Kentaro Mori Department of Neurosurgery, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 3598513, Japan
a r t i c l e
i n f o
Article history: Received 7 February 2015 Accepted 18 April 2015 Available online xxxx Keywords: Electrocardiography Insular cortex Intracerebral hemorrhage Stroke
a b s t r a c t This study investigated the prevalence and type of electrocardiography (ECG) abnormalities, and their possible association with the clinical/radiological findings in 118 consecutive patients with non-traumatic, non-neoplastic intracerebral hemorrhage (ICH). ECG frequently demonstrates abnormalities in patients with ischemic stroke and subarachnoid hemorrhage, but little is known of ECG changes in ICH patients. Clinical and radiological information was retrospectively reviewed. ECG recordings that were obtained within 24 hours of the initial hemorrhage were analyzed. Sixty-six patients (56%) had one or more ECG abnormalities. The most frequent was ST depression (24%), followed by left ventricular hypertrophy (20%), corrected QT interval (QTc) prolongation (19%), and T wave inversion (19%). The logistic regression analysis demonstrated the following: insular involvement was an independent predictive factor of ST depression (p < 0.001; odds ratio OR 10.18; 95% confidence interval [CI] 2.84–36.57); insular involvement (p < 0.001; OR 23.98; 95% CI 4.91–117.11) and presence of intraventricular hemorrhage (p < 0.001; OR 8.72; 95% CI 2.69–28.29) were independent predictive factors of QTc prolongation; deep hematoma location (p < 0.001; OR 19.12; 95% CI 3.82–95.81) and hematoma volume >30 ml (p = 0.001; OR 6.58; 95% CI 2.11–20.46) were independent predictive factors of T wave inversion. We demonstrate associations between ECG abnormalities and detailed characteristics of ICH. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Electrocardiography (ECG) often demonstrates abnormalities including ST elevation, ST depression, inverted T wave, and corrected QT interval (QTc) prolongation in patients with central nervous system disorders in the absence of known organic heart disease. In fact, such ECG abnormalities are frequently found in patients with acute stroke, especially ischemic stroke and subarachnoid hemorrhage (SAH) [1–5]. However, in intracerebral hemorrhage (ICH), little is understood about the relationship between ECG abnormalities and radiological findings, including hematoma size, location, and extension [5–8]. We investigated the prevalence and type of ECG abnormalities in patients with ICH, and the possible associations with clinical/radiological findings. 2. Materials and methods 2.1. Patients This study was conducted with the approval of the Ethics Committee of the National Defense Medical College. Written ⇑ Corresponding author. Tel.: +81 42 995 1656; fax: +81 42 996 5207. E-mail address: [email protected] (S. Takeuchi).
informed consent was not required because of the retrospective nature of the investigation. A total of 161 patients were admitted to our hospital with ICH not related to a vascular abnormality or aneurysm, trauma, or neoplasm, between 2004 and 2014. CT scans or magnetic resonance angiography (MRA) were performed for all patients to exclude a ruptured aneurysm or other vascular abnormalities. All patients underwent 12 lead ECG examinations on admission. Twenty patients were excluded from the study because of unavailable clinical and/or radiological findings. In addition, another 23 patients were excluded because of cardiac pacemakers, a past history of ischemic heart disease and/or arrhythmia, or an interval between the onset and admission of >24 hours. Therefore, this study included 118 patients in total. The data were retrospectively reviewed to establish the clinical characteristics such as age, sex, level of consciousness, hematoma location, side of hematoma, hematoma volume, midline shift, presence of intraventricular hemorrhage (IVH), presence of SAH, presence of insular involvement, and ECG findings. The level of consciousness was assessed using the Glasgow coma scale at admission. The hematoma location was classified into lobar, deep (basal ganglia and thalamus), cerebellum, and brainstem. Hematoma volume was calculated by the ABC/2 formula, where A and B are the perpendicular maximal diameters of the lesion, and C is the total length in the vertical plane [9].
http://dx.doi.org/10.1016/j.jocn.2015.04.028 0967-5868/Ó 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
2
S. Takeuchi et al. / Journal of Clinical Neuroscience xxx (2015) xxx–xxx
2.2. ECG analyses ECG recordings that were obtained within 24 hours of the initial hemorrhage were analyzed by the same observer, who was unaware of the clinical details. The ECG abnormalities were defined according to the following established criteria [7]: (a) rhythm: sinus rhythm, atrial fibrillation, atrial flutter, atrial and ventricular premature complexes (ectopic beats); (b) heart rate: sinus bradycardia (<60/min), sinus tachycardia (>100/min); (c) PR interval: short <120 ms, prolonged >200 ms; (d) QRS width: prolonged >120 ms; (e) QT interval: measured using the tangent method [10], where QT was corrected for heart rate (QTc) using the Bazett formula, and was considered prolonged at >450 ms in females and >440 ms in males; (f) ST segment depression: horizontal or down sloping ST segment with (>0.05 mV) or without ST-J depression (measured 80 ms after the J point); (g) ST segment elevation: upward convexity of the ST segment (>0.1 mV) with or without ST-J elevation, measured 80 ms after the J point; (h) prominent U wave > 25% of the highest T wave in precordial leads; (i) left ventricular hypertrophy (LVH): Sokolow voltage criteria (S V1/V2 + R V5/V6 > 35 mV). 2.3. Statistical analyses Univariate analyses were used to identify the possible relationships between each ECG abnormality and clinical characteristics, using the chi-squared test or Fisher’s extract test. An ECG abnormality was analyzed if it was observed in more than 10% of patients. Multivariate logistic regression analyses were used to determine the factors that were independently associated with ECG abnormality. All variables with a significance level of p < 0.15 in the univariate analysis were included as independent variables in the logistic regression analyses. Only those variables with p < 0.15 in two-tailed tests were retained within the model. Odds ratios (OR) and 95% confidence intervals (CI) were reported. A value of p < 0.05 was considered to be statistically significant. All statistical analyses were performed using SPSS (version 11.0; IBM Corporation, Armonk, NY, USA). 3. Results The clinical and radiological findings are summarized in Table 1. Of the 118 patients, 65 were men (55%) and 53 women, with a mean age of 63.5 years (range: 32–91). The Glasgow coma scale score was 68 in 28 patients (24%). The past history included hypertension in 71 patients (60%), hypercholesterolemia in 23 (19%), and diabetes mellitus in 13 (11%). The hematoma was located in a deep region in 64 patients (54%), the lobar region in 33 (28%), the cerebellum in 10 (8%), and the brainstem in 11 (9%), and it was on the right side in 43 (45%). The mean hematoma volume was 28.3 ml, and was >30 ml in 40 patients (34%). The mean midline shift was 3.2 mm, and was >5 mm in 28 patients (24%). The associated lesions included IVH in 32 patients (27%) and SAH in 13 (11%). Insular involvement of the hematoma was observed in 13 patients (11%). One or more ECG abnormalities were detected in 66 patients (56%). The most frequently observed ECG abnormality was ST depression (24%), followed by LVH (20%), QTc prolongation (19%), and T wave inversion (19%). No life threatening arrhythmias or ECG changes were seen. Three patients developed ischemic heart diseases, including angina (n = 2) and myocardial infarction (n = 1). The ECG findings within 24 hours after the onset were ST depression in one patient and LVH in two.
Table 1 Baseline characteristics of 118 patients with intracerebral hemorrhage Characteristics Age Mean years ± SD 660 years, n (%) Male, n (%) GCS 6 8, n (%) Location of hematoma, n (%) Lobar Deep Cerebellum Brainstem Right sided hematoma, n (%)* Hematoma volume Mean ml ± SD >30 ml, n (%) Midline shift, n (%) Mean mm ± SD >5 mm, n (%) IVH, n (%) SAH, n (%) Insular involvement, n (%) ECG abnormalities, n (%) ST depression LVH QTc prolongation Inverted T wave Sinus tachycardia Atrial fibrillation CRBBB ST elevation PVC Sinus bradycardia
Patient data 63.5 ± 14.7 47 (40) 65 (55) 28 (24) 33 64 10 11 43
(28) (54) (8) (9) (45)
28.3 ± 25.3 40 (34) 3.2 ± 8.3 28 (24) 32 (27) 13 (11) 13 (11) 66 (56) 28 (24) 24 (20) 23 (19) 23 (19) 11 (9) 9 (8) 7 (6) 7 (6) 5 (4) 5 (4)
* Data from 95 patients, excluding 23 who had a hematoma located in bilateral hemispheres (n = 2), cerebellum (n = 10), or brainstem (n = 11). CRBBB = complete right bundle branch block, ECG = electrocardiography, GCS = Glasgow coma scale, IVH = intraventricular hemorrhage, LVH = left ventricular hypertrophy, PVC = premature ventricular contraction, QTc = corrected QT interval, SAH = subarachnoid hemorrhage, SD = standard deviation.
3.1. Analysis of ST depression The univariate analysis showed that ST depression was associated with deep hematoma location (p = 0.004) and ICH involvement of the insular cortex (p < 0.001; Table 2). The logistic regression analysis demonstrated that insular involvement (p < 0.001; OR 10.18; 95% CI 2.84–36.57) was an independent predictive factor of ST depression.
3.2. Analysis of LVH The univariate analysis showed that LVH was not associated with any clinicoradiological characteristics (Table 2). The logistic regression analysis also demonstrated that there was no independent predictive factor of LVH.
3.3. Analysis of QTc prolongation The univariate analysis showed that QTc prolongation was associated with deep hematoma location, midline shift >5 mm, presence of IVH, and ICH involvement of the insular cortex (p = 0.011, p = 0.026, p < 0.001, and p < 0.001, respectively; Table 2). The logistic regression analysis demonstrated that insular involvement (p < 0.001; OR 23.98; 95% CI 4.91–117.11) and the presence of IVH (p < 0.001; OR 8.72; 95% CI 2.69–28.29) were independent predictive factors of QTc prolongation.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
* Data from 95 patients, excluding 23 who had a hematoma located in both hemispheres (n = 2), cerebellum (n = 10), or brainstem (n = 11). These patients consisted of 18 with and 77 without ST depression; 18 with and 77 without LVH; 22 with and 73 without QTc prolongation; 21 with and 74 without an T wave inversion. GCS = Glasgow coma scale, IVH = intraventricular hemorrhage, LVH = left ventricular hypertrophy, QTc = corrected QT interval, SAH = subarachnoid hemorrhage.
0.179 0.434 0.129 0.129 20 (21) 24 (25) 8 (8) 8 (8) 19 (21) 27 (30) 11 (12) 4 (4) (32) (18) (7) (32) 9 5 2 9
0.308 0.235 0.523 <0.001
5 (21) 10 (42) 4 (17) 5 (21)
23 (24) 22 (23) 9 (10) 8 (9)
0.794 0.120 0.298 0.136
10 (43) 15 (65) 4 (17) 10 (43)
18 (19) 17 (18) 9 (9) 3 (3)
0.026 <0.001 0.278 <0.001
8 8 5 5
(35) (35) (22) (22)
0.003 26 (27) 14 (61) 0.329 30 (32) 10 (43) 1.000 32 (34) 8 (33) 28 (30) 12 (43)
0.262
0.118 <0.001 0.206 0.119 0.329 30 44 10 11 38 29 (32) 42 (47) 8 (9) 11 (12) 37 (48) 4 (14) 22 (79) 2 (7) 0 (0) 9 (50)
0.091 0.004 1.000 0.064 1.000
5 (21) 17 (71) 2 (8) 0 (0) 6 (33)
28 (30) 47 (50) 8 (9) 11 (12) 40 (52)
0.454 0.107 1.000 0.117 0.195
3 (13) 18 (78) 2 (9) 0 (0) 10 (45)
30 (32) 46 (48) 8 (8) 11 (12) 36 (49)
0.118 0.011 1.000 0.119 0.811
3 (13) 20 (87) 0 (0) 0 (0) 8 (38)
(32) (46) (11) (12) (51)
0.236 0.162 0.179 95 35 (37) 49 (52) 20 (21) 90 35 (39) 47 (52) 25 (28) 28 12 (43) 18 (64) 3 (11)
Patients, n Age 6 60 years, n (%) Male sex, n (%) GCS 6 8, n (%) Hematoma location, n (%) Lobar Deep Cerebellum Brainstem Right sided hematoma, n (%)* Hematoma volume>30 ml, n (%) Midline shift > 5 mm, n (%) IVH, n (%) SAH, n (%) Insular involvement, n (%)
0.826 0.285 0.077
24 7 (29) 15 (63) 2 (8)
94 40 (43) 50 (53) 26 (28)
0.254 0.494 0.060
23 10 (43) 14 (61) 4 (17)
95 37 (39) 51 (54) 24 (25)
0.813 0.642 0.587
23 12 (52) 16 (70) 8 (35)
p value Without ST depression With ST depression Characteristics
Table 2 Results of the univariate analyses of ECG abnormalities
p value
With LVH
Without LVH
p value
With QTc prolongation
Without QTc prolongation
p value
With T wave inversion
Without T wave inversion
S. Takeuchi et al. / Journal of Clinical Neuroscience xxx (2015) xxx–xxx
3
3.4. Analysis of T wave inversion The univariate analysis showed that T wave inversion was associated with a deep hematoma location (p < 0.001) and hematoma volume >30 ml (p = 0.003; Table 2). The logistic regression analysis demonstrated that a deep hematoma location (p < 0.001; OR 19.12; 95% CI 3.82–95.81) and hematoma volume >30 ml (p = 0.001; OR 6.58; 95% CI 2.11–20.46) were independent predictive factors of T wave inversion. 4. Discussion The present study showed that ECG abnormalities are common (56%) in patients with ICH, and that the most frequently observed ECG abnormalities were ST depression, LVH, QTc prolongation, and an T wave inversion. Some previous studies also found that these abnormalities were frequently observed in the acute phase of ICH [7,11]. The mechanism by which stroke, including ICH, leads to ECG abnormalities remains unclear. However, there is some evidence that stroke-related ECG changes are a manifestation of autonomic dysregulation, caused by overactivity of the sympathetic limb of the autonomic nervous system [12]. In addition, the insular cortex may be important for the regulation of sympathetic activity. In fact, both experimental investigations and clinical studies suggest that a tissue defect or irritation in the insular region may lead to abnormal cardiac function [13–17], possibly mediated through reduced inhibition of the sympathetic activity, resulting in an increased release of catecholamines [14,18]. Given this known role of the insular cortex in autonomic nervous system function [11], we assessed the effect of insular involvement as well as other characteristics. The present study also revealed that QTc prolongation was independently associated with the presence of IVH and insular involvement, consistent with the previous reports [7]. Both this and previous studies [7,11] have not identified a patient with critical arrhythmia in acute stroke patients. However, a prolonged QTc interval is well known to lead to an increased risk of ventricular tachyarrhythmia, particularly polymorphic ventricular tachycardia (torsade de pointes) [2–4]. In turn, polymorphic arrhythmia may rapidly develop into ventricular fibrillation and cause sudden death. Therefore, careful monitoring is essential if QTc prolongation is detected. ST changes and an T wave inversion are considered to be representative ECG changes of ischemic heart disease, and consequently cause considerable diagnostic problems in the acute setting, possibly misleading physicians in the initial diagnosis. This misdiagnosis may result in adverse consequences, especially in patients with ICH because treatment of ischemic heart disease consists of anticoagulants, antiplatelet therapy, or fibrinolytic agents with or without angioplasty [7]. In particular, an T wave inversion is non-specific, so a neurogenic cause is important to consider in order to avoid inappropriate therapies [19]. It is unclear whether ST changes are associated with ICH characteristics such as insular involvement [7,13]. No association was found between ST changes and the clinical characteristics of 31 ICH patients [7], whereas ST elevation was associated with insular involvement in 179 stroke patients, including 17 ICH patients [13]. Our study showed that ST depression was independently associated with insular involvement. The discrepancy between these studies may be due to differences in the number and population of patients. We could not investigate the predictive factors of ST elevation because of the small sample size and these points need further investigation.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
4
S. Takeuchi et al. / Journal of Clinical Neuroscience xxx (2015) xxx–xxx
The present study also revealed that T wave inversion was independently associated with a deep hematoma location and large hematoma volume. In a previous study, T wave inversion was detected in five of 31 patients, and the univariate analysis showed no association with any characteristics, due to the small sample size [7]. The present study included a larger number of patients, and adopted a multivariable analysis. Therefore, we believe our results are more reliable for this point. The present and previous studies have shown that LVH is one of the most frequent ECG abnormalities, but the occurrence of LVH may not be considered as a new development in all patients with ICH because LVH may reflect the long term effects of chronic hypertension [8]. The absence of any relationship between LVH and the clinicoradiological findings supports this interpretation. Several experimental and clinical studies have shown a right sided dominance for sympathetic cardiovascular effects [13– 15,17,18,20]. In our study, the frequency of ECG abnormalities was comparable between the patients with right and left hemispheric lesions, consistent with previous studies [11,20]. This point needs further investigation. There are several serious limitations to our study. As a single center retrospective study, it was subject to observational and assessment bias. It also had a small sample size, which reduced the statistical power and increased the chance of type 2 errors. Furthermore, ECG data prior to ICH were not available for comparison, and 20 patients were excluded because of unavailable data. We did not assess the relationship between the ECG abnormalities and outcomes because no patient developed a critical arrhythmia. Also, follow-up ECG were not obtained, resulting in a lack of data on whether the ECG changes were transient or permanent. Despite these limitations, the present findings allow for some useful clinical speculations. 5. Conclusion We studied the relationship between ECG abnormalities and the detailed characteristics of ICH in a large series of patients, and showed that insular involvement was an independent predictive factor of ST depression, insular involvement and the presence of IVH were independent predictive factors of QTc prolongation, and a deep hematoma location and hematoma volume >30 ml were independent predictive factors of T wave inversion.
6. Conflicts of Interest/Disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication. References [1] Christensen H, Fogh Christensen A, Boysen G. Abnormalities on ECG and telemetry predict stroke outcome at 3 months. J Neurol Sci 2005;234:99–103. [2] Lanzino G, Kongable GL, Kassell NF. Electrocardiographic abnormalities after nontraumatic subarachnoid hemorrhage. J Neurosurg Anesthesiol 1994;6:156–62. [3] Andreoli A, di Pasquale G, Pinelli G, et al. Subarachnoid hemorrhage: frequency and severity of cardiac arrhythmias. A survey of 70 cases studied in the acute phase. Stroke 1987;18:558–64. [4] Machado C, Baga JJ, Kawasaki R, et al. Torsade de pointes as a complication of subarachnoid hemorrhage: a critical reappraisal. J Electrocardiol 1997;30:31–7. [5] Junttila E, Vaara M, Koskenkari J, et al. Repolarization abnormalities in patients with subarachnoid and intracerebral hemorrhage: predisposing factors and association with outcome. Anesth Analg 2013;116:190–7. [6] Popescu D, Laza C, Mergeani A, et al. Lead electrocardiogram changes after supratentorial intracerebral hemorrhage. Maedica (Buchar) 2012;7:290–4. [7] van Bree MD, Roos YB, van der Bilt IA, et al. Prevalence and characterization of ECG abnormalities after intracerebral hemorrhage. Neurocrit Care 2010;12:50–5. [8] Arruda WO, de Lacerda Júnior FS. Electrocardiographic findings in acute cerebrovascular hemorrhage. A prospective study of 70 patients. Arq Neuropsiquiatr 1992;50:269–74. [9] Kothari RU, Brott T, Broderick JP, et al. The ABCs of measuring intracerebral hemorrhage volumes. Stroke 1996;27:1304–5. [10] Lepeschkin E, Surawicz B. The measurement of the Q-T interval of the electrocardiogram. Circulation 1952;6:378–88. [11] Tatschl C, Stöllberger C, Matz K, et al. Insular involvement is associated with QT prolongation: ECG abnormalities in patients with acute stroke. Cerebrovasc Dis 2006;21:47–53. [12] Samuels MA. The brain-heart connection. Circulation 2007;116:77–84. [13] Christensen H, Boysen G, Christensen AF, et al. Insular lesions, ECG abnormalities, and outcome in acute stroke. J Neurol Neurosurg Psychiatry 2005;76:269–71. [14] Cechetto DF. Experimental cerebral ischemic lesions and autonomic and cardiac effects in cats and rats. Stroke 1993;24:I6–9 [discussion I10–2]. [15] Colivicchi F, Bassi A, Santini M, et al. Cardiac autonomic derangement and arrhythmias in right-sided stroke with insular involvement. Stroke 2004;35:2094–8. [16] Oppenheimer SM, Wilson JX, Guiraudon C, et al. Insular cortex stimulation produces lethal cardiac arrhythmias: a mechanism of sudden death? Brain Res 1991;550:115–21. [17] Oppenheimer SM. Neurogenic cardiac effects of cerebrovascular disease. Curr Opin Neurol 1994;7:20–4. [18] Hachinski VC, Oppenheimer SM, Wilson JX, et al. Asymmetry of sympathetic consequences of experimental stroke. Arch Neurol 1992;49:697–702. [19] Mandrioli J, Zini A, Cavazzuti M, et al. Neurogenic T wave inversion in pure left insular stroke associated with hyperhomocysteinemia. J Neurol Neurosurg Psychiatry 2004;75:1788–9. [20] Afsar N, Fak AS, Metzger JT, et al. Acute stroke increases QT dispersion in patients without known cardiac diseases. Arch Neurol 2003;60:346–50.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Electrocardiograph abnormalities in intracerebral hemorrhage Satoru Takeuchi ⇑, Kimihiro Nagatani, Naoki Otani, Kojiro Wada, Kentaro Mori Department of Neurosurgery, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 3598513, Japan
a r t i c l e
i n f o
Article history: Received 7 February 2015 Accepted 18 April 2015 Available online xxxx Keywords: Electrocardiography Insular cortex Intracerebral hemorrhage Stroke
a b s t r a c t This study investigated the prevalence and type of electrocardiography (ECG) abnormalities, and their possible association with the clinical/radiological findings in 118 consecutive patients with non-traumatic, non-neoplastic intracerebral hemorrhage (ICH). ECG frequently demonstrates abnormalities in patients with ischemic stroke and subarachnoid hemorrhage, but little is known of ECG changes in ICH patients. Clinical and radiological information was retrospectively reviewed. ECG recordings that were obtained within 24 hours of the initial hemorrhage were analyzed. Sixty-six patients (56%) had one or more ECG abnormalities. The most frequent was ST depression (24%), followed by left ventricular hypertrophy (20%), corrected QT interval (QTc) prolongation (19%), and T wave inversion (19%). The logistic regression analysis demonstrated the following: insular involvement was an independent predictive factor of ST depression (p < 0.001; odds ratio OR 10.18; 95% confidence interval [CI] 2.84–36.57); insular involvement (p < 0.001; OR 23.98; 95% CI 4.91–117.11) and presence of intraventricular hemorrhage (p < 0.001; OR 8.72; 95% CI 2.69–28.29) were independent predictive factors of QTc prolongation; deep hematoma location (p < 0.001; OR 19.12; 95% CI 3.82–95.81) and hematoma volume >30 ml (p = 0.001; OR 6.58; 95% CI 2.11–20.46) were independent predictive factors of T wave inversion. We demonstrate associations between ECG abnormalities and detailed characteristics of ICH. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Electrocardiography (ECG) often demonstrates abnormalities including ST elevation, ST depression, inverted T wave, and corrected QT interval (QTc) prolongation in patients with central nervous system disorders in the absence of known organic heart disease. In fact, such ECG abnormalities are frequently found in patients with acute stroke, especially ischemic stroke and subarachnoid hemorrhage (SAH) [1–5]. However, in intracerebral hemorrhage (ICH), little is understood about the relationship between ECG abnormalities and radiological findings, including hematoma size, location, and extension [5–8]. We investigated the prevalence and type of ECG abnormalities in patients with ICH, and the possible associations with clinical/radiological findings. 2. Materials and methods 2.1. Patients This study was conducted with the approval of the Ethics Committee of the National Defense Medical College. Written ⇑ Corresponding author. Tel.: +81 42 995 1656; fax: +81 42 996 5207. E-mail address: [email protected] (S. Takeuchi).
informed consent was not required because of the retrospective nature of the investigation. A total of 161 patients were admitted to our hospital with ICH not related to a vascular abnormality or aneurysm, trauma, or neoplasm, between 2004 and 2014. CT scans or magnetic resonance angiography (MRA) were performed for all patients to exclude a ruptured aneurysm or other vascular abnormalities. All patients underwent 12 lead ECG examinations on admission. Twenty patients were excluded from the study because of unavailable clinical and/or radiological findings. In addition, another 23 patients were excluded because of cardiac pacemakers, a past history of ischemic heart disease and/or arrhythmia, or an interval between the onset and admission of >24 hours. Therefore, this study included 118 patients in total. The data were retrospectively reviewed to establish the clinical characteristics such as age, sex, level of consciousness, hematoma location, side of hematoma, hematoma volume, midline shift, presence of intraventricular hemorrhage (IVH), presence of SAH, presence of insular involvement, and ECG findings. The level of consciousness was assessed using the Glasgow coma scale at admission. The hematoma location was classified into lobar, deep (basal ganglia and thalamus), cerebellum, and brainstem. Hematoma volume was calculated by the ABC/2 formula, where A and B are the perpendicular maximal diameters of the lesion, and C is the total length in the vertical plane [9].
http://dx.doi.org/10.1016/j.jocn.2015.04.028 0967-5868/Ó 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
2
S. Takeuchi et al. / Journal of Clinical Neuroscience xxx (2015) xxx–xxx
2.2. ECG analyses ECG recordings that were obtained within 24 hours of the initial hemorrhage were analyzed by the same observer, who was unaware of the clinical details. The ECG abnormalities were defined according to the following established criteria [7]: (a) rhythm: sinus rhythm, atrial fibrillation, atrial flutter, atrial and ventricular premature complexes (ectopic beats); (b) heart rate: sinus bradycardia (<60/min), sinus tachycardia (>100/min); (c) PR interval: short <120 ms, prolonged >200 ms; (d) QRS width: prolonged >120 ms; (e) QT interval: measured using the tangent method [10], where QT was corrected for heart rate (QTc) using the Bazett formula, and was considered prolonged at >450 ms in females and >440 ms in males; (f) ST segment depression: horizontal or down sloping ST segment with (>0.05 mV) or without ST-J depression (measured 80 ms after the J point); (g) ST segment elevation: upward convexity of the ST segment (>0.1 mV) with or without ST-J elevation, measured 80 ms after the J point; (h) prominent U wave > 25% of the highest T wave in precordial leads; (i) left ventricular hypertrophy (LVH): Sokolow voltage criteria (S V1/V2 + R V5/V6 > 35 mV). 2.3. Statistical analyses Univariate analyses were used to identify the possible relationships between each ECG abnormality and clinical characteristics, using the chi-squared test or Fisher’s extract test. An ECG abnormality was analyzed if it was observed in more than 10% of patients. Multivariate logistic regression analyses were used to determine the factors that were independently associated with ECG abnormality. All variables with a significance level of p < 0.15 in the univariate analysis were included as independent variables in the logistic regression analyses. Only those variables with p < 0.15 in two-tailed tests were retained within the model. Odds ratios (OR) and 95% confidence intervals (CI) were reported. A value of p < 0.05 was considered to be statistically significant. All statistical analyses were performed using SPSS (version 11.0; IBM Corporation, Armonk, NY, USA). 3. Results The clinical and radiological findings are summarized in Table 1. Of the 118 patients, 65 were men (55%) and 53 women, with a mean age of 63.5 years (range: 32–91). The Glasgow coma scale score was 68 in 28 patients (24%). The past history included hypertension in 71 patients (60%), hypercholesterolemia in 23 (19%), and diabetes mellitus in 13 (11%). The hematoma was located in a deep region in 64 patients (54%), the lobar region in 33 (28%), the cerebellum in 10 (8%), and the brainstem in 11 (9%), and it was on the right side in 43 (45%). The mean hematoma volume was 28.3 ml, and was >30 ml in 40 patients (34%). The mean midline shift was 3.2 mm, and was >5 mm in 28 patients (24%). The associated lesions included IVH in 32 patients (27%) and SAH in 13 (11%). Insular involvement of the hematoma was observed in 13 patients (11%). One or more ECG abnormalities were detected in 66 patients (56%). The most frequently observed ECG abnormality was ST depression (24%), followed by LVH (20%), QTc prolongation (19%), and T wave inversion (19%). No life threatening arrhythmias or ECG changes were seen. Three patients developed ischemic heart diseases, including angina (n = 2) and myocardial infarction (n = 1). The ECG findings within 24 hours after the onset were ST depression in one patient and LVH in two.
Table 1 Baseline characteristics of 118 patients with intracerebral hemorrhage Characteristics Age Mean years ± SD 660 years, n (%) Male, n (%) GCS 6 8, n (%) Location of hematoma, n (%) Lobar Deep Cerebellum Brainstem Right sided hematoma, n (%)* Hematoma volume Mean ml ± SD >30 ml, n (%) Midline shift, n (%) Mean mm ± SD >5 mm, n (%) IVH, n (%) SAH, n (%) Insular involvement, n (%) ECG abnormalities, n (%) ST depression LVH QTc prolongation Inverted T wave Sinus tachycardia Atrial fibrillation CRBBB ST elevation PVC Sinus bradycardia
Patient data 63.5 ± 14.7 47 (40) 65 (55) 28 (24) 33 64 10 11 43
(28) (54) (8) (9) (45)
28.3 ± 25.3 40 (34) 3.2 ± 8.3 28 (24) 32 (27) 13 (11) 13 (11) 66 (56) 28 (24) 24 (20) 23 (19) 23 (19) 11 (9) 9 (8) 7 (6) 7 (6) 5 (4) 5 (4)
* Data from 95 patients, excluding 23 who had a hematoma located in bilateral hemispheres (n = 2), cerebellum (n = 10), or brainstem (n = 11). CRBBB = complete right bundle branch block, ECG = electrocardiography, GCS = Glasgow coma scale, IVH = intraventricular hemorrhage, LVH = left ventricular hypertrophy, PVC = premature ventricular contraction, QTc = corrected QT interval, SAH = subarachnoid hemorrhage, SD = standard deviation.
3.1. Analysis of ST depression The univariate analysis showed that ST depression was associated with deep hematoma location (p = 0.004) and ICH involvement of the insular cortex (p < 0.001; Table 2). The logistic regression analysis demonstrated that insular involvement (p < 0.001; OR 10.18; 95% CI 2.84–36.57) was an independent predictive factor of ST depression.
3.2. Analysis of LVH The univariate analysis showed that LVH was not associated with any clinicoradiological characteristics (Table 2). The logistic regression analysis also demonstrated that there was no independent predictive factor of LVH.
3.3. Analysis of QTc prolongation The univariate analysis showed that QTc prolongation was associated with deep hematoma location, midline shift >5 mm, presence of IVH, and ICH involvement of the insular cortex (p = 0.011, p = 0.026, p < 0.001, and p < 0.001, respectively; Table 2). The logistic regression analysis demonstrated that insular involvement (p < 0.001; OR 23.98; 95% CI 4.91–117.11) and the presence of IVH (p < 0.001; OR 8.72; 95% CI 2.69–28.29) were independent predictive factors of QTc prolongation.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
* Data from 95 patients, excluding 23 who had a hematoma located in both hemispheres (n = 2), cerebellum (n = 10), or brainstem (n = 11). These patients consisted of 18 with and 77 without ST depression; 18 with and 77 without LVH; 22 with and 73 without QTc prolongation; 21 with and 74 without an T wave inversion. GCS = Glasgow coma scale, IVH = intraventricular hemorrhage, LVH = left ventricular hypertrophy, QTc = corrected QT interval, SAH = subarachnoid hemorrhage.
0.179 0.434 0.129 0.129 20 (21) 24 (25) 8 (8) 8 (8) 19 (21) 27 (30) 11 (12) 4 (4) (32) (18) (7) (32) 9 5 2 9
0.308 0.235 0.523 <0.001
5 (21) 10 (42) 4 (17) 5 (21)
23 (24) 22 (23) 9 (10) 8 (9)
0.794 0.120 0.298 0.136
10 (43) 15 (65) 4 (17) 10 (43)
18 (19) 17 (18) 9 (9) 3 (3)
0.026 <0.001 0.278 <0.001
8 8 5 5
(35) (35) (22) (22)
0.003 26 (27) 14 (61) 0.329 30 (32) 10 (43) 1.000 32 (34) 8 (33) 28 (30) 12 (43)
0.262
0.118 <0.001 0.206 0.119 0.329 30 44 10 11 38 29 (32) 42 (47) 8 (9) 11 (12) 37 (48) 4 (14) 22 (79) 2 (7) 0 (0) 9 (50)
0.091 0.004 1.000 0.064 1.000
5 (21) 17 (71) 2 (8) 0 (0) 6 (33)
28 (30) 47 (50) 8 (9) 11 (12) 40 (52)
0.454 0.107 1.000 0.117 0.195
3 (13) 18 (78) 2 (9) 0 (0) 10 (45)
30 (32) 46 (48) 8 (8) 11 (12) 36 (49)
0.118 0.011 1.000 0.119 0.811
3 (13) 20 (87) 0 (0) 0 (0) 8 (38)
(32) (46) (11) (12) (51)
0.236 0.162 0.179 95 35 (37) 49 (52) 20 (21) 90 35 (39) 47 (52) 25 (28) 28 12 (43) 18 (64) 3 (11)
Patients, n Age 6 60 years, n (%) Male sex, n (%) GCS 6 8, n (%) Hematoma location, n (%) Lobar Deep Cerebellum Brainstem Right sided hematoma, n (%)* Hematoma volume>30 ml, n (%) Midline shift > 5 mm, n (%) IVH, n (%) SAH, n (%) Insular involvement, n (%)
0.826 0.285 0.077
24 7 (29) 15 (63) 2 (8)
94 40 (43) 50 (53) 26 (28)
0.254 0.494 0.060
23 10 (43) 14 (61) 4 (17)
95 37 (39) 51 (54) 24 (25)
0.813 0.642 0.587
23 12 (52) 16 (70) 8 (35)
p value Without ST depression With ST depression Characteristics
Table 2 Results of the univariate analyses of ECG abnormalities
p value
With LVH
Without LVH
p value
With QTc prolongation
Without QTc prolongation
p value
With T wave inversion
Without T wave inversion
S. Takeuchi et al. / Journal of Clinical Neuroscience xxx (2015) xxx–xxx
3
3.4. Analysis of T wave inversion The univariate analysis showed that T wave inversion was associated with a deep hematoma location (p < 0.001) and hematoma volume >30 ml (p = 0.003; Table 2). The logistic regression analysis demonstrated that a deep hematoma location (p < 0.001; OR 19.12; 95% CI 3.82–95.81) and hematoma volume >30 ml (p = 0.001; OR 6.58; 95% CI 2.11–20.46) were independent predictive factors of T wave inversion. 4. Discussion The present study showed that ECG abnormalities are common (56%) in patients with ICH, and that the most frequently observed ECG abnormalities were ST depression, LVH, QTc prolongation, and an T wave inversion. Some previous studies also found that these abnormalities were frequently observed in the acute phase of ICH [7,11]. The mechanism by which stroke, including ICH, leads to ECG abnormalities remains unclear. However, there is some evidence that stroke-related ECG changes are a manifestation of autonomic dysregulation, caused by overactivity of the sympathetic limb of the autonomic nervous system [12]. In addition, the insular cortex may be important for the regulation of sympathetic activity. In fact, both experimental investigations and clinical studies suggest that a tissue defect or irritation in the insular region may lead to abnormal cardiac function [13–17], possibly mediated through reduced inhibition of the sympathetic activity, resulting in an increased release of catecholamines [14,18]. Given this known role of the insular cortex in autonomic nervous system function [11], we assessed the effect of insular involvement as well as other characteristics. The present study also revealed that QTc prolongation was independently associated with the presence of IVH and insular involvement, consistent with the previous reports [7]. Both this and previous studies [7,11] have not identified a patient with critical arrhythmia in acute stroke patients. However, a prolonged QTc interval is well known to lead to an increased risk of ventricular tachyarrhythmia, particularly polymorphic ventricular tachycardia (torsade de pointes) [2–4]. In turn, polymorphic arrhythmia may rapidly develop into ventricular fibrillation and cause sudden death. Therefore, careful monitoring is essential if QTc prolongation is detected. ST changes and an T wave inversion are considered to be representative ECG changes of ischemic heart disease, and consequently cause considerable diagnostic problems in the acute setting, possibly misleading physicians in the initial diagnosis. This misdiagnosis may result in adverse consequences, especially in patients with ICH because treatment of ischemic heart disease consists of anticoagulants, antiplatelet therapy, or fibrinolytic agents with or without angioplasty [7]. In particular, an T wave inversion is non-specific, so a neurogenic cause is important to consider in order to avoid inappropriate therapies [19]. It is unclear whether ST changes are associated with ICH characteristics such as insular involvement [7,13]. No association was found between ST changes and the clinical characteristics of 31 ICH patients [7], whereas ST elevation was associated with insular involvement in 179 stroke patients, including 17 ICH patients [13]. Our study showed that ST depression was independently associated with insular involvement. The discrepancy between these studies may be due to differences in the number and population of patients. We could not investigate the predictive factors of ST elevation because of the small sample size and these points need further investigation.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028
4
S. Takeuchi et al. / Journal of Clinical Neuroscience xxx (2015) xxx–xxx
The present study also revealed that T wave inversion was independently associated with a deep hematoma location and large hematoma volume. In a previous study, T wave inversion was detected in five of 31 patients, and the univariate analysis showed no association with any characteristics, due to the small sample size [7]. The present study included a larger number of patients, and adopted a multivariable analysis. Therefore, we believe our results are more reliable for this point. The present and previous studies have shown that LVH is one of the most frequent ECG abnormalities, but the occurrence of LVH may not be considered as a new development in all patients with ICH because LVH may reflect the long term effects of chronic hypertension [8]. The absence of any relationship between LVH and the clinicoradiological findings supports this interpretation. Several experimental and clinical studies have shown a right sided dominance for sympathetic cardiovascular effects [13– 15,17,18,20]. In our study, the frequency of ECG abnormalities was comparable between the patients with right and left hemispheric lesions, consistent with previous studies [11,20]. This point needs further investigation. There are several serious limitations to our study. As a single center retrospective study, it was subject to observational and assessment bias. It also had a small sample size, which reduced the statistical power and increased the chance of type 2 errors. Furthermore, ECG data prior to ICH were not available for comparison, and 20 patients were excluded because of unavailable data. We did not assess the relationship between the ECG abnormalities and outcomes because no patient developed a critical arrhythmia. Also, follow-up ECG were not obtained, resulting in a lack of data on whether the ECG changes were transient or permanent. Despite these limitations, the present findings allow for some useful clinical speculations. 5. Conclusion We studied the relationship between ECG abnormalities and the detailed characteristics of ICH in a large series of patients, and showed that insular involvement was an independent predictive factor of ST depression, insular involvement and the presence of IVH were independent predictive factors of QTc prolongation, and a deep hematoma location and hematoma volume >30 ml were independent predictive factors of T wave inversion.
6. Conflicts of Interest/Disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication. References [1] Christensen H, Fogh Christensen A, Boysen G. Abnormalities on ECG and telemetry predict stroke outcome at 3 months. J Neurol Sci 2005;234:99–103. [2] Lanzino G, Kongable GL, Kassell NF. Electrocardiographic abnormalities after nontraumatic subarachnoid hemorrhage. J Neurosurg Anesthesiol 1994;6:156–62. [3] Andreoli A, di Pasquale G, Pinelli G, et al. Subarachnoid hemorrhage: frequency and severity of cardiac arrhythmias. A survey of 70 cases studied in the acute phase. Stroke 1987;18:558–64. [4] Machado C, Baga JJ, Kawasaki R, et al. Torsade de pointes as a complication of subarachnoid hemorrhage: a critical reappraisal. J Electrocardiol 1997;30:31–7. [5] Junttila E, Vaara M, Koskenkari J, et al. Repolarization abnormalities in patients with subarachnoid and intracerebral hemorrhage: predisposing factors and association with outcome. Anesth Analg 2013;116:190–7. [6] Popescu D, Laza C, Mergeani A, et al. Lead electrocardiogram changes after supratentorial intracerebral hemorrhage. Maedica (Buchar) 2012;7:290–4. [7] van Bree MD, Roos YB, van der Bilt IA, et al. Prevalence and characterization of ECG abnormalities after intracerebral hemorrhage. Neurocrit Care 2010;12:50–5. [8] Arruda WO, de Lacerda Júnior FS. Electrocardiographic findings in acute cerebrovascular hemorrhage. A prospective study of 70 patients. Arq Neuropsiquiatr 1992;50:269–74. [9] Kothari RU, Brott T, Broderick JP, et al. The ABCs of measuring intracerebral hemorrhage volumes. Stroke 1996;27:1304–5. [10] Lepeschkin E, Surawicz B. The measurement of the Q-T interval of the electrocardiogram. Circulation 1952;6:378–88. [11] Tatschl C, Stöllberger C, Matz K, et al. Insular involvement is associated with QT prolongation: ECG abnormalities in patients with acute stroke. Cerebrovasc Dis 2006;21:47–53. [12] Samuels MA. The brain-heart connection. Circulation 2007;116:77–84. [13] Christensen H, Boysen G, Christensen AF, et al. Insular lesions, ECG abnormalities, and outcome in acute stroke. J Neurol Neurosurg Psychiatry 2005;76:269–71. [14] Cechetto DF. Experimental cerebral ischemic lesions and autonomic and cardiac effects in cats and rats. Stroke 1993;24:I6–9 [discussion I10–2]. [15] Colivicchi F, Bassi A, Santini M, et al. Cardiac autonomic derangement and arrhythmias in right-sided stroke with insular involvement. Stroke 2004;35:2094–8. [16] Oppenheimer SM, Wilson JX, Guiraudon C, et al. Insular cortex stimulation produces lethal cardiac arrhythmias: a mechanism of sudden death? Brain Res 1991;550:115–21. [17] Oppenheimer SM. Neurogenic cardiac effects of cerebrovascular disease. Curr Opin Neurol 1994;7:20–4. [18] Hachinski VC, Oppenheimer SM, Wilson JX, et al. Asymmetry of sympathetic consequences of experimental stroke. Arch Neurol 1992;49:697–702. [19] Mandrioli J, Zini A, Cavazzuti M, et al. Neurogenic T wave inversion in pure left insular stroke associated with hyperhomocysteinemia. J Neurol Neurosurg Psychiatry 2004;75:1788–9. [20] Afsar N, Fak AS, Metzger JT, et al. Acute stroke increases QT dispersion in patients without known cardiac diseases. Arch Neurol 2003;60:346–50.
Please cite this article in press as: Takeuchi S et al. Electrocardiograph abnormalities in intracerebral hemorrhage. J Clin Neurosci (2015), http://dx.doi.org/ 10.1016/j.jocn.2015.04.028