- Email: [email protected]
Pre-existing structural abnormalities of the limbic system in transient global amnesia
Journal of Clinical Neuroscience 22 (2015) 843–847
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Pre-existing structural abnormalities of the limbic system in transient global amnesia Kang Min Park a,1, Yong Hee Han b,1, Tae Hyung Kim c, Chi Woong Mun b,c, Kyong Jin Shin a, Sam Yeol Ha a, JinSe Park a, Sung Eun Kim a,⇑ a b c
Department of Neurology, Haeundae Paik Hospital, Inje University College of Medicine, Haeundae-ro 875, Haeundae-gu, Busan 612-896, Republic of Korea Department of Biomedical Engineering/u-HARC, Inje University, Gimhae, Republic of Korea Department of Health Science and Technology, Inje University, Gimhae, Republic of Korea
a r t i c l e
i n f o
Article history: Received 17 September 2014 Accepted 8 November 2014
Keywords: Amnesia Disease susceptibility Limbic system Morphometry
a b s t r a c t This study aimed to investigate the clinical and radiological findings in patients with transient global amnesia and to evaluate structural abnormalities using voxel-based morphometry. The subjects were diagnosed with transient global amnesia. For the voxel-based morphometry analyses, Statistical Parametric Mapping, running on the MATLAB platform (MathWorks, Natick, MA, USA), was employed to analyze the structural differences between patients with transient global amnesia and control subjects. Eighty patients met the inclusion criteria. Twenty-three patients (29%) were men, and 57 patients (71%) were women. There were significantly more women among the transient global amnesia patients compared with the general Korean population. MRI revealed hippocampal cavities in 41 patients (51%), and the incidence of such cavities was significantly different from that of the control subjects (24%). There were no differences in the clinical factors between the patients with and without hippocampal cavities. Diffusion-weighted imaging was performed in 54 patients, and 13 patients (24%) exhibited high signal intensity in the hippocampus. There were also no differences in the clinical factors between the patients with and without high signal intensities in the hippocampus on diffusion-weighted imaging. Twenty-six patients underwent three-dimensional volumetric T1-weighted imaging that produced results suitable for voxel-based morphometry, and these patients presented with gray matter volume reductions in the hippocampus, cingulum, and cerebellum. There were significant structural differences in the limbic structures between patients with transient global amnesia and the control subjects that might have contributed to vulnerability of the memory pathways of the patients with transient global amnesia. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Transient global amnesia (TGA) is a disorder that is characterized by the sudden onset of transient memory impairment with anterograde and partial retrograde amnesia that lasts for several hours followed by complete resolution [1]. Data from clinical and psychometric studies have suggested explicit involvements of brain structures that are implicated in memory processing, such as the Papez circuit, in patients with TGA [2–4]. Altered perfusion or hypofunction of the temporal area has also been demonstrated in single-photon emission computed tomography (SPECT) and positron emission tomography studies [5,6]. Additionally, the
⇑ Corresponding author. Tel.: +82 51 797 1195; fax: +82 51 797 1196. 1
E-mail address: [email protected] (S.E. Kim). These authors have contributed equally to the manuscript.
http://dx.doi.org/10.1016/j.jocn.2014.11.017 0967-5868/Ó 2015 Elsevier Ltd. All rights reserved.
published studies that have used the diffusion-weighted imaging (DWI) form of MRI have demonstrated that signal changes in the lateral hippocampus are present in patients with TGA [7,8]. Nevertheless, there is no direct evidence of structural abnormalities in areas implicated in memory processing in patients with TGA. MRI-based measures of atrophy are regarded as valid markers of disease state and progression [9]. Voxel-based morphometry (VBM) was designed to increase the sensitivity of comparisons of the local compositions of different brain tissue types, while discounting positional and other large-scale volumetric differences in gross anatomy [10,11]. VBM is a fully automated, computerized, quantitative MRI analysis, therefore it is not biased toward any particular structure. However, limited studies have been conducted using VBM in patients with TGA. This study aimed to investigate the clinical and radiological findings in patients with TGA, and to evaluate the structural abnormalities using VBM analysis.
844
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
2. Materials and methods 2.1. Patient subjects This study was conducted with the approval of the Institutional Review Board of Inje University. The study was performed using consecutive patients from a single, tertiary referral hospital. The subjects were diagnosed with TGA between March 2010 and December 2013. The inclusion criteria were patients diagnosed with TGA at Haeundae Paik Hospital, and patients with no structural brain lesions on MRI. The diagnoses of TGA were made according to the criteria of Hodges and Warlow [1]. The exclusion criteria were patients with a history of neurological or psychiatric disease due to the potential association with brain atrophy, or patients with a history of drug or alcohol abuse, because this might have interfered with cognitive performance. All patients underwent a standard neurological examination and a structured interview to obtain information on vascular risk factors (diabetes, hypertension and dyslipidemia) and characteristics of the episode (duration, precipitating event, and associated symptoms). The imaging evaluation involved DWI and three-dimensional (3D) volumetric T1-weighted images. 2.2. Control subjects The control group consisted of 29 age and sex-matched healthy subjects. All subjects had a normal neurological examination and no exclusion factors in their medical history. All control subjects had normal MRI scans. Twelve subjects were men, and the mean age was 59.0 ± a standard deviation of 11.9 years.
Jacobian determination, and spatial smoothing with 6 mm Gaussian kernel. For the group analysis, two sample t-tests, patients versus normal controls, were performed using a general linear model to identify whole brain gray matter alterations with significant differences. Significance was defined at the level of p < 0.05, after correcting for the false discovery rate to compensate for type 1 errors and applying an extent threshold of 400 voxels. To consider other effects such as age, sex, and total intracranial volume, the two sample t-tests were repeated once with these factors, which were entered as covariates. 2.5. Statistical analysis We compared the clinical factors between the patients with and without signal changes in the hippocampus on DWI [12] and investigated the incidence of hippocampal cavities in patients with TGA [13]. The hippocampal cavities were detected based on the agreement of two investigators. Additionally, we compared the clinical factors between the patients with and without hippocampal cavities. These comparisons were made using chi-squared tests or Fisher’s exact tests for categorical variables, and Student’s t-tests or Mann–Whitney U tests for numerical variables. The categorical variables were presented as frequencies and percentages. The numerical variables with normal distribution were presented as the mean ± standard deviation, and those without normal distribution were described as the median with the 95% confidence interval (CI) and ranges. 3. Results
2.3. MRI data acquisition All scans were performed on a 3.0 Tesla MRI scanner (AchievaTx; Phillips Healthcare, Best, The Netherlands) equipped with an eight channel head coil. All subjects underwent conventional brain MRI protocols that included axial and coronal two-dimensional T2-weighted images, that were obtained with a turbo spin echo sequence (repetition time [TR]/echo time [TE] = 3000/80 ms, slice thickness = 5 mm, echo train length = 14, field of view [FOV] = 210 mm, and matrix size = 512 512) and axial and coronal two-dimensional T1-weighted images that were obtained with an inversion recovery sequence (inversion time [TI] = 800 ms, TR/TE = 2000/10 ms, slice thickness = 5 mm, echo train length = 7, FOV = 210 mm, and matrix size = 512 512). DWI was performed with a single-shot spin echo-based echo planer image with the following parameters: TR/TE = 6000/83.03 ms, slice thickness = 5 mm, echo train length = 53, FOV = 240 mm, matrix size = 256 256, and b-value = 2000 s/mm2. The 3D T1-weighted images were obtained with a turbo field echo sequence with the following parameters: TI = 1300 ms, TR/TE = 8.6/3.96 ms, flip angle = 8°, and a 1 mm3 isotropic voxel size. Sagittal-oriented high-resolution contiguous 3D T1-weighted images were obtained. To speed up data acquisition, sensitivity encoding parallel imaging with an acceleration factor of two was applied. 2.4. MRI data processing and VBM analysis VBM, based on 3D T1-weighted images, was used to analyze volumetric differences. Image processing was performed using the VBM8 toolbox (Structural Brain Mapping Group, University of Jena, Thuringia, Germany), implemented in Statistical Parametric Mapping 8 (SPM 8, Wellcome Trust Centre for Neuroimaging, University College London, London, UK). The VBM data were processed in the standard manner including the following steps: spatial normalization to the Montreal Neurological Institute template, gray matter segmentation, intensity modulation using
The numbers of patients who underwent structured interviews to obtain the clinical factors and who were subjected to radiological investigations are reported in Table 1. 3.1. Clinical factors Eighty patients met the inclusion criteria. Twenty-three patients (29%) were men, and 57 patients (71%) were women. There were significantly more women among the TGA patients compared with the general Korean population (p < 0.001; 5,811,891 men [44%] and 7,397,720 women [56%] out of 12,209,611 Korean subjects between 50 and 79 years old based on the 2010 data from the Korean Statistical Information Service). The mean age of onset was 60.3 ± 7.3 years. Among the patients with TGA, the presenting TGA event was the first such event in 77 patients, and three patients reported previous episodes of TGA. The median duration of amnesia was 5 hours (95% CI 4.0–6.0 hours, range 0.5–23 hours). Thirty-five (44%) of the 80 patients had a history of vascular risk factors, with hypertension in 27, diabetes in 10, and dyslipidemia in eight patients. Forty-four patients (55%) had a history of precipitating events, with emotional stress in 26, physical effort in nine, and water contact/temperature change in nine patients. Eleven patients (14%) had associated symptoms, with headache in four, general weakness in three, nausea in three, and dizziness in one. Table 1 Numbers of patients with transient global amnesia who underwent structured interviews to obtain clinical factors, radiological investigations, and electroencephalography Medical examination Structured interview Conventional MRI including T1 and T2-weighted imaging Diffusion-weighted imaging Three-dimensional volumetric T1-weighted imaging
n 80 (100%) 80 (100%) 54 (66%) 26 (33%)
845
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
3.2. Radiological findings MRI of the TGA patients revealed hippocampal cavities in 41 patients (51%) which were right localized in 18, left localized in 14, and bilateral in nine patients. The cavities were single, had rounded or crescent appearances and were located around the CA1 area. The incidence of hippocampal cavities was significantly different between the TGA and control subjects (41/80 versus 7/29, respectively, p = 0.0159). Of the 80 patients with TGA, DWI was performed in 54 patients, and 13 patients (24%) exhibited high signal intensities on DWI. All of the lesions on DWI were located in the hippocampus, and were right localized in five patients, left localized in six patients, and bilateral in two patients. The median time delay from symptom onset to DWI was 7 hours (95% CI 6.0– 11.0 hours, range 2–480 hours). There were no differences in age of onset (59.6 versus 61.1 years, p = 0.3555), sex (male 14/41 versus female 9/39, p = 0.3974), median duration of amnesia (5.0 versus 5.0, p = 0.7214), presence of vascular risk factors (20/41 versus 15/39, p = 0.4811), presence of precipitating events (19/41 versus 25/39, p = 0.1703), or presence of associated symptoms (7/41 versus 4/39, p = 0.5753) between the patients with and without hippocampal cavities, respectively. In addition, there were no differences in age of onset (57.6 versus 61.4 years, p = 0.1093), male sex (2/13 versus 13/41, p = 0.3112), median duration of amnesia (11.5 versus 4.0, p = 0.0739), presence of vascular risk factors (4/13 versus 21/41, p = 0.2228), presence of precipitating events (10/13 versus
22/41, p = 0.1988), presence of associated symptoms (3/13 versus 6/41, p = 0.6702), or median time delay from symptom onset to DWI (10 versus 7 hours, p = 0.1658) between patients with and without signal changes in the hippocampus on DWI, respectively. 3.3. VBM Among the 80 patients with TGA, 26 patients underwent 3D volumetric T1-weighted imaging suitable for VBM. Seven patients (27%) were men, and the mean age of onset was 60.7 ± 7.4 years. The VBM analysis comparing these 26 TGA patients and 29 control subjects revealed that TGA patients exhibited significant reductions in the gray matter volumes of the left hippocampus, left amygdala, bilateral cingulum, and right posterior lobe of the cerebellum. However, there were no structures with increased gray matter volumes in TGA patients compared with the control subjects. Furthermore, after adjusting for age, sex, and intracranial volume, the patients with TGA still exhibited significant reductions in the gray matter volumes of the left hippocampus, bilateral cingulum, and right posterior lobe of the cerebellum. However, the significance of the reduction in the gray matter volume of the amygdala disappeared (Fig. 1, Table 2, Supp. Fig. 1). 4. Discussion The main finding of our study was that patients with TGA exhibited significant reductions in the gray matter volumes of
Fig. 1. (A) Sagittal, (B) coronal and (C) axial views of the result of voxel-based morphometry analysis in grayscale on a glass-brain render of the brain after adjusting for age, sex, and intracranial volume. Patients with transient global amnesia exhibit significant reductions in gray matter volumes of the left hippocampus, bilateral cingulum, and right cerebellum at a statistically significant level, p < 0.05 (false discovery rate corrected). The red arrows indicate the center point. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this paper.) Table 2 Brain regions displaying reductions in gray matter volumes by voxel-based morphometry in the patients with transient global amnesia, compared with control subjects after adjusting for age, sex, and intracranial volume Region
Side
MNI coordinates
Cerebellum, posterior lobe Cingulum Hippocampus Cingulum
Right Right Left Left
(6,
MNI = Montreal Neurological Institute.
73.5, 18) (6, 45, 9) ( 21, 13.5, 16.5) ( 4.5, 18, 43.5)
Peak intensity
Cluster size (voxels)
7.7406 7.1054 6.2612 4.6584
13554 1323 1979 541
846
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
the hippocampus, cingulum, and cerebellum. We also found that there were significantly more women in the TGA group and that there were no differences in the clinical factors between the patients with and without signal changes in the hippocampus on DWI. Additionally, we confirmed the increased incidence of hippocampal cavities among the patients with TGA, and there were no differences in the clinical factors between the patients with and without hippocampal cavities. Although TGA is well-characterised clinically, the etiology and pathogenesis of TGA remains unclear. Several hypotheses, including arterial transient ischemic attacks, epileptic seizures, transient mesial temporal ischemia induced by venous congestion, and migraine have been postulated, but definite evidence supporting these mechanisms is lacking [14–15]. Similarly, the findings of this study do not provide direct evidence regarding the pathogenesis of TGA. However, we postulate that pre-existing abnormalities in limbic structures might contribute to an increased vulnerability to TGA events. This assumption is supported by the observation that patients who have experienced previous TGA events exhibit a higher frequency of recurrence compared with the de novo incidence of TGA across the lifetime; the rate of annual recurrence in patients with TGA is between 6 and 10%, whereas the annual incidence in the general population is between 0.003 and 0.008% [2]. Although no MRI scans were conducted in the pre-symptomatic period, we assume that the reductions in the volumes of the gray matter in the limbic structures were pre-existing. The range of the volume reductions was multifocal and extensive, therefore, volume reductions might not have developed suddenly after the onset of the TGA symptoms. Additionally, no previous reports have demonstrated any acute structural changes on brain MRI following the onset of symptoms, with the exception of small diffusion restrictions in the hippocampus. Altogether, these findings suggest that the differences in the MRI results were due to events that occurred prior to the onset of TGA, rather than as consequences of TGA. We also infer that precipitating events might trigger an episode of TGA in patients with more vulnerable memory pathways due to pre-existing abnormalities in the limbic structures. The hippocampus has a critical role in memory, particularly episodic memory. Semantic memory primarily activates the frontal and temporal cortexes, whereas episodic memory activity, at least initially, is concentrated in the hippocampus [16,17]. Neuropsychological examinations performed during the acute phase have revealed that TGA is a selective disorder of episodic memory that results in no impairments of other components of memory [2,3]. Additionally, during TGA episodes, patients are characterized as being unable to encode new information into long-term memory, and the hippocampus is known to play a role in the formation of long-term memory from short-term memory [2,3]. Thus, it can be inferred that memory disturbance in patients with TGA originates from impairments of the hippocampus. We found that patients with TGA exhibited reductions in hippocampal gray matter volumes, and this finding supports the involvement of the hippocampus in the pathogenesis of TGA. Additionally, our findings are consistent with the results of previous functional studies that have used SPECT or positron emission tomography and indicated that hippocampal abnormalities exist in TGA [18]. However, we cannot exclude the effects of the hippocampal cavities on the reductions in hippocampal volumes. The Monroe– Kellie doctrine states that losses of parenchymal tissue result from increases in the adjacent CSF space. Considering this, increased incidence of hippocampal cavities might induce hippocampal tissue loss. However, in our study, hippocampal cavities were found in the right side of 18 TGA patients and the left side of 14 TGA patients, and hippocampal cavities were detected in four control subjects on the right side and three control subjects on the left side. Although, hippocampal volume reductions were detected
only on the left side in patients with TGA, the incidence of hippocampal cavities was not higher on the left side, compared to the right side (p = 1.0 by Fisher’s exact test). Thus, hippocampal volume reductions cannot be explained solely by the effects of hippocampal cavities. There is evidence that the cerebellum is also a limbic structure [19–21]. Several previous studies have demonstrated that cerebellar infarction without cortical lesions results in cognitive changes, and these findings indicate the relevance of the connecting loops between the cerebellum and the contralateral cortex, to cognitive processing [20–21]. It is also reported that some patients with TGA display cerebellar hypoperfusion on SPECT [19]. Schmahmann et al. described the so-called cerebellar cognitive affective syndrome with deficits in executive functioning, language, and spatial cognition, and behavioral changes in patients with structural cerebellar lesions [22]. Interestingly, these previous studies highlight the dominant role of the posterior lobe of the cerebellum in cognitive and affective processing [21,22], and our study revealed reductions in the gray matter volumes of the posterior lobes of the cerebellum in patients with TGA. The dominant role of the posterior lobe of the cerebellum in limbic function might be explained by evidence regarding the linkages of the parietal and paralimbic association areas with the posterior lobe of the cerebellum [21]. Several investigators have used DWI to further examine the etiology of TGA, and they have reported variable results. One study reported that high signal intensity on DWI was detected in seven of 10 patients with TGA, whereas another study did not find abnormal DWI signals in any of eight patients with TGA [4,12]. In our study, we found that only 24% of the patients with TGA exhibited high signal intensity on DWI. These conflicting results might be due to differences in the timing of examination and differences in MRI protocols, including parameters such as slice thickness [23]. Hippocampal lesions on DWI are rarely noted in the hyperacute phase, but all such lesions become visible at 48 hours [24]. The detection rates of hippocampal lesions can be as high as 88% when more appropriate MRI protocols are utilized [8,23]. Among our patients, 37 of the 54 who underwent DWI (69%) did so less than 24 hours after symptom onset, and all of our patients underwent DWI utilizing a slice thickness of 5 mm. The short time from symptom onset to DWI and the slice thickness utilized might have produced a lower hippocampal abnormality detection rate on DWI in our study. Additionally, we compared the clinical factors between patients with and without hippocampal lesions on DWI and found no differences. These findings were consistent with those of a previous report that indicated no differences in long-term cognitive outcomes between patients with and without hippocampal lesions [25]. Nakada et al. reported that the incidence of hippocampal cavities was 100% in patients with TGA, and an incidence of 31% in control subjects [13]. Additionally, another previous study reported hippocampal cavities in 60% of patients with TGA and no differences in cognitive performance between patients with and without hippocampal cavities [26]. In our study, we found hippocampal cavities in 51% of the patients with TGA, similar to the results of Nakada et al. [13], and we also confirmed an increased incidence of hippocampal cavities in patients with TGA.
5. Conclusion We demonstrated that there were significant structural differences in the limbic structures of patients with TGA and control subjects and that these reductions in limbic gray matter volumes might contribute to greater vulnerability of the memory pathways in patients with TGA. These findings suggest that the occurrence of
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
TGA events is predisposition.
associated
with
an
underlying anatomical
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. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jocn.2014.11.017. References [1] Hodges JR, Warlow CP. The aetiology of transient global amnesia. A casecontrol study of 114 cases with prospective follow-up. Brain 1990;113: 639–57. [2] Bartsch T, Deuschl G. Transient global amnesia: functional anatomy and clinical implications. Lancet Neurol 2010;9:205–14. [3] Eustache F, Desgranges B, Laville P, et al. Episodic memory in transient global amnesia: encoding, storage, or retrieval deficit? J Neurol Neurosurg Psychiatry 1999;66:148–54. [4] Gass A, Gaa J, Hirsch J, et al. Lack of evidence of acute ischemic tissue change in transient global amnesia on single-shot echo-planar diffusion-weighted MRI. Stroke 1999;30:2070–2. [5] Gonzalez-Martinez V, Comte F, de Verbizier D, et al. Transient global amnesia: concordant hippocampal abnormalities on positron emission tomography and magnetic resonance imaging. Arch Neurol 2010;67:510. [6] Baron JC, Petit-Taboue MC, Le Doze F, et al. Right frontal cortex hypometabolism in transient global amnesia. A PET study. Brain 1994;117: 545–52. [7] Bartsch T, Alfke K, Stingele R, et al. Selective affection of hippocampal CA-1 neurons in patients with transient global amnesia without long-term sequelae. Brain 2006;129:2874–84. [8] Kim J, Kwon Y, Yang Y, et al. Clinical experience of modified diffusion-weighted imaging protocol for lesion detection in transient global amnesia: an 8-year large-scale clinical study. J Neuroimaging 2014;24:331–7. [9] Frisoni GB, Fox NC, Jack Jr CR, et al. The clinical use of structural MRI in Alzheimer disease. Nat Rev Neurol 2010;6:67–77.
847
[10] Ashburner J, Friston KJ. Voxel-based morphometry–the methods. Neuroimage 2000;11:805–21. [11] Ashburner J, Friston KJ. Why voxel-based morphometry should be used. NeuroImage 2001;14:1238–43. [12] Strupp M, Bruning R, Wu RH, et al. Diffusion-weighted MRI in transient global amnesia: elevated signal intensity in the left mesial temporal lobe in 7 of 10 patients. Ann Neurol 1998;43:164–70. [13] Nakada T, Kwee IL, Fujii Y, et al. High-field, T2 reversed MRI of the hippocampus in transient global amnesia. Neurology 2005;64:1170–4. [14] Schreiber SJ, Doepp F, Klingebiel R, et al. Internal jugular vein valve incompetence and intracranial venous anatomy in transient global amnesia. J Neurol Neurosurg Psychiatry 2005;76:509–13. [15] Olesen J, Jorgensen MB. Leao’s spreading depression in the hippocampus explains transient global amnesia. A hypothesis. Acta Neurol Scand 1986;73: 219–20. [16] Knierim JJ, Lee I, Hargreaves EL. Hippocampal place cells: parallel input streams, subregional processing, and implications for episodic memory. Hippocampus 2006;16:755–64. [17] Jeneson A, Squire LR. Working memory, long-term memory, and medial temporal lobe function. Learn Mem 2012;19:15–25. [18] Takeuchi R, Yonekura Y, Matsuda H, et al. Resting and acetazolamidechallenged technetium-99m-ECD SPECT in transient global amnesia. J Nucl Med 1998;39:1360–2. [19] Yang Y, Kim JS, Kim S, et al. Cerebellar hypoperfusion during transient global amnesia: an MRI and oculographic study. J Clin Neurol 2009;5:74–80. [20] Hokkanen LS, Kauranen V, Roine RO, et al. Subtle cognitive deficits after cerebellar infarcts. Eur J Neurol 2006;13:161–70. [21] Exner C, Weniger G, Irle E. Cerebellar lesions in the PICA but not SCA territory impair cognition. Neurology 2004;63:2132–5. [22] Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain 1998;121:561–79. [23] Choi BS, Kim JH, Jung C, et al. High-resolution diffusion-weighted imaging increases lesion detectability in patients with transient global amnesia. AJNR Am J Neuroradiol 2012;33:1771–4. [24] Sedlaczek O, Hirsch JG, Grips E, et al. Detection of delayed focal MR changes in the lateral hippocampus in transient global amnesia. Neurology 2004;62:2165–70. [25] Uttner I, Prexl S, Freund W, et al. Long-term outcome in transient global amnesia patients with and without focal hyperintensities in the CA1 region of the hippocampus. Eur Neurol 2012;67:155–60. [26] Uttner I, Weber S, Freund W, et al. Hippocampal cavities are not associated with cognitive impairment in transient global amnesia. Eur J Neurol 2011;18: 882–7.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Pre-existing structural abnormalities of the limbic system in transient global amnesia Kang Min Park a,1, Yong Hee Han b,1, Tae Hyung Kim c, Chi Woong Mun b,c, Kyong Jin Shin a, Sam Yeol Ha a, JinSe Park a, Sung Eun Kim a,⇑ a b c
Department of Neurology, Haeundae Paik Hospital, Inje University College of Medicine, Haeundae-ro 875, Haeundae-gu, Busan 612-896, Republic of Korea Department of Biomedical Engineering/u-HARC, Inje University, Gimhae, Republic of Korea Department of Health Science and Technology, Inje University, Gimhae, Republic of Korea
a r t i c l e
i n f o
Article history: Received 17 September 2014 Accepted 8 November 2014
Keywords: Amnesia Disease susceptibility Limbic system Morphometry
a b s t r a c t This study aimed to investigate the clinical and radiological findings in patients with transient global amnesia and to evaluate structural abnormalities using voxel-based morphometry. The subjects were diagnosed with transient global amnesia. For the voxel-based morphometry analyses, Statistical Parametric Mapping, running on the MATLAB platform (MathWorks, Natick, MA, USA), was employed to analyze the structural differences between patients with transient global amnesia and control subjects. Eighty patients met the inclusion criteria. Twenty-three patients (29%) were men, and 57 patients (71%) were women. There were significantly more women among the transient global amnesia patients compared with the general Korean population. MRI revealed hippocampal cavities in 41 patients (51%), and the incidence of such cavities was significantly different from that of the control subjects (24%). There were no differences in the clinical factors between the patients with and without hippocampal cavities. Diffusion-weighted imaging was performed in 54 patients, and 13 patients (24%) exhibited high signal intensity in the hippocampus. There were also no differences in the clinical factors between the patients with and without high signal intensities in the hippocampus on diffusion-weighted imaging. Twenty-six patients underwent three-dimensional volumetric T1-weighted imaging that produced results suitable for voxel-based morphometry, and these patients presented with gray matter volume reductions in the hippocampus, cingulum, and cerebellum. There were significant structural differences in the limbic structures between patients with transient global amnesia and the control subjects that might have contributed to vulnerability of the memory pathways of the patients with transient global amnesia. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Transient global amnesia (TGA) is a disorder that is characterized by the sudden onset of transient memory impairment with anterograde and partial retrograde amnesia that lasts for several hours followed by complete resolution [1]. Data from clinical and psychometric studies have suggested explicit involvements of brain structures that are implicated in memory processing, such as the Papez circuit, in patients with TGA [2–4]. Altered perfusion or hypofunction of the temporal area has also been demonstrated in single-photon emission computed tomography (SPECT) and positron emission tomography studies [5,6]. Additionally, the
⇑ Corresponding author. Tel.: +82 51 797 1195; fax: +82 51 797 1196. 1
E-mail address: [email protected] (S.E. Kim). These authors have contributed equally to the manuscript.
http://dx.doi.org/10.1016/j.jocn.2014.11.017 0967-5868/Ó 2015 Elsevier Ltd. All rights reserved.
published studies that have used the diffusion-weighted imaging (DWI) form of MRI have demonstrated that signal changes in the lateral hippocampus are present in patients with TGA [7,8]. Nevertheless, there is no direct evidence of structural abnormalities in areas implicated in memory processing in patients with TGA. MRI-based measures of atrophy are regarded as valid markers of disease state and progression [9]. Voxel-based morphometry (VBM) was designed to increase the sensitivity of comparisons of the local compositions of different brain tissue types, while discounting positional and other large-scale volumetric differences in gross anatomy [10,11]. VBM is a fully automated, computerized, quantitative MRI analysis, therefore it is not biased toward any particular structure. However, limited studies have been conducted using VBM in patients with TGA. This study aimed to investigate the clinical and radiological findings in patients with TGA, and to evaluate the structural abnormalities using VBM analysis.
844
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
2. Materials and methods 2.1. Patient subjects This study was conducted with the approval of the Institutional Review Board of Inje University. The study was performed using consecutive patients from a single, tertiary referral hospital. The subjects were diagnosed with TGA between March 2010 and December 2013. The inclusion criteria were patients diagnosed with TGA at Haeundae Paik Hospital, and patients with no structural brain lesions on MRI. The diagnoses of TGA were made according to the criteria of Hodges and Warlow [1]. The exclusion criteria were patients with a history of neurological or psychiatric disease due to the potential association with brain atrophy, or patients with a history of drug or alcohol abuse, because this might have interfered with cognitive performance. All patients underwent a standard neurological examination and a structured interview to obtain information on vascular risk factors (diabetes, hypertension and dyslipidemia) and characteristics of the episode (duration, precipitating event, and associated symptoms). The imaging evaluation involved DWI and three-dimensional (3D) volumetric T1-weighted images. 2.2. Control subjects The control group consisted of 29 age and sex-matched healthy subjects. All subjects had a normal neurological examination and no exclusion factors in their medical history. All control subjects had normal MRI scans. Twelve subjects were men, and the mean age was 59.0 ± a standard deviation of 11.9 years.
Jacobian determination, and spatial smoothing with 6 mm Gaussian kernel. For the group analysis, two sample t-tests, patients versus normal controls, were performed using a general linear model to identify whole brain gray matter alterations with significant differences. Significance was defined at the level of p < 0.05, after correcting for the false discovery rate to compensate for type 1 errors and applying an extent threshold of 400 voxels. To consider other effects such as age, sex, and total intracranial volume, the two sample t-tests were repeated once with these factors, which were entered as covariates. 2.5. Statistical analysis We compared the clinical factors between the patients with and without signal changes in the hippocampus on DWI [12] and investigated the incidence of hippocampal cavities in patients with TGA [13]. The hippocampal cavities were detected based on the agreement of two investigators. Additionally, we compared the clinical factors between the patients with and without hippocampal cavities. These comparisons were made using chi-squared tests or Fisher’s exact tests for categorical variables, and Student’s t-tests or Mann–Whitney U tests for numerical variables. The categorical variables were presented as frequencies and percentages. The numerical variables with normal distribution were presented as the mean ± standard deviation, and those without normal distribution were described as the median with the 95% confidence interval (CI) and ranges. 3. Results
2.3. MRI data acquisition All scans were performed on a 3.0 Tesla MRI scanner (AchievaTx; Phillips Healthcare, Best, The Netherlands) equipped with an eight channel head coil. All subjects underwent conventional brain MRI protocols that included axial and coronal two-dimensional T2-weighted images, that were obtained with a turbo spin echo sequence (repetition time [TR]/echo time [TE] = 3000/80 ms, slice thickness = 5 mm, echo train length = 14, field of view [FOV] = 210 mm, and matrix size = 512 512) and axial and coronal two-dimensional T1-weighted images that were obtained with an inversion recovery sequence (inversion time [TI] = 800 ms, TR/TE = 2000/10 ms, slice thickness = 5 mm, echo train length = 7, FOV = 210 mm, and matrix size = 512 512). DWI was performed with a single-shot spin echo-based echo planer image with the following parameters: TR/TE = 6000/83.03 ms, slice thickness = 5 mm, echo train length = 53, FOV = 240 mm, matrix size = 256 256, and b-value = 2000 s/mm2. The 3D T1-weighted images were obtained with a turbo field echo sequence with the following parameters: TI = 1300 ms, TR/TE = 8.6/3.96 ms, flip angle = 8°, and a 1 mm3 isotropic voxel size. Sagittal-oriented high-resolution contiguous 3D T1-weighted images were obtained. To speed up data acquisition, sensitivity encoding parallel imaging with an acceleration factor of two was applied. 2.4. MRI data processing and VBM analysis VBM, based on 3D T1-weighted images, was used to analyze volumetric differences. Image processing was performed using the VBM8 toolbox (Structural Brain Mapping Group, University of Jena, Thuringia, Germany), implemented in Statistical Parametric Mapping 8 (SPM 8, Wellcome Trust Centre for Neuroimaging, University College London, London, UK). The VBM data were processed in the standard manner including the following steps: spatial normalization to the Montreal Neurological Institute template, gray matter segmentation, intensity modulation using
The numbers of patients who underwent structured interviews to obtain the clinical factors and who were subjected to radiological investigations are reported in Table 1. 3.1. Clinical factors Eighty patients met the inclusion criteria. Twenty-three patients (29%) were men, and 57 patients (71%) were women. There were significantly more women among the TGA patients compared with the general Korean population (p < 0.001; 5,811,891 men [44%] and 7,397,720 women [56%] out of 12,209,611 Korean subjects between 50 and 79 years old based on the 2010 data from the Korean Statistical Information Service). The mean age of onset was 60.3 ± 7.3 years. Among the patients with TGA, the presenting TGA event was the first such event in 77 patients, and three patients reported previous episodes of TGA. The median duration of amnesia was 5 hours (95% CI 4.0–6.0 hours, range 0.5–23 hours). Thirty-five (44%) of the 80 patients had a history of vascular risk factors, with hypertension in 27, diabetes in 10, and dyslipidemia in eight patients. Forty-four patients (55%) had a history of precipitating events, with emotional stress in 26, physical effort in nine, and water contact/temperature change in nine patients. Eleven patients (14%) had associated symptoms, with headache in four, general weakness in three, nausea in three, and dizziness in one. Table 1 Numbers of patients with transient global amnesia who underwent structured interviews to obtain clinical factors, radiological investigations, and electroencephalography Medical examination Structured interview Conventional MRI including T1 and T2-weighted imaging Diffusion-weighted imaging Three-dimensional volumetric T1-weighted imaging
n 80 (100%) 80 (100%) 54 (66%) 26 (33%)
845
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
3.2. Radiological findings MRI of the TGA patients revealed hippocampal cavities in 41 patients (51%) which were right localized in 18, left localized in 14, and bilateral in nine patients. The cavities were single, had rounded or crescent appearances and were located around the CA1 area. The incidence of hippocampal cavities was significantly different between the TGA and control subjects (41/80 versus 7/29, respectively, p = 0.0159). Of the 80 patients with TGA, DWI was performed in 54 patients, and 13 patients (24%) exhibited high signal intensities on DWI. All of the lesions on DWI were located in the hippocampus, and were right localized in five patients, left localized in six patients, and bilateral in two patients. The median time delay from symptom onset to DWI was 7 hours (95% CI 6.0– 11.0 hours, range 2–480 hours). There were no differences in age of onset (59.6 versus 61.1 years, p = 0.3555), sex (male 14/41 versus female 9/39, p = 0.3974), median duration of amnesia (5.0 versus 5.0, p = 0.7214), presence of vascular risk factors (20/41 versus 15/39, p = 0.4811), presence of precipitating events (19/41 versus 25/39, p = 0.1703), or presence of associated symptoms (7/41 versus 4/39, p = 0.5753) between the patients with and without hippocampal cavities, respectively. In addition, there were no differences in age of onset (57.6 versus 61.4 years, p = 0.1093), male sex (2/13 versus 13/41, p = 0.3112), median duration of amnesia (11.5 versus 4.0, p = 0.0739), presence of vascular risk factors (4/13 versus 21/41, p = 0.2228), presence of precipitating events (10/13 versus
22/41, p = 0.1988), presence of associated symptoms (3/13 versus 6/41, p = 0.6702), or median time delay from symptom onset to DWI (10 versus 7 hours, p = 0.1658) between patients with and without signal changes in the hippocampus on DWI, respectively. 3.3. VBM Among the 80 patients with TGA, 26 patients underwent 3D volumetric T1-weighted imaging suitable for VBM. Seven patients (27%) were men, and the mean age of onset was 60.7 ± 7.4 years. The VBM analysis comparing these 26 TGA patients and 29 control subjects revealed that TGA patients exhibited significant reductions in the gray matter volumes of the left hippocampus, left amygdala, bilateral cingulum, and right posterior lobe of the cerebellum. However, there were no structures with increased gray matter volumes in TGA patients compared with the control subjects. Furthermore, after adjusting for age, sex, and intracranial volume, the patients with TGA still exhibited significant reductions in the gray matter volumes of the left hippocampus, bilateral cingulum, and right posterior lobe of the cerebellum. However, the significance of the reduction in the gray matter volume of the amygdala disappeared (Fig. 1, Table 2, Supp. Fig. 1). 4. Discussion The main finding of our study was that patients with TGA exhibited significant reductions in the gray matter volumes of
Fig. 1. (A) Sagittal, (B) coronal and (C) axial views of the result of voxel-based morphometry analysis in grayscale on a glass-brain render of the brain after adjusting for age, sex, and intracranial volume. Patients with transient global amnesia exhibit significant reductions in gray matter volumes of the left hippocampus, bilateral cingulum, and right cerebellum at a statistically significant level, p < 0.05 (false discovery rate corrected). The red arrows indicate the center point. (For interpretation of the references to colors in this figure legend, the reader is referred to the web version of this paper.) Table 2 Brain regions displaying reductions in gray matter volumes by voxel-based morphometry in the patients with transient global amnesia, compared with control subjects after adjusting for age, sex, and intracranial volume Region
Side
MNI coordinates
Cerebellum, posterior lobe Cingulum Hippocampus Cingulum
Right Right Left Left
(6,
MNI = Montreal Neurological Institute.
73.5, 18) (6, 45, 9) ( 21, 13.5, 16.5) ( 4.5, 18, 43.5)
Peak intensity
Cluster size (voxels)
7.7406 7.1054 6.2612 4.6584
13554 1323 1979 541
846
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
the hippocampus, cingulum, and cerebellum. We also found that there were significantly more women in the TGA group and that there were no differences in the clinical factors between the patients with and without signal changes in the hippocampus on DWI. Additionally, we confirmed the increased incidence of hippocampal cavities among the patients with TGA, and there were no differences in the clinical factors between the patients with and without hippocampal cavities. Although TGA is well-characterised clinically, the etiology and pathogenesis of TGA remains unclear. Several hypotheses, including arterial transient ischemic attacks, epileptic seizures, transient mesial temporal ischemia induced by venous congestion, and migraine have been postulated, but definite evidence supporting these mechanisms is lacking [14–15]. Similarly, the findings of this study do not provide direct evidence regarding the pathogenesis of TGA. However, we postulate that pre-existing abnormalities in limbic structures might contribute to an increased vulnerability to TGA events. This assumption is supported by the observation that patients who have experienced previous TGA events exhibit a higher frequency of recurrence compared with the de novo incidence of TGA across the lifetime; the rate of annual recurrence in patients with TGA is between 6 and 10%, whereas the annual incidence in the general population is between 0.003 and 0.008% [2]. Although no MRI scans were conducted in the pre-symptomatic period, we assume that the reductions in the volumes of the gray matter in the limbic structures were pre-existing. The range of the volume reductions was multifocal and extensive, therefore, volume reductions might not have developed suddenly after the onset of the TGA symptoms. Additionally, no previous reports have demonstrated any acute structural changes on brain MRI following the onset of symptoms, with the exception of small diffusion restrictions in the hippocampus. Altogether, these findings suggest that the differences in the MRI results were due to events that occurred prior to the onset of TGA, rather than as consequences of TGA. We also infer that precipitating events might trigger an episode of TGA in patients with more vulnerable memory pathways due to pre-existing abnormalities in the limbic structures. The hippocampus has a critical role in memory, particularly episodic memory. Semantic memory primarily activates the frontal and temporal cortexes, whereas episodic memory activity, at least initially, is concentrated in the hippocampus [16,17]. Neuropsychological examinations performed during the acute phase have revealed that TGA is a selective disorder of episodic memory that results in no impairments of other components of memory [2,3]. Additionally, during TGA episodes, patients are characterized as being unable to encode new information into long-term memory, and the hippocampus is known to play a role in the formation of long-term memory from short-term memory [2,3]. Thus, it can be inferred that memory disturbance in patients with TGA originates from impairments of the hippocampus. We found that patients with TGA exhibited reductions in hippocampal gray matter volumes, and this finding supports the involvement of the hippocampus in the pathogenesis of TGA. Additionally, our findings are consistent with the results of previous functional studies that have used SPECT or positron emission tomography and indicated that hippocampal abnormalities exist in TGA [18]. However, we cannot exclude the effects of the hippocampal cavities on the reductions in hippocampal volumes. The Monroe– Kellie doctrine states that losses of parenchymal tissue result from increases in the adjacent CSF space. Considering this, increased incidence of hippocampal cavities might induce hippocampal tissue loss. However, in our study, hippocampal cavities were found in the right side of 18 TGA patients and the left side of 14 TGA patients, and hippocampal cavities were detected in four control subjects on the right side and three control subjects on the left side. Although, hippocampal volume reductions were detected
only on the left side in patients with TGA, the incidence of hippocampal cavities was not higher on the left side, compared to the right side (p = 1.0 by Fisher’s exact test). Thus, hippocampal volume reductions cannot be explained solely by the effects of hippocampal cavities. There is evidence that the cerebellum is also a limbic structure [19–21]. Several previous studies have demonstrated that cerebellar infarction without cortical lesions results in cognitive changes, and these findings indicate the relevance of the connecting loops between the cerebellum and the contralateral cortex, to cognitive processing [20–21]. It is also reported that some patients with TGA display cerebellar hypoperfusion on SPECT [19]. Schmahmann et al. described the so-called cerebellar cognitive affective syndrome with deficits in executive functioning, language, and spatial cognition, and behavioral changes in patients with structural cerebellar lesions [22]. Interestingly, these previous studies highlight the dominant role of the posterior lobe of the cerebellum in cognitive and affective processing [21,22], and our study revealed reductions in the gray matter volumes of the posterior lobes of the cerebellum in patients with TGA. The dominant role of the posterior lobe of the cerebellum in limbic function might be explained by evidence regarding the linkages of the parietal and paralimbic association areas with the posterior lobe of the cerebellum [21]. Several investigators have used DWI to further examine the etiology of TGA, and they have reported variable results. One study reported that high signal intensity on DWI was detected in seven of 10 patients with TGA, whereas another study did not find abnormal DWI signals in any of eight patients with TGA [4,12]. In our study, we found that only 24% of the patients with TGA exhibited high signal intensity on DWI. These conflicting results might be due to differences in the timing of examination and differences in MRI protocols, including parameters such as slice thickness [23]. Hippocampal lesions on DWI are rarely noted in the hyperacute phase, but all such lesions become visible at 48 hours [24]. The detection rates of hippocampal lesions can be as high as 88% when more appropriate MRI protocols are utilized [8,23]. Among our patients, 37 of the 54 who underwent DWI (69%) did so less than 24 hours after symptom onset, and all of our patients underwent DWI utilizing a slice thickness of 5 mm. The short time from symptom onset to DWI and the slice thickness utilized might have produced a lower hippocampal abnormality detection rate on DWI in our study. Additionally, we compared the clinical factors between patients with and without hippocampal lesions on DWI and found no differences. These findings were consistent with those of a previous report that indicated no differences in long-term cognitive outcomes between patients with and without hippocampal lesions [25]. Nakada et al. reported that the incidence of hippocampal cavities was 100% in patients with TGA, and an incidence of 31% in control subjects [13]. Additionally, another previous study reported hippocampal cavities in 60% of patients with TGA and no differences in cognitive performance between patients with and without hippocampal cavities [26]. In our study, we found hippocampal cavities in 51% of the patients with TGA, similar to the results of Nakada et al. [13], and we also confirmed an increased incidence of hippocampal cavities in patients with TGA.
5. Conclusion We demonstrated that there were significant structural differences in the limbic structures of patients with TGA and control subjects and that these reductions in limbic gray matter volumes might contribute to greater vulnerability of the memory pathways in patients with TGA. These findings suggest that the occurrence of
K.M. Park et al. / Journal of Clinical Neuroscience 22 (2015) 843–847
TGA events is predisposition.
associated
with
an
underlying anatomical
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. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jocn.2014.11.017. References [1] Hodges JR, Warlow CP. The aetiology of transient global amnesia. A casecontrol study of 114 cases with prospective follow-up. Brain 1990;113: 639–57. [2] Bartsch T, Deuschl G. Transient global amnesia: functional anatomy and clinical implications. Lancet Neurol 2010;9:205–14. [3] Eustache F, Desgranges B, Laville P, et al. Episodic memory in transient global amnesia: encoding, storage, or retrieval deficit? J Neurol Neurosurg Psychiatry 1999;66:148–54. [4] Gass A, Gaa J, Hirsch J, et al. Lack of evidence of acute ischemic tissue change in transient global amnesia on single-shot echo-planar diffusion-weighted MRI. Stroke 1999;30:2070–2. [5] Gonzalez-Martinez V, Comte F, de Verbizier D, et al. Transient global amnesia: concordant hippocampal abnormalities on positron emission tomography and magnetic resonance imaging. Arch Neurol 2010;67:510. [6] Baron JC, Petit-Taboue MC, Le Doze F, et al. Right frontal cortex hypometabolism in transient global amnesia. A PET study. Brain 1994;117: 545–52. [7] Bartsch T, Alfke K, Stingele R, et al. Selective affection of hippocampal CA-1 neurons in patients with transient global amnesia without long-term sequelae. Brain 2006;129:2874–84. [8] Kim J, Kwon Y, Yang Y, et al. Clinical experience of modified diffusion-weighted imaging protocol for lesion detection in transient global amnesia: an 8-year large-scale clinical study. J Neuroimaging 2014;24:331–7. [9] Frisoni GB, Fox NC, Jack Jr CR, et al. The clinical use of structural MRI in Alzheimer disease. Nat Rev Neurol 2010;6:67–77.
847
[10] Ashburner J, Friston KJ. Voxel-based morphometry–the methods. Neuroimage 2000;11:805–21. [11] Ashburner J, Friston KJ. Why voxel-based morphometry should be used. NeuroImage 2001;14:1238–43. [12] Strupp M, Bruning R, Wu RH, et al. Diffusion-weighted MRI in transient global amnesia: elevated signal intensity in the left mesial temporal lobe in 7 of 10 patients. Ann Neurol 1998;43:164–70. [13] Nakada T, Kwee IL, Fujii Y, et al. High-field, T2 reversed MRI of the hippocampus in transient global amnesia. Neurology 2005;64:1170–4. [14] Schreiber SJ, Doepp F, Klingebiel R, et al. Internal jugular vein valve incompetence and intracranial venous anatomy in transient global amnesia. J Neurol Neurosurg Psychiatry 2005;76:509–13. [15] Olesen J, Jorgensen MB. Leao’s spreading depression in the hippocampus explains transient global amnesia. A hypothesis. Acta Neurol Scand 1986;73: 219–20. [16] Knierim JJ, Lee I, Hargreaves EL. Hippocampal place cells: parallel input streams, subregional processing, and implications for episodic memory. Hippocampus 2006;16:755–64. [17] Jeneson A, Squire LR. Working memory, long-term memory, and medial temporal lobe function. Learn Mem 2012;19:15–25. [18] Takeuchi R, Yonekura Y, Matsuda H, et al. Resting and acetazolamidechallenged technetium-99m-ECD SPECT in transient global amnesia. J Nucl Med 1998;39:1360–2. [19] Yang Y, Kim JS, Kim S, et al. Cerebellar hypoperfusion during transient global amnesia: an MRI and oculographic study. J Clin Neurol 2009;5:74–80. [20] Hokkanen LS, Kauranen V, Roine RO, et al. Subtle cognitive deficits after cerebellar infarcts. Eur J Neurol 2006;13:161–70. [21] Exner C, Weniger G, Irle E. Cerebellar lesions in the PICA but not SCA territory impair cognition. Neurology 2004;63:2132–5. [22] Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain 1998;121:561–79. [23] Choi BS, Kim JH, Jung C, et al. High-resolution diffusion-weighted imaging increases lesion detectability in patients with transient global amnesia. AJNR Am J Neuroradiol 2012;33:1771–4. [24] Sedlaczek O, Hirsch JG, Grips E, et al. Detection of delayed focal MR changes in the lateral hippocampus in transient global amnesia. Neurology 2004;62:2165–70. [25] Uttner I, Prexl S, Freund W, et al. Long-term outcome in transient global amnesia patients with and without focal hyperintensities in the CA1 region of the hippocampus. Eur Neurol 2012;67:155–60. [26] Uttner I, Weber S, Freund W, et al. Hippocampal cavities are not associated with cognitive impairment in transient global amnesia. Eur J Neurol 2011;18: 882–7.