Sports

Amygdalar Atrophy in Early Alzheimer's Disease

Description
Amygdalar Atrophy in Early Alzheimer's Disease
Categories
Published
of 14
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
   Send Orders for Reprints to reprints@benthamscience.net   Current Alzheimer Research,  2014  , 11,  000-000   1   1567-2050/14 $58.00+.00 © 2014 Bentham Science Publishers   Amygdalar Atrophy in Early Alzheimer’s Disease Yanica Klein-Koerkamp 1 , Rolf A. Heckemann 2,3,4 , Kylee T. Ramdeen 3 , Olivier Moreaud 1,6 , Sandrine Keignart 6 , Alexandre Krainik  7,8,9 , Alexander Hammers 3,4 , Monica Baciu 1,10 , Pascal Hot 5,*  and for the Alzheimer’s Disease Neuroimaging Initiative ** 1 Université Grenoble Alpes, Laboratoire de Psychologie et Neurocognition (CNRS UMR-5105), BP 47, 38040 Grenoble Cedex 9, France; 2  MedTech West at Sahlgrenska University Hospital, Gothenburg University, Sweden; 3 The Neurodis  Foundation (Fondation Neurodis), CERMEP – Imagerie du vivant, Lyon, France; 4  Division of Brain Sciences, Faculty of Medicine, Imperial College London, UK; 5 Université de Savoie, Laboratoire de Psychologie et Neurocognition (CNRS UMR-5105), BP 1104, 73011 Chambéry Cedex, France; 6  Centre Mémoire de Ressources et de Recherche, Pole de Psychiatrie et Neurologie, CHU Grenoble, Grenoble, France; 7   Inserm, U836, Grenoble, France; 8 Team 5 “Functional neuroimaging and brain perfusion” of Grenoble Institute of Neuroscience, INSERM/CEA, Joseph - Fourier University; 9  Joint Service Unit (JSU), UMS 3552, ‘IRMaGe’, CNRS/INSERM, Joseph-Fourier University, Grenoble  Institute of Neuroscience, Grenoble, France; 10  Institut Universitaire de France, 75005 Paris, France Abstract:  Current research suggests that amygdalar volumes in patients with Alzheimer’s disease (AD) may be a relevant measure for its early diagnosis. However, findings are still inconclusive and controversial, partly because studies did not focus on the earliest stage of the disease. In this study, we measured amygdalar atrophy in 48 AD patients and 82 healthy controls (HC) by using a multi-atlas procedure, MAPER. Both hippocampal and amygdalar volumes, normalized by in-tracranial volume, were significantly reduced in AD compared with HC. The volume loss in the two structures was of similar magnitude (~24%). Amygdalar volume loss in AD predicted memory impairment after we controlled for age, gen-der, education, and, more important, hippocampal volume, indicating that memory decline correlates with amygdalar at-rophy over and above hippocampal atrophy. Amygdalar volume may thus be as useful as hippocampal volume for the di-agnosis of early AD. In addition, it could be an independent marker of cognitive decline. The role of the amygdala in AD should be reconsidered to guide further research and clinical practice. Keywords:  Automatic segmentation, brain, hippocampus, MRI, neuropsychology. INTRODUCTION The neuropathology of Alzheimer’s disease (AD) is characterized by neuronal loss, first affecting the medial temporal lobe (MTL) [1, 2]. In particular, subregions of the hippocampus [3-6] and entorhinal cortex [7] undergo mas- sive pathological changes, leading to progressive memory impairments [8-11]. Several studies suggest that hippocam- pal atrophy is the best neuroimaging-derived marker of dis-ease and disease progression. However, hippocampal atro- phy is associated with a range of other neurological patholo-gies [12-15], thus limiting its specificity to AD. With ad-vances in automated volumetric segmentation, structures that were previously difficult to assess are now more reliably *Address correspondence by this authors at the Laboratoire de Psychologie et Neurocognition (LPNC, UMR CNRS 5105), UFR LLSH, Université de Savoie, Domaine Universitaire de Jacob-Bellecombette, B.P. 1104 73011 Chambéry Cedex, France; Tel: (+33)4 79 75 85 66; Fax: (+33)4 76 82 78 34; E-mail: pascal.hot@univ-savoie.fr **Data used in the preparation of this article were obtained from the Alz-heimer’s Disease Neuroimaging Initiative (ADNI) database (adni.loni.ucla.edu). The investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not par-ticipate in the analysis or writing of this report. A complete listing of ADNI investigators can be found at http://adni.loni.ucla.edu/wpcontent/ uploads/how_to_apply/ADNI_Acknowledgement_List.pdf. segmented and evaluated. This has led to the discovery of other structures that undergo change in the course of demen-tia, notably, other limbic structures close to the hippocam- pus. In particular, the amygdala has recently received in-creased attention. Tables 1  and 2  list studies that focus on amygdalar atrophy in mild to moderate AD (Table 1 ) and moderate to severe AD (Table 2 ), along with factors that influence (1) the magnitude of atrophy measured (i.e., par-ticipants’ characteristics and segmentation procedures); and (2) the association with clinical features (i.e., information related to correlation analysis between cognitive scores and neuroanatomical volumes). Given our interest in amygdalar volume, we have included only those studies that consider data for this structure. Reduction of amygdalar volume com- pared to elderly HC was a robust finding in  post mortem studies [16-20] and in groups that included mildly and se-verely affected AD patients compared with elderly HC, as shown in Tables 1  and 2  (Clinical Dementia Rating or CDR scores ranging from 0.5 to 3 [4]; Mini-Mental Score Exami-nation or MMSE ranging from 5 to 21 [21]; MMSE ranging from 2 to 27 [22]; MMSE ranging from 11 to 25 [23]). This observation suggests that the diagnosis of AD may be im- proved if amygdalar volume is considered in addition to hip- pocampal volume [24]. However, findings are more contra-dictory for earlier stages of the disease, such as AD patients  2 Current Alzheimer Research,  2014  , Vol. 11, No. 2    Klein-Koerkamp et al. Table 1. Summary of research on amygdalar atrophy in mild to moderate AD. Disease Severity Population Characteristics Segmentation Characteristics Volume data Hippocampal Volume Amygdalar Volume Hippocampal vs. Amygdalar Atrophy Correlation Analysis of Volumes with Cognitive Score Regression Analysis of Volumes with Cognitive Score Reference Mild AD = 11; MMSE: HC = 29 (1); AD = 21.5 (5) Manual. MRI: 1.5T - slice thickness: 5 mm ICV-normalized volumes HC = AD (20% loss) HC > AD (33% loss) A > H AD; H and A: no significant corre-lation with MMSE. AD + HC; A: correla-tion MMSE [31] Mild HC = 8; AD = 18. Age: HC = 69.2 (2.7); AD = 72.4 (1.5); MMSE: HC = 28.7 (0.4); AD = 22.3 (0.9; 13-27) Semi-automatic. MRI: 5 mm ICV-normalized volumes HC > AD (30% loss) HC > AD (37% loss) A = H AD + HC; H: correlation MMSE, memory (Mattis, Wechsler, Grober Buschke Test, Verbal, Intrusion); A: no significant correlations [25] Mild HC = 40; AD = 24. Age: HC = 79 (3.5); AD = 78.4 (3.2); MMSE: HC = 28.9 (1); AD = 22.1 (1.9) Manual. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes; age as covariate  HC > AD (18% loss) AD + HC; A: correlation MMSE [44] Mild HC = 7; AD = 8. Age: HC = 70; AD = 72; MMSE: AD = 23.9 (17-29) Manual. MRI: 1.5T - slice thickness: 1.5 mm Raw vol-umes HC > AD HC = AD (p=0.06) A < H [84] Mild HC = 21; AD = 13. Age: HC = 69.3 (6.8); AD = 71.2 (8.3); MMSE: HC = 29.7 (0.2; 28-30); AD = 23.7 (2.7; 20-28) Manual. MRI: 1.5T et 0.5T - slice thickness: 5 mm ICV-normalized volumes; age as covariate  HC > AD (19% loss) HC > AD (33% loss) [32] Mild HC = 34; AD = 54. Age: HC = 72 (4); AD = 70 (8); MMSE: HC = 28.4 (1.3); AD = 21.7 (3.7) Manual. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes HC > AD (21% loss) AD; A: no sig-nificant correla-tion with MMSE [45] Mild HC = 16; AD = 32. Age: HC = 70 (5); MA = 69 (8); MMSE: HC =28.6 (1.4); AD = 22.8 (3.7) Manual. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes HC > AD (38% loss) HC = AD (16% loss) AD; A: no signifi-cant correlations (MMSE, verbal memory). H: correlation (MMSE, verbal memory) [30] Mild HC = 94; AD = 118. Age: HC = 74 (5); AD = 75 (6); MMSE: HC = 29 (1); AD = 21 (5) Automatic. MRI: 1.5T - slice thickness: 1.3 mm ICV-normalized volumes HC > AD HC > AD [85] Mild HC 1 = 87; AD 1 = 90. HC 2 = 193; AD 2 = 174. Age: HC 1= 77.7 (7.9); AD 1 = 77.2 (6.7); HC 2= 75.6 (5.1); AD 2 = 75.5 (7.3); MMSE: HC 1 = 28.9 (1.2); AD 1 = 24.6 (3.9); HC 2 = 29.1 (1); AD 2 = 23.3 (2) Automatic. MRI: 1.5T - slice thickness: 1 mm  Normalized volumes HC > AD (18.3% loss AD 1; 19.1% loss AD 2) HC > AD (19.3% loss AD 1; 18.5% loss AD 2) A = H AD; H and A: correlation with MMSE and CDR AD 2; volumes, age,  sex, education as covariates : H corre-lations MMSE and CDR. A: no signifi-cant correlations [26]   Amygdalar Atrophy in AD Current Alzheimer Research,  2014  , Vol. 11, No. 2 3   (Table 1) contd…. Disease Severity Population Characteristics Segmentation Characteristics Volume data Hippocampal Volume Amygdalar Volume Hippocampal vs. Amygdalar Atrophy Correlation Analysis of Volumes with Cognitive Score Regression Analysis of Volumes with Cognitive Score Reference Mild HC = 19; AD = 20. Age: HC = 72.5 (7.8); AD = 72.7 (9.1); MMSE: HC = 29.1 (1); AD = 22 (4.3; 13-28) Manual. MRI: 1T - slice thick-ness: 1.3 mm Raw vol-umes (spa-tial normali-zation on srcinal images) HC > AD (26-28% loss in right and left, respectively) HC > AD (19-24% loss in right and left, respectively) Right: A < H Left: A = H AD; H: no signifi-cant correlation (language, visuo-spatial, executive functions MMSE). A: correlation memory AD + HC; volumes and MMSE diagno- sis as covariate:  no significant correla-tions [27] Mild HC = 20; AD = 20. Age: HC = 66 (6.7); AD = 70 (8.6); MMSE: HC = 27.6 (2.06); AD = 23.3 (2.56; 20-29) Manual. MRI: 1.5T - slice thickness: 2 mm ICV-normalized volumes HC > AD (30% loss) HC > AD (29.5% loss) A = H AD; H and A: correlation mem-ory [28] Mild to moderate HC = 15; AD = 12. Age: HC = 77 (10.6); AD = 78.4 (10); MMSE: AD = 20.8 (3.7) Manual. MRI: 1.5T - slice thickness: 1.5 mm CV-ormalized olumes for id-sagittal   ntracranial area HC > AD [86] Mild to moderate HC = 27; AD = 46; Age: HC = 71.1 (7.3); AD = 68.2 (7.9); MMSE: HC = 27.7 (2); 3 groups: AD 1 (CDR 0.5) MMSE = 23.2 (3.7); AD 2 (CDR 1) MMSE = 20.2 (2.7); AD 3 (CDR 2-3) MMSE = 12.2 (3.4) Semi-automatic. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes; age, sex, and educa-tion as covariate  HC = AD (CDR 0,5); HC > AD (CDR 1; poste-rior hippocam- pus); HC > AD (CDR 1; anterior hippocampus) HC > AD (CDR 2-3) HC > AD (CDR 0.5); HC > AD (CDR 1); HC > AD (CDR 2-3) Volumes as covariate:  AD (CDR 0, 5) + HC; A: correlations (verbal, visual, and Wechsler memory scores). AD (CDR 1) + HC; A and H correlations (verbal, visual, and Wechsler memory scores)   [4] Mild to moderate HC = 126; AD 1 = 36 AD 2 = 43; AD 3 = 15. Age: HC = 71.1 (7.3); AD 1 = 72.9; AD 2 = 73.5; AD 3 = 75.9. MMSE: HC = 28.6; AD 1 (CDR 0.5) = 21.7; AD 2 (CDR 1) = 18.9; AD 3 (CDR 2) = 16 Manual. MRI: 1.5T - slice thickness: 1.6 mm ICV-normalized volumes, then W scores (normal deviates:  percentiles relative to HC adjusted for age, gender, education, and duration of disease) D + HC; W scores   olumes, age, sex, and ducation, diagnosis as ovariates . H correla-on dementia scores oston Naming Test, echsler, memory, erbal auditory learning: no significant orrelations. AD; H: orrelation Wechsler, erbal auditory learningemory; A: no signifi-ant correlations [10] Mild to moderate HC = 57; Mild = 66; Moderate = 79. Age: HC= 66.1 (8.3); Mild AD = 75.2 (7); Moder-ate AD = 73.4 (8.6). MMSE: HC = 29 (28-30); Mild AD = 23 (20-25); Moderate AD = 19 (16-22) Automatic. MRI: 1.5T - slice thickness: 1.5 mm Raw vol-umes -  sex, education, and ICV as covariate  HC > mild > moderate HC > mild = moderate AD + HC; H and A: correlation (attention, lan-guage, visuo-spatial, memory, executive func-tions, dementia scores) AD + HC:  volumes, age, sex, education, and ICV as covari-ate: H correlation (memory, visuo-spatial, and execu-tive functions, dementia scores). A: no significant corre-lations [23]  Note: Disease severity was coded as follows: Mean MMSE between 21 and 26: Mild. Mean MMSE between 13 and 20: Moderate. AD = HC means that volumes between both groups are equivalent. HC > AD means that volumes of HC are higher than AD patients. H: hippocampal volume, A: amygdalar volume, HC: healthy control individuals, AD: Alz-heimer’s disease patients.  with mild to moderate disease (see Table 1 , MMSE scores ranging from 13 to 29 or CDR scores ranging from 0.5 to 1) [4, 10, 23, 25-28]. In particular, volume loss in mild to mod-erate AD patients varies from 15-16% [29, 30] to 33-37% [25, 31, 32]. The variation in findings may result from methodological issues related to amygdalar segmentation. Because of the numerous cortical and subcortical nuclei of the amygdala, its  proximity to the hippocampus, and the similarity of neigh- boring tissues, the boundaries of the amygdala are difficult  4 Current Alzheimer Research,  2014  , Vol. 11, No. 2    Klein-Koerkamp et al. Table 2. Summary of research on amygdalar atrophy in moderate to severe AD. Disease Severity Population Char-acteristics Segmentation Characteristics Volume Data Hippocampal Volume Amygdalar Volume Hippocampal vs. Amygdalar Atrophy Correlation Analysis of Volumes with Cognitive Score Regression Analysis of Volumes with Cognitive Score Reference Moderate HC = 42; AD = 56. Age: HC = 73.2 (6.7); AD = 71.2 (8.6); MMSE: HC = 29 (1); AD = 18.3 (4.3) Manual. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes HC > AD (17% loss) HC > AD (23% loss) A = H AD; H: no corre-lation (language, memory, orienta-tion, praxis, MMSE). A: correlation (memory, orien-tation, MMSE) [37] Moderate HC = 10; AD = 10. Age: HC = 56 (11); AD = 57 (9); MMSE: HC = 29 (1); AD = 15 (6) Manual. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes HC > AD (16.4% loss) HC > AD (15% loss) A = H [29] Moderate HC = 19; AD = 19. Age: HC = 73.6 (5.5); AD = 76.1 (5.7); MMSE: HC = 28.6 (1.1; 27-30); AD = 13.1 (3.8; 5-21) Manual. MRI: 3T - slice thickness: 1 mm ICV-normalized volumes;   sex and education as covariate  HC > AD (22% loss) [21] Moderate HC = 18; AD = 27; Age: HC = 69.5 (6.4); AD = 71 (7.5); MMSE: HC = 30; AD = 19 (3.6) Manual. MRI: 1.5T - slice thickness: 1 mm ICV-normalized volumes HC > AD (35% loss) HC > AD (16% loss) A < H AD; H and A: correlation MMSE [38] Moderate HC = 126; AD = 94. Age: HC = 79 (6.73); AD = 73 (8); MMSE: HC = 28 (1.26); AD = 17.8 (4.94) Manual. MRI: 1.5T - slice thickness: 1.6 mm Raw vol-umes;  ICV as covariate  HC > AD HC > AD A < H [3] Moderate HC = 22; AD = 31; Age: HC = 67.7 (7.9); AD = 68 (6.8); MMSE: AD: 17.2 (3.2,14-23); HC: 28.8 (1.1, 25-30) Manual. MRI: 0.5T - slice thickness: 5 mm  Normalized volumes HC > AD AD; age and A volume as covariate : A: no significant correlation (MMSE, cognitive battery) [87] Moderate HC = 12; AD = 46. Age: HC = 66.2 (4.9); AD = 70.3 (7.1); MMSE: HC = 28 (1); AD = 19.6 (3.5; 12-26) Manual. MRI: 1.5T - slice thickness: 1.5 mm  Normalized volumes HC > AD (15% loss) HC > AD (18.5% loss) AD; H (right) and A: correlation (Wechsler visuo-spatial memory). A: correlation (Wechsler verbal memory) AD; volumes , age,  sex, and education as covariates : A corre-lation (Wechsler visuo-spatial and verbal memory subset). H: no sig-nificant correlation [5] Moderate HC = 27; AD = 36. Age: HC = 72 (4.2); AD = 73 (8.9); MMSE: HC = 28; AD = 17.1 (5.2; 2-27) Semi-automatic. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes HC > AD (24% loss) HC > AD (21% loss) [22]    Amygdalar Atrophy in AD Current Alzheimer Research,  2014  , Vol. 11, No. 2 5   (Table 2) contd….   Disease Severity Population Char-acteristics Segmentation Characteristics Volume Data Hippocampal Volume Amygdalar Volume Hippocampal vs. Amygdalar Atrophy Correlation Analysis of Volumes with Cognitive Score Regression Analysis of Volumes with Cognitive Score Reference Moderate HC = 20; AD = 20; MMSE: MA = 18.8 (5.7; <26) Manual. MRI: 1.5T - slice thickness: 2 mm ICV-normalized volumes HC > AD (20% loss) HC > AD (33% loss) AD; volumes, age as covariate:  H no significant correla-tion; A (Right) correlation (ADAS non-cognitive score); AD; cognitive scores, age as covariates:  H correlation (ADAS non-cognitive score); A correlation (ADAS-non-cognitive score, MMSE) [88] Moderate HC= 17; AD = 20. Age: HC = 63.6 (10.5); AD = 63.8 (9.1); MMSE: HC = 28.9 (1.3); AD = 20.3 (5.1) Semi-automatic. MRI: 1.5T - slice thickness: 1.5 mm ICV-normalized volumes HC > AD (20.5% loss) [73] Moderate to severe HC = 57; Moder-ate = 79; Severe = 34. Age: HC= 66.1 (8.3); Moderate AD = 73.4 (8.6); Severe AD = 71.3 (9.6). MMSE: HC = 29 (28-30); Moderate AD = 19 (16-22); Severe AD = 12.5 (11-18) Automatic. MRI: 1.5T - slice thickness: 1.5 mm Raw vol-umes -  sex, education, and ICV as covariate  HC > moderate > severe HC > mod-erate = severe AD + HC; H and A: correlation (attention, lan-guage, visuo-spatial, memory, executive func-tions, dementia scores) AD + HC:  volumes, age, sex, education, and ICV as covari-ate: H correlation (memory, visuo-spatial, and executive functions, dementia scores). A: no sig-nificant correlations [23]  Note: Disease severity was coded as follows: Mean MMSE between 13 and 20: Moderate. Mean MMSE between 3 and 12: Severe. AD = HC means that volumes between both groups are equivalent. HC > AD means that volumes of HC are higher than AD patients. H: hippocampal volume, A: amygdalar volume, HC: healthy control individuals, AD: Alz-heimer’s disease patients.  to identify [33, 34]. Various approaches have been imple-mented to assess MTL atrophy in AD. One of these is based on voxel-based morphometry, which investigates amygdala volume change in AD within the MTL [35, 36]. While this technique allows group comparisons between AD and HC, it cannot provide the absolute volume of the structure at the individual level. The second technique uses manual segmen-tation of the amygdala on magnetic resonance imaging (MRI) [28, 31, 37, 38]. Although it remains the gold stan-dard, this approach is impractical for cohorts beyond a cer-tain size, because it requires a large amount of expert interac-tion for each image [33]. The third technique consists of semi-automatic (18) or automatic segmentation of the amygdala and hippocampus [26] on the basis of single-subject digital atlases. A limitation of these techniques is that single-subject atlases do not sufficiently take into account the neuroanatomical variation between subjects. This prob-lem affects amygdala segmentation in particular, because the size and shape of the amygdala varies substantially even within demographically homogeneous groups of healthy subjects. This limitation can be addressed by probabilistic seeding followed by region-growing techniques [39] or by using multiple atlases [40, 41]. In combination with tissue  probability maps to enhance registration, multi-atlas ap- proaches are particularly suitable for subjects with neurode-generative disease [42, 43]. We therefore used multi-atlas  propagation with enhanced registration (MAPER) [42] to achieve the first objective of this study, which was to per-form amygdalar volumetry in patients with mild to moderate AD. The relation between amygdalar atrophy and the severity of the disease is still a matter of debate (see Tables 1  and 2 ). A correlation between amygdalar atrophy and cognitive im- pairment in AD has been found in some studies [5, 26, 27, 37, 38], but not in others [25, 30, 31]. In most studies, the clinical significance of amygdalar atrophy in AD was esti-mated without controlling for hippocampal atrophy [25, 30, 31, 37, 38, 44, 45], which adds to the difficulty of interpret-ing the results. Considering the strong functional connec-tivity between the hippocampus and the amygdala, specifi-cally in the context of memory [46, 47], it seems especially important to correct for hippocampal atrophy when assessing the specificity of the relationship between amygdalar atrophy and memory decline in AD. Furthermore, the majority of  prior studies combined AD and HC groups for correlation
Search
Similar documents
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x