Patterns of Brain Iron Accumulation in Vascular Dementia and Alzheimer’s Dementia Using Quantitative Susceptibility Mapping Imaging

Abstract

Background:

Emerging evidence suggests that the excessive accumulation of iron in subcortical and deep gray matter has been related to dementia. However, the presence and pattern of iron accumulation in vascular dementia (VaD) and Alzheimer’s disease (AD) are rarely investigated.

Objective:

To examine and compare the pattern and presence of brain iron accumulation of VaD and AD using quantitative susceptibility mapping (QSM).

Materials and Methods:

Twelve patients with VaD, 27 patients with AD, and 18 control subjects were recruited in this institutional review-board approved study. Susceptibility maps were reconstructed from a three-dimensional multiecho spoiled gradient-echo sequence. Four regions of interest were drawn manually on QSM images, namely the globus pallidus, putamen, caudate nucleus, and pulvinar nucleus of the thalamus. Comparisons of patient demographics, and iron concentrations among the VaD, AD, and control subjects were assessed using analysis of variance and post-hoc analyses. The relationships of age and cognitive state with susceptibility values were assessed using partial correlation analysis.

Results:

In VaD and AD, overall susceptibility values were higher than those of control subjects. A significant difference in susceptibility values was found in the putamen and caudate nucleus (p < 0.001 and p = 0.002, respectively). However, susceptibility values did not differ between VaD and AD. Age and cognitive deficit severity were not related to susceptibility values in the VaD and AD groups.

Conclusion:

Increased iron deposition in the putamen and caudate nucleus in VaD and AD patients was not associated with age or the severity of cognitive deficits. Further evaluations are needed to determine the temporal changes in iron load and their diagnostic role in dementia pathology.

Keywords

Cognitive decline imaging iron

INTRODUCTION

Iron contributes to many biological processes [1], including transport of oxygen [2], regulation of protein expression [3, 4], and cell growth and differentiation [5]. In the brain, iron plays a central role in brain development, neurotransmitter systems, and myelin synthesis and is also required for the generation of reactive oxygen species [6, 7], aggregation of α-synuclein [8], and functioning of iron-dependent enzymes [9]. Throughout normal aging, iron accumulates in deep gray matter brain structures such as the basal ganglia, hippocampus, cerebellar nuclei, and subcortical brain regions [10]. In a variety of neurodegenerative diseases, including Alzheimer’s disease (AD) and Parkinson’s disease, abnormal excessive accumulation of iron in subcortical and deep gray matter nuclei has been reported [7, 11]. Regional iron accumulation in the brain is associated with selective cognitive decline and clinical findings in patients with or without dementia [12, 13]. It has been hypothesized that iron-related oxidative stress precedes the neurodegenerative process and excessive iron accumulation may attenuate neuronal function or disrupt connectivity; however, the mechanism is still unclear [14].

In contrast to AD, the presence and pattern of iron accumulation in vascular dementia (VaD) has rarely been investigated. As small-vessel disease contributes to the process of increased iron accumulation [15], it seems that more iron accumulation should be detected in VaD than in controls. However, a study reported no differences in iron content in VaD compared with controls [16]. Moreover, most studies of iron accumulation are limited by their use of nonspecific measures of iron, such as susceptibility-weighted magnetic resonance imaging (MRI) or T2^*-weighted MRI, rather than specific measures of iron in the brain.

It has recently been proposed that an iron-specific measuring method using quantitative susceptibility mapping (QSM) provides the highest sensitivity and specificity for detecting brain iron compared to other quantitative magnetic resonance (MR) methods [17 –19].

We hypothesized that VaD would show a different iron load pattern compared with AD and control subjects. Accordingly, we aimed to examine and compare the patterns of brain iron accumulation in VaD and AD, using QSM of MR.

MATERIAL AND METHODS

This retrospectively analysis of acquired data was approved by our institutional review board. The requirement for written informed consent was waived.

Participants

A total of 68 subjects, with or without subjective memory complaints, who visited the Memory Clinic of Konkuk University Medical Center and had undergone brain MRI at 3T between June 2012 and Nov 2012 were considered forthis study.

We assessed all available information, such as basic demographic characteristics, vascular factors (including history of hypertension, diabetes mellitus, and dyslipidemia), results of laboratory tests, Mini-Mental State Examination (MMSE), global cognitive assessment (Clinical Dementia Rating Sum of Boxes, CDRSOB), and brain imaging data. Comprehensive neuropsychological assessments were also included for evaluation of multiple cognitive domains. The tasks consisted of the modified Korean version of the Hopkins Verbal Learning Test, the Digit Span Forward and Backward Test, the Rey-Osterrieth Complex Figure Test, the Korean version of the Boston Naming Test, the Stroop Test, and the Word Fluency and Rey-Osterrieth Complex Figure Copy Test. The results of laboratory tests were used to exclude other medical conditions that are associated with dementia-like symptoms [20]. The vascular risk factors were selected based on a previous study [13], and assessed through patients’ histories and measurements at the examination.

The diagnoses of dementia, VaD, and probable AD were based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, the criteria suggested by the National Institute of Neurological Disorders and Stroke of the Association Internationale pour la Recherche et l’Enseignement en Neurosciences (NINDS-AIREN) [21], the criteria of the National Institute of Neurological and Communicative Disorders and Stroke, and the Alzheimer Patients and Related Disorders Association (NINCDS-ADRDA) [22], respectively. After excluding patients with calcification in basal ganglia, which have dark spots with negative susceptibility, 41 patients, 12 with VaD and 29 with probable AD were included. Two patients with AD did not have phase data available, leading to a final sample of 12 patients with VaD and 27 patients with AD.

Twenty-three control subjects, who were referred for MR imaging as part of a routine medical checkup between the same periods, were included with the QSM sequences added on to their routine protocol. All eligible subjects were thoroughly examined by an experienced neurologist in order to exclude any subjects who showed focal neurologic deficits. All MR images were also carefully reviewed by an experienced neuroradiologist.

Among the 23 subjects, five were excluded due to unavailable gradient-echo phase data. Accordingly, 18 subjects were finally included for analysis. The indications for MRI included headache (n = 7), anxiety (n = 1), and a part of health screening (n = 10).

MR image acquisition

MR images were obtained at the Konkuk University Medical Center using a 3T scanner (Signa HDxT; GE Medical Systems, Milwaukee, WI, USA) with an 8-channel head coil. The routine MRI protocol included the following sequences: 1) axial and sagittal T1-weighted inversion-recovery (TR/TE/TI, 2468/12/920 ms, respectively; section thickness, 5 mm; matrix, 512×224); 2) axial T2-weighted fast spin-echo (TR/effective TE, 4000/106 ms; section thickness, 5 mm; matrix 384×384); 3) axial fluid-attenuated inversion-recovery (FLAIR, TR/TE/TI, 11000/105/2600 ms; section thickness, 5 mm; matrix, 384×384). For T2^* weighted imaging, we acquired a multi-echo axial three-dimensional gradient echo sequence (based on susceptibility-weighted angiography sequence [SWAN]) with the following parameters: TR/TE = 37 ms/3.5 ms (8 multiple echoes), echo spacing = 4.09 ms, FA = 20, bandwidth =±41.67 kHZ, FOV = 240×240 mm², matrix = 256×256, in-plane resolution = 0.938×0.938 mm², slice number = 56, slice thickness = 2.5 mm, and acquisition time = 3 min 32 s.

The QSM images were pre-processed and analyzed using the morphology-enabled dipole inversion (MEDI) tool box and a standardized algorithm suggested by Liu et al. [23, 24]. QSM images were reconstructed using the MEDI approach, which inverts an estimated magnetic field to generate a magnetic susceptibility distribution that is structurally consistent with an anatomic prior, which is derived from the magnitude image obtained during the same imaging. First, a nonlinear fit to the multi-echo data was performed to estimate the magnetic field inhomogeneity, followed by a magnitude-guided phase unwrapping [25]. The background field was then removed by applying a projection onto the dipole field (PDF) [26]. Finally, the remaining field was inverted to calculate the quantitative susceptibility map (QSM) [27].

Data processing and analysis

All image analyses were performed using freely available software called medical imaging processing, analysis, and visualization (MIPAV) (http://mipav.cit.nih.gov/). Four regions of interest (ROIs) were drawn manually on the QSM images for each case, namely the globus pallidus (GP), putamen (PUT), caudate nucleus (CN), and pulvinar nucleus of the thalamus (PULV). The ROIs for each structure were drawn by researchers blinded to the diagnoses and demographics. The ROIs for the caudate head, putamen, and globus pallidus were semi-automatically drawn by using the level-set method of MIPAV on the axial QSM image, 12 mm above the commissural line (Fig. 1). The demarcated boundaries of GP, PUT, and CN in the axial T1-weighted images were cross-checked for the accurate selection of ROIs on QSM images. In contrast, the boundary of the thalamic pulvinar nucleus cannot be identified on T1WI but can be easily identified on QSM images. Therefore, we operationally defined the pulvinar nucleus by using the level-set volume of interest (VOI) tool of MIPAV. The average susceptibility value of each region was calculated.

An experienced rater performed the entire imaging analysis under the supervision of a neuroradiologist. To assess measurement reliability, we selected the MR images of 18 normal subjects. For those subjects, another rater independently measured the ROIs for each anatomical target. Inter-rater reliability among all regions was 0.960 (95% confidence interval: 0.944–0.971, p < 0.001).

Statistical analysis

SPSS (v. 17.0; SPSS Inc., Chicago, IL, USA) was used for statistical analyses. Normality of the data was tested before performing parametric analysis with the Kolmogorov-Smirnov test. Comparisons of patient demographics, vascular factors, MMSE, CDRSOB, and iron concentrations among the VaD, AD, and normal control subjects were assessed by ANCOVA.

Pearson correlations were used to assess the relationship between iron accumulation and age and the clinical scales (i.e., MMSE and CDRSOB). Thereafter, partial correlation analyses were used to exclude the effect of age. p < 0.05 was considered the threshold of significance.

RESULTS

Demographic characteristics of the three groups are summarized in Table 1. There was a significant difference in age among the three groups (p < 0.001). The normal control group was younger than the other two groups. The ratio of females to males was relatively high in the AD group; however, there was no statistical difference among the groups. The other covariates, including years of education, vascular factors, MMSE, and CDRSOB did not differ among the three groups.

The susceptibility values of the three groups are summarized in Table 2. In VaD and AD, the overall susceptibility value was higher than that of control subjects (Fig. 2). A significant difference in susceptibility values was found in the PUT and CN (PUT: 94.81±29.70, 98.90±33.63, 58.48±24.01, p < 0.001; CN: 85.90±13.69, 83.44±22.44, 63.96±16.38, p =0.002). However, susceptibility values did not differ between VaD and AD.

Age was not related to susceptibility values in any brain regions for the VaD group, but in the AD group age was associated with decreased susceptibility values in CN (r = –0.291, p = 0.031); however, significance disappeared after controlling for MMSE (r = –0.250, p = 0.069) or CDRSOB (r = –0.282, p = 0.060). In control subjects, age was associated with increased susceptibility values in CN and PUT (r = 0.532, p = 0.023 and r = 0.678, p = 0.002).

In both VaD and AD groups, there was no relationship between susceptibility values and MMSE. In the AD group, CDRSOB was related to susceptibility values in CN (r = –0.299, p = 0.044); however, significance disappeared after controlling for age.

DISCUSSION

In the present study, we observed that the susceptibility values in VaD and AD were higher than in control subjects. However, age, MMSE, and global cognitive assessments were not related to iron accumulation in VaD or AD groups.

In the VaD group, iron was highly accumulated in the putamen and caudate nucleus compared to controls. Previous studies have used susceptibility-weighted imaging (SWI) to reveal a higher prevalence of iron signals in the putamen and caudate nucleus in patients with cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), and in patients with subcortical ischemic vascular dementia [15, 28]. QSM images in the present study showed similar patterns; the patterns occur not only in small-vessel disease, but also in VaD patients. The mechanism that leads to iron accumulation in the specific brain areas with vascular damage remains to be determined. The iron accumulation might be the result of the small-vessel disease, or alternatively, one of the causes of vascular damage. The demyelination of white matter and atrophy of the cerebral cortex due to vascular damage can lead to a decreased demand for iron from storage locations such as the putamen and caudate nucleus [15, 29]. Alternatively, iron accumulation could cause white-matter disruption and atrophy by inducing endothelial cell damage [30]. Our findings contrast with a study showing non-accumulation of iron in vascular dementia [16]. However, half of the patients who were defined as the VaD group in that study had cerebral amyloid angiopathy, and there is a high probability that the majority of the VaD patients had mixed AD pathology. Moreover, the postmortem brain study method used has the issue that in vivo iron can be lost during the preparation process.

Patients with AD had a similar iron accumulation pattern to those with VaD, consisting of an increase in the putamen and caudate nucleus. This finding is in accordance with a previous report that used phase-corrected imaging to reveal a high iron load in the putamen and caudate nucleus in AD [31]. In transgenic mice models of AD, iron accumulation overlaps with areas containing amyloid plaques, and total iron accumulation is associated with early plaque formation [32]. However, given the presence of iron overload in both AD and VaD in our study, iron accumulation does not seem to reflect a disease-specific pathology, such as amyloid accumulation, in AD. That similar iron accumulation patterns were observed both in AD and VaD supports the more commonly held idea that iron accumulation reflects the status of neurodegeneration, regardless of the underlying insult [7].

Interestingly, there was a different relationship between iron and age in VaD compared with AD and controls. Normally, iron concentration is correlated with age [13 , 33]. Age-related iron accumulation might be induced by increased permeability of the blood-brain barrier, inflammation, redistribution of iron within the brain, and changes in iron homeostasis [34]. As iron accumulates, excessive oxidative stress and accumulation of neurotoxic proteins result in neuronal losses [7, 34]. However, in VaD patients, susceptibility in all four regions did not show any correlation with age. Age-independent iron accumulation in VaD patients might be associated with a plateau effect due to VaD-associated pathology or the limited number of subjects in this study.

AD showed a paradoxical decrease in iron accumulation in the caudate head with increasing age in the present study; however, significance disappeared after controlling for cognition. It is known that iron load increases with age. Although, in detail, mean value of iron accumulation was not different between the two groups, the AD group showed more various values in law data compared to VaD. Contrary to a plateau effect seen in VaD, it might be a progressive ‘degenerative’ character of AD. As the pathology of AD is progressing, the iron accumulation could be also proceeding, although this was not a longitudinal study. The positive relationship between iron accumulation and age in the normal control group corroborates previous studies [13 , 35].

Several studies have evaluated the relationship between cognition and iron accumulation. Liem et al. reported more iron accumulation in the putamen and caudate nucleus in symptomatic versus asymptomatic CADASIL patients, but not in the thalamus [15].Liu et al. could not find lesions reflecting the cognitive assessments in subcortical ischemic vascular dementia [28]. In terms of thalamic pathology, these studies evaluated the whole thalamus, not the pulvinar nucleus. Instead, we focused on the pulvinar nucleus, which is a candidate area associated with decreased cognition in AD and multiple sclerosis patients [35, 36]. In our previous study, we explored iron accumulation in the deep gray matter of AD patients by using T2^* mapping [35]. We did not find a correlation between pulvinar T2^* and MMSE, despite the presence of a modest correlation between pulvinar hypointensity and MMSE. Based on this, we assumed that the inherent limitations of the T2^* map probably prevented us from finding the putative correlation between iron imaging markers and MMSE, and that QSM would reveal the correlation.

However, our current results contradict our assumption of a possible correlation between iron markers and MMSE. As the pulvinar nucleus has wide connections with the visual and limbic cortices, which participate in visual attention and social cognition [37], we evaluated CDRSOB, which assesses not only global cognition, but also hygiene and social functioning of patients with dementia [38]. While the MMSE assesses only cognition, CDRSOB appraises global cognition and daily life activities, which require more complex cognitive functions, including visual attention and social cognitive skills. We hypothesized that CDRSOB scores would be related to pulvinar iron accumulation. However, the results did not support our hypothesis.

Likewise, neither MMSE nor CDRSOB scores were related to the iron load in any regions in the AD group, after controlling for age. Our findings contradict the previous studies on mouse and human brains of AD. The iron burden of AD facilitates oxidative stress in the brain, promotes amyloid-β protein (Aβ) aggregation [1, 39], and facilitates hyperphosphorylation of tau [1, 40]. Mouse models have revealed that iron-induced amyloid-β protein precursor misprocessing hastens cognitive decline through inordinate extrasynaptic N-methyl-D-aspartate receptor (NMDAR) activation [41]. This has been also confirmed in human studies, which report increased brain iron accumulation in the early stages of AD, across the entire age range [33, 42]. This lack of relationship between iron accumulation and cognitive decline in both AD and VaD in our current study may be related to the relatively small group size that likely rendered this study relatively underpowered in reliably assessing the potential effect of iron accumulation on cognitive decline. Therefore, further studies should include larger sample sizes to address this issue. Alternatively, iron accumulation may be one of the earliest changes in disease progression and may not have a linear correlation with cognitive decline.

Our study has the strength that we used a relatively accurate method available to evaluate iron deposition within the brain, namely QSM. QSM is a recently-introduced MR imaging technique, which uses both phase and magnitude images for iron quantification [17 , 43]. QSM is superior to R2^* mapping for detecting a significant increase in iron. R2^* mapping is readily confounded by the heterogeneity of the field, depending on orientation, field strength, and the distribution of susceptibility sources [17 , 43]. In contrast, QSM provides reliable iron quantification by avoiding such local field heterogeneities [17 , 43]. Moreover, QSM has improved contrast of the subcortical nuclei as compared with other sequences, such as T2-weighted, T2^*-weighted, R2^* mapping, susceptibility-weighted, and phase images [19, 25]. By using another novel iron-focused MRI sequence, we could further investigate the iron status of dementia due to any causes.

Nevertheless, our study has some limitations. First, we lacked pathologically confirmed patients. However, we meticulously diagnosed the patients with comprehensive neuropsychological assessments and 3T MRI, which likely ruled out patients with mixed AD pathology. The relatively small number of patients with VaD and normal control subjects could be another limitation of our study. Age differences between controls and dementia groups are the weakest point of the study though we attempt to correct for this. Moreover, a selective ROI analysis could be a concern, because Acosta-Cabronero et al. reported substantial iron overload in amygdala, and posterior cerebral regions as well [44]. However, as the most iron abundant regions are the GP, PUT, CN, and pulvinar nucleus of the thalamus, we targeted these areas to investigate the images more effectively. Finally, our study was limited by the use of cross-sectional data, which makes it difficult to investigate the dynamic course of iron deposition. A longitudinal study with a larger population would help elucidate the mechanisms and possible diagnostic roles of iron deposition in dementia.

CONCLUSIONS

The present study explored the pattern and presence of brain iron accumulation in VaD and AD using QSM of MR. We found increased iron deposition in the putamen and caudate nucleus in patients with VaD and AD, but the iron-load of the deep gray matter of the brain was not related to the severity of cognitive decline. Further evaluations are needed to determine the temporal changes in iron load and their diagnostic role in dementia pathology.

ACKNOWLEDGMENTS

This study was supported by a grant from the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (HI12C0713). The authors thank Chung-Hwan Kang, RT and Radiology team for their support and guidance in the imaging data acquisition. Most importantly, the authors thank all those who participated in the study for their dedication to helping research in dementia.

Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/15-1037r1).

References

Ayton

, Lei

, Bush

(2013) Metallostasis in Alzheimer’s disease. Free Radic Biol Med 62, 76–89.

Sanders-Loehr

(1988) Involvement of oxo-bridged binuclear iron centers in oxygen transport, oxygen reduction, and oxygenation. Prog Clin Biol Res 274, 193–209.

Friedlich

, Tanzi

, Rogers

(2007) The 5’-untranslated region of Parkinson’s disease alpha-synuclein messengerRNA contains a predicted iron responsive element. Mol Psychiatry 12, 222–223.

Rogers

, Randall

, Cahill

, Eder

, Huang

, Gunshin

, Leiter

, McPhee

, Sarang

, Utsuki

, Greig

, Lahiri

, Tanzi

, Bush

, Giordano

, Gullans

(2002) An iron-responsive element type II in the 5’-untranslated region of the Alzheimer’s amyloid precursor protein transcript. J Biol Chem 277, 45518–45528.

Landschulz

, Thesleff

, Ekblom

(1984) A lipophilic iron chelator can replace transferrin as a stimulator of cell proliferation and differentiation. J Cell Biol 98, 596–601.

Schipper

(2004) Brain iron deposition and the free radical-mitochondrial theory of ageing. Ageing Res Rev 3, 265–301.

Zecca

, Youdim

, Riederer

, Connor

, Crichton

(2004) Iron, brain ageing and neurodegenerative disorders. Nat Rev Neurosci 5, 863–873.

Kostka

, Hogen

, Danzer

, Levin

, Habeck

, Wirth

, Wagner

, Glabe

, Finger

, Heinzelmann

, Garidel

, Duan

, Ross

, Kretzschmar

, Giese

(2008) Single particle characterization of iron-induced pore-forming alpha-synuclein oligomers. J Biol Chem 283, 10992–11003.

Namgaladze

, Hofer

, Ullrich

(2002) Redox control of calcineurin by targeting the binuclear Fe(2+)-Zn(2+) center at the enzyme active site. J Biol Chem 277, 5962–5969.

10.

Hallgren

, Sourander

(1958) The effect of age on the non-haemin iron in the human brain. J Neurochem 3, 41–51.

11.

Brass

, Chen

, Mulkern

, Bakshi

(2006) Magnetic resonance imaging of iron deposition in neurological disorders. Top Magn Reson Imaging 17, 31–40.

12.

Sullivan

, Adalsteinsson

, Rohlfing

, Pfefferbaum

(2009) Relevance of iron deposition in deep gray matter brain structures to cognitive and motor performance in healthy elderly men and women: Exploratory findings. Brain Imaging Behav 3, 167–175.

13.

Ghadery

, Pirpamer

, Hofer

, Langkammer

, Petrovic

, Loitfelder

, Schwingenschuh

, Seiler

, Duering

, Jouvent

, Schmidt

, Fazekas

, Mangin

, Chabriat

, Dichgans

, Ropele

, Schmidt

(2015) R2* mapping for brain iron: Associations with cognition in normal aging. Neurobiol Aging 36, 925–932.

14.

Bilgic

, Pfefferbaum

, Rohlfing

, Sullivan

, Adalsteinsson

(2012) MRI estimates of brain iron concentration in normal aging using quantitative susceptibility mapping. Neuroimage 59, 2625–2635.

15.

Liem

, Lesnik Oberstein

, Versluis

, Maat-Schieman

, Haan

, Webb

, Ferrari

, van Buchem

, van der Grond

(2012) 7 T MRI reveals diffuse iron deposition in putamen and caudate nucleus in CADASIL. J Neurol Neurosurg Psychiatry 83, 1180–1185.

16.

De Reuck

, Deramecourt

, Auger

, Durieux

, Cordonnier

, Devos

, Defebvre

, Moreau

, Caparros-Lefebvre

, Leys

, Maurage

, Pasquier

, Bordet

(2014) Iron deposits in post-mortem brains of patients with neurodegenerative and cerebrovascular diseases: A semi-quantitative 7.0 T magnetic resonance imaging study. Eur J Neurol 21, 1026–1031.

17.

Wang

, Liu

(2015) Quantitative susceptibility mapping (QSM): Decoding MRI data for a tissue magnetic biomarker. Magn Reson Med 73, 82–101.

18.

Liu

, Li

, Tong

, Yeom

, Kuzminski

(2015) Susceptibility-weighted imaging and quantitative susceptibility mapping in the brain. J Magn Reson Imaging 42, 23–41.

19.

Barbosa

, Santos

, Tumas

, Liu

, Zheng

, Haacke

, Salmon

(2015) Quantifying brain iron deposition in patients with Parkinson’s disease using quantitative susceptibility mapping, R2 and R2*. Magn Reson Imaging 33, 559–565.

20.

Moon

, Moon

, Kim

, Han

(2014) Regional atrophy of the insular cortex is associated with neuropsychiatric symptoms in Alzheimer’s disease patients. Eur Neurol 71, 223–229.

21.

Roman

, Tatemichi

, Erkinjuntti

, Cummings

, Masdeu

, Garcia

, Amaducci

, Orgogozo

, Brun

, Hofman

(1993) Vascular dementia: Diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 43, 250–260.

22.

McKhann

, Drachman

, Folstein

, Katzman

, Price

, Stadlan

(1984) Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34, 939–944.

23.

Liu

, Liu

, de Rochefort

, Spincemaille

, Khalidov

, Ledoux

, Wang

(2011) Morphology enabled dipole inversion (MEDI) from a single-angle acquisition: Comparison with COSMOS in human brain imaging. Magn Reson Med 66, 777–783.

24.

Liu

, Liu

, de Rochefort

, Ledoux

, Khalidov

, Chen

, Tsiouris

, Wisnieff

, Spincemaille

, Prince

, Wang

(2012) Morphology enabled dipole inversion for quantitative susceptibility mapping using structural consistency between the magnitude image and the susceptibility map. Neuroimage 59, 2560–2568.

25.

Liu

, Eskreis-Winkler

, Schweitzer

, Chen

, Kaplitt

, Tsiouris

, Wang

(2013) Improved subthalamic nucleus depiction with quantitative susceptibility mapping. Radiology 269, 216–223.

26.

Liu

, Khalidov

, de Rochefort

, Spincemaille

, Liu

, Tsiouris

, Wang

(2011) A novel background field removal method for MRI using projection onto dipole fields (PDF). NMR Biomed 24, 1129–1136.

27.

Liu

, Wisnieff

, Lou

, Chen

, Spincemaille

, Wang

(2013) Nonlinear formulation of the magnetic field to source relationship for robust quantitative susceptibility mapping. Magn Reson Med 69, 467–476.

28.

Liu

, Li

, Yang

, Gui

, Zhao

, Evans

, Yin

, Wang

(2015) Characterizing brain iron deposition in subcortical ischemic vascular dementia using susceptibility-weighted imaging: An in vivo MR study. Behav Brain Res 288, 33–38.

29.

Dietrich

, Bradley

Jr (1988) Iron accumulation in the basal ganglia following severe ischemic-anoxic insults in children. Radiology 168, 203–206.

30.

Won

, Lee

, Park

, Gwag

, Lee

(2011) Iron mediates endothelial cell damage and blood-brain barrier opening in the hippocampus after transient forebrain ischemia in rats. Exp Mol Med 43, 121–128.

31.

Zhu

, Zhong

, Wang

, Zhan

, Wang

, Qi

, Wang

, Lei

(2009) Quantitative MR phase-corrected imaging to investigate increased brain iron deposition of patients with Alzheimer disease. Radiology 253, 497–504.

32.

Smith

, Hirai

, Hsiao

, Pappolla

, Harris

, Siedlak

, Tabaton

, Perry

(1998) Amyloid-beta deposition in Alzheimer transgenic mice is associated with oxidative stress. J Neurochem 70, 2212–2215.

33.

Graham

, Nasaruddin

, Carey

, Holscher

, McGuinness

, Kehoe

, Love

, Passmore

, Elliott

, Meharg

, Green

(2014) Age-associated changes of brain copper, iron, and zinc in Alzheimer’s disease and dementia with Lewy bodies. J Alzheimers Dis 42, 1407–1413.

34.

Ward

, Zucca

, Duyn

, Crichton

, Zecca

(2014) The role of iron in brain ageing and neurodegenerative disorders. Lancet Neurol 13, 1045–1060.

35.

Moon

, Kim

, Roh

, Choi

, Han

(2012) Fluid-attenuated inversion recovery hypointensity of the pulvinar nucleus of patients with Alzheimer disease: Its possible association with iron accumulation as evidenced by the t2(*) map. Korean J Radiol 13, 674–683.

36.

Modica

, Zivadinov

, Dwyer

, Bergsland

, Weeks

, Benedict

(2015) Iron and volume in the deep gray matter: Association with cognitive impairment in multiple sclerosis. AJNR Am J Neuroradiol 36, 57–62.

37.

Benarroch

(2015) Pulvinar: Associative role in cortical function and clinical correlations. Neurology 84, 738–747.

38.

Coley

, Andrieu

, Jaros

, Weiner

, Cedarbaum

, Vellas

(2011) Suitability of the Clinical Dementia Rating-Sum of Boxes as a single primary endpoint for Alzheimer’s disease trials. Alzheimers Dement 7, 602–610.

39.

Liu

, Moloney

, Meehan

, Morris

, Thomas

, Serpell

, Hider

, Marciniak

, Lomas

, Crowther

(2011) Iron promotes the toxicity of amyloid beta peptide by impeding its ordered aggregation. J Biol Chem 286, 4248–4256.

40.

Yamamoto

, Shin

, Hasegawa

, Naiki

, Sato

, Yoshimasu

, Kitamoto

(2002) Iron (III) induces aggregation of hyperphosphorylated tau and its reduction to iron (II) reverses the aggregation: Implications in the formation of neurofibrillary tangles of Alzheimer’s disease. J Neurochem 82, 1137–1147.

41.

Becerril-Ortega

, Bordji

, Freret

, Rush

, Buisson

(2014) Iron overload accelerates neuronal amyloid-beta production and cognitive impairment in transgenic mice model of Alzheimer’s disease. Neurobiol Aging 35, 2288–2301.

42.

House

, St Pierre

, Kowdley

, Montine

, Connor

, Beard

, Berger

, Siddaiah

, Shankland

, Jin

(2007) Correlation of proton transverse relaxation rates (R2) with iron concentrations in postmortem brain tissue from alzheimer’s disease patients. Magn Reson Med 57, 172–180.

43.

Haacke

, Liu

, Buch

, Zheng

, Wu

, Ye

(2015) Quantitative susceptibility mapping: Current status and future directions. Magn Reson Imaging 33, 1–25.

44.

Acosta-Cabronero

, Williams

, Cardenas-Blanco

, Arnold

, Lupson

, Nestor

(2013) In vivo quantitative susceptibility mapping (QSM) in Alzheimer’s disease. Plos One 8, e81093.