You are here

Article Text

Article menu
Movement disorders
Research paper
Iron accumulation in the basal ganglia in Huntington's disease: cross-sectional data from the IMAGE-HD study
  1. Juan F Domínguez D1,
  2. Amanda C L Ng2,
  3. Govinda Poudel1,3,4,
  4. Julie C Stout1,
  5. Andrew Churchyard5,
  6. Phyllis Chua1,
  7. Gary F Egan1,3,
  8. Nellie Georgiou-Karistianis1
  1. 1School of Psychological Sciences, Monash University, Clayton, Victoria, Australia
  2. 2Department of Anatomy and Neuroscience, The University of Melbourne, Parkville, Victoria, Australia
  3. 3Monash Biomedical Imaging (MBI), Monash University, Melbourne, Victoria, Australia
  4. 4VLSCI Life Sciences Computation Centre, Melbourne, Victoria, Australia
  5. 5Department of Neurology, Monash Medical Centre, Clayton, Victoria, Australia
  1. Correspondence to Professor Nellie Georgiou-Karistianis, School of Psychological Sciences, Monash University, Clayton, VIC 3800, Australia; nellie.georgiou-karistianis{at}monash.edu

Abstract

Objectives To measure iron accumulation in the basal ganglia in Huntington's disease (HD) using quantitative susceptibility mapping (QSM), and to ascertain its relevance in terms of clinical and disease severity.

Methods In this cross-sectional investigation, Embedded Image weighted imaging was undertaken on 31 premanifest HD, 32 symptomatic HD and 30 control participants as part of the observational IMAGE-HD study. Group differences in iron accumulation were ascertained with QSM. Associations between susceptibility values and disease severity were also investigated.

Results Compared with controls, both premanifest and symptomatic HD groups showed significantly greater iron content in pallidum, putamen and caudate. Additionally, iron accumulation in both putamen and caudate was significantly associated with disease severity.

Conclusions These findings provide the first evidence that QSM is sensitive to iron deposition in subcortical target areas across premanifest and symptomatic stages of HD. Such findings could open up new avenues for biomarker development and therapeutic intervention.

  • HUNTINGTON'S
  • IRON DEPOSITION
  • MRI
  • NEUROPATHOLOGY
  • MOVEMENT DISORDERS

https://doi.org/10.1136/jnnp-2014-310183

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Iron accumulation has been implicated in the pathogenesis of a raft of neurodegenerative conditions, including Huntington's disease (HD).1–3 Increased iron levels can potentiate free radicals, which contribute to neuronal loss via oxidative stress.2 ,4 In HD, while dysregulation of iron metabolism is thought to be generalised,1 the basal ganglia and, in particular, the striatum (primary target of neurodegeneration), exhibit the highest increase in iron deposition.3 ,5 Neuroimaging studies of iron in HD using a range of approaches (including T2 hypointensities, R2 and Embedded Image relaxometry, magnetic field correlation (MFC) and field-dependent R2 increase (FDRI)) have confirmed greater iron deposition in the basal ganglia during premanifest and/or symptomatic stages, although increases in all basal ganglia structures have not been consistently reported in all stages across studies.3 ,6–10 Iron increases have also been observed in cortical regions in symptomatic patients at an advanced disease stage (frontal, precentral and paracentral, parietal, cingulate, precuneus cortices).8 Moreover, reduced iron has been reported in cortical areas (premotor and parietotemporo-occipital cortices) during the early symptomatic stages, and in white matter regions (frontal lobe white matter, and genu and body of corpus callosum), during both early and advanced symptomatic stages.3 ,8 ,11 ,12

    Quantitative susceptibility mapping (QSM) is a novel and validated postprocessing method capable of quantifying iron content with high sensitivity and specificity from the magnetic susceptibility of tissue samples in gradient-echo MRI.13–17 Other techniques for measuring iron infer iron content from inhomogeneity in the B0 field (eg, phase-based methods), or changes in Embedded Image that are induced by inhomogeneities in the B0 field (eg, Embedded Image and Embedded Image-based methods). Iron has high magnetic susceptibility, and therefore induces perturbations in the B0 field. However, susceptibility effects on the B0 field are non-local, as is evidenced in susceptibility-induced air-tissue interface signal loss artefacts. QSM estimates susceptibility measures and is thus not affected by the non-local nature of B0 inhomogeneities. QSM, therefore, holds great potential for obtaining measures of iron accumulation that can be used as biomarkers of tissue composition in HD. Importantly, such methods may also reveal treatment effects and offer new avenues to explore neuroprotective treatments.18 ,19

    In this investigation, we used QSM for the first time to measure iron deposition in the basal ganglia in premanifest and symptomatic HD. Magnetic susceptibility values were obtained from pallidum, putamen and caudate, subcortical structures affected early during the disease process and also known for their high levels of iron concentration relative to other brain regions in healthy individuals.14 ,20 Susceptibility values were also measured in the thalamus as a control structure relatively spared in HD. In line with the literature, we predicted that significantly increased iron deposition would be observed in the pallidum, putamen and caudate across both HD groups compared with controls. No group differences were expected in the thalamus. Additionally, we hypothesised that iron increase in these structures (pallidum, putamen and caudate) would be associated with clinical and disease severity.

    Participants and methods

    Participants

    One hundred and eight participants (36 premanifest HD (pre-HD), 36 symptomatic HD (symp-HD) and 36 healthy controls matched to the pre-HD group) were recruited as part of the observational Australian-based IMAGE-HD study21–27 between September 2008 and October 2009. Participants underwent T1-weighted and Embedded Image-weighted imaging. Embedded Image-weighted scans were available for 31 pre-HD, 35 symp-HD and 31 controls. Additionally, three symp-HD participants and one control were excluded due to artefacts in the Embedded Image-weighted scans. Results are therefore presented for 31 pre-HD, 32 symp-HD and 30 controls. Participants with the mutant Huntingtin gene (ie, with an expansion of the cytosine-adenine-guanine (CAG) repeat established prior to enrolment in the study) were clinically assessed by a neurologist (AC and PC) with the Unified Huntington's Disease Rating Scale (UHDRS)28 total motor score (TMS). Individuals with a UHDRS TMS ≤5 were included in the pre-HD group, and those with UHDRS TMS >5 were included in the symp-HD group as per our previous reports.21 ,22 CAG repeat length ranged from 39 to 50 (42.4±2.0 for pre-HD; 43.2±2.5 for symp-HD). The pre-HD group had an average estimated years to clinical onset29 of 14.6±5.9 years, while the symp-HD group had an estimated average years since diagnosis of 1.9±1.6 years. See table 1 for demographic and clinical characteristics.

    View this table:
    Table 1

    Demographic and clinical characteristics for study participants

    All participants underwent a rigorous screening process prior to recruitment. Participants were free from brain injury, neurological and/or severe diagnosed psychiatric conditions (eg, bipolar, psychosis), other than HD. Written informed consent was obtained from each participant in accordance with the Helsinki Declaration. The study was approved by the Monash University and Melbourne Health Human Research Ethics committees.

    MRI acquisition

    MRI of the entire brain was performed with a Siemens Magnetom Trio Tim 3 Tesla scanner at the Murdoch Childrens Research Institute (Royal Children’s Hospital, Victoria, Australia). Phase images were acquired using a Embedded Image-weighted gradient-recalled echo (GRE) sequence (repetition time (TR)=32 ms, time to echo (TE)=23 ms, α=15°, 88 slices, field of view (FOV)=384×512, voxel size=0.45 mm×0.45 mm×1.5 mm). High-resolution T1-weighted images were also acquired (TR=1900 ms, TE=2.59 ms, α=9°, 192 slices, FOV=320×320, voxel size=0.9 mm×0.8 mm× 0.8 mm).

    Quantitative susceptibility mapping

    For each participant, the phase from each individual channel was unwrapped using Laplacian-based phase unwrapping.30 The reconstructed phase image was then calculated by averaging the unwrapped channel phase images. V-SHARP31 filtering was applied to remove the bias field, and then QSM was calculated using the LSQR method.17 ,32

    QSM yields relative rather than absolute susceptibility measures. Susceptibility values measured in all structures were therefore referenced to that of cerebrospinal fluid sampled from the lateral ventricles as per previous studies.17

    Segmentation of subcortical structures

    FMRIB's Software Library (FSL, V.4.1.61) was used to delineate pallidum, putamen, caudate and thalamus (figure 1). Each participant's T1-weighted image was first linearly realigned to the respective magnitude image with FLIRT (FSL's linear registration tool). Subcortical masks were then generated from this image using FIRST (FSL’s automated segmentation and registration tool). Two spherical masks of radius 3 voxels (volume 81 voxels) located in the left and right lateral ventricles (centred at MNI coordinates (−5, 1, 16) and (6, 1, 16), 1 mm MNI space) were linearly (FLIRT), then non-linearly (FNIRT, FSL's non-linear registration tool) registered onto the T1-weighted image after realignment to the magnitude image. All masks were then binarised and used to obtain magnetic susceptibility values from the respective QSM images. Segmentations and registrations were visually inspected to ensure the masks accurately enabled structure-wide measurements.

    Figure 1
    Figure 1

    Automated identification of subcortical structures across groups in quantitative susceptibility mapping (QSM) images. The process is illustrated on a control participant (A–E). After calculating QSM (A), the T1-wighted image was linearly registered to the magnitude image, resulting in T1-MAG (B); subcortical structures were segmented from this T1-MAG image (C and D). Masks generated were then used to obtain susceptibility measures from the respective QSM map (E). Examples include one pre-Huntington's disease (HD) and one symp-HD participant, provided in (F) and (G), respectively. Segmented putamen (magenta), caudate (green), pallidum (purple) and thalamus (yellow) are rendered in three-dimension in (C) viewed from above (top), profile (middle) and from below (bottom) (CD, caudate; PT, putamen; PD, pallidum; TH, thalamus).

    Statistical analysis

    Susceptibility values across the subcortical structures were statistically compared between groups using linear regression (controlling for hemisphere, sex and age). Contrasts specified t tests for differences between the HD groups (pre-HD and symp-HD) and controls. To evaluate their validity, we also carried out linear fittings between our iron measurements and measurements reported in previous studies in controls, pre-HD and symp-HD and estimated the associated coefficient of determination (R2). Partial correlations between susceptibility values in the subcortical structures (averaged across hemispheres) and UHDRS TMS (with disease burden score, sex and age as covariates), as well as disease burden score (with sex and age as covariates) were performed. Statistical significance for all analyses was defined using a probability threshold <0.05 corrected for multiple tests. We report results from bootstrapped regressions performed on the basis of 5000 permutations for analyses where normality assumptions were violated. Stata V.1233 was used for statistical analysis.

    Results

    Compared with controls, both pre-HD and symp-HD groups showed significantly greater iron deposition in the pallidum, putamen and caudate after adjusting for hemisphere, sex and age (see table 2 and figure 2 for adjusted results; see online supplementary table S1 for unadjusted results). Evaluation of iron content in controls, as reflected in susceptibility values against reported iron content measured through biochemical techniques and QSM in previous studies,16 ,17 ,20 ,34 ,35 revealed excellent linear approximations: coefficient of determination (R2) >0.90 (see online supplementary table S2 and supplementary figure S1A,B). Iron content in the HD groups, as previously estimated through relaxometry,7 MFC9 and FDRI,3 similarly exhibited good to excellent linear approximations with the iron content in the present study (pre-HD R2=0.77 and 0.80; symp-HD R2=0.80–0.97) (see online supplementary table S2 and supplementary figure S1C,D). We also found statistically significant positive correlations between susceptibility values and disease burden scores in all HD participants (pre-HD and symp-HD together) in the putamen (r=0.39, p<0.001) and caudate (r=0.32, p<0.01) (see figure 3).

    View this table:
    Table 2

    Magnetic susceptibility (ppm) in subcortical structures and significant group differences

    Figure 2
    Figure 2

    Adjusted mean (±SEM) magnetic susceptibility values in pre-Huntington's disease (HD), symp-HD and controls in subcortical regions of interest. Susceptibility values are relative to cerebrospinal fluid; ppm, parts per million. Statistical significance of differences between pre-HD/symp-HD and control groups at a Bonferroni-corrected threshold of α=0.05: *p≤0.05, **p≤0.01, ***p≤0.001.

    Figure 3
    Figure 3

    Associations between susceptibility values in the putamen and caudate with disease burden score. Susceptibility values are relative to cerebrospinal fluid; ppm, parts per million. Statistical significance for correlations at a Bonferroni-corrected threshold of α=0.05: *p≤0.05, **p≤0.01.

    Discussion

    This is the first study in HD using QSM to measure iron levels in the basal ganglia. Significantly increased iron deposition was found in the pallidum, putamen and caudate, both before and after symptom onset. Consistent with our predictions, no differences in iron content were observed in the thalamus. Moreover, we found that iron increase in putamen and caudate was associated with disease severity.

    Our findings are consistent with the available imaging studies on iron concentration in HD using various neuroimaging approaches.3 ,6–10 ,12 In these previous studies, greater iron accumulation has been collectively reported in pallidum, putamen and caudate in both pre-HD and symp-HD. However, findings have not been fully consistent across studies. For example, Dumas et al9 (using MFC) found increased iron in pallidum, putamen and caudate in symp-HD but not pre-HD; Jurgens et al10 (T2 hypointensities) found increased iron in pre-HD in pallidum but not in putamen or caudate; Rosas et al8 (field mapping evolution) found increased iron in pallidum and putamen in pre-HD and symp-HD and in caudate in symp-HD only; Vymazal et al6 (R2 relaxometry) found increased iron in symp-HD in pallidum but not in putamen or caudate; and Sánchez-Castañeda et al12 (Embedded Image relaxometry) found increased iron in pallidum and putamen in pre-HD and symp-HD and in caudate increased iron in pre-HD only. By contrast with the above studies, IMAGE-HD is the first to report increased iron deposition in pallidum, putamen and caudate across both the premanifest and symptomatic stages. Moreover, iron accumulation in putamen and caudate was found to significantly correlate with disease burden score. Participant numbers and demographic and clinical characteristics in the present study are, overall, comparable with other studies. This suggests that QSM is a sensitive method of quantifying iron accumulation in HD across all stages of the disease. However, establishing which method of iron quantification is most sensitive (QSM, Embedded Image or Embedded Image based) will require future studies to directly compare measurements taken from the same cohort.

    The normal Huntingtin protein (Htt) has been shown to be involved in iron homoeostasis.36 The increased magnetic susceptibility we report in basal ganglia nuclei in HD groups, may be, in part, a result of the role of mutant Htt (mHtt) in iron homoeostasis. mHtt Impairs iron homoeostasis in a number of ways, eventually resulting in iron increases in tissue. For example, ferritin (iron-binding protein) degradation is inhibited by mHtt, which leads to more iron transport.36 Moreover, mHtt acts downstream to stop cellular uptake of iron, causing excess iron in the extracellular space.36 Owing to a lack of iron uptake, iron-dependent enzymes inside cells cannot function normally, resulting in deviations from the normal metabolic pathways. This can cause excess free radical production, most notable in dopaminergic tracts of the basal ganglia.36 Free radicals trigger the release of more iron, thus ensuring that a cycle of iron production and cell death will continue with disease progression.36 Elevated iron in the basal ganglia in HD has also been shown to result, paradoxically, from the brain's homoeostatic attempts at remyelination in response to fibre loss.3 Most seriously affected are striatopallidal connections, which are selectively lost likely as a result of the death of striatal spiny neurons.37 Not only is iron required for remyelination, but oligodendrocytes, which are involved in myelination and repair, have the highest iron concentration of all cell types, and are found in elevated numbers in the basal ganglia well before symptom onset.3

    When interpreting the results from this study, it is important to consider that the GRE data were acquired for susceptibility weighted imaging (SWI), and consequently, acquisition involved a single echo as per standard SWI protocols. Ideally, a dual echo acquisition would produce reliable coil phase offset correction.38 QSM processing therefore followed a previously published method,31 ,39 adapted for retrospective analysis of single-echo GRE data. In-depth analysis of the different phase processing methods and their effects on QSM results has not been undertaken and is outside the scope of this investigation. However, the susceptibility values reported for controls were within the range of variation previously reported in other QSM studies,16 ,17 ,35 ,38 reflected the same rank ordering by brain structure (from highest to lowest: pallidum, putamen, caudate and thalamus) and obtained good to excellent linear approximation not only with measures from other QSM16 ,17 ,35 ,40 but also biochemical studies.20 ,34 This was also the case for pre-HD and symp-HD groups, with reference to iron content measured with MRI techniques other than QSM.3 ,7 ,9 Further studies that employ dual-echo acquisition of GRE data are, however, required to further validate the present findings. Also important to take into account when evaluating our results is that the symp-HD group was significantly older than controls. This poses a limitation for assessing differences between these groups, despite the fact that age was included as a covariate in all analyses. Another relevant consideration is the potential effect on susceptibility measures of myelin breakdown and zinc accumulation in the basal ganglia, both of which have been previously reported in HD.3 ,8 ,37 Myelin and zinc are diamagnetic, and therefore have an effect that is opposite to the paramagnetic effect of iron. This means that myelin breakdown, associated with striatal and pallidal fibre loss, may have an additive effect on the susceptibility values, whereas increased zinc may have a subtractive effect. While the magnetic susceptibility values we report are in agreement with previous studies, myelin loss and zinc accumulation in HD may possibly have a small effect (which may anyway cancel out).

    In conclusion, we have shown that iron deposition in HD can be sensitively measured using QSM before and after symptom onset, and that increased iron in the basal ganglia is related to disease severity. Future studies are required, however, not only to replicate findings but to determine whether this method identifies meaningful change over time in the context of biomarker development. Our findings open up new avenues for therapeutic investigation, notably with conservative iron chelation,41–43 aimed at reducing iron accumulation in this disease, and where MRI morphometry methods could be used to assess their effects.

    Acknowledgments

    The authors would like to acknowledge the contribution of all the participants who took part in this study. They are also grateful to the CHDI Foundation, Inc (USA), for their support in funding this research. The authors also thank the Royal Children's Hospital for the use of their 3 T MR scanner. GFE is a principal NHMRC research fellow.

    References

    1. ↵
      1. Berg D,
      2. Youdim MB
      . Role of iron in neurodegenerative disorders. Top Magn Reson Imaging 2006;17:5–17. doi:10.1097/01.rmr.0000245461.90406.ad
      OpenUrlCrossRefPubMed
    2. ↵
      1. Bartzokis G,
      2. Cummings J,
      3. Perlman S, et al
      . Increased basal ganglia iron levels in Huntington disease. Arch Neurol 1999;56:569–74. doi:10.1001/archneur.56.5.569
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Bartzokis G,
      2. Lu PH,
      3. Tishler TA, et al
      . Myelin breakdown and iron changes in Huntington's disease: pathogenesis and treatment implications. Neurochem Res 2007;32:1655–64. doi:10.1007/s11064-007-9352-7
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Browne SE,
      2. Beal MF
      . Oxidative damage in Huntington's disease pathogenesis. Antioxid Redox Signal 2006;8:2061–73. doi:10.1089/ars.2006.8.2061
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Dexter DT,
      2. Jenner P,
      3. Schapira AH, et al
      . Alterations in levels of iron, ferritin, and other trace metals in neurodegenerative diseases affecting the basal ganglia. The Royal Kings and Queens Parkinson's Disease Research Group. Ann Neurol 1992;32(Suppl):S94–100. doi:10.1002/ana.410320716
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Vymazal J,
      2. Klempir J,
      3. Jech R, et al
      . MR relaxometry in Huntington's disease: correlation between imaging, genetic and clinical parameters. J Neurol Sci 2007;263:20–5. doi:10.1016/j.jns.2007.05.018
      OpenUrlCrossRefPubMed
    7. ↵
      1. Sánchez-Castañeda C,
      2. Cherubini A,
      3. Elifani F, et al
      . Seeking Huntington disease biomarkers by multimodal, cross-sectional basal ganglia imaging. Hum Brain Mapp 2013;34:1625–35. doi:10.1002/hbm.22019
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Rosas HD,
      2. Chen YI,
      3. Doros G, et al
      . Alterations in brain transition metals in Huntington disease: an evolving and intricate story. Arch Neurol 2012;69:887–93. doi:10.1001/archneurol.2011.2945
      OpenUrlCrossRefPubMed
    9. ↵
      1. Dumas EM,
      2. Versluis MJ,
      3. van den Bogaard SJ, et al.
      , TRACK-HD investigators. Elevated brain iron is independent from atrophy in Huntington's disease. Neuroimage 2012;61:558–64. doi:10.1016/j.neuroimage.2012.03.056
      OpenUrlCrossRefPubMed
    10. ↵
      1. Jurgens CK,
      2. Jasinschi R,
      3. Ekin A, et al
      . MRI T2 hypointensities in basal ganglia of premanifest Huntington's disease. PLoS Curr 2010;2:pii: RRN1173. doi:10.1371/currents.RRN1173
      OpenUrlCrossRefPubMed
    11. ↵
      1. Di Paola M,
      2. Phillips OR,
      3. Sanchez-Castaneda C, et al
      . MRI measures of corpus callosum iron and myelin in early Huntington's disease. Hum Brain Mapp 2014;35:3143–51. doi:10.1002/hbm.22391
      OpenUrlCrossRefPubMed
    12. ↵
      1. Sánchez-Castañeda C,
      2. Squitieri F,
      3. Di Paola M, et al
      . The role of iron in gray matter degeneration in Huntington's disease: a magnetic resonance imaging study. Hum Brain Mapp 2015;36:50–66. doi:10.1002/hbm.22612
      OpenUrlCrossRefPubMed
    13. ↵
      1. Langkammer C,
      2. Krebs N,
      3. Goessler W, et al
      . Susceptibility induced gray-white matter MRI contrast in the human brain. Neuroimage 2012;59:1413–19. doi:10.1016/j.neuroimage.2011.08.045
      OpenUrlCrossRefPubMed
    14. ↵
      1. Deistung A,
      2. Schafer A,
      3. Schweser F, et al
      . Toward in vivo histology: a comparison of quantitative susceptibility mapping (QSM) with magnitude-, phase-, and R2*-imaging at ultra-high magnetic field strength. Neuroimage 2013;65:299–314. doi:10.1016/j.neuroimage.2012.09.055
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. de Rochefort L,
      2. Liu T,
      3. Kressler B, et al
      . Quantitative susceptibility map reconstruction from MR phase data using Bayesian regularization: validation and application to brain imaging. Magn Reson Med 2010;63:194–206. doi:10.1002/mrm.22187
      OpenUrlPubMed
    16. ↵
      1. Lim IA,
      2. Faria AV,
      3. Li X, et al
      . Human brain atlas for automated region of interest selection in quantitative susceptibility mapping: application to determine iron content in deep gray matter structures. Neuroimage 2013;82:449–69. doi:10.1016/j.neuroimage.2013.05.127
      OpenUrlCrossRefPubMed
    17. ↵
      1. Wharton S,
      2. Bowtell R
      . Whole-brain susceptibility mapping at high field: a comparison of multiple- and single-orientation methods. Neuroimage 2010;53:515–25. doi:10.1016/j.neuroimage.2010.06.070
      OpenUrlCrossRefPubMed
    18. ↵
      1. Kozauer N,
      2. Katz R
      . Regulatory innovation and drug development for early-stage Alzheimer's disease. N Engl J Med 2013;368:1169–71. doi:10.1056/NEJMp1302513
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Tabrizi SJ,
      2. Langbehn DR,
      3. Leavitt BR, et al
      . Biological and clinical manifestations of Huntington's disease in the longitudinal TRACK-HD study: cross-sectional analysis of baseline data. Lancet Neurol 2009;8:791–801. doi:10.1016/S1474-4422(09)70170-X
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Hallgren B,
      2. Sourander P
      . The effect of age on the non-haemin iron in the human brain. J Neurochem 1958;3:41–51. doi:10.1111/j.1471-4159.1958.tb12607.x
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Georgiou-Karistianis N,
      2. Gray MA,
      3. Domínguez D JF, et al
      . Automated differentiation of pre-diagnosis Huntington's disease from healthy control individuals based on quadratic discriminant analysis of the basal ganglia: the IMAGE-HD study. Neurobiol Dis 2013;51:82–92. doi:10.1016/j.nbd.2012.10.001
      OpenUrlCrossRefPubMed
    22. ↵
      1. Domínguez D JF,
      2. Egan GF,
      3. Gray MA, et al
      . Multi-modal neuroimaging in premanifest and early Huntington's disease: 18 month longitudinal data from the IMAGE-HD study. PLoS ONE 2013;8:e74131. doi:10.1371/journal.pone.0074131
      OpenUrlCrossRefPubMed
    23. ↵
      1. Georgiou-Karistianis N,
      2. Poudel GR,
      3. Domínguez D JF, et al
      . Functional and connectivity changes during working memory in Huntington's disease: 18 month longitudinal data from the IMAGE-HD study. Brain Cogn 2013;83:80–91. doi:10.1016/j.bandc.2013.07.004
      OpenUrlCrossRefPubMed
    24. ↵
      1. Georgiou-Karistianis N,
      2. Stout JC,
      3. Domínguez D JF, et al
      . Functional magnetic resonance imaging of working memory in Huntington's disease: cross-sectional data from the IMAGE-HD study. Hum Brain Mapp 2014;35:1847–64.. doi:10.1002/hbm.22296
      OpenUrlCrossRefPubMed
    25. ↵
      1. Poudel GR,
      2. Stout JC,
      3. Domínguez DJ, et al
      . White matter connectivity reflects clinical and cognitive status in Huntington's disease. Neurobiol Dis 2014;65C:180–7. doi:10.1016/j.nbd.2014.01.013
      OpenUrlCrossRefPubMed
    26. ↵
      1. Poudel GR,
      2. Stout JC,
      3. Gray MA, et al
      . Functional changes during working memory in Huntington's disease: 30 month longitudinal data from the IMAGE-HD study. Brain Struct Funct 2015;220:501–12. doi:10.1007/s00429-013-0670-z
      OpenUrlCrossRefPubMed
    27. ↵
      1. Poudel GR,
      2. Egan GF,
      3. Churchyard A, et al
      . Abnormal synchrony of resting state networks in premanifest and symptomatic Huntington disease: the IMAGE-HD study. J Psychiatry Neurosci 2014;39:87–96.
      OpenUrlPubMed
    28. ↵
      Huntington Study Group. Unified Huntington's disease rating scale: reliability and consistency. Mov Disord 1996;11:136–42. doi:10.1002/mds.870110204
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Langbehn DR,
      2. Brinkman RR,
      3. Falush D, et al
      . A new model for prediction of the age of onset and penetrance for Huntington's disease based on CAG length. Clin Genet 2004;65:267–77. doi:10.1111/j.1399-0004.2004.00241.x
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Schofield MA,
      2. Zhu Y
      . Fast phase unwrapping algorithm for interferometric applications. Opt Lett 2003;28:1194–6. doi:10.1364/OL.28.001194
      OpenUrlCrossRefPubMed
    31. ↵
      1. Wu B,
      2. Li W,
      3. Guidon A, et al
      . Whole brain susceptibility mapping using compressed sensing. Magn Reson Med 2012;67:137–47. doi:10.1002/mrm.23000
      OpenUrlCrossRefPubMed
    32. ↵
      1. Li W,
      2. Wu B,
      3. Liu C
      . Quantitative susceptibility mapping of human brain reflects spatial variation in tissue composition. Neuroimage 2011;55:1645–56. doi:10.1016/j.neuroimage.2010.11.088
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      StataCorp. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP, 2011.
    34. ↵
      1. Langkammer C,
      2. Krebs N,
      3. Goessler W, et al
      . Quantitative MR imaging of brain iron: a postmortem validation study. Radiology 2010;257:455–62. doi:10.1148/radiol.10100495
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Bilgic B,
      2. Pfefferbaum A,
      3. Rohlfing T, et al
      . MRI estimates of brain iron concentration in normal aging using quantitative susceptibility mapping. Neuroimage 2012;59:2625–35. doi:10.1016/j.neuroimage.2011.08.077
      OpenUrlCrossRefPubMed
    36. ↵
      1. Rivera-Mancia S,
      2. Perez-Neri I,
      3. Rios C, et al
      . The transition metals copper and iron in neurodegenerative diseases. Chem Biol Interact 2010;186:184–99. doi:10.1016/j.cbi.2010.04.010
      OpenUrlCrossRefPubMed
    37. ↵
      1. Douaud G,
      2. Behrens TE,
      3. Poupon C, et al
      . In vivo evidence for the selective subcortical degeneration in Huntington's disease. Neuroimage 2009;46:958–66. doi:10.1016/j.neuroimage.2009.03.044
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Robinson S,
      2. Grabner G,
      3. Witoszynskyj S, et al
      . Combining phase images from multi-channel RF coils using 3D phase offset maps derived from a dual-echo scan. Magn Reson Med 2011;65:1638–48. doi:10.1002/mrm.22753
      OpenUrlCrossRefPubMed
    39. ↵
      1. Li W,
      2. Wu B,
      3. Batrachenko A, et al
      . Differential developmental trajectories of magnetic susceptibility in human brain gray and white matter over the lifespan. Hum Brain Mapp 2014;35:2698–713. doi:10.1002/hbm.22360
      OpenUrlCrossRefPubMed
    40. ↵
      1. Langkammer C,
      2. Liu T,
      3. Khalil M, et al
      . Quantitative susceptibility mapping in multiple sclerosis. Radiology 2013;267:551–9. doi:10.1148/radiol.12120707
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      1. Mounsey RB,
      2. Teismann P
      . Chelators in the treatment of iron accumulation in Parkinson's disease. Int J Cell Biol 2012;2012:983245. doi:10.1155/2012/983245
      OpenUrlPubMed
    42. ↵
      1. Devos D,
      2. Moreau C,
      3. Devedjian JC, et al
      . Targeting chelatable iron as a therapeutic modality in Parkinson's disease. Antioxid Redox Signal 2014;21:195–210. doi:10.1089/ars.2013.5593
      OpenUrlCrossRefPubMedWeb of Science
    43. ↵
      1. Wilson RB
      . Therapeutic developments in Friedreich ataxia. J Child Neurol 2012;27:1212–16. doi:10.1177/0883073812449691
      OpenUrlAbstract/FREE Full Text
    View Abstract

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors JFDD was involved in drafting/revising the manuscript for content; study concept or design; analysis or interpretation of data and statistical analysis. ACLN was involved in drafting/revising the manuscript for content; study concept or design; analysis or interpretation of data and statistical analysis; quantitative susceptibility mapping. GP was involved in drafting/revising the manuscript for content; analysis or interpretation of data and acquisition of data. JCS was involved in drafting/revising the manuscript for content; study concept or design; analysis or interpretation of data and obtaining funding. AC was involved in study concept or design; obtaining funding and acquisition of data. PC was involved in study concept or design; obtaining funding and acquisition of data. GFE was involved in drafting/revising the manuscript for content; study concept or design; analysis or interpretation of data; obtaining funding and study supervision or coordination. NG-K was involved in drafting/revising the manuscript for content; study concept or design; analysis or interpretation of data; obtaining funding and study supervision or coordination.

    • Funding This work was supported by the CHDI Foundation Inc., New York, USA [A—3433]; and the National Health and Medical Research Council (NHMRC), Australia [606650].

    • Competing interests None declared.

    • Ethics approval Monash University Research Ethics Committee.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement The CHDI Foundation will be provided with all raw data on completion of the IMAGE-HD study.