In vivo assessment of brain microstructure in patients with Huntington's disease using quantitative MRI.
Journal of neurology · 2026
Abstract originale
OBJECTIVES: Using multiparametric mapping (MPM), we assessed the integrity of cortical and subcortical regions across different stages of Huntington's disease (HD), relating the derived findings to genetic modifiers and clinical measures. METHODS: Sixteen premanifest patients with HD (PwHDpre), ten manifesting PwHD (PwHDman), and forty-two healthy controls (HC) underwent clinical examinations and a 3 T imaging. We derived quantitative MPM maps to perform voxel-based quantification (VBQ) and voxel-based morphometry (VBM). We assessed group differences by combining PwHDpre and PwHDman (PwHDtotal) and by comparing PwHDpre and PwHDman with age- and sex-matched HC. Using correlation analysis, we investigated the link between MPM-derived findings, CAG repeat length, and clinical scores. RESULTS: PwHDman presented with reductions in all four quantitative MPM maps across the bilateral caudate nucleus and the frontal, parietal, temporal, and occipital gyri. We observed spatial overlap between VBQ- and VBM-derived findings, demonstrating widespread volume loss in PwHDman. PwHDpre showed reductions in certain quantitative maps within the frontal and temporal regions. Reduced MPM values correlated with clinical scores of HD severity. INTERPRETATION: The observed reductions across four MPM modalities in PwHDman are consistent with demyelination and neuronal loss, as supported by widespread volume loss. The reductions observed in PwHDpre indicate early demyelination and inflammation. In this study, we demonstrated the utility and clinical relevance of MPM for detecting microstructural alterations across different stages of HD. MPM can serve as a non-invasive biomarker of neurodegeneration in HD, providing insight into the pathological events underlying the observed neuronal loss.
Parole chiave
Leggi il testo integrale originale
Testo estratto dalla pubblicazione: sono esclusi immagini, media, supplementi e parti soggette a diritti distinti.
In vivo assessment of brain microstructure in patients with Huntington’s disease using quantitative MRI
Abstract
Objectives
Using multiparametric mapping (MPM), we assessed the integrity of cortical and subcortical regions across different stages of Huntington’s disease (HD), relating the derived findings to genetic modifiers and clinical measures.
Methods
Sixteen premanifest patients with HD (PwHDpre), ten manifesting PwHD (PwHDman), and forty-two healthy controls (HC) underwent clinical examinations and a 3 T imaging. We derived quantitative MPM maps to perform voxel-based quantification (VBQ) and voxel-based morphometry (VBM). We assessed group differences by combining PwHDpre and PwHDman (PwHDtotal) and by comparing PwHDpre and PwHDman with age- and sex-matched HC. Using correlation analysis, we investigated the link between MPM-derived findings, CAG repeat length, and clinical scores.
Results
PwHDman presented with reductions in all four quantitative MPM maps across the bilateral caudate nucleus and the frontal, parietal, temporal, and occipital gyri. We observed spatial overlap between VBQ- and VBM-derived findings, demonstrating widespread volume loss in PwHDman. PwHDpre showed reductions in certain quantitative maps within the frontal and temporal regions. Reduced MPM values correlated with clinical scores of HD severity. Interpretation. The observed reductions across four MPM modalities in PwHDman are consistent with demyelination and neuronal loss, as supported by widespread volume loss. The reductions observed in PwHDpre indicate early demyelination and inflammation. In this study, we demonstrated the utility and clinical relevance of MPM for detecting microstructural alterations across different stages of HD. MPM can serve as a non-invasive biomarker of neurodegeneration in HD, providing insight into the pathological events underlying the observed neuronal loss.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s00415-026-14244-8.
Keywords: Multiparametric mapping (MPM), Huntington’s disease (HD), Quantitative MRI, In vivo histology, Basal ganglia
Introduction
Huntington’s disease (HD) is an autosomal-dominant neurodegenerative disease caused by an expanded CAG repeat in the huntingtin (HTT) gene, with longer repeat lengths inversely correlating with earlier onset and greater disease severity [1, 2]. Although the causes of HD are well characterized, no disease-modifying treatments have been approved and only symptomatic strategies are available [3]. Furthermore, patients with HD (PwHD) exhibit highly heterogeneous symptomatology, spanning across motor, cognitive, and psychiatric domains, with emerging studies demonstrating symptom worsening with disease progression. Additionally, PwHD subgroups experience a more pronounced decline in specific domains, for reasons not yet understood [4, 5]. Such heterogeneous clinical presentation and disease progression, along with variable responses to symptomatic treatment, demonstrate the need for improved biomarkers to enhance patient stratification and monitoring.
Magnetic resonance imaging (MRI) is the most widely used neuroimaging tool for assessing basal ganglia (BG), a hallmark of HD pathology that correlates with motor symptom severity and disease duration [5, 6]. However, MRI is limited to detecting macroscopic structural changes and falls short of probing underlying molecular and pathogenic mechanisms [7]. Quantitative MRI (qMRI), such as multiparametric mapping (MPM), offers a more sensitive alternative by assessing qualitative measures of tissue biophysical properties and enabling evaluation of brain microstructural and cellular-level composition [8]. By modeling signals from multi-contrast gradient-echo acquisitions, MPM quantifies the longitudinal relaxation rate (R1), effective transverse relaxation rate (R2*), proton density (PD), and magnetization transfer saturation (MTsat) [9, 10]. As R1 and MTsat are sensitive to myelin, R2* to iron, and PD to the MR-visible mobile proton pool and tissue composition, MPM enables in vivo assessment of these tissue characteristics and the underlying pathogenic processes, yielding insights previously obtainable only with histology. Because histology assesses the end-stage of brain microstructure, its clinical relevance is limited, making MPM a better clinical alternative [8].
Despite its strong, clinically relevant potential, MPM use in HD research and clinical practice remains underexplored. Therefore, we used MPM to assess microstructural differences in the cortex, white matter (WM), and BG among manifesting PwHD (PwHDman), premanifest PwHD (PwHDpre), and healthy controls (HC). Given evidence of widespread cortical neurodegeneration in HD, we hypothesized that MPM insights would reveal cortical and BG atrophy and demyelination associated with clinical measures of HD severity [11, 12].
Methods
Experimental design and participants
We recruited PwHD from outpatient clinics at the Center for Rare Diseases and the inpatient clinics of the Department of Neurology, University Medical Center Schleswig–Holstein (Campus Lübeck). We selected PwHD based on a genetically confirmed diagnosis with fully penetrant, ≥ 40 CAG repeats in the HTT gene [13]. We classified participants with genetically confirmed HD as premanifest in the absence of unequivocal motor signs and as manifest when unequivocal HD-related motor signs were present. By contrast, we recruited healthy controls from pre-existing research cohorts at our institution. We determined eligibility as healthy controls from medical history and screening information collected at cohort enrollment; we included individuals only if they had no history of genetic or idiopathic neurological disease, alcohol or substance abuse, structural brain disease, or MRI contraindications. All PwHD underwent clinical examinations, including the Unified Huntington’s Disease Rating Scale, Total Motor Score (UHDRS-TMS), and cognitive screening with the Montreal Cognitive Assessment (MoCA) [14, 15]. MoCA was not administered to healthy controls. Additionally, we report Total Functional Capacity (TFC) and CAG-Age Product (CAP) scores. CAP scores were calculated using the previously reported formula [16],
CAP = Age × (CAG repeat length – 33.66)
The study adhered to the Declaration of Helsinki and was approved by the Committee for the Ethics of Safety- and Security-Relevant Research at the University of Luebeck (AZ-18–196). All participants provided informed written consent prior to study participation.
Because PwHDpre and PwHDman differed in mean age and sex distribution, we stratified the corresponding HC groups accordingly, ensuring each HC group matched its comparison group in age, sex distribution, and sample size. We derived the HCpre group, which we compared with PwHDpre, and the HCman group, which we compared with PwHDman. HC groups comprise only healthy participants; the “pre” and “man” labels refer only to the matched PwHD group.
Image acquisition, voxel-based quantification, and voxel-based morphometry
MRI sessions were conducted using a 3 T Siemens Magnetom Skyra scanner equipped with a 64-channel head coil. Using a multi-echo 3D fast low-angle shot, we acquired T1, PD, and magnetization transfer (MT) weightings [17]. We measured all sequences in coronal orientation with 240 contiguous slices, 1.0 mm isotropic resolution, a field of view of 256 × 176 mm2 (phase FoV 68.8%), GRAPPA acceleration factor 2 with 24 reference lines, and partial Fourier 6/8 in the slice direction. The received bandwidth for all echoes was 440 Hz/px. We enabled inline distortion correction (2D mode) and prescan normalization. The T1-weighted acquisition (TR = 19.0 ms, flip angle = 20°, six echoes with TE = 2.20/4.70/7.20/9.70/12.20/15.00 ms) required about 8 min. The PD-weighted acquisition (TR = 24.0 ms, flip angle = 6°, eight echoes with TE = 2.20/4.70/7.20/9.70/12.20/15.00/17.70/20.00 ms) required approximately 7 min. The magnetic transfer-weighted acquisition (TR = 37.0 ms, flip angle = 6°, six echoes with TE = 2.20/4.70/7.20/9.70/12.20/15.00 ms, with an off-resonance magnetization transfer pre-pulse enabled) required about 11 min. A neuroradiologist reviewed the neuroimaging data to exclude concurrent conditions. We visually assessed artifacts (e.g., head motion) and performed quantitative evaluations, including counting artifact voxels (exclusion threshold: artifacts exceeding two standard deviations above the mean).
Voxel-based quantification (VBQ) analysis was performed in the hMRI toolbox within the Statistical Parametric Mapping 25 framework (SPM25, Wellcome Trust Center for Neuroimaging, London) in MATLAB software (R2025a, The MathWorks Inc., Natick, MA, USA) [8]. We reconfigured the toolbox with a custom-written, scanner-specific settings script in MATLAB. We created and processed R1, R2*, PD, and MTsat maps from unprocessed, multi-echo T1-, PD-, and MT-weighted images, using the default hMRI pipeline with ‘UNICORT’ setting for B1 correction and unified segmentation for radiofrequency sensitivity bias correction [8]. We used the hMRI toolbox pipeline for processing, comprising segmentation (into gray matter (GM) and WM); nonlinear spatial registration to MNI space; and tissue-weighted smoothing [18]. We set a voxel size of 1 × 1 × 1 × mm3 and Gaussian smoothing with a 6 × 6 × 6 mm FWHM kernel.
For voxel-based morphometry (VBM) analysis, we utilized the Computational Anatomy Toolbox 12 (CAT12.9) and R1 maps generated during VBQ analysis [19].
Statistical analyses
Using SPM25, we conducted VBQ and VBM second-level analyses involving a two-sample t-test. Age and sex were included as covariates for VBQ, and total intracranial volume was additionally included for VBM analyses. The cluster-forming threshold was set to p < 0.001 (uncorrected), with a minimum number of voxels of 20; additional cluster-level family-wise error (FWE) correction of p < 0.05 was applied. We tested two contrasts: clusters with higher values in PwHD (PwHD > HC) and clusters with lower values in PwHD (PwHD < HC). Using MarSBaR (v0.45), we extracted clusters with significant group-level differences and visualized them using Python (v3.13.5, Python Software Foundation) using libraries such as Nilearn (v.0.12.0) [20, 21]. We performed two sets of group comparisons. The two-group comparison comprised all PwHD participants (PwHDtotal; PwHDpre and PwHDman) versus age- and sex-matched HCtotal. The four-group comparison comprised PwHDpre versus HCpre; and PwHDman versus HCman.
We performed separate analyses for the whole brain and the BG; VBQ and VBM; each MPM map; and GM and WM, using explicit masks. We used tissue probability masks incorporated within the SPM25 environment, and the BG mask was created using the ATAG atlas [22]. Significant clusters were anatomically labeled using the Neuromorphometrics atlas for GM and the JHU White-Matter Tractography Atlas for WM [23]. Clusters within each modality and atlas group were corrected for multiple comparisons using family-wise error (FWE) corrections. Clusters surviving the FWE-corrected threshold of pFWE < 0.05 were considered significant and are reported in this work.
Correlation analyses were performed using Pearson’s or Spearman’s correlation tests, depending on the distribution of the data. Pearson’s correlation coefficient (r) was used when the analyzed variables were approximately normally distributed and their relationship was approximately linear. Spearman’s rank correlation coefficient (rho) was used when these assumptions were not met. For every correlation, we report the analyzed group, sample size, the coefficient, and the corrected p-value. Multiple-comparison correction was performed using the Benjamini–Hochberg false discovery rate (FDR) procedure across all tests of cross-modal overlap (n = 21) and associations with clinical measures (n = 115). Disease duration was defined as the time elapsed since clinical motor onset and was analyzed only in PwHDman (n = 10). PwHDpre were excluded from analyses involving disease duration because clinical motor onset had not yet occurred.
Results
Demographic and Clinical Assessment
We included 26 PwHDtotal (age: 47.1 ± 14.9 years; 15 females) and 42 HCtotal (age: 49.0 ± 18.0 years; 22 females). Mean age (t = 0.467, df = 60, p = 0.3173) and sex distribution (chi-squared = 0.056, df = 1, p = 0.8130) did not differ between groups. Within PwHDtotal, 16 participants were PwHDpre (age: 39.3 ± 12.3 years; 9 females) and 10 were PwHDman (age: 59.8 ± 8.5 years; 6 females), with no age or sex differences relative to their corresponding HC groups (PwHDpre: age, t = 0.227, df = 31, p = 0.9820; sex distribution, chi-square = 0.022, df = 1, p = 0.8812; PwHDman: age, t = 0.314, df = 17, p = 0.7573; sex distribution, chi-square = 0.000, df = 1, p = 1.000). The average CAG repeat length was 42 ± 2 in PwHDpre and 41 ± 1 in PwHDman. Mean disease duration and UHDRS-TMS scores in PwHDman were 7.7 ± 5.5 years and 42.40 ± 19.72, respectively. Mean TFC scores were 12.7 ± 0.6 in PwHDpre and 8.0 ± 2.5 in PwHDman. CAP scores were 321.6 ± 78.7 in PwHDpre and 461.1 ± 66.0 in PwHDman. Average MoCA scores were 26.56 ± 2.59 in PwHDpre and 19.70 ± 5.38 in PwHDman (Table 1). UHDRS-TMS, TFC, and CAP were not done in healthy controls. MoCA was not done in healthy controls and is therefore reported as unavailable.
Table 1Demographic and clinical characteristics
Two-group comparison
PwHDtotal HCtotal p-value
number 26 42
male/female 11/15 20/22 0.8130 1
age (years) 47.1 ± 14.9 49.0 ± 18.0 0.3173 2
mHTT
(no. of CAG repeats)
42 ± 2 n.a
UHDRS-TMS 17.50 ± 23.35 n.a
TFC 10.9 ± 2.8 n.a
CAP 375.2 ± 100.4 n.a
Disease duration (years) 7.7 ± 5.5 (n = 10) n.a
MoCA 23.92 ± 5.10 n.av
Four-group comparison
PwHDpre HCpre p-value PwHDman HCman p-value
number 16 16 10 10
male/female 7/9 7/9 0.88121 4/6 4/6 1.0000 1
age (years) 39.2 ± 12.3 39.3 ± 14.5 0.9820 2 59.8 ± 8.5 60.9 ± 7.1 0.7573 2
mHTT
(no. of CAG repeats)
42 ± 2 n.a 41 ± 1 n.a
UHDRS-TMS 1.94 ± 1.98 n.a 42.40 ± 19.72 n.a
TFC 12.7 ± 0.6 n.a 8.0 ± 2.5 n.a
CAP 321.6 ± 78.7 n.a 461.1 ± 66.0 n.a
Disease duration (years) n.a n.a 7.7 ± 5.5 n.a
MoCA 26.56 ± 2.59 n.av 19.70 ± 5.38 n.av
Values are mean ± standard deviation. Superscripts indicate: 1 Pearson chi-square test; 2 two-sample t-test. HCpre and HCman comprise exclusively healthy participants; “pre” and “man” identify the PwHD group to which the controls were matched. Disease duration is not defined for PwHDpre; the PwHDtotal disease-duration value therefore reflects PwHDman only (n = 10). MoCA was not done in healthy controls and is reported as unavailable. CAP: CAG-Age Product, HC: HC healthy controls, mHTT: number of CAG repeats in the expanded huntingtin gene, MoCA: Montreal Cognitive Assessment. n.a.: not applicable. n.av.: not available, PwHDman: manifesting participants with Huntington’s disease, PwHDpre: premanifest participants with Huntington’s disease, PwHDtotal: all premanifest and manifesting participants with Huntington’s disease, TFC: Total Functional Capacity, UHDRS-TMS: Unified Huntington’s Disease Rating Scale Total Motor Score
Widespread microstructural cortical and subcortical reductions in patients with Huntington’s disease
We observed constant reductions in t-values across all four MPM modalities in PwHDtotal (Fig. 1, Tables S1-4). Identified clusters included bilateral middle temporal gyri (left: R1: T = 5.43, pFWE = 0.000; R2*: T = 3.93, pFWE = 0.005; PD: T = 5.64, pFWE = 0.000. Right: R1: T = 5.33, pFWE = 0.001; MTsat: T = 4.83, pFWE = 0.000), superior temporal gyri (left: R2*: T = 4.99; MTsat: T = 6.00; right: R2*: T = 4.82; MTsat: T = 4.52; all pFWE = 0.000), left posterior orbital gyrus (R1: T = 3.87, pFWE = 0.000; R2*: T = 4.16, pFWE = 0.000: MTsat: T = 4.38, pFWE = 0.002), left fusiform gyrus (R1: T = 6.68, pFWE = 0.041; R2*: T = 6.62, pFWE = 0.010; MTsat: T = 6.48, pFWE = 0.023), left supramarginal gyrus (R2*: T = 5.19, pFWE = 0.000; MTsat: T = 4.76, pFWE = 0.021) and left postcentral gyrus (R1: T = 4.14, pFWE = 0.000, R2*: T= 5.78, pFWE = 0.000; MTsat: T = 4.28, pFWE = 0.001). Additionally, reductions in PD were observed in bilateral caudate nuclei (left: T = 4.77; right: T = 4.50; all pFWE = 0.000). VBM analysis revealed only clusters with lower t-values in PwHDtotal, including left caudate nucleus (T = 5.60, pFWE = 0.000) and right middle frontal gyrus (T = 4.99, pFWE = 0.024), which overlapped with VBQ-identified clusters (Table S5).
Within BG, t-values were lower in PwHDtotal across all MPM modalities in the bilateral caudate nuclei (left: R1: T = 3.63, pFWE = 0.041; R2*: T = 3.86, pFWE = 0.017; MTsat: T = 4.33, pFWE = 0.000. Right: R1: T = 4.07, pFWE = 0.000; R2*: T = 4.40, pFWE = 0.000; PD: T = 3.86, pFWE = 0.040; MTsat: T = 4.14, pFWE = 0.000, Figures S1-2 and Table S6). VBM-derived clusters with lower t-values were identified in the bilateral caudate nuclei (left: T = 5.27; right: T = 4.59; all pFWE = 0.000).
Altered R2*, R1, and MTsat in premanifest patients with Huntington’s disease
PwHDpre presented with differences in R1, R2*, and MTsat (Fig. 2; Table S7), including lower R1 values in the left superior frontal gyrus (T = 5.53, pFWE = 0.028) and left middle temporal gyrus (T = 3.95, pFWE = 0.031); lower R2* values in the right middle frontal gyrus (T = 5.26, pFWE < 0.001) and right superior parietal lobule (T = 4.84, pFWE < 0.001); and lower MTsat values in the right superior frontal gyrus (T = 5.82, pFWE = 0.001) and right posterior cingulate gyrus (T = 4.45, pFWE = 0.014). VBM analysis showed that PwHDpre had lower t-values in cerebellar vermal lobules (T = 28.76, pFWE < 0.001) and the right middle orbital gyrus (T = 23.26, pFWE = 0.007) than HCpre. No significant group differences were observed in basal-ganglia VBQ or VBM analyses.
Multimodal reductions in cortical and subcortical structures in manifesting patients with Huntington’s disease
All identified clusters presented with lower t-values in PwHDman (Fig. 3; Tables S8-11). The left superior temporal gyrus showed a consistent reduction across all MPM modalities (R1: T = 10.65; R2*: T = 5.09; all pFWE = 0.000; PD: T = 6.96; MTsat: T = 10.72; all pFWE = 0.016). Multiple clusters demonstrated reduced R1 and R2* values, including the bilateral postcentral gyrus (left: R1: T = 8.21, pFWE = 0.000; R2*: T = 6.67, pFWE = 0.002. Right: R1: T = 6.71, pFWE = 0.028; R2*: T = 6.60, pFWE = 0.000), the left middle cingulate gyrus (R1 = 8.20, pFWE = 0.003; R2*: T = 8.10, pFWE = 0.002), and the left inferior occipital gyrus (R1 = 4.77; R2*: T = 9.14, all pFWE = 0.000). Lower MTsat (all pFWE = 0.000) values were detected in cerebellar structures, including left cerebellum exterior (T = 4.65) and vermal lobules VII-VIII (T = 7.16). All VBM-identified clusters overlapped with VBQ-derived clusters, including within the frontal (right superior frontal gyrus: T = 7.33, pFWE = 0.001; right middle frontal gyrus: T = 6.45, pFWE = 0.023; and left middle frontal gyrus: T = 6.42, pFWE = 0.000) and parietal (right superior parietal lobule: T = 7.33, pFWE = 0.023; and right postcentral gyrus: T = 10.46, pFWE = 0.000) lobes (Table S12). Furthermore, the largest VBM-identified clusters in the right fusiform gyrus (T = 16.32, pFWE = 0.000) overlapped with lower R2* values (T = 5.51, pFWE = 0.033).
The BG analysis showed reduced R2* and MTsat values only in PwHDman (Figure S3 and Table S13). The right caudate nucleus showed lower R2* (T = 4.75, pFWE = 0.001) and MTsat values (T = 5.23, pFWE = 0.000), while the left caudate nucleus showed reduced MTsat values only (T = 6.06, pFWE = 0.000). All VBQ-derived clusters overlapped with VBM-derived reductions (all pFWE = 0.000) in the right (T = 5.37) and left putamen (T = 5.07) (Fig. 4).
Overlapping and correlated cross-modal group differences within the same clusters
In PwHDtotal (n = 26), reductions across MPM modalities overlapped within the left fusiform gyrus, left middle temporal gyrus, right thalamus, and left middle occipital gyrus (Table S14). Cross-modal correlations used subject-level cluster means; the correction status of all p-values in this paragraph is FDR-corrected. Within clusters showing group-level reductions, R1 values correlated positively with R2* values in the left fusiform gyrus (rho = 0.953, pFDR = 0.0002), right thalamus (rho = 0.955, pFDR = 0.0002), left caudate nucleus (r = 0.690, pFDR = 0.0002), right caudate nucleus (r = 0.694, pFDR = 0.0002), and left middle occipital gyrus (r = 0.551, pFDR = 0.0043). R1 values also correlated positively with MTsat values in the left fusiform gyrus (r = 0.836, pFDR = 0.0002), right thalamus (r = 0.707, pFDR = 0.0002), left caudate nucleus (rho = 0.910, pFDR = 0.0002), and right caudate nucleus (rho = 0.934, pFDR = 0.0002). R1 and PD values correlated positively in the left middle temporal gyrus (rho = 0.966, pFDR = 0.0002). R2* and MTsat values correlated positively in the left fusiform gyrus (rho = 0.892, pFDR = 0.0002), right thalamus (rho = 0.691, pFDR = 0.0002), left caudate nucleus (rho = 0.753, pFDR = 0.0002), and right caudate nucleus (rho = 0.837, pFDR = 0.0002).
In PwHDman (n = 10), R1 and R2* values correlated positively in the left superior temporal gyrus (rho = 0.660, pFDR = 0.0438), left postcentral gyrus (rho = 0.878, pFDR = 0.0021), left middle cingulate gyrus (rho = 0.987, pFDR = 0.0002), right postcentral gyrus (rho = 0.927, pFDR = 0.0004), and left inferior occipital gyrus (rho = 0.745, pFDR = 0.0203). In the left superior temporal gyrus, R1 and MTsat values also correlated positively (rho = 0.672, pFDR = 0.0410).
Associations between MPM-derived values and clinical variables
Clinical correlations were calculated at the subject level in PwHDtotal (n = 26); the correction status of all p-values in this paragraph is FDR-corrected (Table S15). R2* values within clusters characterized by group-level reductions correlated positively with MoCA scores in the right middle occipital gyrus (rho = 0.927, pFDR = 0.0005) and left precuneus (rho = 0.884, pFDR = 0.0144). Thus, higher R2* values were associated with higher MoCA scores. CAG repeat length correlated positively with R1 and R2* values in the left middle cingulate gyrus (R1: rho = 0.636, pFDR = 0.0492; R2*: rho = 0.636, pFDR = 0.0492), left supramarginal gyrus (rho = 0.532, pFDR = 0.0111), and right precentral gyrus (rho = 0.395, pFDR = 0.0103), and MTsat values in the left superior temporal gyrus (rho = 0.399, pFDR = 0.0473).
VBQ- and VBM-derived white matter microstructural and volumetric changes
We observed multimodal MPM reductions in PwHDtotal and PwHDpre. No significant white-matter VBM differences were detected in either group (Figures S4-7 and Tables S16-19). Within the PwHDman group, we detected MPM reductions (all pFWE = 0.000), within the splenium (R1: T = 21.58; R2*: T = 57.4; MTsat: T = 5.09) and genu of the corpus callosum (R1: T = 6.56; R2*: T = 6.16; MTsat: T = 5.83) (Figure S4 and Table S16). Consistent with GM findings, MPM-derived clusters overlapped with VBM-derived reductions in t-values in PwHDman in posterior corona radiata (left: T = 7.52, pFWE = 0.011; right: T = 5.65, pFWE = 0.015), right superior corona radiata (T = 6.16, pFWE = 0.016), sagittal stratum (T = 6.26, pFWE = 0.000), and left cingulum (T = 6.11, pFWE = 0.000). Some of the clusters with identified VBQ reductions cross-modally overlapped (Figure S7).
Discussion
Utilizing MPM, we detected microstructural alterations in PwHD. PwHDtotal showed widespread cortical and subcortical reductions across all MPM parameters, whereas PwHDman showed more spatially constrained changes with stronger group effects. Reductions in R1 and MTsat are consistent with lower myelin and macromolecular content, and reduced R2* may reflect lower iron content and/or demyelination. Proton density provides a semiquantitative measure of the MR-visible mobile proton pool and is influenced by both tissue water and macromolecular composition. The observed reductions in PD should therefore not be interpreted as direct evidence of reduced tissue water content. Although changes in tissue water may contribute to PD alterations, their direction cannot be inferred from PD measurements alone. The cross-modal overlap of MPM-derived findings was spatially consistent with VBM-derived cortical and subcortical gray- and white-matter volume loss.
Bilateral caudate nuclei consistently demonstrated reductions across all MPM maps in PwHDman, while reductions in R2* and MTsat in PwHDpre suggest sensitivity to pathology before motor manifestation. Although R2* was reduced at the group level in the identified regions, the positive within-group correlations indicate that participants with higher individual R2* values had higher MoCA scores. Subject-level correlations also linked regional MPM values with CAG repeat length. Whole-brain VBM in PwHDpre showed significant reduction in gray matter volume in the right middle orbital gyrus. Within the basal ganglia, we observed no MPM- or VBM-derived differences in PwHDpre, possibly because of limited sensitivity to subtle changes or limited statistical power (n = 16) [24]. Alternatively, the basal ganglia are functionally segregated into motor, associative, and limbic circuits, and analyses using functional atlases may provide greater sensitivity [25]. Although quantitative susceptibility mapping studies have demonstrated iron accumulation in the basal ganglia in PwHD, we observed reduced R2*, which may reflect lower iron content [26]. QSM and R2* differ in their sensitivity to iron and myelin, limiting direct comparison [27, 28]. Overlapping reductions across MPM modalities and VBM are consistent with combined demyelination, altered proton density, and loss of neuronal structure, which may also reduce the cellular infrastructure available for iron accumulation.
Spatially restricted clusters with higher R2* and MTsat in PwHDpre may reflect local iron accumulation and neuroinflammation, indicating early-stage alterations with regional susceptibility. MPM-derived alterations were observed in the superior and middle frontaly gyrus, and middle temporal gyrus. These findings were accompanied by VBM-derived volume loss in the right middle orbital gyrus [11, 12].
The two- and four-group comparisons revealed distinct spatial patterns in MPM-related alterations in PwHDtotal and PwHDman, with PwHDtotal showing the core disease-related microstructural alterations, whereas PwHDman exhibited spatially confined clusters with larger effect sizes, reflecting disease-stage vulnerability. We observed consistent reductions across all four MPM modalities in the bilateral caudate nucleus, thalamus, middle forntal gyrus, superior temporal gyrus, postcentral gyri, and occipital gyri, aligning with available evidence and adding insight into the biophysical changes and mechanisms driving the observed neurodegeneration [29, 30]. Furthermore, our findings are supported by longitudinal and postmortem studies demonstrating progressive neurodegeneration with disease progression [31–33].
Although MPM alterations in PwHDman were associated with clinical measures, no single modality could be attributed to a specific clinical measure. Given the multisystem and progressive nature of HD, the identified changes are best understood as contributors rather than sole causes of the clinical presentation [34–36].
The similar pattern of reductions in WM across MPM modalities demonstrates widespread microstructural alterations in both tissue types. Although WM degeneration has been shown to precede GM loss, we did not replicate this finding and detected no white-matter volumetric loss in PwHDpre [37]. Because diffusion tensor imaging (DTI) is the most widely used technique for assessing WM integrity and relies on fundamentally different physical principles than MPM, cross-comparing findings is challenging. However, observed reduction in R2* in PwHDpre and PwHDman may suggest that myelin changes may occur before neuronal loss in later stages of HD.
Our study has several limitations, including the cross-sectional design. We aimed to demonstrate the utility of the MPM in HD research, and future studies should include longitudinal trajectories to assess the disease-stage-dependent progression of MPM-derived findings. Additionally, stratifying disease stage into PwHDpre and PwHDman cohorts reduced statistical power and our ability to detect more subtle microstructural changes, particularly in PwHDpre. The narrow range of CAG repeat lengths within our cohort may limit our ability to detect significant correlations between repeat length and MPM-derived metrics. Therefore, future studies should aim to include larger, more heterogeneous cohorts of PwHD. The measure of disease severity used has its own limitations, including intra- and inter-rater variability and subjective bias, while MoCA tests assess the general state of cognition with no domain-specific insights [38–40]. Further studies may use a multi-domain composite UHDRS-TMS and a more representative panel of cognitive and neuropsychological assessments, as well as integrate genetic expression profiling and post-mortem investigation, to gain greater insight into the association between MPM-derived findings and specific domains of symptomatology and underlying cellular compositions [36, 41].
In conclusion, using MPM, we identified widespread cortical and subcortical microstructural alterations across HD stages, with reductions in tissue biophysical properties consistent with neurodegeneration. These observations were spatially overlapping with volumetric reductions and were associated with clinical measures in PwHD, supporting the utility of MPM for in vivo assessment of disease-related pathology in HD. Given the high volume of insights generated in a single scan, MPM has great potential to improve diagnostics and patient stratification, as well as personalized treatment approaches, though this requires confirmation in future longitudinal studies.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We would like to acknowledge Jenny Schmalfeld for her assistance in participant recruitment and study coordination.
Author contributions
MMP: Conceptualization, Methodology, Data curation, Formal analysis, Visualization, Writing the original draft. JLA: Investigation, Data curation, Writing review and editing. MGKM: Investigation, Data curation, Writing review and editing. JH: Investigation, Data curation, Writing review and editing. JU: Investigation, Data curation, Writing review and editing. LvW: Investigation, Data curation, Writing review and editing. AM: Resources, Clinical expertise, Writing review and editing. SL: Investigation, Data curation, Writing review and editing. NB: Supervision, Project administration, Funding acquisition, Writing review and editing. JP: Conceptualization, Methodology, Supervision, Writing review and editing.
Funding
Open Access funding enabled and organized by Projekt DEAL. This work was supported by the funding provided by the Deutsche Forschungsgemeinschaft (DFG; Research Unit FOR2488, N.B.) and by the JPND Consortium Control-PD (N.B.).
Data availability
The data generated in this study is not publicly available. De-identified data may be provided by the corresponding authors upon reasonable request, subject to institutional approval.
Declarations
Conflicts of interest
The authors have no financial disclosure or conflict of interest to report.
Ethical approval statement
The study adhered to the Declaration of Helsinki and was approved by the Committee for the Ethics of Safety- and Security-Relevant Research at the University of Luebeck (AZ-18–196). All participants provided informed written consent prior to study participation.
References
1.Walker FO (2007) Huntington’s disease. Lancet 369:218–228. 10.1016/S0140-6736(07)60111-1
2.Langbehn DR, Hayden MR, Paulsen JS, PREDICT‐HD Investigators of the Huntington Study Group (2010) CAG‐repeat length and the age of onset in Huntington disease (HD): a review and validation study of statistical approaches. Am J Med Genet B Neuropsychiatr Genet 153B:397–408. 10.1002/ajmg.b.30992
3.Tong H, Yang T, Xu S, Li X, Liu L, Zhou G et al (2024) Huntington’s disease: complex pathogenesis and therapeutic strategies. Int J Mol Sci 25:3845. 10.3390/ijms25073845
4.Achenbach J, Von Hein SM, Saft C (2020) Functional and cognitive capacity differ in dystonic motor subtypes when compared to choreatic and hypokinetic‐rigid motor subtypes in Huntington’s disease. Brain Behav 10:e01704. 10.1002/brb3.1704
5.Cao L, Yin J, Du G, Yang Q, Huang Y (2024) Identifying and verifying Huntington’s disease subtypes: clinical features, neuroimaging, and cytokine changes. Brain Behav 14:e3469. 10.1002/brb3.3469
6.Hanrahan J, Locke DP, Cahill LS (2024) Magnetic resonance imaging to detect structural brain changes in Huntington’s disease: a review of data from mouse models. J Huntington Dis 13:279–299. 10.3233/JHD-240045
7.Aqel S, Ahmad J, Saleh I, Fathima A, Al Thani AA, Mohamed WMY et al (2025) Advances in Huntington’s disease biomarkers: a 10-year bibliometric analysis and a comprehensive review. Biology (Basel) 14:129. 10.3390/biology14020129
8.Tabelow K, Balteau E, Ashburner J, Callaghan MF, Draganski B, Helms G et al (2019) hMRI – a toolbox for quantitative MRI in neuroscience and clinical research. Neuroimage 194:191–210. 10.1016/j.neuroimage.2019.01.029
9.Callaghan MF, Helms G, Lutti A, Mohammadi S, Weiskopf N (2015) A general linear relaxometry model of R1 using imaging data: general linear relaxometry model of R1. Magn Reson Med 73:1309–1314. 10.1002/mrm.25210
10.Lorio S, Tierney TM, McDowell A, Arthurs OJ, Lutti A, Weiskopf N et al (2019) Flexible proton density (PD) mapping using multi-contrast variable flip angle (VFA) data. Neuroimage 186:464–475. 10.1016/j.neuroimage.2018.11.023
11.Estevez-Fraga C, Altmann A, Parker CS, Scahill RI, Costa B, Chen Z et al (2023) Genetic topography and cortical cell loss in Huntington’s disease link development and neurodegeneration. Brain 146:4532–4546. 10.1093/brain/awad275
12.Wang X, Li Y, Li B, Shang H, Yang J (2024) Gray matter alterations in Huntington’s disease: a meta-analysis of VBM neuroimaging studies. J Neurosci Res 102:e25366. 10.1002/jnr.25366
13.Caron NS, Wright GE, Hayden MR. Huntington Disease. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, (1993) editors. GeneReviews®, Seattle (WA): University of Washington, Seattle
14.Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I et al (2005) The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53:695–699. 10.1111/j.1532-5415.2005.53221.x
15.Disorders M (1996) Unified Huntington’s disease rating scale: Reliability and consistency. Mov Disord 11:136–142. 10.1002/mds.870110204
16.Zhang Y, Long JD, Mills JA, Warner JH, Lu W, Paulsen JS (2011) Indexing disease progression at study entry with individuals at‐risk for Huntington disease. Am J Med Genet B Neuropsychiatr Genet 156:751–763. 10.1002/ajmg.b.31232
17.Weiskopf N, Suckling J, Williams G, Correia MM, Inkster B, Tait R et al (2013) Quantitative multi-parameter mapping of R1, PD*, MT, and R2* at 3T: a multi-center validation. Front Neurosci. 10.3389/fnins.2013.00095
18.Draganski B, Ashburner J, Hutton C, Kherif F, Frackowiak RSJ, Helms G et al (2011) Regional specificity of MRI contrast parameter changes in normal ageing revealed by voxel-based quantification (VBQ). Neuroimage 55:1423–1434. 10.1016/j.neuroimage.2011.01.052
19.Gaser C, Dahnke R, Thompson PM, Kurth F, Luders E (2024) The alzheimer’s disease neuroimaging initiative CAT: a computational anatomy toolbox for the analysis of structural MRI data. GigaScience. 10.1093/gigascience/giae049
20.Brett M, Jean-Luc A, Valabregue R, Poline J-B (2002) Region of interest analysis using an SPM toolbox. Neuroimage 16:769–1198. 10.1016/s1053-8119(02)90013-3
21.Abraham A, Pedregosa F, Eickenberg M, Gervais P, Mueller A, Kossaifi J et al (2014) Machine learning for neuroimaging with scikit-learn. Front Neuroinformatics. 10.3389/fninf.2014.00014
22.Keuken MC, Bazin P-L, Backhouse K, Beekhuizen S, Himmer L, Kandola A et al (2017) Effects of aging on T1, T2*, and QSM MRI values in the subcortex. Brain Struct Funct 222:2487–2505. 10.1007/s00429-016-1352-4
23.Mori S, Oishi K, Jiang H, Jiang L, Li X, Akhter K et al (2008) Stereotaxic white matter atlas based on diffusion tensor imaging in an ICBM template. Neuroimage 40:570–582. 10.1016/j.neuroimage.2007.12.035
24.Van Eimeren T, Giehl K, Reetz K, Sampaio C, Mestre TA (2023) Neuroimaging biomarkers in Huntington’s disease: preparing for a new era of therapeutic development. Parkinsonism Relat Disord 114:105488. 10.1016/j.parkreldis.2023.105488
25.Lanciego JL, Luquin N, Obeso JA (2012) Functional neuroanatomy of the basal ganglia. Cold Spring Harb Perspect Med 2:a009621–a009621. 10.1101/cshperspect.a009621
26.Yao J, Morrison MA, Jakary A, Avadiappan S, Rowley P, Glueck J et al (2024) Altered iron and microstructure in Huntington’s disease subcortical nuclei: insight from 7T MRI. J Magn Reson Imaging 60:1484–1499. 10.1002/jmri.29195
27.Hametner S, Endmayr V, Deistung A, Palmrich P, Prihoda M, Haimburger E et al (2018) The influence of brain iron and myelin on magnetic susceptibility and effective transverse relaxation - a biochemical and histological validation study. Neuroimage 179:117–133. 10.1016/j.neuroimage.2018.06.007
28.Lewis MM, Du G, Baccon J, Snyder AM, Murie B, Cooper F et al (2018) Susceptibility MRI captures nigral pathology in patients with parkinsonian syndromes. Mov Disord 33:1432–1439. 10.1002/mds.27381
29.Rosas HD, Salat DH, Lee SY, Zaleta AK, Pappu V, Fischl B et al (2008) Cerebral cortex and the clinical expression of Huntington’s disease: complexity and heterogeneity. Brain 131:1057–1068. 10.1093/brain/awn025
30.Jech R, Klempíř J, Vymazal J, Židovská J, Klempířová O, Růžička E et al (2007) Variation of selective gray and white matter atrophy in Huntington’s disease. Mov Disord 22:1783–1789. 10.1002/mds.21620
31.Coppen EM, Jacobs M, Van Den Berg-Huysmans AA, Van Der Grond J, Roos RAC (2018) Grey matter volume loss is associated with specific clinical motor signs in Huntington’s disease. Parkinsonism Relat Disord 46:56–61. 10.1016/j.parkreldis.2017.11.001
32.Labuschagne I, Cassidy AM, Scahill RI, Johnson EB, Rees E, O’Regan A et al (2016) Visuospatial processing deficits linked to posterior brain regions in premanifest and early stage Huntington’s disease. J Int Neuropsychol Soc 22:595–608. 10.1017/S1355617716000321
33.Lila E, Hunt D, Child DD, Latimer C, Le B, Davis M et al (2025) Asymmetric brain atrophy in Huntington’s disease: a postmortem MRI study. J Huntingt Dis 14:132–139. 10.1177/18796397251333334
34.Johnson EB, Ziegler G, Penny W, Rees G, Tabrizi SJ, Scahill RI et al (2021) Dynamics of cortical degeneration over a decade in Huntington’s disease. Biol Psychiatry 89:807–816. 10.1016/j.biopsych.2020.11.009
35.Frost C, Mulick A, Scahill RI, Owen G, Aylward E, Leavitt BR et al (2017) Design optimization for clinical trials in early‐stage manifest Huntington’s disease. Mov Disord 32:1610–1619. 10.1002/mds.27122
36.Schobel SA, Palermo G, Auinger P, Long JD, Ma S, Khwaja OS et al (2017) Motor, cognitive, and functional declines contribute to a single progressive factor in early HD. Neurology 89:2495–2502. 10.1212/WNL.0000000000004743
37.Aylward EH, Anderson NB, Bylsma FW, Wagster MV, Barta PE, Sherr M et al (1998) Frontal lobe volume in patients with Huntington’s disease. Neurology 50:252–258. 10.1212/WNL.50.1.252
38.Bates GP, Dorsey R, Gusella JF, Hayden MR, Kay C, Leavitt BR et al (2015) Huntington disease Nat Rev Dis Primer 1:15005. 10.1038/nrdp.2015.5
39.Arney K (2018) Improved metrics for Huntington’s disease trials. Nature 557:S46–S47. 10.1038/d41586-018-05179-w
40.Winder JY, Roos RAC, Burgunder J, Marinus J, Reilmann R (2018) Interrater reliability of the Unified Huntington’s Disease Rating Scale‐Total Motor Score certification. Mov Disord Clin Pract 5:290–295. 10.1002/mdc3.12618
41.Edwards LJ, McColgan P, Helbling S, Zarkali A, Vaculčiaková L, Pine KJ et al (2023) Quantitative MRI maps of human neocortex explored using cell type-specific gene expression analysis. Cereb Cortex 33:5704–5716. 10.1093/cercor/bhac453
Associated Data
Supplementary Materials
Data Availability Statement
The data generated in this study is not publicly available. De-identified data may be provided by the corresponding authors upon reasonable request, subject to institutional approval.
Come citare questo studio
Marta M Pokotylo, Joke-Lina Aßmann, Maximilian G Ködderitzsch Mertins, Julia Henkel, Jan Uter, Leon van Well, Alexander Münchau, Sebastian Loens, Martin Göttlich, Norbert Brüggemann, Jannik Prasuhn. (2026). In vivo assessment of brain microstructure in patients with Huntington's disease using quantitative MRI.. Journal of neurology. DOI: 10.1007/s00415-026-14244-8
Consulta la pubblicazione originale ↗