NeuroImage: Clinical 28 (2020) 102415 Contents lists available at ScienceDirect NeuroImage: Clinical journal homepage: www.elsevier.com/locate/ynicl Structural brain correlates of dementia in Huntington’s disease T a,b,c,d,e a,b,c a,b,c,e Saul Martinez-Horta , Frederic Sampedro , Andrea Horta-Barba , Jesús Perez-Pereza,b,c,d,e, Javier Pagonabarragaa,b,c,d,e, Beatriz Gomez-Ansonf, Jaime Kulisevskya,b,c,d,e, ⁎ a Movement Disorders Unit, Neurology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain b Biomedical Research Institute (IIB-Sant Pau), Barcelona, Spain c Centro de Investigación Biomédica en Red-Enfermedades Neurodegenerativas (CIBERNED), Spain d Autonomous University of Barcelona, Spain e European Huntington’s Disease Network (EHDN), Spain f Neuroradiology, Radiology Department, Hospital de la Santa Creu i Sant Pau, Autonomous University of Barcelona, Barcelona, Spain ARTICLE INFO ABSTRACT Keywords: Background: Huntington’s disease (HD) is a fatal genetic neurodegenerative disorder with no effective treatment Huntington’s disease currently available. Progressive basal ganglia and whole-brain atrophy and concurrent cognitive deterioration Dementia are prototypical aspects of HD. However, the specific patterns of brain atrophy underlying cognitive impairment Neuropsychology of different severity in HD are poorly understood. The aim of this study was to investigate the specific structural Mild cognitive impairment brain correlates of major cognitive deficits in HD and to explore its association with neuropsychological in- Movement disorders dicators. Participants: Thirty-five symptomatic early-to-mild HD patients and 15 healthy controls (HC) with available T1- MRI imaging were included in this study. Methods: In this cross-sectional study, HD patients were classified as patients with (HD-Dem) and without (HD- ND) major cognitive impairment in the range of dementia. This classification was based on previously validated PD-CRS cutoff scores for HD. Differences in brain atrophy across groups were studied by means of grey-matter volume voxel-based morphometry (GMV-VBM) and cortical thickness (Cth). Voxelwise and vertexwise general linear models were used to assess the group comparisons, controlling for the effects of age, sex, education, CAG repeat length and severity of motor symptoms. Clusters surviving p < 0.05 and family-wise error (FWE) cor- rection were considered statistically significant. In order to characterize the impact on cognitive performance of the observed brain differences across groups, GMV and Cth values in the set of significant regions were computed and correlated with specific neuropsychological tests. Results: All groups had similar sociodemographic profiles, and the HD groups did not significantly differ in terms of CAG repeat length. Compared to HC, both HD groups exhibited significant atrophy in multiple subcortical and parietal brain regions. However, compared to HC and HD-ND groups, HD-Dem patients showed a more pro- minent pattern of reduced GMV and cortical thinning. Importantly, this thinning was restricted to regions of the parietal-temporal and occipital cortices. Furthermore, these brain alterations were further associated with poorer cognitive performance in tasks assessing frontal-executive and attention domains as well as memory, language and constructional abilities. Conclusions: Major cognitive impairment in the range of dementia in HD is associated with brain and cognitive alterations exceeding the prototypical frontal-executive deficits commonly recognized in HD. The observed posterior-cortical damage identified by MRI and its association with memory, language, and visuoconstructive dysfunction suggest a strong involvement of extra-striatal atrophy in the onset of severe cognitive dysfunction in HD patients. Critically, major cognitive impairment in this sample was not associated with CAG repeat length, age or education. This finding could support a possible involvement of additional neuropathological mechanisms aggravating cognitive deterioration in HD. ⁎ Corresponding author at: Movement Disorders Unit, Neurology Department, Hospital de la Santa Creu i Sant Pau, Mas Casanovas, 90 - 08041 Barcelona, Spain. E-mail address: jkulisevsky@santpau.cat (J. Kulisevsky). https://doi.org/10.1016/j.nicl.2020.102415 Received 27 April 2020; Received in revised form 1 September 2020; Accepted 3 September 2020 Available online 09 September 2020 2213-1582/ © 2020 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 1. Introduction with HD. Although this approach may have limitations, it distinguishes patients with normal or mild global cognitive defects from those with Huntington’s disease (HD) is a monogenic, autosomal dominant, severe cognitive impairment in the range of dementia. neurodegenerative disease caused by the an abnormal CAG expansion In the present study, our main aim was to explore specific structural on the HTT gene (Walker, 2007). Its inevitably progressive course is brain differences underlying major cognitive impairment in the range of characterized by progressive motor abnormalities and neuropsychiatric dementia in HD and to explore their association with poorer cognitive symptoms, early cognitive deterioration, and the development of de- performance. mentia (Ross et al., 2014; Peavy et al., 2010). The key neuropathological hallmark of HD is progressive atrophy of 2. Methods the basal ganglia, which may be detected up to 15 years before motor symptoms appear and the diagnosis is made. Compelling evidence, 2.1. Participants however, shows that HD affects not only basal ganglia but the whole- brain(Rosas et al., 2008). Cross-sectional and longitudinal studies have We included thirty-five symptomatic gene-mutation carriers with illustrated that even in the premanifest stage of the disease, grey matter CAG repeat length > 38 regularly attending the outpatient HD-Clinic volume (GMV) reductions, cortical thinning (Cth), and reduced brain of the Movement Disorders Unit at Hospital de la Santa Creu i Sant Pau metabolism affect multiple extra-striatal regions. These regions com- in Barcelona, and fifteen age- and education-matched healthy controls prise extensive territories of the parietal, temporal and occipital lobes (HC). HC were non-consanguineous relatives of the HD participants. that also contribute to the clinical expression of the disease(Rosas et al., None of the HCs had a history of neurologic, psychiatric or un- 2008, 2005; Nopoulos et al., 2010; Tabrizi et al., 2009; Coppen et al., compensated systemic diseases. Similarly, all HD participants were free 2018; Rub et al., 2015; Kuwert et al., 1990). of any neurological disorder other than HD and had no history of brain From a neurocognitive perspective, HD has been described as a surgery, traumatic brain injury, epilepsy, drug abuse, or un- prototypical frontal-subcortical dementia due to striatal, but also tha- compensated systemic disease. lamic degeneration(Kassubek et al., 2005), with frontal-executive dis- Written informed consent was obtained from all participants and all turbances, attention deficits, and processing speed reduction as the procedures were performed in accordance with the standards of the most consistently affected cognitive domains(Snowden, 2017). In- Ethics Committee at Hospital de la Santa Creu i Sant Pau in Barcelona volvement of the caudate nucleus and putamen in cognitive aspects of and in accordance with the 1964 Declaration of Helsinki and its later HD is characteristic(Tabrizi et al., 2009; Dogan et al., 2013). However, amendments. many studies in HD have shown that as the neuropathological changes extend to multiple extra-striatal regions, cognitive domains other than 2.2. Assessments frontal-executive functions become affected. These include alterations in visuomotor integration(Gomez-Anson et al., 2009; Say et al., 2011), HD patients were classified as with severe or without severe cog- episodic and autobiographical memory(Carmichael et al., 2019; Harris nitive impairment using the HD-specific PD-CRS cutoff scores(Martinez- et al., 2019), visual perception(Martinez-Horta et al., 2019; Coppen Horta et al., 2020). These cutoff scores were determined in HD popu- et al., 2019), mental rotation(Wolf et al., 2014; Labuschagne et al., lation in a previous study using as gold standard the presence of major 2016) and language production and organization(Hinzen et al., 2018; cognitive and functional impairment. Thus, in accordance with this Chan et al., 2019; Chenery et al., 2002; Podoll et al., 1988), suggesting previous study we used the PD-CRS total score ≤ 64 to classify patients that the cognitive phenotype of HD cannot be attributed solely to as with major cognitive impairment in the range of dementia(Martinez- frontal-striatal dysfunction. Horta et al., 2020). This instrument assesses nine subtests: immediate Longitudinal studies have demonstrated that CAG repeat length and delayed verbal memory, confrontation naming, attention, working plays a key role in the linear decline of cognitive capacity throughout memory, unprompted draw of a clock, copy of a clock, alternating disease progression(Tabrizi et al., 2012; Paulsen et al., 2014; verbal fluency, and action verbal fluency. The total score is computed Consortium, 2015; Langbehn et al., 2019). However, age-at-onset by summing all the raw scores obtained in each subtest, but specific (AAO) and the rate of progression of cognitive deterioration varies scores can be grouped to compute a “frontal-subcortical” score and a greatly among individuals even when age, education level and CAG “posterior-cortical” score(Pagonabarraga et al., 2008). In the present repeat length are equivalent(Consortium, 2015; Braisch et al., 2019). study, the PD-CRS total score was used to group patients according to This heterogeneity suggests that, besides CAG repeat length, other their cognitive status. Accordingly, patients who obtained a PD-CRS mechanisms such as environmental and genetic variables also con- total score > 64 were included in the “non-demented” group (HD-ND) tribute to the neuropathological and clinical progression of cognitive whereas patients who obtained a PD-CRS total score ≤ 64 were con- deterioration in HD. It has recently been shown, for example, that ge- sidered as with major cognitive impairment in the range of dementia netic variability in chromosome 15, contributes to age at onset of motor (HD-Dem). The non-demented group included a slight proportion of and cognitive symptoms (Consortium, 2015). Other mechanisms, such participants with mild cognitive deficits whereas in the demented group as MAPT haplotypes and their role in the expression of TAU, have also all participants exhibited severe cognitive deficits. Accordingly, this been associated with cognitive progression in HD (Vuono et al., 2015). approach allowed us to compare participants with severe cognitive Studies addressing the mechanisms involved in the differential ex- impairment vs those with minor or absent cognitive impairment ac- pression of cognitive deterioration in HD are scarce, probably due to the cording to the PD-CRS. Additionally, we administered the Mini-Mental lack of specific diagnostic criteria to differentiate HD patients with mild State Examination (MMSE) screening test for comparative analyses with cognitive impairment from those with severe or major cognitive im- the classification done with the PD-CRS. pairment in the range of dementia, and even from those with normal We also administered the composite cognitive score of the Unified cognition (Mestre et al., 2018). The specific patterns of brain alterations Huntington’s Disease Rating Scale (UHDRS cogscore) and other mea- underlying major cognitive impairment in HD thus remain elusive, sures commonly used in cognitive assessment protocols in HD partly because of the difficulty in appropriately grouping patients ac- (Landwehrmeyer et al., 2017). Accordingly, assessments included the cording to cognitive status. To overcome this limitation, we recently Symbol Digit Modalities Test (SDMT), the phonetic fluency test with validated the Parkinson’s Disease – Cognitive Rating Scale (PD-CRS) as letters F, A and S, the Stroop color-naming, word-reading and inter- a screening tool for global cognition in HD(Martinez-Horta et al., 2020). ference tests, the semantic fluency test (animals) and parts A and B of We showed that the HD-specific PD-CRS cutoff scores have a good ca- the Trail Making Test (TMT). Raw scores for all these measures were pacity to discriminate between categories of cognitive status in patients adjusted for age and education level and then converted to z-scores. A 2 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 Table 1 Clinical and Sociodemographic characteristics of the sample. Controls HD HD-ND HD-Dem P a Age 45.9 ± 8 51.8 ± 12 50.85 ± 11 53.2 ± 14 0.096; b0.156; c0.102; d0.590 a Gender (f/m) 4/11 23/12 12/8 11/4 0.012; b0.052; c0.013; d0.324 a Education 12.3 ± 2 11.5 ± 4 11.2 ± 5 12 ± 4 0.376; b0.346; c0.778; d0.615 d CAG – 43.7 ± 3 43.3 ± 3 44 ± 3 0.424 DBS1 – 492 ± 84 469 ± 80 523 ± 83 d 0.061 UHDRS-TMS2 – 28.5 ± 17 22.5 ± 18 37.2 ± 11 d < 0.01 TFC3 – 11.2 ± 2 12.1 ± 1 10 ± 2 d < 0.01 a PD-CRS total score 106.7 ± 10.7 70.17 ± 21 84.3 ± 15 51.2 ± 10 < 0.001; b < 0.001; c < 0.001; d < 0.001 d UHDRS cogscore – 170.2 ± 78 212 ± 72 114 ± 47 < 0.001 a MMSE 29.3 ± 0.8 25.2 ± 3.8 27.8 ± 1.8 21.6 ± 2.6 < 0.001; b < 0.005; c < 0.001; d < 0.001 1 Disease burden score; 2Unified Huntington’s disease rating scale – Total motor score; 3Total functional capacity. a HC vs HD b HC vs HD-ND c HC vs HD-Dem d HD-ND vs HD-Dem global composite z-score was also calculated combining all these mea- 2.4. Statistical analysis sures. Screening for global cognitive status was also performed in all HC Socio-demographic and clinical variables were subjected to ANOVA using the PD-CRS and the MMSE. In this group, all participants ob- between the three groups. Post-hoc independent t-test comparisons tained a PD-CRS > 81 (mean = 106.7 ± 10.7) and a MMSE > 27 were performed between the three groups for continuous variables and (mean = 29.3 ± 0.8), indicating absence of cognitive impairment. χ2 for categorical variables. To calculate the effect size of the differ- For the HD participants, a neurologist specialized in HD (JPP) rated ences observed between cognitive groups we used Cohen’s d coefficient the severity of motor symptoms using the Unified Huntington’s Disease (d values: 0–0.2, small effect size; 0.6, moderate effect size; ≥ 0.8, large Rating Scale – Total motor score (UHDRS-TMS)(Hs, 1996). All HD effect size). participants obtained a diagnostic confidence level = 4, indicating that Voxelwise and vertexwise measures derived from VBM and Cth motor abnormalities were unequivocal signs of HD with a confidence analyses were introduced into a generalized lineal model (GLM) to level of 99%. All HD patients were classified as having early stage or compare these measures across groups, using age, sex, education, CAG mild disease according to a total functional capacity score repeat length and UHDRS-TMS as covariates. The following pairwise (TFC) > 6(Shoulson and Fahn, 1979). The disease burden score (DBS), group comparisons were performed: HC > HD-ND, HC > HD-Dem, an index assumed to reflect lifelong exposure to mutant huntingtin, was and HD-ND > HD-Dem. The set of clusters surviving p < 0.05 and calculated using the following formula based on age and CAG repeat family-wise error (FWE) correction for multiple-comparison (cluster- length: DBS = age × (CAG-35.5)(Penney et al., 1997). We also re- level Bonferroni correction for VBM and Monte-Carlo simulation with corded socio-demographic and clinical data, including age, sex, edu- 10,000 repeats for Cth) were considered statistically significant. cation and global cognitive functioning. To investigate the clinical translation of the imaging findings, we computed quantitative volumetric and mean Cth information at the identified clusters where we observed significant differences across 2.3. Neuroimaging acquisition and preprocessing groups. Using linear regression analysis, we then studied the association of these imaging measures with the different cognitive variables within Volumetric MRI was available for all participants. 3D-T1 images an exploratory analysis, controlling again for the effect of potential were acquired on a 3 T Philips Achieva using an MP-RAGE sequence confounders such as age, education, CAG repeat length and UHDRS- (TR/TE = 6.74/3.14 ms, flip-angle = 8°, field of view = 23 cm, ma- TMS, for which a p-value < 0.05 was considered significant. trix = 256x256 and slice thickness = 1 mm). We applied voxel-based morphometry (VBM) and cortical thickness 3. Results (Cth) neuroimaging preprocessing procedures. A standard VBM pipe- line using the Statistical Parametrical Mapping software package 3.1. Clinical and sociodemographic data (SPM12, http://www.fil.ion.ucl.ac.uk/spm) was performed(Martinez- Horta et al., 2019). Briefly, GMV tissue probability maps were com- The sample consisted of 35 HD patients (mean age = 51.8 ± 12; puted from T1-MRI scans. These maps were then normalized to the mean CAG = 43.7 ± 2.8; mean years of education = 11.5 ± 4.5; Montreal Neurological Institute (MNI) space by applying the DARTEL mean TFC = 11.2 ± 2) and 15 HC (mean age = 45.8 ± 8; mean algorithm. The resulting normalized GMV maps were then smoothed years of education = 12.3 ± 1.6). According to the PD-CRS score, using an isotropic spatial filter of 8x8x8mm full-width at half-maximum n = 20 were included in the HD-ND group (mean PD-CRS total (FWHM) to reduce inter-individual variability. score = 84.3 ± 15; mean age = 50.8 ± 11; mean CAG = 43.3 ± 3) Cth analysis was performed using the FreeSurfer 6.0 software and n = 15 in the HD-Dem group (mean PD-CRS total score = 51.2 ± package (https://surfer.nmr.mgh.harvard.edu). The specific methods 10; mean age = 53.2 ± 14; mean CAG = 44 ± 3). As summarized in used for cortical reconstruction of T1-MRI brain images have been fully table 1, no significant differences were found between HC, HD-ND and described elsewhere(Fischl and Dale, 2000). In short, optimized surface HD-Dem groups regarding age and education. In the HD groups, no deformation models following intensity gradients accurately identify significant differences were found between HD-ND and HD-Dem re- white matter and gray matter boundaries in the cerebral cortex, from garding age, gender, education, CAG repeat length or DBS, but the HD- which cortical thickness is computed at each vertex of the resulting Dem group showed a significantly higher UHDRS-TMS, lower TFC, surface. The resulting cortical surfaces are normalized to average space lower UHDRS cogscore, lower PD-CRS total score, and lower MMSE and smoothed using a Gaussian kernel of 15 mm FWHM. score. In the HD-ND group, the mean MMSE score was 27.85 ± 1.8 (no impairment) while in the HD-Dem group it was 21.6 ± 2.6 (below the 3 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 Table 2 Performance in neuropsychological measures. HD-ND HD-Dem P d PD-CRS Total 84.3 ± 16 51.2 ± 10 < 0.001 1.543 PD-CRS frontal-subcortical 57.2 ± 14 28.9 ± 7 < 0.001 1.560 PD-CRS posterior-cortical 27.1 ± 3 22.2 ± 6 < 0.005 0.993 Immediate verbal memory 8.1 ± 1 5.5 ± 2 < 0.001 1.168 Naming 17.6 ± 2 14.5 ± 4 < 0.005 0.946 Sustained attention 6.7 ± 2 2.2 ± 2 < 0.001 1.418 Working memory 5.1 ± 1 3 ± 2 < 0.005 1.084 Clock drawing 8.3 ± 2 6.2 ± 3 < 0.01 0.866 Clock copying 9.4 ± 1 7.6 ± 2 < 0.005 0.957 Delayed verbal memory 6.2 ± 2 3.3 ± 2 < 0.001 1.203 Alternating verbal fluency 9.6 ± 5 3.7 ± 2 < 0.001 1.202 Action verbal fluency 12.9 ± 6 5.5 ± 1 < 0.001 1.251 SDMT 31.8 ± 15 15.9 ± 6 < 0.001 1.391 Z-score −1.1 ± 1.3 −2.2 ± 1 < 0.05 1.039 2 % cases Z-score < -1.5 SD 35% 66.7% χ = 0.65 Stroop color-naming 51.2 ± 16 28.2 ± 12 < 0.001 1.626 Z-score −1.3 ± 1.1 −2.6 ± 0.5 < 0.001 1.272 % cases Z-score < -1.5 SD 55% 93.3% χ2 < 0.05 Stroop word-reading 76.6 ± 25 41.3 ± 20 < 0.001 1.559 Z-score −1.3 ± 1.3 −2.4 ± 0.8 < 0.01 1.143 % cases Z-score < -1.5 SD 50% 80% χ2 = 0.07 Stroop interference 28 ± 11 14.5 ± 8 < 0.001 1.493 Z-score −1.3 ± 1 −2.2 ± 0.7 < 0.005 0.948 2 % cases Z-score < -1.5 SD 50% 80% χ = 0.07 FAS 24.2 ± 13 14.5 ± 11 < 0.05 0.805 Z-score −1.6 ± 0.9 −2.1 ± 1 0.116 0.561 % cases Z-score < -1.5 SD 60% 80% χ2 = 0.18 Semantic fluency 14.8 ± 4 9 ± 3 < 0.001 1.640 Z-score −1.3 ± 0.8 −2.3 ± 0.5 < 0.005 0.994 2 % cases Z-score < -1.5 SD 30% 80% χ < 0.005 TMT-A 57.2 ± 32 118.8 ± 62 < 0.005 1.248 Z-score −0.7 ± 1.4 −2.1 ± 1 < 0.005 1.416 % cases Z-score < -1.5 SD 0% 100% χ2 < 0.001 TMT-B 149.6 ± 80 221.8 ± 34 < 0.01 1.174 Z-score −1 ± 1.5 −2.3 ± 0.8 < 0.01 1.250 2 % cases Z-score < -1.5 SD 40% 80% χ = 0.05 Composite Z-score −1.2 ± 0.9 −2.3 ± 0.7 < 0.005 1.078 % cases Z-score < -1.5 SD 35% 73.3% χ2 = 0.05 general < 24 cutoff for dementia). Table 3. Group comparisons between HD-ND and HD-Dem groups showed that HD-Dem scores were significantly lower on all the cognitive tests. 3.2. Neuropsychological performance in HD groups We found a large effect size in all groups according to d Cohen’s coef- ficient > 0.8. The HD-ND group obtained a mean PD-CRS total score of 84.3 ± 15 which, according to PD-CRS criteria, situates this group in the range between cognitive normality and mild cognitive deficits. 3.3. GMV differences between groups Looking at the prevalence of cases scoring above the proposed cutoff score for mild cognitive impairment in the HD-ND group (PD- The voxelwise group comparison revealed significant GMV differ- CRS ≤ 81), we saw that 45% of the patients scored for mild cognitive ences when comparing the HD-ND and HD-Dem groups with the HC impairment while 55% scored for cognitive normality. None of the group, and also when comparing the HD-ND group with the HD-Dem defects, however, were severe enough to fulfill the criteria for dementia group. Specifically, compared to the HC group, HD-ND showed sig- per specific PD-CRS criteria in HD. nificantly lower GMV in large cortico-subcortical clusters. The set of As reported in table 2, after adjustment for age and education, all significant clusters included the bilateral caudate nucleus and putamen, the measures obtained through the UHDRS cogscore and through the the bilateral insula, the bilateral inferior orbital prefrontal cortex other Enroll-HD assessments were significantly different between (oPFC), and also the right rolandic operculum, the right mid-occipital groups in all the cases with the exception of the phonetic verbal flu- gyrus, the right superior parietal gyrus, and the right mid-frontal and ency. In the HD-ND group, all the obtained z-scores were in the lower mid- and superior oPFC. Comparing HC and HD-Dem, we found marked range of normality, but above the critical cutoff of −1.5 SD with the GMV differences bilaterally in the whole basal ganglia and in the in- exception of the phonetic verbal fluency (-1.6 ± 0.9). In this group the sular cortex, in the left superior and medial frontal cortex, and in sev- mean composite z-score was −1.2 ± 0.9 with 35% of cases scoring eral posterior-cortical clusters, including the right superior and inferior below −1.5 SD. Conversely, in the HD-Dem group all measures scored occipital gyrus and the right lingual gyrus. below −2 SD and the mean composite z-score was −2.3 ± 0.7 with When we compared the HD-ND and HD-Dem group, the latter 73.3% of cases scoring below −1.5 SD. In congruence with the range of showed significantly reduced GMV in parietal-temporal regions, speci- PD-CRS scores obtained by the HD-ND group, performance below −1.5 fically in the bilateral anterior and posterior insular cortex, the superior SD was found in more than half of the non-demented cases for the temporal gyrus, and the left supramarginal gyrus (Fig. 1). Stroop color-naming, Stroop word-reading, Stroop interference, and for the phonetic verbal fluency (FAS). In the HD-Dem group, performance below −1.5 SD was found in > 80% of cases in most of the tasks 4 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 Table 3 predominantly affecting fronto-temporal and parietal regions of the left Cluster description table of the VBM-GMV analyses. hemisphere and temporo-occipital regions of the right hemisphere. In MNI coordinates (x, y, Cluster size T value the left hemisphere, significant differences involved the inferior tem- z) poral and the fusiform gyrus, the supramarginal and inferior parietal HD-ND > HD-Dem gyrus, the precentral and postcentral gyrus, the caudal and rostral mid- frontal gyrus, the insula, the pars triangularis, and the rostral middle Right insula 48–8 5 2298 5.22 Right posterior insula 36–27 17 3.79 frontal gyrus. Conversely, in the right hemisphere, differences were Right superior temporal 45–24 3 3.54 eminently circumscribed to the temporal pole, the superior and inferior gyrus temporal gyrus, the fusiform gyrus and the lateral occipital gyrus Left superior temporal gyrus −56–6 −2 2568 4.31 (Fig. 2). Left supramarginal gyrus −45–26 24 4.10 Left posterior insula −36–23 23 4.82 HC > HD-ND Left insula −31 9 10 6473 7 3.5. Cognitive-imaging regression analysis: Left inferior oPFC −41 38–9 6.33 Left caudate/putamen −12 3 20 6.13 Linear regression analysis showed that after adjusting for age, Right insula 31 14 8 6871 6.17 Right caudate/putamen 9 14–1 6.15 education, CAG repeat length and UHDRS-TMS, GMV at multiple Right rolandic operculum 36–15 20 5.75 clusters obtained in the VBM analysis contributed strongly to global Right inferior oPFC 30 30–8 5.69 cognitive performance. PD-CRS total score performance was strongly Right mid-occipital 36–68 26 1115 5.15 associated with GMV in the basal ganglia (β = 0.456; P = 0.017), the Right parietal superior 26–69 51 4.23 frontal lobes (β = 0.586; P = 0.005), the temporal lobes (β = 0.587; Right mid-frontal 33 59 15 1330 4.84 Right superior oPFC 29 51–3 4.54 P = 0.004), the insular cortex (β = 0.621; P = 0.002), the mid cin- Right mid- oPFC 53 42–6 4.24 gulate (β = 0.431; P = 0.036), and the occipital lobe (β = 0.441; HC > HD-Dem P = 0.043). PD-CRS performance was also associated with GMV in the Bilateral caudate/putamen/ −20 14 6 37,308 8.18 parietal lobe (β = 0.373; P = 0.043). However, this association was insula also influenced by the UHDRS-TMS (β = -0.390; P = 0.038) (Fig. 3). Right superior occipital 21–90 17 3970 6.21 Right lingual gyrus 17–86 −11 6 Following the same linear regression analysis approach, we looked Right inferior occipital gyrus 30–87 −6 5.81 at specific clusters of GMV and their association with PD-CRS total Right precentral gyrus 45 0 35 857 5.18 score and performance in each subtest as well as with performance in Left frontal superior medial −6 24 42 2957 4.89 all the other cognitive measures administered. Given the high number SMA −9 21 51 4.02 Left superior frontal −17 44 38 4.77 of correlation analyses performed between each cluster and each PD- CRS subtest, these results are reported in supplementary data. The PD- FWE corrected (p < 0.05) CRS total score was associated with GMV in the left (β = 0.491; P = 0.013) and right putamen (β = 0.556; P = 0.006), with the left 3.4. Cth differences between groups (β = 0.523; P = 0.011) and right inferior oPFC (β = 0.531; P = 0.009), with the left (β = 0.602; P = 0.003) and right insular The vertexwise comparison between HC and HD-ND groups showed cortex (β = 0.632; P = 0.002), with the left (β = 0.666; P = 0.001) a pattern of cortical thinning bilaterally in multiple fronto-temporal and and right superior temporal gyrus (β = 0.495; P = 0.011), with the left parieto-occipital regions in the HD-ND group. These regions were the (β = 0.563; P = 0.009) and right inferior temporal gyrus (β = 0.548; rostral mid-frontal and superior frontal gyrus, the orbital PFC, the su- P = 0.010), with the right cuneus (β = 0.508; P = 0.019), with the left perior temporal gyrus, the supramarginal, inferior and superior parietal mid oPFC (β = 0.515; P = 0.020), with the right mid-temporal pole gyrus, the lateral occipital gyrus, the precuneus, the precentral gyrus, (β = 0.473; P = 0.014), with the left (β = 0.648; P = 0.002) and right and the pars triangularis. Compared to the HC group, the HD-Dem mid-frontal gyrus (β = 0.579; P = 0.007), with the left (β = 0.591; group showed a similar but more significant thinning pattern than that P = 0.003) and right-mid temporal gyrus (β = 0.567; P = 0.006), with in the HD-ND group. Differences were also found in the post-central the right inferior parietal lobe (β = 0.482; P = 0.007), with the right gyrus, the inferior and the mid-temporal gyrus, and the fusiform gyrus. rectus gyrus (β = 0.573; P = 0.006), with the right superior oPFC Comparing the HD-ND and HD-Dem groups we found that the HD- (β = 0.501; P = 0.019), and with the left (β = 0.520; P = 0.013 and Dem group was characterized by a pattern of cortical thinning right medial superior PFC (β = 0.545; P = 0.011). Focusing in the other cognitive measures, we observed that SDMT Fig. 1. Regions showing lower GMV (A) and lower Cth (B) in the HD-ND and in the HD-Dem groups compared to the HC group. No regions showed a significant increase in grey matter volume. For depicting purposes, the image is shown at p < 0.001. 5 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 Fig. 2. Regions showing lower GMV (A) and lower Cth (B) in the HD-Dem group than in the HD-ND group. No regions showed a significant increase in grey matter volume. For depicting purposes, the image is shown at p < 0.005. performance correlated with GMV in the left mid-temporal gyrus verbal memory was associated with the left inferior temporal cortex (β = 0.515; P < 0.005), the left superior temporal gyrus (β = 0.409; (β = 0.435; P < 0.05). Confrontation naming (β = 0.465; P < 0.05), P < 0.05) and the inferior frontal gyrus (β = 0.450; P < 0.05). sustained attention, and (β = 0.396; P < 0.05), and action verbal Semantic fluency correlated strongly with GMV in the right insular fluency (β = 0.446; P < 0.05) where associated with the right fusi- cortex (β = 0.493; P < 0.01), the left superior temporal gyrus form gyrus. Working memory was also associated with the left supra- (β = 0.500; P < 0.01), the right inferior temporal gyrus (β = 0.489; marginal gyrus (β = 0.535; P < 0.005), and the left fusiform gyrus P < 0.01), the bilateral mid-temporal pole (β = 0.522; P < 0.005), (β = 0.541; P < 0.005). Focusing on the cogscore and related mea- and the right mid-frontal gyrus (β = 0.463; P < 0.05). sures, total cogscore was not associated with Cth. The Stroop color- In the Cth analyses, multiple regions overlapping the VBM-GMV naming was associated with the superior temporal gyrus (β = 0.395; findings were strongly associated with cognitive performance by means P < 0.05), and with the right lateral occipital gyrus (β = 0.449; of the PD-CRS total score. The most significant clusters were found at P < 0.05), and the semantic verbal fluency was associated with the the level of posterior-cortical regions, specifically at the fusiform gyrus right fusiform gyrus (β = 0.407; P < 0.05). in the temporo-occipital region (β = 0.491; P = 0.008). The PD-CRS posterior-cortical score was associated with cortical thickness in the 4. Discussion: right lateral occipital (β = 0.392; P < 0.05) and in the right fusiform gyrus (β = 0.512; P < 0.01). Focusing on PD-CRS subtests, immediate In the present study, major cognitive impairment in the range of Fig. 3. Linear regression analysis depicting correlations between GMV/Cth clusters and PD-CRS total score. 6 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 dementia -as determined using the specific PD-CRS cutoff scores for HD- Previous studies have highlighted the participation of extra-striatal is associated with brain and cognitive differences that exceed the pro- GMV atrophy and cortical thinning in the clinical expression of HD totypical pattern of frontal-executive, attention and processing speed (Rosas et al., 2008, 2005; Nopoulos et al., 2010; Tabrizi et al., 2009). deficits attributed to frontal-striatal atrophy, and critically involve When addressing cognitive aspects, these previous studies focused on more severe atrophy of posterior-cortical brain regions. specific measures of verbal fluency and psychomotor processing speed Our results highlight that severe cognitive impairment in the range but not on multiple measures of global cognitive performance. Al- of dementia may occur in the early stages of HD and with relative in- though they performed correlational analysis between brain structure dependence of the CAG repeat length, DBS, age and education. Notably, and cognitive performance, they did not compared patients according this more severe form of cognitive impairment is associated with a to the severity of cognitive deterioration. Nevertheless, cortical thick- widespread pattern of cortical thinning and whole-brain atrophy that ness of the pre-central gyrus, the superior temporal gyrus, the superior involves multiple cortical and subcortical clusters. The imaging data, frontal gyrus, the lingual gyrus, the precuneus and the cuneus were highlights that, as previously reported, GMV atrophy is mostly ascribed found to be associated with verbal fluency performance. Cortical to the basal ganglia, frontal and occipital lobe, and Cth involves the thickness of the pre-central gyrus, the bilateral paracentral lobule and temporal and parieto-occipital regions(Rosas et al., 2008, 2005; Tabrizi the occipital cortex were also associated with psychomotor processing et al., 2009). Importantly, group comparisons showed that a more speed(Rosas et al., 2008; Nopoulos et al., 2010). Regarding the impact prominent decrease of GMV and Cth is present in those participants of cortical changes over functional capacity, the most significant asso- exhibiting severe cognitive deterioration even when controlling for the ciations were found between TFC and the motor cortex, the superior effect of age, CAG, education of UHDRS-TMS. Accordingly, more ad- parietal, and the cuneus(Rosas et al., 2008). Similarly, other studies vanced disease stage or higher DBS cannot explain these differences. highlighted the involvement of multiple areas related to executive Our findings are in accordance with previous works showing that functions and sensorimotor and visuospatial processing in cognitive the development of cognitive deterioration in HD cannot be solely at- performance in HD(Garcia-Gorro et al., 2019). Interestingly, more re- tributed to basal ganglia atrophy(Rosas et al., 2008; Nopoulos et al., cent data also point to hippocampal-dependent memory deficits in HD 2010; Tabrizi et al., 2009; Coppen et al., 2018; Podoll et al., 1988; Say that cannot be solely explained as a function of degeneration in the et al., 2011; Carmichael et al., 2019; Harris et al., 2019; Martinez-Horta basal ganglia(Carmichael et al., 2019; Harris et al., 2019). et al., 2019; Wolf et al., 2014; Labuschagne et al., 2016; Hinzen et al., Keeping in mind that the two HD groups were matched for age, 2018; Chan et al., 2019). Structures of the basal ganglia, such as the gender, education, CAG and disease burden and that healthy controls caudate nucleus and putamen, but also the insular cortex, the PFC and were also matched in terms of age, and education, our findings suggest the occipital and parietal cortex, strongly differentiated non-demented that mechanisms other than only those promoted as a function of CAG HD patients from healthy controls. However, the most representative repeat length, DBS, age or education level contribute to a differential brain changes differentiating patients with major cognitive impairment course and expression of cognitive deterioration in HD. Although mu- from non-demented HD patients were found at the level of parieto- tant huntingtin aggregation is the primary mechanism leading to HD temporal regions, including, bilaterally, the anterior and posterior in- (Langbehn et al., 2019; Penney et al., 1997; Wild et al., 2015), multiple sular cortex, the superior temporal gyrus, and the left supramarginal mechanisms are also known to participate in the neuropathology of HD gyrus. In terms of cortical thinning, our results pointed in the same (Rub et al., 2016). However, how these additional mechanisms con- direction, supporting the critical participation of decreased cortical tribute to variability in the phenotypic expression of HD is not fully thinning in fronto-temporal and parietal regions of the left hemisphere elucidated. Of these mechanisms, inflammatory, autoimmune activity and in temporo-occipital regions of the right hemisphere in the more and TAU pathology have been suggested to play a cardinal role in severe forms of cognitive deterioration in HD. neurodegeneration and clinical expression of HD(Rocha et al., 2016). The involvement of cortical atrophy in HD, especially in parietal The question regarding the possible participation of these mechanisms and occipital regions, is a well-known finding supported by several in the acceleration of cognitive deterioration and related neurodegen- imaging studies(Rosas et al., 2008, 2005; Kuwert et al., 1990; Tabrizi eration in HD is only partially understood. TAU pathology and the re- et al., 2009; Coppen et al., 2018). However, the functional translation lated MAPT H2 haplotype have been associated with the rate of cog- of these cortical changes is partially understood. In this sense, our data nitive decline in HD(Vuono et al., 2015; Fernandez-Nogales et al., supports for the first time that more aggressive cortical atrophy is, at 2014). Whether these mechanisms contribute to the extent of brain least, critically associated with the presentation of a significantly more differences observed in our sample is as yet unknown. In this sense, sever profile of multi-domain cognitive and functional affectation in further studies must clarify the role of TAU pathology and other me- HD. chanisms on neurodegeneration and related cognitive deficits in HD. As expected, both the measures of GMV and cortical thinning cor- Moreover, how and when these mechanisms start promoting clinical related with multiple cognitive variables. However, among the various and brain changes and why they affect people with equivalent CAG and measures obtained for cognitive performance, those focusing on ex- DBS in a different manner merits further in-depth research. In any case, ecutive functions, memory, processing speed, language and construc- the present study adds novel evidence on the heterogeneity of HD, and tional abilities were those better characterizing patients with dementia. supports the need identifying mechanisms participating in this clinical These data suggest that the neurocognitive profile differs between HD heterogeneity. This is of major importance taking into account that patients with and those without major cognitive impairment in the ongoing and imminent clinical trials on HD are focused on huntingtin- range of dementia according to the PD-CRS classification. Whereas HD lowering strategies but did not take into account other potential factors patients with normal-to-mild cognitive deficits exhibit a prototypical contributing to disease severity(Tabrizi et al., 2019). frontal-executive dysfunction profile, HD patients with major cognitive In any case, pinpointing the pattern of brain changes and the cog- impairment exhibit a cortical-subcortical profile with deficits extending nitive profile of major cognitive impairment in the range of dementia in beyond executive functions and involving amnestic difficulties, con- HD has valuable implications that merit further research in bigger structional apraxia, confrontation naming deficits and reduced se- samples and in a longitudinal design. mantic abilities. However, these findings emerged performing com- The main limitation of the present study is the lack of a gold-stan- parisons with the subtests comprising the PD-CRS, it is, the instrument dard to classify patients according to cognitive status, but the method used to classify patients. Thus, further studies should explore the spe- we used to classify patients has recently been shown to be reliable in cific neuropsychological correlates of major cognitive impairment and this population(Martinez-Horta et al., 2020). Despite this limitation, it related brain differences in HD using additional comprehensive neu- is reasonable to assume that the measure we used for classification ropsychological assessment. differentiates patients with normal cognitive function or mild cognitive 7 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 defects from those with severe cognitive alterations. However, the doi.org/10.1016/j.nicl.2020.102415. concept of dementia used in the present work must be taken cautiously because of the absence of HD-specific clinical diagnostic criteria for References dementia. Moreover, the use of a functional assessment other than the TFC, specifically addressing cognitive-related functional deficits, must Walker, F.O., 2007. Huntington's disease. Lancet 369 (9557), 218–228. be considered in further studies. Other obvious limitations to take in Ross, C.A., Aylward, E.H., Wild, E.J., Langbehn, D.R., Long, J.D., Warner, J.H., Scahill, R.I., Leavitt, B.R., Stout, J.C., Paulsen, J.S., Reilmann, R., Unschuld, P.G., Wexler, A., consideration are the small sample size and the cross-sectional design of Margolis, R.L., Tabrizi, S.J., 2014. Huntington disease: natural history, biomarkers the study. Accordingly, will be required to confirm our findings in and prospects for therapeutics. Nature reviews. Neurology 10 (4), 204–216. further studies addressing this question in a bigger sample and in a Peavy, G.M., Jacobson, M.W., Goldstein, J.L., Hamilton, J.M., Kane, A., Gamst, A.C., Lessig, S.L., Lee, J.C., Corey-Bloom, J., 2010. Cognitive and functional decline in longitudinal setting. Huntington's disease: dementia criteria revisited. Movement disorders : official journal of the Movement Disorder Society 25 (9), 1163–1169. 5. Conclusions Rosas, H.D., Salat, D.H., Lee, S.Y., Zaleta, A.K., Pappu, V., Fischl, B., Greve, D., Hevelone, N., Hersch, S.M., 2008. Cerebral cortex and the clinical expression of Huntington's disease: complexity and heterogeneity. Brain : a journal of neurology 131 (Pt 4), Overall, significant patterns of cortical thinning and reduced GMV 1057–1068. in parieto-temporal and occipital regions are associated with more se- Nopoulos, P.C., Aylward, E.H., Ross, C.A., Johnson, H.J., Magnotta, V.A., Juhl, A.R., Pierson, R.K., Mills, J., Langbehn, D.R., Paulsen, J.S., 2010. Cerebral cortex structure vere cognitive deterioration in HD. The addition of cortical-instru- in prodromal Huntington disease. Neurobiology of disease 40 (3), 544–554. mental, semantic and amnestic-like deficits to the prototypical frontal- Rosas, H.D., Hevelone, N.D., Zaleta, A.K., Greve, D.N., Salat, D.H., Fischl, B., 2005. executive neurocognitive profile of HD is a differentiating characteristic Regional cortical thinning in preclinical Huntington disease and its relationship to cognition. Neurology 65 (5), 745–747. of major cognitive impairment in the range of dementia in HD. Early Tabrizi, S.J., Langbehn, D.R., Leavitt, B.R., Roos, R.A., Durr, A., Craufurd, D., Kennard, C., detection of brain changes and cognitive parameters associated with Hicks, S.L., Fox, N.C., Scahill, R.I., Borowsky, B., Tobin, A.J., Rosas, H.D., Johnson, severe forms of cognitive deterioration in HD may contribute to the H., Reilmann, R., Landwehrmeyer, B., Stout, J.C., 2009. Biological and clinical manifestations of Huntington's disease in the longitudinal TRACK-HD study: cross- early identification of individuals at greater risk of a more rapid and sectional analysis of baseline data. The Lancet. Neurology 8 (9), 791–801. aggressive course of cognitive impairment. Identifying the mechanisms Coppen, E.M., Grond, J.V., Hafkemeijer, A., Barkey Wolf, J.J.H., Roos, R.A.C., 2018. that contribute to this more aggressive cognitive deterioration should Structural and functional changes of the visual cortex in early Huntington's disease. Hum. Brain Mapp. 39 (12), 4776–4786. be of major relevance in the planning of future clinical trials. Rub, U., Seidel, K., Vonsattel, J.P., Lange, H.W., Eisenmenger, W., Gotz, M., Del Turco, D., Bouzrou, M., Korf, H.W., Heinsen, H., 2015. Huntington's Disease (HD): CRediT authorship contribution statement Neurodegeneration of Brodmann's Primary Visual Area 17 (BA17). Brain Pathol 25 (6), 701–711. Kuwert, T., Lange, H.W., Langen, K.J., Herzog, H., Aulich, A., Feinendegen, L.E., 1990. Saul Martinez-Horta: Conceptualization, Data curation, Formal Cortical and subcortical glucose consumption measured by PET in patients with analysis, Funding acquisition, Investigation, Methodology, Project ad- Huntington's disease. Brain : a journal of neurology 113 (Pt 5), 1405–1423. Kassubek, J., Juengling, F.D., Ecker, D., Landwehrmeyer, G.B., 2005. Thalamic atrophy in ministration, Resources, Writing - original draft, Writing - review & Huntington's disease co-varies with cognitive performance: a morphometric MRI editing. Frederic Sampedro: Data curation, Formal analysis, analysis. Cereb. Cortex 15 (6), 846–853. Methodology, Writing - original draft, Writing - review & editing. Snowden, J.S., 2017. The Neuropsychology of Huntington's Disease. Archives of clinical Andrea Horta-Barba: Data curation, Formal analysis, Project admin- neuropsychology : the official journal of the National Academy of Neuropsychologists 32 (7), 876–887. istration, Writing - review & editing. Jesús Perez-Perez: Dogan, I., Eickhoff, S.B., Schulz, J.B., Shah, N.J., Laird, A.R., Fox, P.T., Reetz, K., 2013. Conceptualization, Data curation, Funding acquisition, Investigation, Consistent neurodegeneration and its association with clinical progression in Methodology, Project administration. Javier Pagonabarraga: Huntington's disease: a coordinate-based meta-analysis. Neuro-degenerative diseases 12 (1), 23–35. Conceptualization, Resources, Writing - original draft, Writing - review Gomez-Anson, B., Alegret, M., Munoz, E., Monte, G.C., Alayrach, E., Sanchez, A., Boada, & editing. Beatriz Gomez-Anson: Methodology, Writing - original M., Tolosa, E., 2009. Prefrontal cortex volume reduction on MRI in preclinical draft, Writing - review & editing. Jaime Kulisevsky: Huntington's disease relates to visuomotor performance and CAG number. Parkinsonism & related disorders 15 (3), 213–219. Conceptualization, Funding acquisition, Investigation, Methodology, Say, M.J., Jones, R., Scahill, R.I., Dumas, E.M., Coleman, A., Santos, R.C., Justo, D., Writing - original draft, Writing - review & editing. Campbell, J.C., Queller, S., Shores, E.A., Tabrizi, S.J., Stout, J.C., 2011. Visuomotor integration deficits precede clinical onset in Huntington's disease. Neuropsychologia 49 (2), 264–270. Declaration of Competing Interest Carmichael, A.M., Irish, M., Glikmann-Johnston, Y., Singh, P., Stout, J.C., 2019. Pervasive autobiographical memory impairments in Huntington's disease. Neuropsychologia 127, 123–130. The authors declare that they have no known competing financial Harris, K.L., Armstrong, M., Swain, R., Erzinclioglu, S., Das, T., Burgess, N., Barker, R.A., interests or personal relationships that could have appeared to influ- Mason, S.L., 2019. Huntington's disease patients display progressive deficits in hip- ence the work reported in this paper. pocampal-dependent cognition during a task of spatial memory. Cortex; a journal devoted to the study of the nervous system and behavior 119, 417–427. Martinez-Horta, S., Horta-Barba, A., Perez-Perez, J., Antoran, M., Pagonabarraga, J., Acknowledgements Sampedro, F., Kulisevsky, J., 2019. Impaired face-like object recognition in pre- manifest Huntington's disease. Cortex; a journal devoted to the study of the nervous system and behavior 123, 162–172. The authors wish to thank all those at the Hospital de la Santa Creu i Coppen, E.M., Jacobs, M., van der Zwaan, K.F., Middelkoop, H.A.M., Roos, R.A.C., 2019. Sant Pau involved in the study. The authors also wish to extend their Visual Object Perception in Premanifest and Early Manifest Huntington's Disease. gratitude to the study participants and their families. Archives of clinical neuropsychology : the official journal of the National Academy of Neuropsychologists 34 (8), 1320–1328. Wolf, R.C., Sambataro, F., Vasic, N., Baldas, E.M., Ratheiser, I., Bernhard Landwehrmeyer, Funding G., Depping, M.S., Thomann, P.A., Sprengelmeyer, R., Sussmuth, S.D., Orth, M., 2014. Visual system integrity and cognition in early Huntington's disease. The European journal of neuroscience 40 (2), 2417–2426. The present study was partially funded by a Spanish Government Labuschagne, I., Cassidy, A.M., Scahill, R.I., Johnson, E.B., Rees, E., O'Regan, A., Queller, Grant (PI17/001885) from the ISCIII (Spain), Fondos FEDER (Spain) S., Frost, C., Leavitt, B.R., Durr, A., Roos, R., Owen, G., Borowsky, B., Tabrizi, S.J., and by the “Human Biology Project Grant” of the Huntington’s Disease Stout, J.C., 2016. Visuospatial Processing Deficits Linked to Posterior Brain Regions in Premanifest and Early Stage Huntington's Disease. Journal of the International Society of America (USA). Neuropsychological Society : JINS 22 (6), 595–608. Hinzen, W., Rossello, J., Morey, C., Camara, E., Garcia-Gorro, C., Salvador, R., de Diego- Balaguer, R., 2018. A systematic linguistic profile of spontaneous narrative speech in Appendix A. Supplementary data pre-symptomatic and early stage Huntington's disease. Cortex; a journal devoted to the study of the nervous system and behavior 100, 71–83. Supplementary data to this article can be found online at https:// Chan, J.C.S., Stout, J.C., Vogel, A.P., 2019. Speech in prodromal and symptomatic 8 S. Martinez-Horta, et al. NeuroImage: Clinical 28 (2020) 102415 Huntington's disease as a model of measuring onset and progression in dominantly Gironell, A., 2008. Parkinson's disease-cognitive rating scale: a new cognitive scale inherited neurodegenerative diseases. Neurosci. Biobehav. Rev. 107, 450–460. specific for Parkinson's disease. Movement disorders : official journal of the Chenery, H.J., Copland, D.A., Murdoch, B.E., 2002. Complex language functions and Movement Disorder Society 23 (7), 998–1005. subcortical mechanisms: evidence from Huntington's disease and patients with non- Landwehrmeyer, G.B., Fitzer-Attas, C.J., Giuliano, J.D., Goncalves, N., Anderson, K.E., thalamic subcortical lesions. International journal of language & communication Cardoso, F., Ferreira, J.J., Mestre, T.A., Stout, J.C., Sampaio, C., 2017. Data Analytics disorders 37 (4), 459–474. from Enroll-HD, a Global Clinical Research Platform for Huntington's Disease. Podoll, K., Caspary, P., Lange, H.W., Noth, J., 1988. Language functions in Huntington's Movement disorders clinical practice 4 (2), 212–224. disease. Brain : a journal of neurology 111 (Pt 6), 1475–1503. Hs, G., 1996. Unified Huntington's Disease Rating Scale: reliability and consistency. Tabrizi, S.J., Reilmann, R., Roos, R.A., Durr, A., Leavitt, B., Owen, G., Jones, R., Johnson, Huntington Study Group, Movement disorders : official journal of the Movement H., Craufurd, D., Hicks, S.L., Kennard, C., Landwehrmeyer, B., Stout, J.C., Borowsky, Disorder Society 11 (2), 136–142. B., Scahill, R.I., Frost, C., Langbehn, D.R., 2012. Potential endpoints for clinical trials Shoulson, I., Fahn, S., 1979. Huntington disease: clinical care and evaluation. Neurology in premanifest and early Huntington's disease in the TRACK-HD study: analysis of 24 29 (1), 1–3. month observational data. The Lancet. Neurology 11 (1), 42–53. Penney Jr., J.B., Vonsattel, J.P., MacDonald, M.E., Gusella, J.F., Myers, R.H., 1997. CAG Paulsen, J.S., Long, J.D., Johnson, H.J., Aylward, E.H., Ross, C.A., Williams, J.K., Nance, repeat number governs the development rate of pathology in Huntington's disease. M.A., Erwin, C.J., Westervelt, H.J., Harrington, D.L., Bockholt, H.J., Zhang, Y., Ann. Neurol. 41 (5), 689–692. McCusker, E.A., Chiu, E.M., Panegyres, P.K., 2014. Clinical and Biomarker Changes in Martinez-Horta, S., Moreu, A., Perez-Perez, J., Sampedro, F., Horta-Barba, A., Premanifest Huntington Disease Show Trial Feasibility: A Decade of the PREDICT-HD Pagonabarraga, J., Gomez-Anson, B., Lozano-Martinez, G.A., Lopez-Mora, D.A., Study. Front. Aging Neurosci. 6, 78. Camacho, V., Fernandez-Leon, A., Carrio, I., Kulisevsky, J., 2019. The impact of bi- Consortium, G.M.o.H.s.D.G.-H., 2015. Identification of Genetic Factors that Modify lingualism on brain structure and function in Huntington's disease. Parkinsonism & Clinical Onset of Huntington's Disease. Cell 162 (3), 516–526. related disorders 60, 92–97. Langbehn, D.R., Stout, J.C., Gregory, S., Mills, J.A., Durr, A., Leavitt, B.R., Roos, R.A.C., Fischl, B., Dale, A.M., 2000. Measuring the thickness of the human cerebral cortex from Long, J.D., Owen, G., Johnson, H.J., Borowsky, B., Craufurd, D., Reilmann, R., magnetic resonance images. PNAS 97 (20), 11050–11055. Landwehrmeyer, G.B., Scahill, R.I., Tabrizi, S.J., 2019. Association of CAG Repeats Garcia-Gorro, C., Llera, A., Martinez-Horta, S., Perez-Perez, J., Kulisevsky, J., Rodriguez- With Long-term Progression in Huntington Disease. JAMA neurology. Dechicha, N., Vaquer, I., Subira, S., Calopa, M., Munoz, E., Santacruz, P., Ruiz-Idiago, Braisch, U., Muche, R., Rothenbacher, D., Landwehrmeyer, G.B., Long, J.D., Orth, M., J., Mareca, C., Beckmann, C.F., de Diego-Balaguer, R., Camara, E., 2019. Specific 2019. Identification of symbol digit modality test score extremes in Huntington's patterns of brain alterations underlie distinct clinical profiles in Huntington's disease. disease, American journal of medical genetics. Part B, Neuropsychiatric genetics : the NeuroImage. Clinical 23, 101900. official publication of the International Society of Psychiatric Genetics 180 (3), Wild, E.J., Boggio, R., Langbehn, D., Robertson, N., Haider, S., Miller, J.R., Zetterberg, H., 232–245. Leavitt, B.R., Kuhn, R., Tabrizi, S.J., Macdonald, D., Weiss, A., 2015. Quantification Vuono, R., Winder-Rhodes, S., de Silva, R., Cisbani, G., Drouin-Ouellet, J., Spillantini, of mutant huntingtin protein in cerebrospinal fluid from Huntington's disease pa- M.G., Cicchetti, F., Barker, R.A., 2015. The role of tau in the pathological process and tients. J. Clin. Investig. 125 (5), 1979–1986. clinical expression of Huntington's disease. Brain : a journal of neurology 138 (Pt 7), Rub, U., Seidel, K., Heinsen, H., Vonsattel, J.P., den Dunnen, W.F., Korf, H.W., 2016. 1907–1918. Huntington's disease (HD): the neuropathology of a multisystem neurodegenerative Mestre, T.A., Bachoud-Levi, A.C., Marinus, J., Stout, J.C., Paulsen, J.S., Como, P., Duff, K., disorder of the human brain. Brain Pathol 26 (6), 726–740. Sampaio, C., Goetz, C.G., Cubo, E., Stebbins, G.T., Martinez-Martin, P., 2018. Rating Rocha, N.P., Ribeiro, F.M., Furr-Stimming, E., Teixeira, A.L., 2016. Neuroimmunology of scales for cognition in Huntington's disease: Critique and recommendations. Huntington's Disease: Revisiting Evidence from Human Studies. Mediators Inflamm. Movement disorders : official journal of the Movement Disorder Society 33 (2), 2016, 8653132. 187–195. Fernandez-Nogales, M., Cabrera, J.R., Santos-Galindo, M., Hoozemans, J.J., Ferrer, I., Martinez-Horta, S., Horta-Barba, A., Perez-Perez, J., Sampedro, F., de Lucia, N., De Rozemuller, A.J., Hernandez, F., Avila, J., Lucas, J.J., 2014. Huntington's disease is a Michele, G., Kehrer, S., Priller, J., Migliore, S., Squitieri, F., Castaldo, A., Mariotti, C., four-repeat tauopathy with tau nuclear rods. Nat. Med. 20 (8), 881–885. Mananes, V., Lopez-Sendon, J.L., Rodriguez, N., Martinez-Descals, A., Garcia-Ruiz, P., Tabrizi, S.J., Leavitt, B.R., Landwehrmeyer, G.B., Wild, E.J., Saft, C., Barker, R.A., Blair, Julio, F., Januario, C., Delussi, M., de Tommaso, M., Noguera, S., Ruiz-Idiago, J., N.F., Craufurd, D., Priller, J., Rickards, H., Rosser, A., Kordasiewicz, H.B., Czech, C., Sitek, E.J., Nuzzi, A., Pagonabarraga, J., Kulisevsky, J., 2020. Utility of the Swayze, E.E., Norris, D.A., Baumann, T., Gerlach, I., Schobel, S.A., Paz, E., Smith, Parkinson's disease-Cognitive Rating Scale for the screening of global cognitive status A.V., Bennett, C.F., Lane, R.M., 2019. Targeting Huntingtin Expression in Patients in Huntington's disease. J. Neurol. with Huntington's Disease. The New England journal of medicine 380 (24), Pagonabarraga, J., Kulisevsky, J., Llebaria, G., Garcia-Sanchez, C., Pascual-Sedano, B., 2307–2316. 9