RESEARCH ARTICLE Selective vulnerability to atrophy in sporadic Creutzfeldt- Jakob disease Kyan Younes1,a , Julio C. Rojas1,a , Amy Wolf1, Goh M. Sheng-Yang1, Matteo Paoletti1,2 , n-Gala4 , Joel H. Gianina Toller1 , Eduardo Caverzasi3 , Maria Luisa Mandelli1 , Ignacio Illa 1 1 1 1 Kramer , Yann Cobigo , Bruce L. Miller , Howard J. Rosen & Michael D. Geschwind1 1 Department of Neurology, Weill Institute for Neurosciences, Memory and Aging Center, University of California, San Francisco (UCSF), San Francisco, California 2 Advanced Imaging and Radiomics Center, Neuroradiology Department, IRCCS Mondino Foundation, Pavia, Italy 3 Department of Neurology, University of California, San Francisco (UCSF), San Francisco, California 4 Department of Neurology, Hospital de la Santa Creu i Sant Pau, Universitat Autonoma de Barcelona, Barcelona, Spain Correspondence Abstract Michael D. Geschwind, University of California, San Francisco (UCSF), Memory Objective: Identification of brain regions susceptible to quantifiable atrophy in and Aging Center, Box 1207, San Francisco, sporadic Creutzfeldt-Jakob disease (sCJD) should allow for improved under- CA 94143-1207, USA. Telephone: 415-476- standing of disease pathophysiology and development of structural biomarkers 2900; Fax 415-476-1816; E-mail: that might be useful in future treatment trials. Although brain atrophy is not michael.geschwind@ucsf.edu usually present by visual assessment of MRIs in sCJD, we assessed whether using voxel-based morphometry (VBM) can detect group-wise brain atrophy in Received: 18 September 2020; Revised: 16 November 2020; Accepted: 4 December sCJD. Methods: 3T brain MRI data were analyzed with VBM in 22 sCJD partic- 2020 ipants and 26 age-matched controls. Analyses included relationships of regional brain volumes with major clinical variables and dichotomization of the cohort Annals of Clinical and Translational according to expected disease duration based on prion molecular classification Neurology 2021; 8(6): 1183–1199 (i.e., short-duration/Fast-progressors (MM1, MV1, and VV2) vs. long-duration/ Slow-progressors (MV2, VV1, and MM2)). Structural equation modeling doi: 10.1002/acn3.51290 (SEM) was used to assess network-level interactions of atrophy between specific a Authors contributed equally to this work. brain regions. Results: sCJD showed selective atrophy in cortical and subcorti- cal regions overlapping with all but one region of the default mode network (DMN) and the insulae, thalami, and right occipital lobe. SEM showed that the effective connectivity model fit in sCJD but not controls. The presence of visual hallucinations correlated with right fusiform, bilateral thalami, and medial orbi- tofrontal atrophy. Interestingly, brain atrophy was present in both Fast- and Slow-progressors. Worse cognition was associated with bilateral mesial frontal, insular, temporal pole, thalamus, and cerebellum atrophy. Interpretation: Brain atrophy in sCJD preferentially affects specific cortical and subcortical regions, with an effective connectivity model showing strength and directionality between regions. Brain atrophy is present in Fast- and Slow-progressors, corre- lates with clinical findings, and is a potential biomarker in sCJD. visual inspection of MRI, except for cases of long dura- Introduction tion. Studies on objective atrophy quantification in sCJD Sporadic Creutzfeldt-Jakob disease (sCJD) is a rapidly are scant, and there are no data on the presence of atro- progressive dementia with underlying neurodegeneration, phy in Slow- versus Fast-progressors based on sCJD gliosis, and vacuolation.1 Although early observations molecular classification subtyping5 or on brain atrophy reported cerebral atrophy in sCJD through clinical neu- patterns associated with some common clinical character- roimaging, this was usually associated with very late dis- istics of sCJD. Previous morphometric studies have ease course and/or long duration.2–4 Currently, cortical focused primarily on isolated brain regions and regions of atrophy in sCJD is not considered as a salient feature on interests,6,7 diffusion tensor imaging metrics in the gray ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1183 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. and white matter,8,9 or cohorts which combined sporadic Although we had initially intended to also examine longi- and genetic forms of prion disease.10,11 Determination of tudinal volume change, of the 14 participants with serial atrophy patterns in sCJD might provide insights into the T1 scans, too few (n = 4) were of sufficient quality for mechanisms of disease progression and identify vulnerable cohort longitudinal analysis, so this was not performed. regions to track disease progression in both Fast- and Twenty-six healthy age- and gender-matched participants, Slow-progressors, which could be targets of future net- from the UCSF MAC Hillblom Healthy Aging Network work-level-based therapies. The primary goal of this project, who had MRIs performed with the same protocol cross-sectional study is to quantify cerebral atrophy in on the same scanner were used as healthy controls (Con- sCJD using 3T brain magnetic resonance imaging (MRI) trols). For some analyses, sCJD participants were grouped and voxel-based morphometry (VBM). Based on the pro- based on their DWI lesion patterns by visual inspection posed mechanisms of transsynaptic propagation of prion into 1) cortical-subcortical, 2) cortical-only, and 3) sub- particles,12 our hypothesis is that neurodegeneration in cortical-only cohorts.17 sCJD causes selective brain network degeneration in anatomically connected regions, as opposed to indiscrimi- Clinical and cognitive evaluation nate global injury that might happen if prion disease was spreading within the brain either randomly or to adjacent sCJD participants had a standardized clinical evaluation regions. Structural equation modeling (SEM) is a statisti- including: neurological history and examination; Mini- cal technique used in functional MRI and positron emis- Mental State Examination (MMSE)18; the Barthel index sion tomography (PET) studies to test whether activity in (measuring activities of daily living function)19; a neu- brain regions correlated with each other indiscriminately ropsychological battery (although we only used MMSE); or if activity in one region influences that in another.13,14 the neuropsychiatric inventory (NPI, 12-item version)20 to We therefore tested this hypothesis-driven approach assess for behavioral symptoms and the presence of halluci- (SEM) to further investigate the relationship between nations; motor evaluation with the Unified Parkinson’s brain regions vulnerable to sCJD pathology beyond sim- Disease Rating Scale – Motor (UPDRS-Motor); and routine ple correlations. electroencephalogram (EEG). Controls underwent neuro- logical history and exam, the clinical dementia rating scale, and a neuropsychological battery. Signs and symptoms Participants and Methods were recorded from patient records and UCSF research visit summaries until the time of the research brain MRI. EEGs Patient selection were classified into three categories: normal, slowing, or Participants or caregivers provided informed consent for periodic sharp wave complexes, with the latter two catego- participation in this study, which was approved by the rized as abnormal for this study.21 CSF total tau (t-tau; per- University of California, San Francisco (UCSF) institu- formed at the U.S. National Prion Disease Pathology tional review board. Participants were evaluated between Surveillance Center (NPDPSC), Cleveland, OH) and neu- January 2010 and August 2013 at the UCSF Memory and ron-specific enolase (NSE; performed at Mayo Laborato- Aging Center (MAC) rapidly progressive dementia ries, Rochester, MN) levels, determined with enzyme- research program. Consecutive sCJD participants who linked immunosorbent assay, and protein 14-3-3 deter- ultimately either met 2007 UCSF clinical criteria for prob- mined by Western Blot (NPDPSC; reported as either posi- able sCJD15 (n = 8, 20%) or who had definite sCJD tive, inconclusive, or negative, with inconclusive (pathology-proven, n = 32, 80%),16 and who had the considered as negative for this study) were available in 17, same standardized MRI protocol of adequate quality (on 13, and 19 sCJD participants, respectively. EEG and CSF the same scanner) from their first UCSF visit were biomarkers were obtained within 24 hours of the brain included in the study. Forty sCJD participants were iden- MRI. We also examined the effect of the prion protein gene tified prior to MRI quality assessment. For the VBM anal- (PRNP) codon 129 polymorphism on other clinical vari- ysis, 18 of these 40 participants were excluded due to ables, including t-tau, NSE, protein 14-3-3, EEG, and clini- severe motion artifact (n = 14) or lack of appropriate cal scales. None of the 40 sCJD participants had a PRNP magnetization-prepared rapid gradient-echo (MPRAGE) mutation, and all had codon 129 polymorphism analysis T1-weighted images (n = 4), leaving 22 participants with (NPDPSC, Case Western Reserve University, Cleveland, sufficient quality scan for VBM analyses. All 40 sCJD par- OH). Pathological confirmation was performed at either or ticipants, however, had sufficient quality diffusion- both UCSF and NPDSPC, and prion typing was performed weighted imaging (DWI) MRIs for visual assessment (e.g., through the NPDPSC. The 32 pathology-proven sCJD par- determination of pattern of involvement by DWI) and ticipants consisted of 18 of 22 patients included in the were used for analyses not requiring T1 sequences. VBM analysis and 14 of 18 in the group excluded from the 1184 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease VMB analysis. One pathology-proven subject without VBM VBM processing and analysis had variably protease-sensitive prionopathy and therefore had no prion typing.22–24 For the MRIs with sufficient quality scans to undergo VBM processing, a general linear model was fit at each voxel using FMRIB Software Library v6.0 (FSL, Created Image acquisition by the Analysis Group, FMRIB, Oxford, UK). All compar- Participants underwent brain MRI at UCSF on a 3T scan- isons included correction for age and total intracranial ner (Siemens TrioTim syngo, Erlangen, Germany) volume (TIV; calculated as the sum of gray matter, white equipped with an eight-channel transmit and receive head matter, and CSF volumes). For comparisons in which we coil using an MPRAGE sequence with the following wanted to capture the effect of the speed of disease pro- parameters: 160 sagittal slices; slice thickness = 1 mm; gression (based on molecular classification dichotomized field of view = 256 mm; matrix = 256 9 240; voxel size into Fast-progressors vs. Slow-progressors), we controlled 1.0 9 1.0 9 1.0 mm3; TR/TE/TI = 2300/2.98/900 ms, flip for disease severity by the Barthel index at the time of the angle = 9°, and a HARDI dataset acquired using a single- MRI. This ensures that the findings are not due to the shot spin-echo echo-planar imaging (EPI) sequence presence of more advanced disease in one group com- including 55 contiguous axial slices acquired in an inter- pared to the other. For comparisons that included clinical leaved order with the following parameters: TR/ severity of symptoms (global cognition based on MMSE, TE = 8000/109 ms; flip angle = 90°; matrix = 100 9 100; motor symptoms based on UPDRS, and behavioral symp- in-plane resolution = 2.2 mm2; slice thickness = 2.2 mm; toms based on the NPI), we controlled for the MRI time- 64 noncollinear diffusion sensitization directions at ratio (the time the MRI was done relative to the total dis- b = 2000 s/mm2, 1 at b = 0; and integrated parallel ease duration) to ensure that the findings are not simply acquisition technique acceleration (IPAT) factor = 2. due to being either early or late in the total disease pro- cess. For comparisons between those subjects with or without certain clinical symptoms (myoclonus, visual hal- Image processing lucinations, and ataxia), we controlled for both disease Image preprocessing for VBM analyses severity (Barthel index) and MRI time-ratio in order to capture the brain regions associated with these symptoms. All T1-weighted images were visually inspected, and images For these categorical VBM comparisons, a t-test was per- with excessive motion or artifact were excluded. Processing formed for each voxel, with accepted threshold p value < for T1-weighted sequences was performed using Statistical .05 after correction for multiple comparisons with the Parametric Mapping (SPM12, Welcome Trust Center for permutation method used by the FMRIB Software Library Neuroimaging, London, UK) running under MATLAB (FSL), and the number of permutations was set at 5,000 R2014b (MathWorks). Images were bias field-corrected using threshold-free cluster enhancement.26,27 Resulting using N3 algorithm and segmented into gray matter, white statistical maps were normalized to Montreal Neurological matter, and CSF compartments using the unified segmenta- Institute space for display.28 tion algorithm from SPM12. A custom template was cre- We ran VBM analyses based on the following clinical ated from the whole cohort population (sCJD participants variables: and Controls) by nonlinear registration template genera- 1 sCJD versus Controls. tion using Large Deformation Diffeomorphic Metric Mapping 2 Fast-progressors versus Slow-progressors based on framework cite.25 The images were then spatially normal- molecular classification (PRNP codon 129 genotype ized to the custom template and then modulated by multi- [MM, MV, or VV] and prion typing [type 1 or type plying the voxel values by the Jacobian determinant derived 2]).5 As the median disease duration of three sCJD from the spatial normalization to preserve the relative vol- molecular subtypes (MM1, MV1, and VV2) is less than umes of gray matter. Finally, the images were smoothed 7 months (Fast-progressors) and for the remaining with an 8-mm full-width-at-half-maximum Gaussian ker- three molecular subtypes (MM2, MV2, and VV1) typi- nel. This parcellation process created raw volumetric values cally is greater than 11 months (Slow-progressors),5,29 for gray matter, white matter, and CSF compartments. The we conducted an analysis dichotomizing the sCJD Anatomical Automated Labeling parcellation method was VBM group into expected Fast-progressors versus used to generate gray matter volumes, which were used in Slow-progressors. For this analysis, we did not include the SEM analysis and some other analyses to compare with sCJD cases with mixed prion types (i.e., 1-2), as such CSF biomarkers (14-3-3, NSE, and total tau). The gray cases often present more heterogeneously, varying along matter images then underwent VBM processing and analy- a spectrum between type 1 and type 2,29,30 nor did we sis as discussed below. ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1185 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. include the single variably protease-sensitive prionopa- SEM to investigate model fit of regional VBM volumetric thy case, who by definition had no prion typing.31 We data. Our hypothesis was that if region atrophy is not also compared MM with MV cases (insufficient num- random, some regions will influence volumetric changes bers for VV). Addtionally, we ran a VBM analysis com- in another region or regions, and a model that includes paring MM2 to MV2 participants (both Slow- the regions most commonly involved in sCJD will pass progressors) given their clinical and pathological differ- the strict significance thresholds of SEM only in sCJD but ences,5 and we had sufficient samples sizes of these two not in regions not commonly involved in sCJD or in sCJD subtypes. Controls. Because previous research that used hypothesis- 3 Global cognitive function dichotomized based on the driven SEM approaches recommends starting with parsi- median MMSE score of 20/30. Higher MMSE group monious models, we tested SEM models in four canonical for participants with score > 20/30 and lower MMSE networks in sCJD participants and Controls first and then group for participants with MMSE score ≤ 20/30. added regions commonly affected in sCJD. 4 The presence or absence of visual hallucinations cap- SEM is a multivariate technique that combines “path tured on the NPI based on having visual hallucination analysis” (a statistic used to evaluate causal models)32 and at any point in the clinical course up until the time of multiple regression to estimate linear relationships brain MRI. between specific variables in an analysis of the covariance 5 The presence or absence of ataxia. among these variables. One might expect cortical volumes 6 The presence of absence of myoclonus. of certain brain regions in sCJD to correlate with each 7 Motor symptoms measured by the UPDRS-Motor score other. This correlation of volumes could occur for at least and dichotomized based on the median UPDRS-Motor three reasons: 1.) regional vulnerability; 2.) prions spread- score (lower ≤ 11, higher > 11; higher score indi- ing transynaptically to functionally connected regions; or cates more motor impairment). 3.) prions spreading to adjacent areas. SEM extends 8 Behavioral changes dichotomized based on the median beyond simple correlations to evaluate the presence of NPI FrequencyxSeverity product score (lower ≤ 20, any directional relationship among brain regions, and it higher > 20; higher score indicates more impairment/ examines the strength as well as the directionality of rela- symptoms). tionships within a network.14 SEM incorporates the resid- ual variance of the independent variables to estimate the dependences among the observed variables allowing for Statistical analyses the validation of a hypothetical model by empirical data Tests of normality for all continuous data in sCJD and using strict goodness-of-fit thresholds.13 The fundamental Controls were conducted with the Shapiro-Wilk test. concept for the application of SEM for brain network Means for continuous clinical variables were compared analysis is effective connectivity, which is defined as the with the Student’s t or the Mann-Whitney U tests, where influence that one neural system exerts over another and appropriate. Because clinical and biomarker variables were vice versa. It differs from functional connectivity in that not normally distributed, correlations were performed with it is not solely defined by statistical dependencies between the Spearman test. False discovery rate (FDR; at p < 0.05) remote events but takes into account the activity-depen- was used to correct for multiple comparisons for all analy- dent, explicit, and directional coupling between them. ses, including non-VBM (demographics, disease duration, Effective connectivity can be estimated from SEM by test- clinical symptoms, cognitive and functional scales, CSF ing whether a theoretical connectivity model seeking to biomarkers, EEG pattern, DWI involvement pattern, codon explain a network of influences can actually fit the influ- 129, molecular classification, MRI-based volume, and com- ences estimated from observed data.14 In our case, the parison of features between the sCJD subgroups included theoretical models that we tested were based on regions vs. those not included in VBM analysis in Table 1) and known to be commonly clinically affected in sCJD (i.e., VBM comparisons. Data analysis was performed with SPSS network of cortico-subcortical regions typically affected (version 23, SPSS/IBM, Chicago, IL). by restricted diffusion, including thalamus and stria- tum)11,17,33 or on known anatomical connectivity in healthy participants (i.e., default mode, executive, sal- Volumetric correlations and structural ience, and motor networks). The DMN analysis was based equation modeling on Buckner et al. 2008, and it included the precuneus, To investigate the general correlation between the volu- angular gyrus (AG), anterior cingulate cortex (ACC), and metric data in the brain regions that showed atrophy in mesial and lateral temporal cortical regions.34 The data sCJD compared with Controls, Pearson correlations were used as input for the SEM model were age- and TIV-cor- performed after correction for age and TIV. We then used rected cortical volumes derived from the VBM analysis.35 1186 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease Table 1. Clinical features of sporadic Creutzfeldt-Jakob patients and Controls. sCJD included sCJD excluded from Controls n = 26 sCJD all n = 40 in VBM n = 22 VBM* n = 18 Age at first evaluation, years, 66  10 63  9 (66, 43-80) 64  10 (68, 43-80) 63  8 (64, 46-75) mean  SD (median, range) (67, 50-77) Sex, female (%) 42 45 36 56 Right-handed (%) 83 94 95 94 Disease duration at the time of MRI, 8  6 (7, 1-32) 8  5 (8, 2-23) 7  8 (6, 1-32) months, M  SD (Md, r) Total disease duration, months, 14  9 (14, 1-32) 16  7 (18, 4-28)5 12  10 (8, 3-38) M  SD (Md, r) Interval from neuroimaging to death, 6  2 (3, 0-21) 8  2 (5, 0-19) 5  2 (2, 2-38) months, M  SD (Md, r) Clinical characteristics1 (%) Cognitive difficulties 90 86 94 Visual disturbance 44 45 41 Ataxia 31 456 12 Hallucinations 36 31 41 Myoclonus 44 31 59 MMSE score, M  SD (Md, r, n) 14  10 (16, 0-29, 38) 18  8 (19, 1-29, 22)7 8  9 (5, 0-29, 16) NPI score, M  SD (Md, r, n) 33  24 (29, 0-93, 33) 27  22 (20, 0-79, 18) 42  25 (35, 8-93, 15) UPDRS motor, M  SD (Md, r, n) 17  16 (18, 0-63, 29) 16  13 (11, 0-42, 15) 26  17 (20, 0-63, 14) Barthel index M  SD (Md, r, n) 66  37 (80, 0-100, 33) 86  23 (95, 15-100, 19)7 40  38 (25, 0-100, 14) CSF t-tau (pg/mL) M  SD (Md, r, n)2 3870  4172 (1800, 2720  3748 (1429, 5497  4517 (4408, 326-15308, 29) 326-15308, 17)6 1022-13597, 12) CSF NSE (ng/mL) M  SD (Md, r, n)3 48  47 (31, 4-180, 26) 46  47 (31,18-178, 13) 49  48 (31,4-180, 13) CSF protein 14-3-3 n = 34 n = 19 n = 15 Positive (%) 44 256 67 Negative (%) 18 356 0 Inconclusive (%) 38 40 33 EEG n = 36 n = 21 n = 15 Periodic epileptiform discharges (PED) (%) 25 19 33 Slowing without PEDs (%) 53 47 60 Normal (%) 22 34 7 Diffusion-weighted image pattern (%) Cortical-subcortical 55 45 67 Cortical-only 28 41 11 Subcortical-only 18 14 22 PRNP gene codon 129 genotype (n (%)) 40 (100) 22 (100) 18 (100) MM (%) 11 (28) 6 (27) 5 (28) MV (%) 22 (55) 14 (64) 8 (44) VV (%) 7 (17) 2 (9) 5 (28) Pathologically confirmed cases (n (%)) 32 (80) 18 (82) 14 (77) Molecular Classification Prion typing not available (n (%)) 10 (25)# 4 (18) 6 (33)# Prion typing available (n (%)) 30 (75)# 18 (81) 12 (66) Fast-progressors (n (%)) 11 (35) (2 MM1, 5 (27) (3 MV1, 2 VV2) 6 (46) (2 MM1, 5 MV1, 4 VV2) 2 MV1, 2 VV2) Total disease duration, 7  4 (7, 3-13) 8  6 (7, 4-13) 6  3 (5.5, 3-9) months, M  SD, (Md, r) Slow-progressors (n (%)) 12 (38) (5 MM2, 9 (50) (4 MM2, 5 MV2) 3 (23) (1 MM2, 6 MV2, 1 VV1) 1 MV2, 1 VV1) Total disease duration, 19  6 (20, 7-27) 20  5 (20, 9-27) 18  5 (23, 7-24) months, M  SD (Md, r) Mixed Classification type (n (%)) 7 (22) (3 MM1-2, 4 (22) (2 MM1-2, 2 MV1-2) 3 (23) (1 MM1-2, 4 MV1-2) 2 MV1-2 ) 9  7 (12, 3-28) 18  9 (18, 10-28) 14  7 (12, 3-28) (Continued) ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1187 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. Table 1 Continued. sCJD included sCJD excluded from Controls n = 26 sCJD all n = 40 in VBM n = 22 VBM* n = 18 Total disease duration, months, M  SD (Md, r) MRI-based volume (corrected for TIV) Whole brain (mm3 9 105) M (SD) 4.7 (0.6) 4.2 (0.4) Gray matter (mm3 9 105) M (SD) 1.8 (0.1) 1.7 (0.1)4 White matter (mm3 9 105) M (SD) 1.4 (0.2) 1.2 (0.1) CSF (mm3 9 105) M (SD) 1.4 (0.2) 1.2 (0.1) CSF, cerebrospinal fluid; TIV, Total Intracranial Volume; MMSE, Mini-mental state examination; M, Mean; d, median; t-tau, total tau; NSE, neu- ronal-specific enolase; r, range, NPI, Neuropsychiatric inventory; UPDRS, Unified Parkinson’s Disease Rating Scale motor; * = excluded due to poor-quality MRI. Percentages might not sum to 100% due to rounding. # = 1 pathology-proven patient had variably protease-sensitive prion dis- ease which by definition has no prion type identified. 1 Includes signs and symptoms up until around the time of UCSF MRI. 2 Abnormal value ≥ 1150 ng/mL. 3 Abnormal value > 30 ng/mL. Comparisons between all sCJD, sCJD included in VBM, sCJD excluded from VBM, and Controls were done for all the variables in the tables and significant results are noted as below 4 Compared to Controls, P < 0.01 5 Compared to sCJD excluded from VBM analysis, P < 0.001 6 Compared to sCJD excluded from VBM analysis, P < 0.05 7 Compared to sCJD excluded from VBM analysis, P < 0.01 The effective connectivity is described by beta coefficients sCJD cohort (n = 40), its various subcohorts, and the (called “path coefficients” in SEM), which were generated Controls cohort (n = 26) are shown in Table 1 and in from interregional volume correlations via a process of the Supplementary Material. The relationship among CSF iterative data fitting using IBM SPSS Amos graphics soft- results and other outcomes or variables is presented in ware. Based on the evidence that prions propagate transy- Supplementary Material. naptically (i.e., via paths),12 it was hypothesized that Comparing volumetric values, derived from image pro- atrophy in sCJD occurs selectively and that volumes in cessing prior to VBM, between the 22 VBM sCJD cases brain regions affected would show a directional relation- and Controls showed significantly less total gray matter ship (i.e., one brain region driving atrophy in another volume in sCJD (p < 0.01, Student’s t-test) but no differ- brain region). Nonetheless, given that SEM is a causal ences in total white matter or CSF (Table 1). modeling statistical tool, we wanted to avoid drawing There was significant difference in t-tau levels using causative biological conclusion, and hence we only focus dichotomization of sCJD as Fast- versus Slow-progressors on the general model fit rather than on the beta coeffi- (based on the molecular classification). Among the cases cients. Theoretical models being tested were graphically with available t-tau and molecular classification (n = 21 represented by nodes (i.e., brain regions) exerting trophic of 40 total sCJD participants), the majority of participants influences through anatomical pathways (arrows), with with nonelevated t-tau were Slow-progressors (total nine the direction of a trophic effect represented by the arrow patients with normal tau: six (66%) Slow-progressors, direction and the connectivity strength (beta coefficient) one (11%) Fast-progressor, and two (22%) mixed prion represented by arrow thickness. Models for right, left, and type), whereas the majority of the participants with ele- whole brain were tested separately. Models with optimal vated t-tau were Fast-progressors (total 12 patients with data fit have low error (root mean square error of elevated tau: nine (75%) Fast-progressors, one (8%) approximation (RMSEA) < .05 and goodness-of-fit index Slow-progressor, and two (16%) mixed) (v2 = 10, (GFI) > .90).14 p = 0.008). In the total sCJD cohort (n = 40), there were no differ- ences in clinical biomarkers (i.e., CSF biomarkers, signs/ Results symptoms, clinical scales, and MMSE; see Methods) based on the three common DWI MRI patterns (cortical-only Fluid, clinical, and neuroimaging biomarkers vs. subcortical-only vs. cortico-subcortical). Because in The demographics, basic clinical features (including CSF our experience, patients with cortico-subcortical DWI biomarker, EEG, and MRI findings), and scores of our involvement tend to have greater functional impairment, 1188 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease in order to increase the possibility of finding an effect of of the executive control (lateral parietal), but none of the the DWI MRI pattern of involvement, we also compared motor network hubs.34, 36-37 this group to a combined group of subcortical-only and VBM analyses revealed atrophy clusters that survived cortical-only involvement. This subcohort with cortical permutations-based correction for multiple comparisons and subcortical involvement on DWI had higher t-tau within sCJD subgroups when dichotomizing based on the (p = 0.014; Supplemental Figure 1D), NSE (p = 0.005; dementia severity (MMSE) or the presence of visual hal- Supplemental Figure 1E), NPI scores (p = 0.021; not lucinations (Fig. 1B,C,G,H). sCJD participants with visual shown), and UPDRS-Motor scores (p = 0.027; not hallucinations (n = 6; 1 MM1, 1 MV1, 2 MV2, 1 VV2, shown), and they were more likely to have a positive 14- and 1 MV without typing available) displayed right more 3-3 (63% vs. 20%, v2 = 6, p = 0.017; not shown) than than left thalamus, bilateral medial orbitofrontal, rectus the combined cortical-only and subcortical-only group. gyri, and right fusiform atrophy compared to those with- There were no statistically significant differences in clini- out visual hallucinations (n = 15; 1 MM1, 3 MV1, 4 cal symptoms, protein 14-3-3, EEG, clinical scales, or MM2, 2 MV2, 1 VV2, and 4 MV without typing avail- biomarkers based on the PRNP codon 129 polymorphism able, one subject excluded because no Barthel score was (data not shown). available to adjust for disease severity – see Methods – The same analyses were performed for the 22 partici- VBM processing and analysis). Participants with more pants included in the VBM analysis and showed relatively severe dementia (MMSE score ≤ 20, n = 12; 1 MM1, 4 similar results (Supplementary Material). In comparing MV1, 3 MM2, 2 MV2, 1 VV2, and 1 MV without typing the subgroup with VBM analysis to the subgroup without available) had significant regional volume loss in several VBM analysis, the VBM-excluded participants had shorter areas including bilateral mesial and inferior frontal, bilat- disease duration, lower MMSE scores, lower Barthel eral insula, thalamus, cerebellum, left orbitofrontal, and scores, higher CSF t-tau concentrations, higher prevalence right mesial temporal regions compared to those with less of positive CSF protein 14-3-3, and lower prevalence of severe dementia (MMSE score > 20, n = 7; 1 MM1, 1 ataxia (Table 1 and Supplementary Material). Most of MM2, 2 MV2, 1 VV2, and 2 MV without typing avail- these findings suggest that the VBM-excluded group had able). No effects were observed when MMSE was used as poorer quality MRIs (unable to be used for VBM analy- a continuous variable; comparing these groups separately sis) due to a higher degree of clinical impairment or dis- against Controls, however, showed significant differences. ease severity. Interestingly, the VBM group (less No volumetric differences were found when comparing impaired) appeared to be much more likely to have only codon 129 MM2 to MV2 (the sample size was small, cortical ribboning, whereas the impaired had subcortical however: 4 MM2 and 5 MV2) or MM to MV genotypes. involvement, either with or without cortical ribboning. Compared to Controls, Fast-progressors (n = 5; 3 MV1 This is consistent with our clinical experience. and 2 VV2) had atrophy in bilateral mesial and lateral frontal, bilateral precuneal, middle temporal, postcentral, and occipitoparietal regions as well as bilateral thalamic Atrophy patterns and brain-phenotype atrophy. Slow-progressors (n = 9; 4 MM2 and 5 MV2) relationships compared to Controls showed a similar pattern of atro- VBM analysis results of the comparisons between 22 sCJD phy as the Fast-progressors compared to Controls, but participants and 26 Controls are shown in 3D brain views did not have atrophy of cerebellum, occipital, and most and in axial views in Figure 1A,F. VBM analysis revealed temporal regions (Fig. 1D, E, I, J). Surprisingly, compar- gray matter reduction in sCJD compared to Controls in ing Fast-progressors directly to Slow-progressors showed multiple cortical regions, including the bilateral frontopo- no significant differences in regional atrophy (not shown). lar, mesial and inferior frontal lobes, mesial and lateral Participants with ataxia (n = 10; 1 MM1, 3 MV1, 2 parietal lobes, bilateral lateral temporal, left mesial tempo- MM2, 2 MV2, and 2 VV2) displayed a trend of posterior ral regions, bilateral insulae, and inferior posterior right (left more than right) cerebellar atrophy compared to occipital regions, as well as bilateral thalami (permutations- participants without ataxia (n = 10; 1 MM1, 1 MV1, 2 based correction for multiple comparisons P < 0.05). Some MM2, 2 MV2, and 4 MV without available type) (uncor- areas of relative sparing of atrophy compared to Controls rected, p < 0.001; not shown), but did not survive multi- included the motor, much of the occipital, and some dor- ple correction. There were no statistically significant solateral and prefrontal cortices as well as bilateral striatum differences when comparing participants with myoclonus and globus pallidi. Of note, no regions showed higher vol- to participants without myoclonus. There were also no umes in sCJD compared to Controls. Interestingly, the statistically significant differences when comparing partici- regions with atrophy in sCJD included all areas of the func- pants with higher UPDRS score (n = 11; 1 MM1, 2 MV1, tionally defined default mode and salience, possibly some 2 MV2, 2 VV2, 1 MM2, 1 MV1-2, and 2 MV without ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1189 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. Figure 1. Regional gray matter atrophy in sporadic Creutzfeldt-Jakob disease. A-E show a 3D rendering, whereas F-J show the same data rendered in axial view. All results shown in color passed permutations-based correction for multiple comparisons p < 0.05. Orientation is neurological (e.g., left side is left brain). Redder colors (A-E) signify higher level of significance (higher t-stat). For F-J (axial views), color bar represents various t-scores. Only regions of t-scores > 2 (i.e., > 2 SD away from the mean) are shown; blue regions color have significantly greater atrophy than the comparison group. Clusters with volume reductions in sCJD compared to Controls were found in the bilateral frontopolar, mesial and inferior frontal, mesial and lateral parietal, bilateral lateral temporal and left mesial temporal, and inferior posterior right occipital regions (A, F). sCJD participants with visual hallucinations had significant volume loss in the bilateral thalami, medial orbitofrontal, rectus gyri, and right fusiform compared to participants without visual hallucinations (B, G). The sCJD group with more severe cognitive impairment (based on dichotomization by the median MMSE score) showed volume reduction in the bilateral mesial and inferior frontal, cerebellum, left orbitofrontal, and right mesial temporal regions compared to the group with less cognitive impairment (C, H). Volume differences between Slow-progressors (based on molecular classification subtype) and Controls were present in the bilateral mesial and lateral frontal, bilateral precuneal, middle temporal, postcentral, and occipitoparietal regions (Slow-progressors = 4 MM2, 5 MV2) (D, J). Volume differences in Fast-progressors, based on molecular classification, and Controls were found in bilateral mesial and lateral frontal, bilateral precuneal, middle temporal, postcentral, and occipitoparietal regions as well as occipital and temporal (Fast-progressors = 3 MV1, 2 VV2) (E, I). No volume differences were found between comparison of Fast-progressors versus Slow-progressors (not shown; see text). 1190 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease typing available) to those with lower UPDRS score (n = allows connectivity between hemispheres. The whole brain 10; 1 MV1, 3 MM2, 3 MV2, 2 MV1-2, and 1 MV without model showed that volumetric changes in the precuneus typing available) or when comparing participants with influenced the volumetric changes in the anterior cingulate higher NPI score (n = 9; 1 MM1, 2 MV2, 3 MM2, 1 cortex, angular gyrus, and mesial and lateral temporal cortex VV2, 1 MV1-2, and 1 MV without typing available) to regions, whereas the connectivity was commensurate those with lower NPI score (n = 8; 1 MV1, 1 MM2, 3 between the precuneus and the thalamus. Furthermore, stri- MV2, 1 MM1-2, 1 MV1-2, and 1 MV without typing atal atrophy appeared to cause precuneus atrophy but not available). Some comparisons did not include all 22 sCJD vice versa (as RMSEA > .05 and GFI < .90, not shown). participants if subjects were missing a variable being con- Thus, coupled (connected) regions within the network chan- trolled for or relevant to that analysis (e.g., Barthel or ged their volumes together (i.e., high covariance), and these sCJD molecular classification). changes were directional within the network. Thus, there was SEM effective connectivity present in sCJD in a model including the DMN plus the striatum and thalamus, with Modeling atrophy covariance the precuneus being the main influencer of directionality. Based on our VBM analysis, the atrophy patterns in sCJD involved all areas of the default mode network (DMN), Discussion except minimal hippocampal involvement, but also involved other regions outside of the DMN. The DMN is a Summary of major findings topographically distant but highly connected network of functionally connected hubs.34 Although we found high This high-resolution 3T MRI VBM study of 22 sCJD cases correlations of the volumetric data between these regions, provides evidence that regional, but not global, atrophy is we could not deduce whether these volumetric changes are a feature of sCJD. Selective regional atrophy was observed happening simultaneously or if one region is influencing in a composite of cortical and deep nuclei regions, the atrophy in another (Supplementary Table 1 shows the including DMN hubs,34 salience network (SN) hubs,37 the Pearson correlation between some of the brain regions). thalami as well as other regions. Specifically, selective Because of this and that many of these same areas are pref- atrophy in sCJD (compared with Controls) was predomi- erentially affected with abnormal diffusion in sCJD,11,17,38 nant in multimodal association regions, including the we investigated whether a structural equation model would bilateral precunei, bilateral lateral parietal, bilateral mesial fit the data and, if so, whether these simultaneously impli- prefrontal cortices, and bilateral lateral temporal (all cated atrophic regions are affected due to intrinsic network DMN regions, with the hippocampus as the only DMN vulnerability independent of connectivity or if one brain region being only minimally involved), and bilateral ante- region could be mathematically, although not necessarily rior cingulate and insula (SN regions), in addition to biologically, influencing atrophy in another region. To test bilateral lateral and inferior prefrontal cortices, and bilat- this hypothesis, we evaluated volumetric interactions in eral thalami (Fig. 1). Notably, in sCJD there was relative brain connectivity network models using SEM. Models sparing of atrophy in the dorsolateral prefrontal cortex with optimal data fit (RMSEA < .05 and GFI > .90) should (part of the executive control network) and no atrophy in fit only in CJD but not Controls, and the beta coefficient the primary motor, sensory, and visual cortices, as well as from region A to region B in the SEM model would mathe- the striatum and cerebellum. We explored whether certain matically indicate a directional relationship.14 Tests of data clinical variables and symptomatology were associated fitness were done first in four functional networks: 1) with regions of atrophy. Both Slow- and Fast-progressors DMN34; 2) motor36; 3) executive control37; and 4) salience (defined based on sCJD molecular classification)5,29 had networks.37 None of these SEM models met the significance significant atrophy compared to Controls but, to our sur- and goodness-of-fit thresholds in either sCJD or Controls prise, were not statistically significantly different from (RMSEA > 0.05 and GFI < 0.90). A model, however, based each other. The sCJD subgroup with visual hallucinations on the DMN model plus the addition of deep nuclei had significantly greater atrophy in the right fusiform regions (striatum and thalamus) – which commonly show gyrus, right more than left thalamus, and bilateral orbito- restricted diffusion in sCJD – resulted in optimal data fit in frontal areas compared to the subgroup without visual sCJD but not the Controls (Fig. 2). For sCJD and Control hallucinations. Although MMSE as a continuous variable groups, left, right, and whole brain volumetric data were did not correlate with atrophy, participants with lower tested separately in the model, with only sCJD showing a MMSE scores (dichotomized by median MMSE of 20) significant effect (Fig. 2) but not the Controls (all had more frontal, temporal, insula, thalamus, and cerebel- RMSEA > 0.05 and all GFI < 0.90). The most meaningful lum atrophy compared with participants with higher model clinically, however, is the whole brain model, as it MMSE scores. Finally, SEM of volumetric brain data ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1191 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. Figure 2. Sporadic Creutzfeldt-Jakob disease selectively changes the effective connectivity between specific cortical and subcortical brain regions that overlap with the default mode network nodes. The figure shows the models of brain effective connectivity when brain volume data are tested in a network of cortical and subcortical regions usually noted by the authors to be commonly affected clinically on diffusion imaging in sCJD, specifically the default mode network plus the striatum and thalamus. Two key take-away points from this figure are (1) the model fit in sCJD but not in Controls, and (2) that the precuneus (PrC) seems to play a central role in influencing volumetric changes in other regions. In the following text, we explain the SEM model and the meaning of the arrows from a mathematical standpoint. The graphs represent anatomical nodes in boxes connected by paths of trophic influence (arrows) that determine the regional volumetric influence on the target nodes. The effective connectivity (i.e., direction of the trophic effect) is represented by the arrow direction. Connectivity strength (i.e., strength of an effect) is represented by path coefficients (i.e., beta coefficient) displayed by the number over each arrow, with higher numbers meaning stronger tropic influence. The thickness of the arrow is a visual representation of the strength of the correlation and the dashed lines representing a negative correlation. Positive values indicate induction of atrophy in the direction of the arrow, whereas negative values indicate induction of increased volume. Goodness-of-fit statistics (GFIs) > .900 are considered significant with the p value equivalent shown by root mean square error of approximation (RMSEA)—only the models in sCJD, and none of the models in Controls, were significant (significant results are indicated with an *). In the whole brain and the right hemisphere models, and partially in the left hemisphere model, the precuneus exerts a large and disproportionate effect on the anterior cingulate (ACC), angular gyrus (AG), and temporal lobe (Temp). For example, in the whole brain model, one-unit change in the precuneus volume results in 1.62, .76, and 1.28 points change in the ACC, AG, and Temp, respectively. Conversely, changes in the ACC, AG, and Temp volumes results in .23, .17, and .21 unit change, respectively, in the precuneus. Interestingly, compared to the tropic influence of the precuneus, the effects were more balanced between the thalamus (Thal) and the precuneus and were unidirectional from the striatum (Str) to the precuneus. Models that included bidirectional effect between the precuneus and the Str did not meet the goodness-of-fit and the statistical significance parameters. This suggests that the striatum influenced atrophy of the precuneus, but not the reverse. L = left hemisphere, R = right hemisphere, C = combined or bilateral structure. revealed significant covariance and trophic (directional) which involves the same multimodal association regions effects among key functional nodes of the DMN (pre- including the cingulate, precuneus, angular gyrus, and cuneus, anterior cingulate cortex, angular gyrus, and superior and middle frontal gyri, with relative sparing of mesial and lateral temporal cortex)34 as well as the thala- primary motor and sensory cortices.11,17,33 We might mus and striatum in sCJD but not in Controls. This expect that areas first involved in sCJD would be the model of effective connectivity showed trophic influence areas to show greatest atrophy later on in the disease. We of subcortical regions on the precuneus and suggests a found five case reports of neurologically normal persons central role for the precuneus in driving volumetric who had brain MRIs for reasons unrelated to sCJD (e.g., changes in other cortical regions. carotid bulb tumor, studies assessing utility of annual MRI for standard of clinical care, etc) and who became symptomatic with sCJD between 3 and 14 months later. Atrophy in sCJD (vs. Controls) and Overall, these cases showed DWI abnormalities (reduced comparison to literature diffusion) in the presymptomatic phase of disease in asso- The pattern of selective brain volume reduction in sCJD ciation cortices including the bilateral temporo-parietal- overlaps with its typical DWI-restricted diffusion pattern, occipital junction, lateral parietal, precunei, and mesial 1192 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease frontal regions,39–43 similar to regions where we detected in Fast-progressors compared to Controls and Slow-pro- atrophy with VBM. Furthermore, our findings are consis- gressors compared to Controls was relatively similar tent with studies in human genetic prion disease showing (Fig. 1). On visual inspection of the VBM data, Fast-pro- atrophy in cortical association regions.10,11 Alner et al. gressors had more atrophic areas than Slow-progressors, 2011, using 1.5T MRI and FreeSurfer imaging analysis, including cerebellum, occipital, and temporal regions. described cortical thickness reduction in 10 patients with These differences, however, did not reach statistical signif- a genetic prion disease (gPrD; 6-OPRI PRNP mutation), icance, which might be due to the relatively small sample predominantly in the precuneus, supramarginal, parietal size. Nevertheless, these findings suggest that brain atro- lobule, and lingual regions.10 In a follow-up study also phy is present, even in Fast-progressors, and although with 1.5T MRI, but using SPM8 VBM analysis, De Vita usually not detectable by visual assessment, it can be et al. 2013 described precuneus, perisylvian, lingula, and quantified using VBM. To our knowledge, VBM analysis basal ganglia volume reduction in nine symptomatic of the Fast- versus Slow-progressors in sCJD based on patients with the same 6-OPRI PRNP mutation.11 Simi- molecular classification has not been reported previously. larly, using 3T VBM analysis in 30 symptomatic predomi- We also identified a number of brain-behavior relation- nantly gPrD subjects (10% sCJD), Caine et al. described ships with VBM in sCJD. Participants with visual halluci- volume reductions in frontal and parietal gray matter vol- nations showed more significant right fusiform, right umes which correlated with predominant frontoparietal more than left thalamus, and bilateral medial orbitofron- dysfunction on neurocognitive testing.44 These three stud- tal volume loss compared with participants without visual ies all showed atrophy in regions for which we also found hallucinations (Fig. 1). The involvement of the fusiform atrophy in our sCJD cohort (Fig. 1). In a study using 3T gyrus is consistent with previous studies showing atrophy VBM analysis, Grau-Rivera et al. 2015 found significant on VBM and functional (fMRI) abnormalities in visual gray matter loss in 15 CJD cases (13 sCJD and 2 genetic, association areas in patients with neurodegenerative con- E200K) specifically in the bilateral thalami, putamen, fusi- ditions such as Parkinson’s disease.47 Similarly, orbito- form gyrus, cerebellum, and left perirolandic cortex.45 frontal cortex involvement in Parkinson’s disease Although we had similar thalamic and fusiform findings, dementia patients experiencing visual hallucinations has we did not observe basal ganglia (striatum and globus been reported in multiple studies, including VBM,48,49 pallidus), perirolandic, or cerebellar volume loss in sCJD selective serotonin 2A receptor ligand F-18 PET,50 and vs. Controls. It is interesting that we did not find striatal diffusion tensor imaging studies.51 Moreover, fluo- volume loss despite visible striatal diffusion reductions rodeoxyglucose-PET (FDG-PET) hypometabolism in the present in 60% of our sCJD participants (Table 1). This orbitofrontal cortex has been found in patients with parallels findings in a study by Seror et al. 2010 with 12 schizophrenia who have visual hallucinations.52 Further- patients with E200K genetic prion disease who had stri- more, recent studies suggest a role for thalamic-DMN atal-restricted diffusion but did not show quantifiable vol- decoupling as the mechanism for visual hallucinations in ume reduction.46 In our experience, for most sCJD cases, synuclinopathies.53 striatal involvement on DWI usually occurs after extensive The sCJD group with more severe cognitive impair- cortical ribboning appears (except in MV2 and VV2 cases ment, based on dichotomization of MMSE scores had which usually have isolated deep nuclei involvement). scattered volume loss in the bilateral mesial orbitofrontal, The fact that we found overlapping but slightly different inferior frontal, posterior central, bilateral middle tempo- cortical areas involved than these other studies might be ral, and bilateral medial precuneus gyri, in addition to due to other studies using mostly genetic cases and few more confluent volume loss in the bilateral insula, thala- sporadic cases or to smaller sample sizes in Navid et al.7 mus, and cerebellum, even after adjusting for the time (n = 11) and Grau-Rivera et al.45 (n = 13) compared point of MRI in the disease course (ratio of disease dura- with 22 sCJD subjects in our study. tion at the time of MRI over total disease duration). When we corrected for Barthel, however, this effect went away, probably because Barthel and MMSE were corre- Atrophy in sCJD subgroups and comparison lated. This atrophy pattern is consistent with the role of to literature the cerebellar-thalamic-cortical connections in a wide As noted above, we predicted that Fast-progressors might range of cognitive functions.54 not have sufficient time to develop detectable atrophy Although it did not survive correction for multiple compared to Slow-progressors (Fast and Slow based on comparisons, sCJD participants with ataxia compared sCJD molecular classification). Contrary to our predic- with those without ataxia showed a trend toward greater tion, however, each of these subgroups showed significant left posterior cerebellar atrophy (not shown), which is atrophy when compared to Controls. The atrophy pattern consistent with a previous finding of a correlation ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1193 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. between cerebellar atrophy and ataxia in E200K genetic that involved DMN regions plus subcortical regions con- prion disease.6 Perhaps a larger sample size is needed to sisting of the thalamus and striatum fit in the sCJD group show this effect in sCJD. but not in the Controls (Fig. 2). In both sCJD and Con- trols, alternative models involving either the DMN with- out the addition of subcortical structures, as well as the Cortico-subcortical regions as the epicenter executive, salience, or motor networks did not fit the in sCJD model. In the cortico-subcortical network model that met From a brain network perspective, our VBM data support the strict significance thresholds of SEM (GFIs>.90 and the idea that frontoparietal and temporal regions, includ- RMSEA<.05), the strongest path effect was from the pre- ing but not limited these areas within the DMN,55 as well cuneus to the anterior cingulate cortex, mesial and lateral as several subcortical structures, are at the epicenter of temporal, and angular gyrus regions in the left, right, and neurodegenerative changes in sCJD. The selective volume combined hemispheres of the sCJD group but not in the reduction we observed in sCJD within most of the DMN Controls. These SEM findings potentially point to a cen- (only minimal hippocampal atrophy) and other regions tral role for the precuneus in sCJD connectivity-mediated may be explained by both distinctive regional neu- neurodegeneration. This finding is consistent with our rometabolic profiles and network connectivity properties clinical experience and literature suggesting that the pre- in this disease. For example, parietal regions such as the cuneus and the adjacent posterior cingulate cortex are precuneus in particular, but also certain occipital and very commonly involved by visual assessment of the temporal regions, display the highest state of oxidative DWI/ADC maps and quantitative mean diffusivity mea- neuronal metabolism in the brain.56,57 In such regions surement in sCJD.17,33 The precuneus exhibited trophic with high baseline metabolism, oxidative stress may play influences on other cortical regions, whereas its trophic a crucial role in prion protein misfolding.58 These results influence on the thalamus was bidirectionally equivalent, are in line with an FDG-PET study of sCJD showing and the striatum exhibited a unidirectional trophic influ- hypometabolism to be common particularly in parietal ence on the precuneus. We speculate that the bidirec- and temporal regions.59 Networks, such as the DMN, tional effects on a group-wise level could be due to the often feature a number of heteromodal association cor- high connectivity between the DMN hubs – once one tices that operate as hub nodes. Such networks are resis- region is affected it begins influencing atrophy in other tant to random node failure but become vulnerable when highly connected regions. Thus, regardless of where a hub failure occurs, thus facilitating transsynaptic degener- patient’s disease begins, involvement eventually converges ation or other forms of network pathological spread.60 on this highly connected network. Some bidirectionality, The biological vulnerability of these hubs is supported by however, was positive in one direction and negative in its implication as a pathogenic substrate in various neu- another, meaning that atrophy in one region is influenc- rodegenerative diseases. For example, Alzheimer’s disease ing increased volume in another region (negative arrow (AD), which like prion disease involves protein misfold- in Fig. 2). We specultate that this might occur because ing, has prominent posterior DMN hypometabolism and astrocytosis, neuroinflammation, and/or possibly edema atrophy.61 Interestingly, the pattern of cortical atrophy in involved in neurodegeneration might produce volumetric sCJD, in part targeting DMN hubs, is reminiscent of the changes that can be interpreted by SEM as increased vol- involvment of the DMN in AD.61,62 This also is consistent umes within the sCJD group, which we simply were not with literature suggesting shared pathogenic mechanisms able to detect in our VBM analyses. and overlap between AD and sCJD.63-66 Furthermore, Despite the striatum not showing atrophy on VBM, the selective vulnerability of networks containing high-degree SEM showed a volumetric trophic effect of the striatum hubs is also characteristic of other neurodegenerative dis- on the precuneus in sCJD, which was not found in Con- orders such as behavioral variant frontotemporal demen- trols. We suspect that, despite the lack of detectable stri- tia, semantic dementia, progressive supranuclear palsy, atal changes on VBM in sCJD, the striatum likely plays and corticobasal syndrome.37,67,68 an important role on the network-level pathogenesis of VBM results implicating the involvement of cortical sCJD, along a continuum of diffusivity changes preceding and subcortical regions in sCJD (through atrophy) were atrophy. These findings point to the importance of sub- supported using SEM in connectivity models that were cortical regions, such as the striatum and thalamus, in the specifically tested for the presence of network-level pathophysiology of sCJD, as similar models including involvement in sCJD. The cortico-subcortical effective only cortical regions were not solvable and only models connectivity in sCJD can be interpreted as a common including both cortical and subcortical regions were. Fur- trophic fate (i.e., volume reduction) for involved cortical thermore, this SEM model provides mathematical evi- and subcortical regions. A structural connectivity model dence for the hypothesis that the selective atrophy pattern 1194 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease in sCJD is mediated by atrophy spread via anatomical only compared cases whom could easily be classified as connections, such as those used for the transsynaptic Fast- or Slow-progressors based on their sCJD molecular spread of prions, rather than selective atrophy due to classificaiton, excluding five cases (four with mixed prion intrinsic regional vulnerability alone, independent of con- types and one variably protease-sensitive prionopathy). nectivity and transsynaptic prion spread. In other words, This dichotomization of sCJD into Fast- vs. Slow-progres- if atrophy is mediated only by intrinsic regional vulnera- sors based on the mean disease duration of each of the bility, then we would expect high volumetric correlations six major molecular subgroups has limitations. For exam- in atrophied regions but without a directional relation- ple, even though on average a molecular classification ship as we found in our SEM analysis. subtype is rather homogenous, an individual patient We also identified several associations between various might be more heterogenous. Additionally, the clinical biomarkers and clinical outcomes in both the entire and biomarker data of the VBM-excluded participants cohort of 40 sCJD participants who had DWI and the suggest that these participants in general had more severe subcohort of 22 sCJD participants who also had sufficient disease, which likely interfered with MRI acquisition and quality T1 MRI volumetric analysis. sCJD participants led to VBM analysis of the relatively less severely affected with both cortical and subcortical DWI involvement had subjects. Lastly, although our cohort might appear to higher CSF biomarkers (t-tau and NSE) and more motor have longer survival than many other large published and behavioral dysfunction compared with the group sCJD cohorts, our 2-day research visit protocol attempts combining cortical-only and subcortical-only involve- to identify the very first symptoms, which we commonly ment. This is consistent with our clinical experience in find to be several weeks and usually months earlier than which we find that patients with both cortical and sub- the medical records or even the family had initially cortical DWI involvement tend to decline more rapidly, reported. This results in our patients’ disease durations particularly compared to those with cortical-only involve- appearing longer than those reported in national surveil- ment. Whole brain volumes inversely correlated with CSF lance cohorts that are largely based on retrospective exter- t-tau levels, which is generally consistent with the mod- nal record review.5,63,76,77 estly high sensitivity and specificity of an elevated t-tau in Our cohort in this study had extensive neurological, sCJD.69 Similarly, t-tau levels were inversely correlated PRNP codon 129 genotyping, and CSF biomarker charac- with total disease duration and MMSE scores (i.e., partic- terization and, to our knowledge, represents the first ipants with lower levels had longer disease durations and quantitative VBM study exclusively in sporadic prion dis- better cognitive performance), which parallels previously ease. The results of this study add a valuable dimension observed changes of t-tau with disease duration70-73 and to the neuroimaging characterization of sCJD and support cognitive function in prion disease.74 the value of unbiased quantification of brain volume as a This study has a number of limitations. Only cross-sec- potential clinical biomarker in sCJD. Trophic SEM analy- tional volumetric quantifications were performed, as we ses of VBM data suggest that atrophy in sCJD might not had too few cases with sufficient quality serial data. Lon- just be due to regional vulnerability, but that there is a gitudinal studies, which are often difficult to obtain in directionality of influence of atrophy in sCJD, which we sCJD, particularly of sufficient quality for volumetric hypothesize might be influenced by propagation of prions analysis, will be needed to better characterize atrophy along synaptic pathways. progression in sCJD. Although our cohort was relatively large for a single-site sCJD imaging cohort, our sample Acknowledgments size nevertheless is relatively small, limiting our ability to identify more interactions of clinical variables and volu- NIH/NIA R01 AG031189 (MDG), P50 AG023501(BLM, metric changes or to compare all six main sCJD molecu- HJR), T32 AG023481 (JCR), AG032289 (JHK), NIH/ lar classification subgroups. The fact that even with our NCRR UCSF-CTSI (UL1 RR024131), JCR is supported by limited sample size, however, we still found any associa- NIA-NIH: K23AG059888. Hellman Family Foundation, tions suggests the strength of our findings. Although our Michael J. Homer Family Fund (MDG), Larry L. Hillblom cohort with DWI (n = 40) included all molecular sub- Network 2018-A-006-NET (JHK), i-PFIS grant (IF15/ types and had a plurality of Fast-progressors (MM1, 00060), Fondo de Investigaciones Sanitarias, Instituto de MV1, and VV2) consistent with the general sCJD popula- Salud Carlos III (I.I-G). Global Brain Health Institute tion,29,75 Fast-progressors were underrepresented in the Atlantic Fellow for Equity in Brain Health (I.I-G). The subcohort included in the VBM analysis because of diffi- authors thank Drs. Stephen J. DeArmond and Henry San- culties obtaining sufficient quality volumetric scans due chez at UCSF and the National Prion Disease Pathology to motion artifact. This may represent sampling bias Surveillance Center (NPDSPC) at Case Western Reserve favoring cases with slower progression. Furthermore, we University for their assistance in pathological ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1195 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. confirmation of many of our sCJD cases. The authors also 3. Hayashi R, Hanyu N, Kuwabara T, Moriyama S. Serial express their gratitude to the patients and their loved computed tomographic and electroencephalographic studies ones for participating in our research program. in Creutzfeldt-Jakob disease. Acta Neurol Scand 1992;85 (3):161–165. https://doi.org/10.1111/j.1600-0404.1992.tb 04020.x. Conflict of Interest 4. Finkenstaedt M, Szudra A, Zerr I, et al. MR imaging of MDG receives/received research support on prion disease Creutzfeldt-Jakob disease. Radiology 1996;199(3):793–798. from the NIH/NIA (R01-AG031189; R56-AG055619; https://doi.org/10.1148/radiology.199.3.8638007. R01AG062562), the Michael J. Homer Family Fund and 5. Parchi P, Giese A, Capellari S, et al. Classification of sporadic Alliance Biosecure. He has consulted for Adept Field Con- Creutzfeldt-Jakob disease based on molecular and phenotypic sulting (Backbay consulting), Advanced Medical Inc., analysis of 300 subjects. Ann Neurol 1999;46(2):224–233. Anderson Boutwell Traylor, Acsel Health LLC, Best Doc- https://www.ncbi.nlm.nih.gov/pubmed/10443888. tors Inc., Blade Therapeutics, Biohaven Pharmaceuticals, 6. Cohen OS, Hoffmann C, Lee H, et al. MRI detection of Bioscience Pharma Partners, LLC (BPP), ClearView the cerebellar syndrome in Creutzfeldt-Jakob disease. HealthCare Partners, Grand Rounds Inc./Second Opinion Cerebellum 2009;8(3):373–381. https://doi.org/10.1007/ Inc., Gerson Lehrman Group (GLG) Inc., Guidepoint s12311-009-0106-8. 7. Navid J, Day GS, Strain J, et al. Structural signature of Global LLC, Market Plus, InThought Consulting, LifeSci sporadic Creutzfeldt-Jakob disease. Eur J Neurol 2019;26 Capital LLC, Maupin Cox Legoy, MEDACorp, Quest (8):1037–1043. https://doi.org/10.1111/ene.13930. Diagnostics, 3M Communications (Microvention Ter- 8. Eisenmenger L, Porter MC, Carswell CJ, et al. Evolution of umo), Smith & Hennessey LLC, TeleDoc Health Inc., and diffusion-weighted magnetic resonance imaging signal Trinity Partners LLC. He has received speaking honoraria abnormality in sporadic Creutzfeldt-Jakob disease with for various medical center lectures and from Oakstone histopathological correlation. JAMA Neurol 2016;73(1):76– Publishing. He has received past research support from 84. https://doi.org/10.1001/jamaneurol.2015.3159. CurePSP, the Tau Consortium, Quest Diagnostics, and 9. Lee H, Cohen OS, Rosenmann H, et al. Cerebral white NIH. Dr. Geschwind serves on the board of directors for matter disruption in Creutzfeldt-Jakob disease. AJNR Am J San Francisco Bay Area Physicians for Social Responsibil- Neuroradiol 2012;33(10):1945–1950. https://doi.org/10. ity and on the editorial board of Dementia & Neuropsy- 3174/ajnr.A3125. chologia. JCR is a site PI for clinical trials supported by 10. Alner K, Hyare H, Mead S, et al. Distinct neuropsychological Eli Lilly and receives support from NIH. IIG is supported profiles correspond to distribution of cortical thinning in by the Rio Hortega grant (CM17/00074) from “Acci on inherited prion disease caused by insertional mutation. J Estrategica en Salud 2013-2016” and the Global Brain Neurol Neurosurg Psychiatry 2012;83(1):109–114. https://doi. Health Institute. org/10.1136/jnnp-2011-300167. 11. De Vita E, Ridgway GR, Scahill RI, et al. Multiparameter Author Contributions MR imaging in the 6-OPRI variant of inherited prion disease. AJNR Am J Neuroradiol 2013;34(9):1723–1730. KY and JR drafted the manuscript and contributed data https://doi.org/10.3174/ajnr.A3504. acquisition, analysis, and interpretation. AW, GY, MP, 12. Stopschinski BE, Diamond MI. The prion model for GT, EC, MM, IG, and YC contributed data acquisition, progression and diversity of neurodegenerative diseases. analysis and interpretation, and critical review of the Lancet Neurol 2017;16(4):323–332. https://doi.org/10.1016/ manuscript. JHK contributed data acquisition, BLM con- S1474-4422(17)30037-6. tributed study concept and design, HR and MG con- 13. Krause JB, Taylor JG, Schmidt D, et al. Imaging and tributed study concept and design, data interpretation, neural modelling in episodic and working memory and critical review of the manuscript. processes. Neural Netw 2000;13(8–9):847–859. https://doi. org/10.1016/s0893-6080(00)00068-x. References 14. de Marco G, Vrignaud P, Destrieux C, et al. Principle of 1. Geschwind MD. Prion diseases. Continuum (Minneap structural equation modeling for exploring functional Minn) 2016;21(16 Neuroinfectious Disease); 1612–38. interactivity within a putative network of interconnected https://doi.org/10.1212/CON.0000000000000251. brain areas. Magn Reson Imaging 2009;27(1):1–12. https:// 2. Kovanen J, Erkinjuntti T, Iivanainen M, et al. Cerebral doi.org/10.1016/j.mri.2008.05.003. MR and CT imaging in Creutzfeldt-Jakob disease. J 15. Geschwind MD, Haman A, Miller BL. Rapidly progressive Comput Assist Tomogr 1985;9(1):125–128. https://doi.org/ dementia. Neurol Clin 2007;25(3):783–807. https://doi.org/ 10.1097/00004728-198501000-00023. 10.1016/j.ncl.2007.04.001. 1196 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease 16. Kretzschmar HA, Ironside JW, DeArmond SJ, Tateishi J. 29. Collins SJ, Sanchez-Juan P, Masters CL, et al. Diagnostic criteria for sporadic Creutzfeldt-Jakob disease. Determinants of diagnostic investigation sensitivities Arch Neurol 1996;53(9):913–920. http://www.ncbi.nlm.nih. across the clinical spectrum of sporadic Creutzfeldt-Jakob gov/htbin-post/Entrez/query?db=m&form=6&dopt= disease. Brain 2006;129(Pt 9):2278–2287. http://www.ncbi. r&uid=8815857. nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed 17. Vitali P, Maccagnano E, Caverzasi E, et al. Diffusion- &dopt=Citation&list_uids=16816392. weighted MRI hyperintensity patterns differentiate CJD 30. Puoti G, Bizzi A, Forloni G, et al. Sporadic human prion from other rapid dementias. Neurology 2011;76(20):1711– diseases: molecular insights and diagnosis. Lancet Neurol 1719. https://doi.org/10.1212/WNL.0b013e31821a4439. 2012;11(7):618–628. https://doi.org/10.1016/S1474-4422 18. Folstein MF, Folstein SE, McHugh PR. “Mini-mental (12)70063-7. state”. A practical method for grading the cognitive state 31. Gambetti P, Puoti G, Zou WQ. Variably protease-sensitive of patients for the clinician. J Psychiatr Res 1975;12 prionopathy: a novel disease of the prion protein. J Mol (3):189–198. https://doi.org/10.1016/0022-3956(75)90026- Neurosci 2011;45(3):422–424. https://doi.org/10.1007/ 6. s12031-011-9543-1. 19. Mahoney FI, Barthel DW. Functional evaluation: the 32. Wright S. The method of path coefficients. Ann Math Stat barther index. Md State Med J 1965;14:61–65. https:// 1934;5(3):161–215. https://doi.org/10.1214/aoms/ www.ncbi.nlm.nih.gov/pubmed/14258950. 1177732676. 20. Cummings JL. The Neuropsychiatric Inventory: assessing 33. Caverzasi E, Mandelli ML, DeArmond SJ, et al. White psychopathology in dementia patients. Neurology 1997;48 matter involvement in sporadic Creutzfeldt-Jakob disease. (5 Suppl 6):S10–S16. https://doi.org/10.1212/wnl.48.5_ Brain 2014;137(Pt 12):3339–3354. https://doi.org/10.1093/ suppl_6.10s. brain/awu298. 21. Tartaglia MC, Johnson DY, Thai JN, et al. Clinical overlap 34. Buckner RL, Andrews-Hanna JR, Schacter DL. The brain’s between Jakob-Creutzfeldt disease and Lewy body disease. default network: Anatomy, function, and relevance to Can J Neurol Sci 2012;39(3):304–310. https://doi.org/10. disease. Ann N Y Acad Sci 2008;1124:1–38. https://doi.org/ 1017/s0317167100013421. 10.1196/annals.1440.011. 22. Zou WQ, Puoti G, Xiao X, et al. Variably protease- 35. Collins DL, Zijdenbos AP, Kollokian V, et al. Design and sensitive prionopathy: a new sporadic disease of the prion construction of a realistic digital brain phantom. IEEE protein. Ann Neurol. 2010;68(2):162–172. http://www.ncb Trans Med Imaging 1998;17(3):463–468. https://doi.org/ i.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed 10.1109/42.712135. &dopt=Citation&list_uids=20695009. 36. DeLong MR, Wichmann T. Basal ganglia circuits as targets 23. Zou WQ, Gambetti P, Xiao X, et al. Prions in variably for neuromodulation in Parkinson disease. JAMA Neurol protease-sensitive prionopathy: an update. Pathogens 2015;72(11):1354–1360. https://doi.org/10.1001/jamane 2013;2(3):457–471. https://doi.org/10.3390/pathoge urol.2015.2397. ns2030457. 37. Seeley WW, Menon V, Schatzberg AF, et al. Dissociable 24. Gambetti P, Cali I, Notari S, et al. Molecular biology and intrinsic connectivity networks for salience processing and pathology of prion strains in sporadic human prion executive control. J Neurosci 2007;27(9):2349–2356. diseases. Acta Neuropathol 2011;121(1):79–90. https://doi. https://doi.org/10.1523/JNEUROSCI.5587-06.2007. org/10.1007/s00401-010-0761-3. 38. Caverzasi E, Henry RG, Vitali P, et al. Application of 25. Ashburner J, Friston KJ. Diffeomorphic registration using quantitative DTI metrics in sporadic CJD. NeuroImage geodesic shooting and Gauss-Newton optimisation. Clin 2014;4:426–435. https://doi.org/10.1016/j.nicl.2014.01. NeuroImage 2011;55(3):954–967. https://doi.org/10.1016/j. 011. neuroimage.2010.12.049. 39. Zanusso G, Camporese G, Ferrari S, et al. Long-term 26. Jenkinson M, Beckmann CF, Behrens TEJ, et al. Fsl. preclinical magnetic resonance imaging alterations in Neuroimage 2012;62(2):782–790. https://doi.org/10.1016/j. sporadic Creutzfeldt-Jakob disease. Ann Neurol 2016;80 neuroimage.2011.09.015. (4):629–632. https://doi.org/10.1002/ana.24757. 27. Winkler AM, Ridgway GR, Webster MA, et al. 40. Iwasaki Y, Mori K, Ito M, et al. An autopsied case of Permutation inference for the general linear model. MM1 + MM2-cortical with thalamic-type sporadic NeuroImage 2014;92:381–397. https://doi.org/10.1016/j.ne Creutzfeldt-Jakob disease presenting with hyperintensities uroimage.2014.01.060. on diffusion-weighted MRI before clinical onset. 28. Fonov VS, Evans AC, McKinstry RC, et al. Unbiased Neuropathology 2017;37(1):78–85. https://doi.org/10.1111/ nonlinear average age-appropriate brain templates from neup.12327. birth to adulthood. NeuroImage 2009;47:S102. https://doi. 41. Satoh K, Nakaoke R, Nishiura Y, et al. Early detection of org/10.1016/s1053-8119(09)70884-5. sporadic CJD by diffusion-weighted MRI before the onset ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1197 Selective Atrophy in Creutzfeldt-Jakob disease K. Younes et al. of symptoms. J Neurol Neurosurg Psychiatry 2011;82 thalamic dysfunction. Mov Disord. 2019;34(8), 1100–1111. (8):942–943. https://doi.org/10.1136/jnnp.2008.155242. https://doi.org/10.1002/mds.27781. 42. Maeda K, Sugihara Y, Shiraishi T, et al. Cortical 54. Schmahmann J. The cerebellar cognitive affective hyperintensity on diffusion-weighted images as the syndrome. Brain 1998;121(4):561–579. https://doi.org/10. presymptomatic marker of sporadic creutzfeldt-Jakob 1093/brain/121.4.561. disease. Intern Med 2019;58(5):727–729. https://doi.org/10. 55. Buckner RL, Andrews-Hanna JR, Schacter DL. The brain’s 2169/internalmedicine.1155-18. default network: anatomy, function, and relevance to 43. Verde F, Ticozzi N, Messina S, et al. MRI abnormalities disease. Ann N Y Acad Sci 2008;1124:1–38. https://doi.org/ found 1 year prior to symptom onset in a case of 10.1196/annals.1440.011. Creutzfeldt-Jakob disease. J Neurol 2016;263(3):597–599. 56. Li Y, Rinne JO, Mosconi L, et al. Regional analysis of https://doi.org/10.1007/s00415-016-8022-6. FDG and PIB-PET images in normal aging, mild cognitive 44. Caine D, Tinelli RJ, Hyare H, et al. The cognitive profile impairment, and Alzheimer’s disease. Eur J Nucl Med Mol of prion disease: a prospective clinical and imaging study. Imaging 2008;35(12):2169–2181. https://doi.org/10.1007/ Ann Clin Transl Neurol 2015;2(5):548–558. https://doi. s00259-008-0833-y. org/10.1002/acn3.195. 57. Shamchi SP, Khosravi M, Taghvaei R, et al. Normal 45. Grau-Rivera O, Calvo A, Bargall o N, et al. Quantitative patterns of regional brain18F-FDG uptake in Normal magnetic resonance abnormalities in creutzfeldt-jakob aging. Hell J Nucl Med 2018;21(3):175–180. https://doi. disease and fatal insomnia. J Alzheimer’s Dis 2016;55 org/10.1967/s002449910902. (1):431–443. https://doi.org/10.3233/JAD-160750. 58. Yuan F, Yang L, Zhang Z, et al. Cellular prion protein 46. Seror I, Lee H, Cohen OS, et al. Putaminal volume and (PrPC) of the neuron cell transformed to a PK-resistant diffusion in early familial Creutzfeldt-Jakob disease. J protein under oxidative stress, comprising main Neurol Sci 2010;288(1–2):129–134. https://doi.org/10.1016/ mitochondrial damage in prion diseases. J Mol Neurosci j.jns.2009.09.019. 2013;51(1):219–224. https://doi.org/10.1007/s12031-013- 47. Ibarretxe-Bilbao N, Junque C, Marti MJ, Tolosa E. 0008-6. Cerebral basis of visual hallucinations in Parkinson’s 59. Henkel K, Zerr I, Hertel A, et al. Positron emission disease: Structural and functional MRI studies. J Neurol tomography with [(18)F]FDG in the diagnosis of Sci 2011;310(1–2):79–81. https://doi.org/10.1016/j.jns.2011. Creutzfeldt-Jakob disease (CJD). J Neurol 2002;249 06.019. (6):699–705. https://doi.org/10.1007/s00415-002-0695-3. 48. Sanchez-Castaneda C, Rene R, Ramirez-Ruiz B, et al. 60. Albert R, Jeong H, Barabasi AL. Error and attack tolerance Frontal and associative visual areas related to visual of complex networks. Struct Dyn Networks hallucinations in dementia with lewy bodies and 2011;9781400841:503–506. https://doi.org/10.1515/ Parkinson’s disease with dementia. Mov Disord 2010;25 9781400841356.503. (5):615–622. https://doi.org/10.1002/mds.22873. 61. Zhou J, Seeley WW. Network dysfunction in Alzheimer’s 49. Rollins CPE, Garrison JR, Simons JS, et al. Meta-analytic disease and frontotemporal dementia: implications for evidence for the plurality of mechanisms in psychiatry. Biol Psychiatry 2014;75(7):565–573. https://doi. transdiagnostic structural MRI studies of hallucination org/10.1016/j.biopsych.2014.01.020. status. EClinicalMedicine 2019;8:57–71. https://doi.org/10. 62. Greicius MD, Srivastava G, Reiss AL, Menon V. Default- 1016/j.eclinm.2019.01.012. mode network activity distinguishes Alzheimer’s disease 50. Ballanger B, Strafella AP, van Eimeren T, et al. Serotonin from healthy aging: evidence from functional MRI. Proc 2A receptors and visual hallucinations in Parkinson Natl Acad Sci U S A. 2004;101(13):4637–42.http://www. disease. Arch Neurol 2010;67(4):416–421. https://doi.org/ ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=Pub 10.1001/archneurol.2010.35. Med&dopt=Citation&list_uids=15070770. 51. Ohtani T, Bouix S, Hosokawa T, et al. Abnormalities in 63. Prusiner SB. Cell biology. A unifying role for prions in white matter connections between orbitofrontal cortex and neurodegenerative diseases. Science (80-) 2012;336 anterior cingulate cortex and their associations with (6088):1511–1513. https://doi.org/10.1126/science.1222951. negative symptoms in schizophrenia: a DTI study. 64. Jucker M, Walker LC. Self-propagation of pathogenic Schizophr Res 2014;157(1–3):190–197. https://doi.org/10. protein aggregates in neurodegenerative diseases. Nature 1016/j.schres.2014.05.016. 2013;501(7465):45–51. https://doi.org/10.1038/nature 52. Silbersweig D. A., Stern E., Frith C., et al. A functional 12481. neuroanatomy of hallucinations in schizophrenia. Nature 65. Tousseyn T, Bajsarowicz K, Sanchez H, et al. Prion 1995;378(6553):176–179. https://doi.org/10.1038/378176a0. Disease Induces Alzheimer disease-like neuropathologic 53. Onofrj M, Espay AJ, Bonanni, L, et al. Hallucinations, changes. J Neuropathol Exp Neurol 2015;74(9):873–888. somatic-functional disorders of PD-DLB as expressions of https://doi.org/10.1097/NEN.0000000000000228. 1198 ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association K. Younes et al. Selective Atrophy in Creutzfeldt-Jakob disease 66. Jaunmuktane Z, Mead S, Ellis M, et al. Evidence for 77. Paterson RW, Torres-Chae CC, Kuo AL, et al. Differential human transmission of amyloid-beta pathology and diagnosis of Jakob-Creutzfeldt disease. Arch Neurol cerebral amyloid angiopathy. Nature 2015;525(7568):247– 2012;69(12):1578–1582. https://doi.org/10.1001/2013.jama 250. https://doi.org/10.1038/nature15369. neurol.79. 67. Gardner RC, Boxer AL, Trujillo A, et al. Intrinsic connectivity network disruption in progressive supranuclear palsy. Ann Neurol 2013;73(5):603–616. Supporting Information https://doi.org/10.1002/ana.23844. Additional supporting information may be found online 68. Fredericks CA, Brown JA, Deng J, et al. Intrinsic in the Supporting Information section at the end of the connectivity networks in posterior cortical atrophy: a role article. for the pulvinar? NeuroImage Clin 2019;21:101628. https://doi.org/10.1016/j.nicl.2018.101628. Supplementary Figure. The relationship between magnetic 69. Forner SA, Takada LT, Bettcher BM, et al. Comparing resonance imaging findings and clinical variables in spo- CSF biomarkers and brain MRI in the diagnosis of radic Creutzfeldt-Jakob disease. Regarding correlations sporadic Creutzfeldt-Jakob disease. Neurol Clin Pract with whole brain volume (A-C), CSF t-tau levels correlated 2015;5(2):116–125. https://doi.org/10.1212/CPJ. negatively with whole brain volume (A), whereas NSE did 0000000000000111. not correlate with whole brain volume (B), and MMSE cor- 70. Sanchez-Juan P, Sanchez-Valle R, Green A, et al. Influence related positively with whole brain volume (C). For D and of timing on CSF tests value for Creutzfeldt-Jakob disease E, we compared CSF t-tau and NSE in cases with both cor- diagnosis. J Neurol 2007;254(7):901–906. http://www.ncbi. tical- and subcortical-restricted diffusion on DWI (cortical nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed and subcortical: C-S) against the combination of cases with &dopt=Citation&list_uids=17385081. cortical-only involvement and subcortical-only involve- 71. Sanchez-Juan P, Green A, Ladogana A, et al. CSF tests in ment (combined cortical-only and subcortical-only: Co-S/ the differential diagnosis of Creutzfeldt-Jakob disease. C) (see text). This was because, our hypothesis was that Neurology 2006;67(4):637–643. http://www.ncbi.nlm.nih. those with cortical and subcortical diffusion would be more gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt= severely affected and more likely to have these biomarkers Citation&list_uids=16924018. positive. As we suspected, sCJD participants with a cortical 72. Skillback T, Rosen C, Asztely F, et al. Diagnostic and subcortical (C-S) pattern of restricted diffusion (gray) performance of cerebrospinal fluid total tau and showed elevated t-tau (C) and NSE (D) levels compared to phosphorylated tau in Creutzfeldt-Jakob disease: results the group of subjects of combined cortical-only and sub- from the Swedish Mortality Registry. JAMA Neurol cortical-only (Co-S/C) restricted diffusion (white box). 2014;71(4):476–483. https://doi.org/10.1001/jamaneurol. Supplementary Material. Biofluid, clinical, and neu- 2013.6455. roimaging biomarker results including CSF biomarker 73. Staffaroni AM, Kramer AO, Casey M, et al. Association of results and comparison of CSF results and other out- blood and cerebrospinal fluid tau level and other comes or variables among entire sCJD cohort, those with biomarkers with survival time in sporadic Creutzfeldt- and those without VBM. Jakob disease. JAMA Neurology. 2019;76(8), 969. https:// Supplementary Table. Pearson correlations between the doi.org/10.1001/jamaneurol.2019.1071. cortical and subcortical regions included in structural 74. Cohen OS, Chapman J, Korczyn AD, et al. CSF tau equation modeling (SEM). This table shows that high correlates with CJD disease severity and cognitive decline. correlations (≥0.5 strong; moderate 0.3-0.49; < 0.3 low) Acta Neurol Scand 2016;133(2):119–123. https://doi.org/ are present in sCJD in some of the brain regions included 10.1111/ane.12441. in the SEM model (significant correlations are shown in 75. Parchi P, Castellani R, Capellari S, et al. Molecular basis bold). This methodology, however, cannot discern if there of phenotypic variability in sporadic Creutzfeldt-Jakob is any effective connectivity and if this effective connectiv- disease. Ann Neurol 1996;39(6):767–778. http://www.ncbi. ity has any preferential directionality. Figure 2 in the nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed main text, however, does show the effective connectivity &dopt=Citation&list_uids=8651649. and any directionality using SEM. * Correlation is signifi- 76. Rabinovici GD, Wang PN, Levin J, et al. First symptom in cant at the 0.05 level (two-tailed). ** Correlation is signif- sporadic Creutzfeldt-Jakob disease. Neurology 2006;66 icant at the 0.01 level (two-tailed). "Control Variables: (2):286–287. https://doi.org/10.1212/01.wnl.0000196440. Gender, Age, TIV, and Barthel" 00297.67. ª 2021 The Authors. Annals of Clinical and Translational Neurology published by Wiley Periodicals LLC on behalf of American Neurological Association 1199