Abstract
Objective
MAPT mutations typically cause behavioral variant frontotemporal dementia with or without parkinsonism. Previous studies have shown that symptomatic MAPT mutation carriers have frontotemporal atrophy, yet studies have shown mixed results as to whether presymptomatic carriers have low gray matter volumes. To elucidate whether presymptomatic carriers have lower structural brain volumes within regions atrophied during the symptomatic phase, we studied a large cohort of MAPT mutation carriers using a voxelwise approach.
Methods
We studied 22 symptomatic carriers (age 54.7 ± 9.1, 13 female) and 43 presymptomatic carriers (age 39.2 ± 10.4, 21 female). Symptomatic carriers’ clinical syndromes included: behavioral variant frontotemporal dementia (18), an amnestic dementia syndrome (2), Parkinson’s disease (1), and mild cognitive impairment (1). We performed voxel‐based morphometry on T1 images and assessed brain volumetrics by clinical subgroup, age, and mutation subtype.
Results
Symptomatic carriers showed gray matter atrophy in bilateral frontotemporal cortex, insula, and striatum, and white matter atrophy in bilateral corpus callosum and uncinate fasciculus. Approximately 20% of presymptomatic carriers had low gray matter volumes in bilateral hippocampus, amygdala, and lateral temporal cortex. Within these regions, low gray matter volumes emerged in a subset of presymptomatic carriers as early as their thirties. Low white matter volumes arose infrequently among presymptomatic carriers.
Interpretation
A subset of presymptomatic MAPT mutation carriers showed low volumes in mesial temporal lobe, the region ubiquitously atrophied in all symptomatic carriers. With each decade of age, an increasing percentage of presymptomatic carriers showed low mesial temporal volume, suggestive of early neurodegeneration.
Introduction
Mutations in the microtubule‐associated protein tau (MAPT) cause behavioral variant frontotemporal dementia (bvFTD) with or without parkinsonism, with some patients meeting criteria for progressive supranuclear palsy (PSP) or corticobasal syndrome. 1 As in sporadic bvFTD, in bvFTD due to MAPT mutations (bvFTD‐MAPT), patients show degeneration in the anterior cingulate cortex, insula, striatum, and the amygdala. 2 , 3 In contrast to sporadic bvFTD, however, bvFTD‐MAPT more prominently targets the mesial temporal lobe and in particular, the hippocampus, a region less typically atrophied in sporadic bvFTD. 2 , 3
Less is known about the timing of brain volume changes during the presymptomatic phase. Early studies with small samples suggested that presymptomatic MAPT mutation carriers did not have apparent gray or white matter volume differences compared to controls. 4 , 5 Studies with larger samples, however, have shown mixed results. A study of 26 symptomatic and presymptomatic MAPT mutation carriers revealed gray matter trajectories that suggested that low hippocampus and amygdala volumes arise years before expected symptom onset. 6 A diffusion tensor imaging study of 30 presymptomatic MAPT mutation carriers found early loss of white matter integrity. 7 In contrast, a study of 23 presymptomatic MAPT mutation carriers did not show significantly low gray matter volumes, although there was a trend toward small regions of low gray matter volume in mesial temporal lobes. 8
Most previous studies of presymptomatic MAPT mutation carriers have reported only group comparisons whose limitation is that they do not account for individual anatomical variation, which could account for mixed results across studies. To address this gap, we studied a multisite cohort of 65 MAPT mutation carriers (22 symptomatic and 43 presymptomatic) by analyzing structural MRI scans with a voxelwise method that detects gray and white matter differences in individual carriers. We hypothesized that: 1) regions of low gray or white matter volume in presymptomatic MAPT mutation carriers would resemble atrophy patterns seen in symptomatic carriers, and 2) a subset of the presymptomatic carriers, that is, those presumably closer to symptom onset, would harbor low gray and white matter volumes beyond expected of their age. Previous literature suggests that specific MAPT mutations manifest different clinical syndromes; 1 therefore, we also explored whether different MAPT mutation subtypes targeted distinct neuroanatomical regions.
Methods
Participants
Study participants were recruited from: the Memory and Aging Center at the University of California, San Francisco (UCSF); Erasmus University Rotterdam; and two multisite genetic FTD research projects, the Advancing Research and Treatment for Frontotemporal Lobar Degeneration (ARTFL) and Longitudinal Evaluation of Familial Frontotemporal Dementia Subjects (LEFFTDS) (Table S1). Clinical diagnoses were rendered by the respective study sites. 9
We studied 22 symptomatic MAPT mutation carriers, 43 presymptomatic MAPT mutation carriers and 107 healthy controls, comprised of 91 noncarrier genetic frontotemporal lobar degeneration family members and 16 unrelated controls (Table 1). These unrelated controls were added to the noncarrier family members to create an even distribution of ages that was similar to the age range of the MAPT mutation carrier cohorts. Among the 22 symptomatic carriers, 18 had a clinical diagnosis of bvFTD. Of the 18 with bvFTD, six carried secondary diagnoses, including PSP syndrome (4); parkinsonism (1); and depression (1). Of the four remaining symptomatic carriers: two had an amnestic dementia syndrome, one of which also had mild behavioral symptoms; one had Parkinson’s disease (PD) with behavioral symptoms not meeting bvFTD criteria; and one had mild cognitive impairment with behavior and language symptoms. All symptomatic carriers were recruited from UCSF, ARTFL, or LEFFTDS (Erasmus University did not recruit symptomatic patients).
Table 1.
Demographics of presymptomatic and symptomatic MAPT mutation carriers.
Presymptomatic | Symptomatic | Healthy Controls | |
---|---|---|---|
n | 43 | 22 | 107 |
Age at MRI scan, years | 39.2 (10.4) | 54.7 (9.1) | 48.9 (13.2) |
M:F, n | 22:21 | 9:13 | 51:56 |
Education, years | 16.1 (2.4) | 15.1 (2.4) | 15.9 (2.3) |
Education, Dutch 7‐point scale 1 | 4.8 (1.7) | n/a | 5.3 (1.1) |
Age at symptom onset | n/a | 45.6 (8.3) | n/a |
CDR® plus NACC FTLD, global score (0‐3) | 0.05 (0.15) | 1.75 (0.98) | 0 (0) |
CDR® plus NACC FTLD, sum of boxes (max = 24) | 0.07 (0.22) | 9.73 (6.13) | 0 (0) |
Unless otherwise indicated, mean values are reported followed by the standard deviation in parentheses.
Erasmus participants were categorized into levels from 1 = less than 6 years of education to 7 = completed a university degree.
Presymptomatic carriers were diagnosed as clinically normal by each respective study site. These participants were required to have a Mini‐Mental State Examination (MMSE) score ≥ 27, and if the participant did not have an MMSE score available, participants had a Montreal Cognitive Assessment (MoCA) 10 score ≥ 21. We chose this threshold for the MoCA score because in participants tested with both measures, an MMSE score of 27 corresponds to a MoCA score of 21. 11 We included three presymptomatic carriers with a Clinical Dementia Rating Scale (CDR®) plus Behavioral and Language Domains from the National Alzheimer's Coordinating Center (NACC) FTLD module (CDR® plus NACC FTLD) 12 global score of 0.5, as this score may not necessarily be indicative of incipient phenoconversion to disease. Nevertheless, we repeated each analysis without the three presymptomatic carriers who had a CDR® plus NACC FTLD global score of 0.5.
Healthy controls were also required to have a MMSE score ≥ 27 or a MoCA score ≥ 21, a CDR® plus NACC FTLD global score of 0, and no significant history of neurological disease. All participants in the study were required to have a MRI brain scan free of structural lesions and prominent white matter disease.
Participants underwent neuropsychological tests at each site. 4 , 13 , 14 The CDR® plus NACC FTLD scale was available for the UCSF and ARTFL/LEFFTDS cohorts. Neuropsychological tests were administered within 90 days of a participant’s MRI scan. We calculated z‐scores for each measure for each participant based on scores of the healthy controls. For variables with skewed distributions, we reflected and log transformed the data to improve the normality of the distribution prior to fitting multiple regression models. Because these z‐scores were not adjusted for demographic characteristics, we compared the z‐scores of these cognitive measures for the presymptomatic, symptomatic, and control groups using multiple regression with age, education, and sex as covariates of no interest. We compared each MAPT mutation carrier group to controls using post hoc group‐wise comparisons.
The Institutional Review Boards at each site approved this study, and participants or their surrogates provided informed consent to participate.
Genetic analysis
Participants were screened for pathogenic MAPT mutations. 15 All MAPT mutation carriers in the study also tested negative for GRN mutations and for the C9orf72 repeat expansion.
Image acquisition and preprocessing
Each participant underwent a T1‐weighted structural MRI scan on a 3T scanner. Participants from UCSF were scanned on a Siemens Tim Trio or a Siemens Prisma scanner (Table S1). All T1‐weighted images were preprocessed with SPM12 (https://www.fil.ion.ucl.ac.uk/spm/). T1‐weighted images were preprocessed using standard spatial normalization in the SPM12 segment module, using the six standard tissue probability maps with a light clean up. Standard affine regularization with the International Consortium for Brain Mapping European brain template and warping regularization with the default parameters were used. Images were then segmented into gray and white matter images, which were smoothed using an 8‐mm full width at half maximum isotropic Gaussian kernel.
Gray and white matter w‐score maps
For each MAPT mutation carrier, we created gray matter and white matter w‐score maps (w‐maps) which represent the difference between the participant’s actual gray or white matter values in each voxel and the expected value based on a group of healthy controls. 16 A map of w‐scores can be conceptualized as a z‐score map of the gray or white matter residuals after adjustment for covariates of no interest via regression. To calculate a w‐score, voxelwise linear regressions were performed on healthy control gray and white matter maps to determine the standard deviation of the residuals. Next, the w‐score for each individual participant was calculated voxelwise using the following formula: w = [(participant’s value – value predicted for participant age)/SD of residuals of the healthy control group]. Covariates of no interest included age, sex, total intracranial volume, site, scanner model, and scan acquisition protocol. Since the w‐score control group is utilized to build a reference range for brain volume, we selected controls to cover the age distribution of the mutation carriers. We prioritized selecting 91 available mutation‐negative family members for our control group and also included 16 unrelated healthy controls because of the importance of an even age distribution in our models.
Gray and white matter relationships with age or symptom severity
We created group mean w‐maps for 1) presymptomatic and 2) symptomatic carriers by averaging participants’ w‐scores voxelwise and thresholded the maps at w≤‐2. 16 , 17 To create frequency gray and white matter w‐maps, we determined the percentage of participants with gray or white matter values of w≤‐2 voxelwise. For presymptomatic carriers, we created group frequency w‐maps binned by age (in decades) to visualize gray and white matter relationships with age. For symptomatic carriers, group frequency w‐maps were binned by symptom severity, defined by the CDR® plus NACC FTLD global score.
Hemispheric symmetry analysis
Previous studies suggest bilateral and symmetric involvement in symptomatic MAPT mutation carriers. 2 , 3 , 6 To determine whether atrophy or low brain volume in symptomatic and presymptomatic MAPT mutation carriers is symmetric at an individual‐subject level, we calculated the difference between each participant’s mean left and right hemisphere w‐scores for gray or white matter.
Correlations between memory measures and hippocampal volume
We performed two analyses to explore associations between memory measures and the volume of the hippocampus, selected as an a priori region because of its importance in memory recall. First, we calculated for each participant the mean gray matter w‐scores for the left and right hippocampi (defined by the Automated Anatomical Labeling atlas 18 ). Verbal memory z‐scores were based on either the California Verbal Learning Test II (CVLT‐II) 10 minute delayed recall (UCSF; ARTFL/LEFFTDS) or the Rey Auditory Verbal Learning Test delayed recall trials (Erasmus). Spatial memory z‐scores were based on the Benson figure recall (UCSF; ARTFL/LEFFTDS). We performed Spearman’s correlations between participants’ hippocampal w‐scores and memory z‐scores. Second, we selected the voxel most frequently showing reduced gray matter across all carriers and created a 4‐mm sphere around that voxel (and its contralateral homolog) to determine whether gray matter volume in this MAPT‐targeted region correlates with verbal and visual memory measures. For these analyses, we correlated left hippocampus w‐scores with verbal memory scores and right hippocampus w‐scores with spatial memory z‐scores given the importance of the left hippocampus for verbal memory 19 and the right hippocampus for spatial memory. 20 Next, we searched voxelwise to identify brain regions where lower memory performance was associated with lower gray matter volume by correlating verbal then spatial memory z‐scores with gray matter w‐maps thresholded at P < 0.05 corrected for family‐wise error.
The above correlations were performed in all carriers combined, symptomatic carriers, and presymptomatic carriers.
MAPT mutation subgroup analysis
Our cohort consisted of participants carrying 13 different MAPT mutations. Because some mutations had low numbers of participants, we examined MAPT mutation subtypes categorized into five groups by either clinical (bvFTD‐type; amnestic‐type; PSP/PD‐type) or neuropathological (four repeat‐type (4R); and Pick bodies type) similarity to explore whether different types of MAPT mutations may target distinct neuroanatomy. The bvFTD group was comprised of carriers with the V337M mutation, a missense mutation located on exon 12 which typically causes bvFTD. 21 The amnestic‐type group consisted of carriers with R406W, a missense mutation on exon 13 which typically manifests with an amnestic syndrome reminiscent of an Alzheimer’s‐like syndrome rather than a bvFTD syndrome. 22 , 23 The PSP/PD‐type group consisted of carriers with mutations in N279K, S305N, and S305S, all exon 10 missense mutations which increase the ratio of 4R to three repeat (3R) tau; these mutations most often cause bvFTD with PSP or parkinsonism. 24 , 25 , 26 , 27 , 28 The 4R‐type group consisted of P301L, S305I, IVS10 + 16C>T, and IVS9‐10G> T, a combination of exonic and splicing mutations which typically present with bvFTD; these mutations increase exon 10 transcription, which increases the ratio of 4R to 3R tau. 15 , 29 , 30 The fifth group was the Pick bodies type group which contained G389R, G272V, S320F, and L315R mutations, all characterized by the inclusion of Pick body or Pick body‐like pathology. 31 , 32 , 33 , 34 Each group consisted of fewer than 10 symptomatic and presymptomatic carriers combined, with the exception of the 4R‐type group, which had 31 carriers. To protect participant anonymity, we chose not to report the specific number within each group. We created mean w‐maps for these five mutation subgroups to assess whether different mutation subtypes target distinct anatomical regions. Due to the limited number of participants in each mutation subgroup, we assessed gray and white matter volume using mean w‐maps instead of frequency w‐maps.
Results
A subset of presymptomatic carriers harbors low mesial temporal lobe volumes
In total, the presymptomatic and symptomatic cohorts included 13 different MAPT mutation subtypes. For most neuropsychological measures, presymptomatic MAPT mutation carriers had no statistically significant differences compared to controls (Tables 2 and 3). The one exception was that presymptomatic carriers had a significantly lower but clinically negligible z‐score of −0.39 on the Multilingual Naming Test (Table 2). With the three presymptomatic carriers with CDR® plus NACC FTLD global score of 0.5 excluded, results were similar with presymptomatic carriers largely showing no differences across measures, with the exception of the Multilingual Naming Test (z = −0.39, range −1.68, 0.38). Symptomatic carriers showed clinically significant impairments in verbal learning and memory as well as naming and semantic verbal fluency, and borderline to low average performance on most other measures, including attention, executive functioning, and visual recall, while Benson Figure Copy was relatively preserved (Table 2).
Table 2.
Neuropsychological testing of North American cohorts.
Test | Presymptomatic MAPT | Symptomatic MAPT | Omnibus Model | ||
---|---|---|---|---|---|
Raw scores: mean(SD) or median[IQR] |
z‐scores: mean(SD) or median[IQR] |
Raw scores: mean(SD) or median[IQR] |
z‐scores: mean(SD) or median[IQR] |
F (df) | |
Global Cognition 4 | |||||
MoCA, total score | 28 [26, 29] | 0.11 [−0.96, 0.64] | 22 [11, 25] | −3.10 [−8.97, −1.49] | 11.24 (5, 79) 1 , 2 |
Executive Functions/Processing Speed/Attention | |||||
Trails A, correct lines per minute | 80.15 (24.58) | 0.24 (0.94) | 37.58 (20.11) | −1.38 (0.77) | 8.67 (5, 88) 1 , 2 |
Trails B, correct lines per minute | 34.42 (10.68) | 0.33 (1.03) | 16.65 (9.88) | −1.39 (0.96) | 8.00 (5, 88) 1 , 2 |
Digit span forward | 9.03 (2.31) | 0.16 (0.98) | 6.50 (2.85) | −0.91 (1.21) | 2.84 (5, 112) 1 , 2 |
Digit span backward | 8.00 (2.80) | 0.21 (1.21) | 5.24 (2.44) | −0.99 (1.05) | 3.45 (5, 113) 1 , 2 |
Semantic fluency, animals in 1 min | 24.21 (6.34) | 0.08 (1.17) | 12.12 (8.42) | −2.14 (1.55) | 12.79 (5, 113) 1 , 2 |
Lexical fluency, words in 1 min | 28.96 (7.04) | 0.10 (0.91) | 18.15 (12.73) | −1.29 (1.64) | 7.10 (5, 81) 1 , 2 |
Memory 4 | |||||
Total recall (California Verbal Learning Test, short form, four learning trials total) | 31 [29, 32.5] | 0.23 [−0.28, 0.61] | 18 [15, 24] | −3.09 [−3.86, −1.56] | 9.67 (5, 101) 1 , 2 |
Delayed free recall (California Verbal Learning Test, short form, 10 min recall) | 7 [6, 9] | −0.35 [−0.95, 0.85] | 0 [0, 5] | −4.54 [−4.54, −1.54] | 12.11 (5, 101) 1 , 2 |
Benson figure 10 min recall, total score | 14 [12, 16] | 0.42 [−0.48, 1.31] | 10 [4, 12] | −1.37 [−4.05, −0.48] | 6.45 (5, 114) 1 , 2 |
Language 4 | |||||
Multilingual Naming Test, total score | 29.5 [27, 31] | −0.39 [−1.67, 0.38] | 26.5 [22.5, 28.5] | −1.92 [−3.97, −0.90] | 7.56 (5, 80) 1 , 2 , 3 |
Visuospatial 4 | |||||
Benson figure copy, total score | 16 [15, 16] | 0.09 [−0.99, 0.09] | 16 [15, 16] | 0.09 [−0.99, 0.09] | 0.60 (5, 114) |
Means and standard deviations (SD) or medians and interquartile ranges (IQR) are reported.
Symptomatic cohort is significantly different from controls at P < 0.05.
Presymptomatic cohort is significantly different from symptomatic cohort at P < 0.05.
Presymptomatic cohort is significantly different from controls at P < 0.05.
Z‐scores estimated from reflected and log‐transformed data were used to fit the regression models.
Table 3.
Neuropsychological testing of Erasmus cohort.
Test | Presymptomatic MAPT | Omnibus Model | |
---|---|---|---|
Raw scores: mean(SD) or median[IQR] |
z‐scores: mean(SD) or median[IQR] |
F (df) | |
Global Cognition 1 | |||
MMSE, total score | 29.5 [29, 30] | 0.28 [−0.28, 0.85] | 3.88 (4, 37) |
Executive Functions/Processing Speed/Attention | |||
Trails A, correct lines per minute | 58.90 (21.66) | 0.30 (1.08) | 4.37 (4, 37) |
Trails B, correct lines per minute | 28.44 (11.99) | 0.46 (1.29) | 4.08 (4, 37) |
Digit span forward | 8.93 (2.56) | −0.04 (1.32) | 1.79 (4, 37) |
Digit span backward | 6.50 (1.74) | 0.05 (0.89) | 2.04 (4, 37) |
Semantic fluency, animals in 1 min | 26.93 (6.17) | 0.52 (1.22) | 1.94 (4, 37) |
Lexical fluency, total D, A, and T words | 36.57 (14.94) | 0.25 (1.30) | 0.72 (4, 37) |
Memory | |||
Total Recall (Rey Auditory Verbal Learning Test, total) | 48.36 (8.95) | 0.54 (0.96) | 3.87 (4, 37) |
Delayed Free Recall (Rey Auditory Verbal Learning Test) | 9.79 (3.75) | 0.34 (1.20) | 2.34 (4, 37) |
Language 1 | |||
Boston Naming Test, total score | 54.5 [50, 58] | 0.07 [−0.92, 0.85] | 1.08 (4, 37) |
Visuospatial 1 | |||
Clock drawing | 12.5 [12, 13] | −0.05 [−0.39, 0.29] | 1.16 (4, 37) |
Means and standard deviations (SD) or medians and interquartile ranges (IQR) are reported.
The Erasmus presymptomatic cohort did not differ from controls for any measure.
Z‐scores estimated from reflected and log‐transformed data were used to fit the regression models.
We found that symptomatic MAPT mutation carriers had severe bilateral mesial temporal atrophy as well as frontotemporal and insula atrophy and white matter involvement in the uncinate fasciculus, anterior corona radiata, corpus callosum, and inferior longitudinal fasciculus (Fig. 1; Fig. S1). Frequency w‐maps revealed that each symptomatic carrier harbored prominent mesial temporal lobe atrophy and about sixty percent had involvement within frontotemporal white matter tracts (Fig. 1). Although the mean w‐maps for presymptomatic carriers did not show any regions with low gray matter (defined as w≤‐2), we found that 20% of presymptomatic carriers had low mesial temporal lobe volumes and low white matter volume arose in approximately 10% (Fig. 1). When calculating the difference between each participant’s mean left and right hemisphere w‐score for gray or white matter, we found that at individual level, both symptomatic and presymptomatic participants had symmetric gray and white matter (maximum difference score w = 0.78 for a symptomatic carrier).
Figure 1.
Gray and white matter frequency w‐maps grouped by clinical stage. Frequency w‐maps show the percentage of subjects within the group, ranging from 1% (magenta) to 100% of subjects (red), who have low gray or white matter volumes with w≤‐2 in a given voxel. Top row, left: All MAPT mutation carriers within the symptomatic group have anteromesial temporal atrophy. Top row, right: Symptomatic carriers have atrophy in the uncinate fasciculi, corpus callosum, and anterior corona radiata. Bottom row, left: Approximately 20% of presymptomatic MAPT mutation carriers show low gray matter volume in the mesial temporal lobes. Bottom row, right: Few carriers show low white matter volume before symptom onset. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain.
We repeated these analyses with presymptomatic carriers who had a CDR® plus NACC FTLD score of 0 only. Gray matter and white matter frequency w‐maps showed similar results (data not shown), supporting that the three presymptomatic carriers with CDR® plus NACC FTLD scores of 0.5 were not driving the overall results.
Mesial temporal lobe and relationships with memory throughout the MAPT lifespan
To explore whether regions of low gray and white matter in presymptomatic MAPT mutation carriers might worsen with age, we created gray and white matter frequency w‐maps for each age decade represented within our data. We found that approximately 20% of presymptomatic MAPT mutation carriers in their thirties showed low mesial temporal volumes (Fig. 2). When examining carriers in their forties and fifties, an increasing percentage of participants had low mesial temporal gray matter with each additional decade. The white matter frequency map (Fig. 1) suggested that low white matter volumes also arise in certain presymptomatic carriers, but in contrast to the gray matter maps, the percentage of participants with low white matter volumes did not clearly increase with each decade.
Figure 2.
Gray and white matter frequency w‐maps of presymptomatic carriers grouped by age decade. Gray matter (left) and white matter (right) frequency w‐maps for presymptomatic carriers divided by age decade (rows). Color bar indicates the percentage of subjects within each age decade with low gray or white matter volumes of w≤‐2 in a given voxel. Low mesial temporal lobe volumes arise in about 20% of MAPT mutation carriers as early as their thirties. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain.
For symptomatic carriers with CDR® plus NACC FTLD global scores of 0.5, mesial and anterior temporal lobe and anterior cingulate atrophy were most common, followed by insula and regions in dorsolateral prefrontal cortex (Fig. 3). With worsening symptom severity, atrophy within these regions grew increasingly more common and extensive. All carriers with a CDR® plus NACC FTLD global score of 1 had mesial temporal lobe atrophy and all carriers with a global score of 2 had regions of anterior cingulate and insular atrophy. By the time carriers reached a global score of 3, all showed widespread frontotemporal, insula, and uncinate fasciculus atrophy.
Figure 3.
Gray and white matter frequency w‐maps of symptomatic carriers grouped by symptom severity. Gray matter (left) and white matter (right) frequency w‐maps for symptomatic carriers binned by symptom severity, as defined by the CDR® plus NACC FTLD global score (FTLD‐CDR). Color bar indicates the percentage of subjects within each symptom severity category with gray or white matter atrophy of w≤‐2 in a given voxel. Mesial temporal lobe atrophy becomes ubiquitous in those with an FTLD‐CDR score of 1. With worsening severity, atrophy within temporal and frontal lobes, insula and anterior cingulate are increasingly more common and extensive. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain.
We examined relationships between memory measures and the hippocampus, selected as an a priori region of interest, and voxelwise gray matter maps. Verbal and spatial memory z‐scores were correlated with hippocampal volumes across the whole MAPT cohort. Across symptomatic and presymptomatic carriers combined, lower scores in verbal recall (CVLT‐II) were associated with lower left hippocampal volume (rho = 0.70, P < 0.001) and lower scores in visual recall were associated with lower right hippocampal volume (rho = 0.44, P < 0.001). To explore whether a MAPT‐targeted gray matter region is also associated with memory measures, we created a spherical region of interest centered on a left hippocampus voxel where carriers most frequently had low gray matter. We also created its contralateral homolog. Lower verbal recall scores were associated with lower volume within this left hippocampal sphere (rho = 0.66, P < 0.001), while lower scores in visual recall were associated with lower right hippocampal sphere volume (rho = 0.50, <0.001). Voxelwise, lower verbal recall was correlated with lower hippocampal volume bilaterally, while lower visual recall was associated with lower volume in a small clusters of the left hippocampus and other scattered voxels (Fig. 4). These three analyses (whole hippocampus, hippocampal sphere, voxelwise) were repeated with symptomatic and presymptomatic carrier subgroups and showed no statistically significant correlations between memory measures and gray matter.
Figure 4.
Voxelwise correlations between memory measures and gray matter. Maps show regions where verbal (top row) or spatial (bottom row) memory z‐scores were significantly correlated with regions of low gray matter volume across all MAPT mutation carriers. Lower scores on a verbal memory measure was associated with low volume in bilateral hippocampi and inferior temporal regions. Lower scores on a spatial memory measure was associated with low gray matter in small scattered clusters including the left hippocampus, right middle frontal gyrus, frontal pole, and left dorsal parietal lobe. Maps were thresholded at P < 0.05, corrected for family‐wise error. Color bars show t‐value ranges. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain.
We repeated each of these analyses by excluding the three presymptomatic carriers with the CDR® plus NACC FTLD score of 0.5 and results were unchanged.
Most MAPT mutation subtypes target frontotemporal cortex
When mean w‐maps were calculated for each MAPT subtype, we found that most of the symptomatic MAPT mutation subtype groups (bvFTD‐type, PSP/PD type, 4R‐type) shared a consistent pattern of fronto‐insulo‐temporal atrophy (Fig. 5). Of those groups, the bvFTD‐type group showed the most severe gray and white matter frontotemporal atrophy, while the PSP/PD type group showed the most severe atrophy within the mesial temporal lobes (Fig. 5). One notable exception to the shared frontotemporal atrophy pattern, however, was the R406W mutation group, which typically results in an amnestic presentation. R406W carriers lacked the frontal gray and white matter atrophy that was characteristic of the other MAPT mutation subtypes, and instead only showed bilateral temporal and insular atrophy. Similar to the bvFTD‐type and 4R‐type mutations, the PSP/PD group showed frontotemporal atrophy, but additionally showed prominent midbrain atrophy. The presymptomatic carriers with PSP/PD‐related mutations had low bilateral anterior cingulate, insula and mesial temporal volumes and were also the only presymptomatic group with voxels of low midbrain volume, albeit in sparse regions. We found no differences in results when the three presymptomatic carriers with a CDR® plus NACC FTLD score of 0.5 were excluded.
Figure 5.
Atrophy patterns by MAPT mutation subtype —mean w‐maps. Mean w‐maps show the mean w‐score in gray and white matter across all subjects grouped by clinical stage. Rows reflect five different mutation subtype groupings. Frontotemporal atrophy patterns are similar across mutation groups, except for the amnestic type R406W mutation group, which lacks frontal atrophy. Mutation subtypes associated with progressive supranuclear palsy or parkinsonism distinctly show midbrain atrophy for both the symptomatic and presymptomatic groups. The decade of the mean age for each group is indicated rather than the specific mean age in order to protect participant anonymity. Group maps are thresholded from w≤‐13 to −2. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain. DD: disease duration.
Discussion
In this study, we took a comprehensive approach to identify individual‐level variation in gray and white matter volume and their relationships with age or symptom severity across the MAPT continuum. We also identified relationships between memory performance and gray matter to explore whether specific MAPT mutation subtypes were associated with distinct atrophy patterns. We found that presymptomatic MAPT mutation carriers have low mesial temporal volumes that are topographically similar to the atrophy pattern seen in symptomatic carriers. Low mesial temporal lobe volumes arose in about 20% of the presymptomatic carriers during their thirties, and regions with low gray matter appeared more common in presymptomatic carriers with each increasing decade. Fewer than 10% of presymptomatic carriers had low white matter values in affected regions. For symptomatic carriers, mesial temporal cortex was ubiquitously atrophied in carriers with a CDR® plus NACC FTLD global score of 1 or greater, with widespread atrophy in frontotemporal cortex in those with more advanced symptoms. In MAPT mutation carriers, poorer memory performance was associated with lower hippocampal volume. We found that atrophy in frontotemporal and insular cortex and the uncinate fasciculi, inferior longitudinal fasciculi, and corpus callosum were commonly observed across different MAPT mutations during the symptomatic phase, with the exception of the R406W mutation carriers who lacked the frontal atrophy characteristic of other MAPT mutation carriers.
A subset of presymptomatic carriers shows low mesial temporal volumes
As in previous studies, we found that frontotemporal cortex is targeted in symptomatic MAPT mutation carriers. 2 , 3 Compared to patients with sporadic bvFTD and those with other FTLD mutations, prominent mesial temporal atrophy is more prevalent in MAPT mutation carriers, 2 , 35 , 36 who feature correspondingly greater memory impairment. 23 , 37 Although mean group w‐maps did not identify regions of low matter in presymptomatic carriers, frequency w‐maps revealed that 20% of this group had low mesial temporal volume. While early studies suggested that gray matter abnormalities were undetectable in presymptomatic MAPT mutation carriers, 4 , 5 recent larger studies suggested that presymptomatic carriers have low hippocampal and amygdalar volumes. 6 , 8 , 17 , 38 Several potential explanations may account for these varied results. First, low gray matter volumes during the presymptomatic phase are subtle and thus are likely to be detectable only with larger cohorts. Second, previous studies of presymptomatic carriers have performed group analyses, which may be insensitive to low brain volumes present in a minority of participants. By examining frequency maps, we were able to identify the subset of carriers with abnormally low gray and white matter volumes, thus circumventing this limitation. Third, the presymptomatic carriers in the present study and previous studies could have differences in the number of years until actual symptom onset. An individual’s age of onset and mean familial age of onset are highly correlated, and perhaps more strongly correlated for MAPT mutation carriers compared to other FTLD mutations. 6 , 39 Although the mean age of the presymptomatic cohort in the present study is similar to previous studies, 5 , 8 , 39 age and mean familial age of onset remain imperfect methods for estimating the actual age of onset. Whether lower hippocampal gray matter volumes portend earlier conversion to the symptomatic phase remains an open question.
In contrast to sporadic bvFTD, memory impairment is a feature of bvFTD due to MAPT mutations. 40 , 41 Indeed, we found that episodic memory, particularly verbal memory, was the most severely affected domain in symptomatic carriers. Despite memory and language impairments and lower than average executive performance, visuoconstructional performance was similar to controls. Our results showed that in all MAPT carriers combined, poorer memory performance was correlated with lower hippocampal volumes. Although hippocampal atrophy has been reported in MAPT, 2 , 3 , 6 , 8 , 17 , 36 , 38 to our knowledge, previous studies have not directly correlated memory performance with hippocampal volume in MAPT mutation carriers. The hippocampus is well‐established for its importance in episodic memory 19 , 20 and our results support that this structure underlies the memory impairment in MAPT. Our subgroup analyses of symptomatic and presymptomatic carriers failed to identify significant correlations between memory measures and hippocampal volume. Smaller subgroup sample sizes, floor effects in the symptomatic carriers’ memory scores, and the limited dynamic range of presymptomatic carriers’ memory scores serve as possible explanations for these results.
Compared to gray matter, low white matter volume arose less frequently among presymptomatic carriers. One possibility is that gray matter decline precedes white matter decline during the presymptomatic phase. Consistent with this notion, a study of five presymptomatic MAPT mutation carriers followed over four years found that upon visual inspection, the extent of gray matter loss appeared more pronounced than white matter integrity loss. 42 Although the present study showed that low white matter volumes as assessed by voxel‐based morphometry were not prevalent in presymptomatic MAPT, diffusion tensor imaging studies have reported lower white matter tract integrity in presymptomatic MAPT compared to controls. 7 , 43 Voxel‐based morphometry may be less sensitive at detecting white matter differences compared to diffusion tensor imaging, thus future studies using multiple imaging modalities are needed.
Regions of low gray and white matter may reflect early neurodegeneration in presymptomatic carriers
Certain presymptomatic carriers in their thirties showed low mesial temporal volumes, consistent with a previous study suggesting that gray matter trajectories indicate low hippocampal and amygdalar volumes arising 15 years before estimated symptom onset. 6 While regions with low gray matter could in part reflect neurodevelopmental differences as has been suggested for the C9orf72 repeat expansion, 44 , 45 our data indicate that presymptomatic carriers more frequently had low volumes within canonical regions that are targeted in MAPT mutations with each age decade. This finding suggests that a subset of presymptomatic carriers may be undergoing incipient neurodegeneration, with more carriers following suit as symptom onset approaches. Interestingly, the frequency of low mesial temporal lobe volumes appeared to outpace other regions with increasing age, underscoring the notion that the mesial temporal lobe is targeted early in MAPT mutation carriers and becomes increasingly ubiquitous with age and throughout the symptomatic phase. A recent study of 14 presymptomatic MAPT mutation carriers studied over a median of 9 years revealed that the rate of temporal lobe gray matter decline outpaced the rate of decline in the frontal and parietal cortices, suggesting that low temporal gray matter volumes in presymptomatic carriers may represent early neurodegeneration. 46
Most MAPT mutation subtypes converge on frontotemporal atrophy
Despite the different mechanisms of each mutation on tau protein biology and neuropathology, patients with different MAPT mutation subtypes converged on a highly similar pattern of frontal and anterior, mesial and lateral temporal atrophy with a corresponding predominance of bvFTD diagnoses. 15 , 47 , 48 Another study found that six MAPT mutation subtypes shared anterior temporal gray matter atrophy. 41 IVS10 + 16, IVS10 + 3, N279K, and S305N additionally showed mesial temporal atrophy, while P301L and V337M had lateral temporal regions targeted, with relative sparing of the mesial temporal lobe. 41 In contrast, our V337M symptomatic carriers showed mesial temporal involvement. Both the present study and previous studies have investigated small numbers of participants for each mutation subtype, such that clinical heterogeneity may influence results.
Two notable exceptions arose when studying MAPT mutation subtypes. R406W is the only known MAPT mutation that causes an amnestic syndrome rather than a behavioral syndrome, 22 , 23 and it typically causes mesial temporal atrophy while sparing the frontal lobes relative to other MAPT mutations. 22 , 23 , 49 , 50 We found that R406W carriers exclusively showed mesial and lateral temporal atrophy, but lacked frontal atrophy, which may explain why these carriers develop an amnestic rather than a bvFTD syndrome. Among the MAPT mutations we studied, N279K, S305N, and S305S cause PSP and parkinsonism. Interestingly, this group was the only MAPT group which had midbrain atrophy, and midbrain atrophy has been associated with patients with PSP. 51 An important caveat to our findings is that variability in age and disease duration within the subtype groups may be contributing to differences in atrophy because the participant numbers are limited for each group. Remarkably, the average disease duration ranged from 2.7 (± 2.7) years for the 4R‐type mutation group to 19.5 (± 10.1) years for V337M carriers, yet despite this wide range, the magnitude of atrophy among the four mutation groups was similar. This long disease duration and relatively old average age of onset suggests that V337M might have a more indolent, protracted disease course compared to the other mutations.
Limitations
This study examined one of the largest MAPT cohorts to date, but sample sizes were still small within most mutation subtypes and larger cohorts are needed to study individual MAPT mutations. To maximize power, we combined different mutation subtypes based on clinical or neuropathological similarity. Although we examined relationships of gray and white matter with age, this study had a cross‐sectional design and longitudinal analyses are needed to map disease trajectories. The extent to which focal regions with low gray and white matter in presymptomatic mutation carriers may be due to neurodevelopmental differences versus early neurodegeneration remains unknown. Yet, the present study and current literature support the notion that low gray and white matter volumes in presymptomatic MAPT mutation carriers represent early neurodegeneration. Longitudinal studies will clarify these questions by mapping trajectories in MAPT mutation carriers from early life through the symptomatic phase.
Conflicts of Interest
SAC, TMF, LCJ, JD, SS, LZ, VES, JSY, WWS, JMP, DHG, HWH, LKF, DEB, MG, YMB, JCF, TF, RHG, DJI, AMK, JHK, WAK, MIL, MFM, AYP, RR, EMR, MCT, NAT, AV, SW, BW, JCVS, and SEL had no disclosures. AMS received research support from the Larry H. Hillblom foundation and support from the NIH. HJR received research support from Biogen Pharmaceuticals, has consulting agreements with Wave Neuroscience and Ionis Pharmaceuticals, and receives research support from the NIH. BFB served as an investigator for clinical trials sponsored by GE Healthcare and Axovant. He received royalties from the publication of a book entitled Behavioral Neurology of Dementia (Cambridge Medicine, 2009, 2017). He served on the Scientific Advisory Board of the Tau Consortium. He received research support from the NIH, the Mayo Clinic Dorothy and Harry T. Mangurian Jr. Lewy Body Dementia Program and the Little Family Foundation. ALB received research support from the NIH, the Tau Research Consortium, the Association for Frontotemporal Degeneration, Bluefield Project to Cure Frontotemporal Dementia, Corticobasal Degeneration Solutions, the Alzheimer's Drug Discovery Foundation, and the Alzheimer's Association. He served as a consultant for Aeton, Abbvie, Alector, Amgen, Arkuda, Ionis, Iperian, Janssen, Merck, Novartis, Samumed, Toyama and UCB, and received research support from Avid, Biogen, BMS, C2N, Cortice, Eli Lilly, Forum, Genentech, Janssen, Novartis, Pfizer, Roche, and TauRx. GC, BCD, KF, JAF, EDH, BLM, AWT received research support from the NIH. HHF reported grants from Toyama Pharmaceuticals, Biohaven Pharmaceuticals, Annovis Bio (QR Pharma) and Vivoryon (Probiodrug). He also reported service agreements through UCSD for consulting with Eisai, Merck, Tau RX, Samus, Arkuda, Samumed, Axon Neurosciences, Roche/Genentech Pharmaceuticals for DMC and DSMB activities; Tau Consortium for Scientific Advisory Board. He reported travel expenses from Vivoryon, Axon Neurosciences and Alion Pharmaceuticals and speaker fees to UCSD from World Events Forum, Optum Health, and Medscape. NG had participated or is currently participating in clinical trials of antidementia drugs sponsored by the following companies: Bristol Myers Squibb, Eli Lilly/Avid Radiopharmaceuticals, Janssen Immunotherapy, Novartis, Pfizer, Wyeth, SNIFF (The Study of Nasal Insulin to Fight Forgetfulness) study, and A4 (The Anti‐Amyloid Treatment in Asymptomatic Alzheimer's Disease) trial. She received research support from Tau Consortium and Association for Frontotemporal Dementia and is funded by the NIH. NRGR received royalties from UpToDate and had participated in multicenter therapy studies sponsored by Biogen, TauRx, AbbVie, Novartis, and Lilly. He received research support from the NIH. GYRH had served as an investigator for clinical trials sponsored by AstraZeneca, Eli Lilly, and Roche/Genentech. He received research support from the Canadian Institutes of Health Research and the Alzheimer Society of British Columbia. KK served on the Data Safety Monitoring Board for Takeda Global Research & Development Center, Inc.; data monitoring boards of Pfizer and Janssen Alzheimer Immunotherapy; research support from the Avid Radiopharmaceuticals, Eli Lilly, the Alzheimer's Drug Discovery Foundation, and the NIH. DIK served as a consultant to Axovant, Abbvie, Janssen Research and Development, and Takeda. He served on the advisory board of Grifols, and received research support from Axovant, Janssen Research and Development, Navidea Biopharmaceuticals, and Acadia. DSK served on the DSMB of the DIAN‐TU study, is a site PI for clinical trials sponsored by Biogen, Lilly and the University of Southern California, and is funded by the NIH. JK had provided expert witness testimony for Teva Pharmaceuticals in Forest Laboratories Inc. et al. v. Teva Pharmaceuticals USA, Inc., Case Nos. 1:14‐cv‐00121 and 1:14‐cv‐00686 (D. Del. filed Jan. 31, 2014 and May 30, 2014) regarding the drug Memantine; for Apotex/HEC/Ezra in Novartis AG et al. v. Apotex Inc., No. 1:15‐cv‐975 (D. Del. filed Oct. 26, 2015, regarding the drug Fingolimod. He had also given testimony on behalf of Puma Biotechnology in Hsingching Hsu et al., vs. Puma Biotechnology, INC., et al. 2018 regarding the drug neratinib. He received research support from the NIH. IL received research support from the NIH, Parkinson Study Group, Parkinson Foundation, Michael J Fox Foundation, AVID Pharmaceuticals, C2N Diagnostics/Abbvie, and Bristol‐Myers Squibb. She was a member of the Biogen and Bristol‐Myers Squibb Advisory Boards, Biotie/Parkinson Study Group Medical Advisory Board, and consultant for Toyama Pharmaceuticals. She received salary from the University of California San Diego and as Editor in Frontiers in Neurology. IRAM received research funding from the Canadian Institutes of Health Research. S.M.M. served as an investigator for clinical trials sponsored by AbbVie, Allon Therapeutics, Biogen, Bristol‐Myers Squibb, C2N Diagnostics, Eisai Inc., Eli Lilly and Co., Genentech, Janssen Pharmaceuticals, Medivation, Merck, Navidea Biopharmaceuticals, Novartis, Pfizer, and TauRx Therapeutics. He received research support from the NIH. CUO received research funding from the NIH, the CIHR, and Biogen, Inc. He was also supported by the Jane Tanger Black Fund for Young‐Onset Dementias, the Nancy H. Hall Fund for Geriatric Psychiatry, and the gift from Joseph Trovato. EDR received research support from the NIH, Bluefield Project to Cure Frontotemporal Dementia, Alzheimer's Association, BrightFocus Foundation, Biogen, Alector, and owns intellectual property related to tau. ZKW was supported by the NIH, Mayo Clinic Center for Regenerative Medicine, the gift from Carl Edward Bolch, Jr., and Susan Bass Bolch, the Sol Goldman Charitable Trust, and Donald G. and Jodi P. Heeringa. He has also received grant funding support from Allergan, Inc. (educational grant), and Abbvie (medication trials).
Author Contributions
SAC, SEL, SS, and WWS contributed to the conception and design of the study. SAC and SEL contributed to drafting the manuscript. SAC, TMF, AMS, JD, LZ, and SEL contributed to data analysis. AMS, LCJ, SS, VES, JSY, WWS, JMP, HJR, BFB, ALB, HWH, LKF, DEB, MG, GC, BC, YMB, KF, HHF, JAF, JCF, TF, RHG, NG, NRGR, GYRH, EDH, DJI, KK, DIK, AMK, DSK, JK, JHK, WAK, MIL, IL, IRAM, MFM, BLM, CUO, AYP, RR, EMR, EDR, MCT, NAT, AWT, AV, SW, BW, ZKW, JCVS, and SEL contributed to data acquisition.
Supporting information
Figure S1. Gray and white matter mean w‐maps of symptomatic carriers. Maps show the mean w‐score of gray and white matter across the symptomatic carriers. Symptomatic carriers have atrophy in frontotemporal cortex, insula, uncinate fasciculus, anterior corona radiata, corpus callosum, and inferior longitudinal fasciculus. Maps are thresholded from w≤‐13 to −2. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain.
Table S1. Participants by Site
Acknowledgements
The manuscript has been reviewed by the Publications Committee of LEFFTDS & ARTFL for scientific content. The authors thank the ARTFL & LEFFTDS consortium, and the National Centralized Repository for Alzheimer’s Disease and Related Dementias. The authors acknowledge the invaluable contributions of the participants at the Memory and Aging Center, Erasmus University Rotterdam, and ARTFL & LEFFTDS, as well as the assistance of the support staff at each of the participating sites.
Funding informationThis work was supported by the Tau Consortium (SEL, GC, DHG, NG); National Institutes of Health [SEL: K23AG039414, R01 AG058233; GC: P50 AG023501, P01 AG019724, U54 NS092089; DHG: AG035610, R01 AG26938; RR: R35 NS097261, UG3 NS103870; EDH: R01 NS076837; BFB: U01 AG045390; ALB: U54NS092089, 2R01AG038791; HJR: U01 AG045390, U54 NS092089, K24 AG045333, R01 AG032306; BLM: P01AG019724, P50AG23501; AWT: P41EB015922; WWS: AG023501, AG19724; NG: K12 HD001459; JSY: K01 AG049152; SS: K08 AG052648; DJI: K23NS088341; MG: P01AG017586, U01AG052943; IL: 5P50AG005131‐33, 2R01AG038791‐06A, U01NS090259, 1U54 NS 092089, U01NS100610, U01NS80818, R25NS098999, P20GM109025; JK: R01EB022055, U01AG045390, U54NS092089, R01CA132870]; the John Douglas French Alzheimer's Foundation [GC, DHG]. Samples from the National Cell Repository for Alzheimer's Disease and Related Dementias (NCRAD), which receives government support under a cooperative agreement grant (U24AG21886) awarded by the National Institute on Aging (NIA), were used in this study. Data collection and dissemination of the data presented in this manuscript were supported by the LEFFTDS & ARTFL Consortium (LEFFTDS: U01 AG045390, funded by the National Institute on Aging and the National Institute of Neurological Diseases and Stroke; ARTFL: U54 NS092089, funded by the National Institute of Neurological Diseases and Stroke and National Center for Advancing Translational Sciences). ZKW is partially supported by the Mayo Clinic Center for Regenerative Medicine, the gifts from The Sol Goldman Charitable Trust, and the Donald G. and Jodi P. Heeringa Family, by the Haworth Family Professorship in Neurodegenerative Diseases fund, and by the Albertson Parkinson's Research Foundation. IRAM is supported by the Canadian Institutes of Health Research (74580). JCVS is supported by the Dioraphte Foundation grant (09‐02‐00), the Association for Frontotemporal Dementias Research Grant 2009, The Netherlands Organization for Scientific Research (NWO) grant (HCMI 056‐13‐018), and ZonMw Memorabel (project number 733050813). The Dutch cohort is supported by ZonMw Memorabel (733050103 & 733050813), Alzheimer Nederland, JPND PreFrontAls consortium, and Bluefield Project. IL is additionally supported by the Parkinson Study Group, Michael J. Fox Foundation, Lewy Body Association, AVID Pharmaceuticals, Abbvie, Biogen and Roche.
Funding Statement
This work was funded by Lewy Body Association grant ; Michael J. Fox Foundation for Parkinson's Research grant ; Biogen grant ; AVID Pharmaceuticals grant ; Roche grant ; Mayo Clinic Center for Regenerative Medicine grant ; The Sol Goldman Charitable Trust grant ; Canadian Institutes of Health Research grant 74580; Stichting Dioraphte grant 09‐02‐00; AbbVie grant ; Tau Consortium grant ; John Douglas French Alzheimer's Foundation grant ; Donald G. and Jodi P. Heeringa Family grant ; National Center for Advancing Translational Sciences grant U54 NS092089; Alzheimer Nederland grant ; Bluefield Project grant ; Haworth Family Professorship in Neurodegenerative Diseases grant ; Albertson Parkinson's Research Foundation grant ; National Institute of Neurological Disorders and Stroke grant U54 NS092089; National Institute on Aging grants U01 AG045390 and U24AG21886; National Institutes of Health grants 2R01AG038791, AG035610, K23AG039414, P01 AG019724, P50 AG023501, R01 AG058233, R01 AG26938, R01 NS076837, R35 NS097261, U01 AG045390, U54 NS092089, U54NS092089, and UG3 NS103870; Parkinson Study Group grant ; JPND PreFrontAls Consortium grant ; Nederlandse Organisatie voor Wetenschappelijk Onderzoek grant HCMI 056‐13‐018; ZonMw grants 733050103 and 733050813.
References
- 1. Ghetti B, Wszolek ZK, Boeve BF, et al. Frontotemporal Dementia and Parkinsonism Linked to Chromosome 17 Neurodegeneration: The Molecular Pathology of Dementia and Movement Disorders. Ltd: John Wiley & Sons, 2011:110–134. 10.1002/9781444341256.ch14. [DOI] [Google Scholar]
- 2. Rohrer JD, Ridgway GR, Modat M, et al. Distinct profiles of brain atrophy in frontotemporal lobar degeneration caused by progranulin and tau mutations. NeuroImage 2010;53:1070–1076. 10.1016/j.neuroimage.2009.12.088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Whitwell JL, Jack CR, Boeve BF, et al. Voxel‐based morphometry patterns of atrophy in FTLD with mutations in MAPT or PGRN. Neurology 2009;72:813–820. 10.1212/01.wnl.0000343851.46573.67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Dopper EGP, Rombouts SARB, Jiskoot LC, et al. Structural and functional brain connectivity in presymptomatic familial frontotemporal dementia. Neurology 2013;80:814–823. 10.1212/WNL.0b013e31828407bc. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Whitwell JL, Josephs KA, Avula R, et al. Altered functional connectivity in asymptomatic MAPT subjects. Neurology 2011;77:866–874. 10.1212/WNL.0b013e31822c61f2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Rohrer JD, Nicholas JM, Cash DM, et al. Presymptomatic cognitive and neuroanatomical changes in genetic frontotemporal dementia in the Genetic Frontotemporal dementia Initiative (GENFI) study: a cross‐sectional analysis. Lancet Neurol 2015;14:253–262. 10.1016/S1474-4422(14)70324-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Jiskoot LC, Bocchetta M, Nicholas JM, et al. Presymptomatic white matter integrity loss in familial frontotemporal dementia in the GENFI cohort: A cross‐sectional diffusion tensor imaging study. Ann Clin Transl Neurol 2018;5:1025–1036. 10.1002/acn3.601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Cash DM, Bocchetta M, Thomas DL, et al. Patterns of gray matter atrophy in genetic frontotemporal dementia: results from the GENFI study. Neurobiol Aging. 2018;62:191–196. 10.1016/j.neurobiolaging.2017.10.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Boeve B, Bove J, Brannelly P, et al. The longitudinal evaluation of familial frontotemporal dementia subjects protocol: Framework and methodology. Alzheimers Dement J Alzheimers Assoc. 2020;16:22–36. 10.1016/j.jalz.2019.06.4947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Nasreddine ZS, Phillips NA, Bédirian V, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53:695–699. 10.1111/j.1532-5415.2005.53221.x. [DOI] [PubMed] [Google Scholar]
- 11. Trzepacz PT, Hochstetler H, Wang S, et al. Alzheimer’s Disease Neuroimaging Initiative. Relationship between the Montreal Cognitive Assessment and Mini‐mental State Examination for assessment of mild cognitive impairment in older adults. BMC Geriatr. 2015;15(1); 10.1186/s12877-015-0103-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Knopman DS, Kramer JH, Boeve BF, et al. Development of methodology for conducting clinical trials in frontotemporal lobar degeneration. Brain 2008;131(11):2957–2968. 10.1093/brain/awn234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Kramer JH. Distinctive neuropsychological patterns in frontotemporal dementia, semantic dementia, and Alzheimer disease. Cogn Behav Neurol Off J Soc Behav Cogn Neurol 2003;16:211–218. [DOI] [PubMed] [Google Scholar]
- 14. Weintraub S, Besser L, Dodge H, et al. Version 3 of the Alzheimer Disease Centers’ Neuropsychological test battery in the uniform data set (UDS). Alzheimer Dis Assoc Disord 2018;32:10–17. 10.1097/WAD.0000000000000223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hutton M, Lendon CL, Rizzu P, et al. Association of missense and 5′‐splice‐site mutations in tau with the inherited dementia FTDP‐17. Nature 1998;393:702–705. 10.1038/31508. [DOI] [PubMed] [Google Scholar]
- 16. La Joie R, Perrotin A, Barré L, et al. Region‐specific hierarchy between atrophy, hypometabolism, and β‐amyloid (Aβ) load in Alzheimer’s disease dementia. J Neurosci Off J Soc Neurosci 2012;32:16265–16273. 10.1523/JNEUROSCI.2170-12.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Olney NT, Ong E, Goh S‐YM, et al. Clinical and volumetric changes with increasing functional impairment in familial frontotemporal lobar degeneration. Alzheimers Dement J Alzheimers Assoc 2020;16:49–59. 10.1016/j.jalz.2019.08.196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Tzourio‐Mazoyer N, Landeau B, Papathanassiou D, et al. Automated Anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single‐subject brain. NeuroImage 2002;15:273–289. 10.1006/nimg.2001.0978. [DOI] [PubMed] [Google Scholar]
- 19. Milner B. Interhemispheric differences in the localization of psychological processes in man. Br Med Bull 1971;27:272–277. 10.1093/oxfordjournals.bmb.a070866. [DOI] [PubMed] [Google Scholar]
- 20. Maguire EA, Frackowiak RSJ, Frith CD. Recalling Routes around London: Activation of the Right Hippocampus in Taxi Drivers. J Neurosci 1997;17:7103–7110. 10.1523/JNEUROSCI.17-18-07103.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Poorkaj P, Bird TD, Wijsman E, et al. Tau is a candidate gene for chromosome 17 frontotemporal dementia. Ann Neurol 1998;43:815–825. 10.1002/ana.410430617. [DOI] [PubMed] [Google Scholar]
- 22. Reed LA, Schelper RL, Solodkin A, et al. Autosomal dominant dementia with widespread neurofibrillary tangles. Ann Neurol 1997;42:564–572. 10.1002/ana.410420406. [DOI] [PubMed] [Google Scholar]
- 23. Ygland E, van Westen D, Englund E, et al. Slowly progressive dementia caused by MAPT R406W mutations: longitudinal report on a new kindred and systematic review. Alzheimers Res Ther 2018;10:2 10.1186/s13195-017-0330-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Delisle MB, Murrell JR, Richardson R, et al. A mutation at codon 279 (N279K) in exon 10 of the Tau gene causes a tauopathy with dementia and supranuclear palsy. Acta Neuropathol (Berl) 1999;98:62–77. [DOI] [PubMed] [Google Scholar]
- 25. Hasegawa M, Smith MJ, Iijima M, et al. FTDP‐17 mutations N279K and S305N in tau produce increased splicing of exon 10. FEBS Lett 1999;443:93–96. 10.1016/S0014-5793(98)01696-2. [DOI] [PubMed] [Google Scholar]
- 26. Iijima M, Tabira T, Poorkaj P, et al. A distinct familial presenile dementia with a novel missense mutation in the tau gene. NeuroReport 1999;10:497–501. [DOI] [PubMed] [Google Scholar]
- 27. Skoglund L, Viitanen M, Kalimo H, et al. The tau S305S mutation causes frontotemporal dementia with parkinsonism. Eur J Neurol 2008;15:156–161. 10.1111/j.1468-1331.2007.02017.x. [DOI] [PubMed] [Google Scholar]
- 28. Stanford PM, Halliday GM, Brooks WS, et al. Progressive supranuclear palsy pathology caused by a novel silent mutation in exon 10 of the tau gene: expansion of the disease phenotype caused by tau gene mutations. Brain J Neurol. 2000;123(Pt 5):880–893. [DOI] [PubMed] [Google Scholar]
- 29. Kovacs GG, Pittman A, Revesz T, et al. MAPT S305I mutation: implications for argyrophilic grain disease. Acta Neuropathol (Berl) 2008;116:103–118. 10.1007/s00401-007-0322-6. [DOI] [PubMed] [Google Scholar]
- 30. Malkani R, D’Souza I, Gwinn‐Hardy K, et al. A MAPT mutation in a regulatory element upstream of exon 10 causes frontotemporal dementia. Neurobiol Dis 2006;22:401–403. 10.1016/j.nbd.2005.12.001. [DOI] [PubMed] [Google Scholar]
- 31. Murrell JR, Spillantini MG, Zolo P, et al. Tau gene mutation G389R causes a tauopathy with abundant pick body‐like inclusions and axonal deposits. J Neuropathol Exp Neurol 1999;58:1207–1226. [DOI] [PubMed] [Google Scholar]
- 32. Bronner IF, ter Meulen BC, Azmani A, et al. Hereditary Pick’s disease with the G272V tau mutation shows predominant three‐repeat tau pathology. Brain 2005;128:2645–2653. 10.1093/brain/awh591. [DOI] [PubMed] [Google Scholar]
- 33. Rosso SM, Herpen EV, Deelen W, et al. A novel tau mutation, S320F, causes a tauopathy with inclusions similar to those in Pick’s disease. Ann Neurol. 2002;51:373–376. 10.1002/ana.10140. [DOI] [PubMed] [Google Scholar]
- 34. Herpen EV, Rosso SM, Serverijnen L‐A, et al. Variable phenotypic expression and extensive tau pathology in two families with the novel tau mutation L315R. Ann Neurol 2003;54:573–581. 10.1002/ana.10721. [DOI] [PubMed] [Google Scholar]
- 35. Whitwell JL, Przybelski SA, Weigand SD, et al. Distinct anatomical subtypes of the behavioural variant of frontotemporal dementia: a cluster analysis study. Brain 2009;132:2932–2946. 10.1093/brain/awp232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Whitwell JL, Weigand SD, Boeve BF, et al. Neuroimaging signatures of frontotemporal dementia genetics: C9ORF72, tau, progranulin and sporadics. Brain J Neurol 2012;135(Pt 3):794–806. 10.1093/brain/aws001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Rascovsky K, Hodges JR, Knopman D, et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain 2011;134:2456–2477. 10.1093/brain/awr179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Domínguez‐Vivero C, Wu L, Lee S, et al. Structural brain changes in pre‐clinical FTD MAPT mutation carriers. J Alzheimers Dis JAD 2020;75:595–606. 10.3233/JAD-190820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Moore KM, Nicholas J, Grossman M, et al. Age at symptom onset and death and disease duration in genetic frontotemporal dementia: an international retrospective cohort study. Lancet Neurol 2020;19:145–156. 10.1016/S1474-4422(19)30394-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Swieten JV, Spillantini MG. Hereditary frontotemporal dementia caused by tau gene mutations. Brain Pathol 2007;17(1):63–73. 10.1111/j.1750-3639.2007.00052.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Whitwell JL, Jack CR, Boeve BF, et al. Atrophy patterns in IVS10+16, IVS10+3, N279K, S305N, P301L, and V337M MAPT mutations. Neurology 2009;73:1058–1065. 10.1212/WNL.0b013e3181b9c8b9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Jiskoot LC, Panman JL, Meeter LH, et al. Longitudinal multimodal MRI as prognostic and diagnostic biomarker in presymptomatic familial frontotemporal dementia. Brain 2019;142:193–208. 10.1093/brain/awy288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Chen Q, Boeve BF, Schwarz CG, et al. Tracking white matter degeneration in asymptomatic and symptomatic MAPT mutation carriers. Neurobiol Aging 2019;83:54–62. 10.1016/j.neurobiolaging.2019.08.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Bertrand A, Wen J, Rinaldi D, et al. Early cognitive, structural, and microstructural changes in Presymptomatic C9orf72 carriers younger than 40 years. JAMA Neurol 2018;75:236–245. 10.1001/jamaneurol.2017.4266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Lee SE, Sias AC, Mandelli ML, et al. Network degeneration and dysfunction in presymptomatic C9ORF72 expansion carriers. NeuroImage Clin 2017;14:286–297. 10.1016/j.nicl.2016.12.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Chen Q, Boeve BF, Senjem M, et al. Rates of lobar atrophy in asymptomatic MAPT mutation carriers. Alzheimers Dement Transl Res Clin Interv 2019;5:338–346. 10.1016/j.trci.2019.05.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Ghetti B, Oblak AL, Boeve BF, et al. Invited review: Frontotemporal dementia caused by microtubule‐associated protein tau gene (MAPT) mutations: a chameleon for neuropathology and neuroimaging. Neuropathol Appl Neurobiol 2015;41:24–46. 10.1111/nan.12213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Spillantini MG, Crowther RA, Kamphorst W, et al. Tau pathology in two Dutch families with mutations in the microtubule‐binding region of tau. Am J Pathol 1998;153:1359–1363. 10.1016/S0002-9440(10)65721-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Ikeuchi T, Imamura T, Kawase Y, et al. Evidence for a common founder and clinical characteristics of Japanese families with the MAPT R406W mutation. Dement Geriatr Cogn Disord EXTRA. 2011;1:267–275. 10.1159/000331243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Jones DT, Knopman DS, Graff‐Radford J, et al. In vivo 18F‐AV‐1451 tau PET signal in MAPT mutation carriers varies by expected tau isoforms. Neurology 2018;90:e947–e954. 10.1212/WNL.0000000000005117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Massey LA, Jäger HR, Paviour DC, et al. The midbrain to pons ratio. Neurology 2013;80:1856–1861. 10.1212/WNL.0b013e318292a2d2. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1. Gray and white matter mean w‐maps of symptomatic carriers. Maps show the mean w‐score of gray and white matter across the symptomatic carriers. Symptomatic carriers have atrophy in frontotemporal cortex, insula, uncinate fasciculus, anterior corona radiata, corpus callosum, and inferior longitudinal fasciculus. Maps are thresholded from w≤‐13 to −2. All maps are shown on the Montreal Neurological Institute template brain with the left side of the axial and coronal slices corresponding to the left side of the brain.
Table S1. Participants by Site