Introduction

Tauopathies represent a spectrum of neurodegenerative diseases unified by the pathologic accumulation of hyperphosphorylated tau protein fragments within the central nervous system (CNS), involving both the neuronal and glial compartments [23]. Tau is an abundant microtubule-associated protein, the product of the microtubule-associated protein tau (MAPT) gene on the long arm of chromosome 17. Six isoforms emerge from alternative mRNA splicing of exons 2, 3 and 10 [21]. Three isoforms contain three microtubule-binding repeats (3R-tau) and three others contain four such repeats (4R-tau). Presence or absence of the amino acid sequence encoded by exon 10 determines tau isoforms with 3 or 4 microtubule-binding repeats, 3R-tau and 4R-tau, respectively [38]. The particular cell type and developmental state regulate isoform properties [25, 44], with 3R- and 4R-tau expressed in the adult brain [21].

Classification of tauopathies is based on the cellular and anatomical distribution of tau-immunoreactive structures, the ultrastructure of tau filaments, and the isoform profiles of pathological tau-bands. Two haplotypes (H1/H2) in MAPT, including two major ancestral H1 haplotypes, may influence the phenotype [9, 15].

3R-tau pathology prevails in several forms of sporadic frontotemporal dementia (FTD). Intraneuronal inclusions of hyperphosphorylated tau lacking the sequence of exon 10 are referred to as Pick bodies that are 3R-tau-positive and 4R-tau-negative [1, 50, 53, 67, 68]. Familial tauopathies, like FTD with parkinsonism linked to chromosome 17 (FTDP-17) [22, 28, 62] and its various subclasses, e.g., pallido-ponto-nigral degeneration (PPND) and familial multiple system tauopathy presenile dementia (MSTD), are caused by mutations in MAPT [21]. Intronic mutations alter the tau protein sequence directly, but more often alternative splicing of exon 10 is due to exonic mutations and typically results in neuronal and glial 4R-tau aggregation. This process leads to abundant neurofibrillary tangles (NFT), neuropil threads (NT), coiled bodies (CB) as well as astrocytic plaques and tuft-shaped astrocytes. These inclusions contain all three repeat isoforms of tau protein [6]. Sporadic tauopathies include Pick disease (mainly featured by 3R-tau inclusions, rarely only with insoluble 4R-tau) [50, 75], and those showing widespread 4R-tau inclusions in the form of glial fibrillary tangles in cortical and subcortical areas in progressive supranuclear palsy (PSP), characterized by tuft-shaped astrocytes and neuronal globose NFTs [15, 16], corticobasal degeneration (CBD) characterized by astrocytic plaques, and the presence of abnormal tau + thread-like processes in both cortical and subcortical gray and white matter [13, 14, 16, 73], whereas in argyrophilic grain disease (AGD) 4R-tau inclusion are restricted to the limbic system [38, 48]. While mixed 3R- and 4R-tau pathology is present in NFTs of AD [11], in tangle dominant dementia (TDD) [29, 30] and in cases of hippocampal sclerosis with dementia but without seizures or hypoxic brain injury [2], hippocampal sclerosis with sole 4R-tau-positive inclusions in the dentate gyrus was also reported [47]. In FTLD with MAPT mutation, cases with different mutations do not have a consistent pattern of pathology or biochemistry [8, 19, 61, 69].

There have been some reports of atypical variants of 4R-tauopathies that cannot be classified into known disease categories [4]. Some were considered primary progressive aphasia or variants of CBD [17, 59, 64], others presenting as sporadic FTD with parkinsonism and/or motor neuron disease [18, 56, 64], leukoencephalopathy with tau deposits in the white matter [52], as frontotemporal lobar degeneration (FTLD) with white matter tauopathy and globular glial inclusions [36], or as pallidoluysian atrophy (PLA) [49]. There are also subforms of the Parkinsonism dementia complex (PDC) of Guam featured by predominantly 4R-tau-positive fine granules in cerebral white matter [72].

A distinct 4R-tauopathy to the best of our knowledge has never been linked to classical Benson’s syndrome, also known as posterior cortical atrophy (PCA, [3]). This rare, clinically homogenous but pathologically heterogenous syndrome [74] is a progressive lobar dementia, presenting with disturbances of visual perception as the most prominent prodromal sign. It is associated with visual agnosia, visual hallucinations [32], visuo-constructive apraxia (Balint’s syndrome), alexia, acalculia, right–left confusion, and agraphia (Gerstmann’s syndrome), language and memory functions tending to be preserved for a long time [34]. Age of onset is around 60 years. Diagnosis is based on neuro-psychological and imaging findings [39, 45].

PCA, sometimes considered as the visual variant of AD [40], is caused by predominant involvement of occipito-parietal brain areas by various neuropathologies. In the majority of cases autopsies revealed Alzheimer-typical pathology in distinct posterior distribution [5, 26, 35], while others reported one autopsy case each of AD, Creutzfeldt-Jakob disease (CJD) and subcortical gliosis [70]. Among nine autopsies, seven were AD, showing higher NFT density in Brodmann areas 17 and 18 than in hippocampus, and two cases of CJD [66]. The largest autopsy series of 21 cases included AD (n = 13), CBD, Lewy body variant of AD (LBV/AD), dementia with Lewy bodies (DLB) and subcortical gliosis (one case each) [58]. While Balint’s syndrome and visuomotor ataxia have been described during the clinical course of CBD without postmortem confirmation [46, 54], two autopsy-proven cases of CBD presented with prominent visuospatial dysfunction resembling PCA, followed by apraxia and parkinsonism later in the course of the disease [65].

Here, we describe the first case of PCA with complete clinico-imaging work-up in whom postmortem examination revealed a probably sporadic 4R-tau pathology with a hitherto unknown neuropathologic phenotype characterized by the presence of severe tauopathy in both the gray and white matter with neuronal and glial inclusions consisting of 4-repeat (4R) tau. Despite some morphologic discrepancies, the neuropathologic findings in this case appear to be consistent with (atypical) CBD.

Clinical case report

BP, a 55-year-old retired right-handed butcher, presented with complaints of an ongoing loss of vision. He had experienced a progressive decline in spatial orientation and visual functions over a period of 2 years. At the time of examination, he was almost unable to recognize objects, faces or large-scale letters, failed to grasp objects on a table in front of him, or to eat from a plate. He also had severe problems with dressing, calculating, writing and the use of everyday objects. In contrast, his ability to identify objects from their typical sound or persons from their voice was well preserved.

Despite his severe visual impairment which left him unable to read the headlines in newspaper or count the number of upheld fingers, BP was still able to take long walks and to ride his bicycle without any assistance.

Neurological examination performed 2½ years after the onset of clinical symptoms demonstrated the full-blown picture of PCA [34, 66]. He had preserved insight in his disorder, partial Balint’s syndrome including optic ataxia and visual agnosia with simultanagnosia and prosopagnosia. He showed agraphia, acalculia, right–left confusion and finger agnosia consistent with complete Gerstmann syndrome. He also presented constructional dyspraxia, ideomotor apraxia and dressing apraxia, alexia and mild hemineglect to the right.

Speech comprehension, expressive language functions as well as repetition of words and sentences were intact; however, speed of information processing was markedly reduced. This was reflected by reduced scores on tests assessing semantic and phonologic word fluency (Regensburger word fluency test: below the 10th percentile for his age group). Perception of colors was difficult to assess as the patient reported an inability to recognize even large objects. Nonetheless the patient was able to sort discs by color, but his exploration strategies were erratic, repetitive and slow, and he was unable to name the color of discs. General knowledge of the colors of natural and artificial objects (e.g. “Imagine a cherry/kiwi/glass of Coca Cola—what color does it have?”) was severely impaired (50% correct answers only).

Only verbal memory could be tested due to severe visual impairment. Episodic memory (immediate and delayed recall of Wechsler Memory Scale, Revised Edition) and list learning (German Version of the Rey Auditory Verbal Learning Test) were markedly affected (below the 5th percentile for his age group). Word recognition was slightly reduced. There was no evidence for abnormal extraocular eye movements, and for pyramidal or extrapyramidal signs on neurologic examination. The patient’s only behavioral change was mild reactive depression. No family members presented with similar symptoms or other neurodegenerative diseases in the past.

Laboratory findings were within normal limits. CSF analysis showed no evidence for inflammation and was negative for protein 14-3-3. Total tau protein was slightly elevated at 275 pg/ml (norm of laboratory < 160–195 pg/ml), phospho-tau (181P) was 35 pg/ml and within normal limits (cut off 61 pg/ml). β-Amyloid 1–42 was elevated to 1,556 pg/ml (norm of laboratory > 487–849 pg/ml).

Brain MRI (1.5T Philips-Interna Scanner, Philips Medical Systems, Best, The Netherlands) performed 2 and 2½ years after the onset of symptoms revealed enlargement of cortical sulci particularly in both parieto-occipital brain regions (focal parietal atrophy) which was clearly disproportionate to frontal brain areas (Fig. 1). There were no abnormalities in corpus callosum, tegmental and hippocampal areas. On FLAIR sequences there was a faint hyperintensity in Broca’s secondary visual association cortex bilaterally. Single-voxel proton-spectroscopy in the area of hyperintensity revealed a substantial increase in myo-inositol.

Fig. 1
figure 1

Brain magnetic resonance imaging. Upper row First two images include visual cortex and visual association cortex. T1-weighted images on the left demonstrate widening of occipital sulci in visual association centers with FLAIR sequences in the middle demonstrating faint hyperintensity in Broca’s secondary visual association cortex bilaterally (arrows). The right picture demonstrates severe widening of cortical sulci in parietal regions which is disproportionate to frontal areas. Lower row Midsagittal section on the left shows no abnormality in corpus callosum and tegmental areas, sagittal view including the left parietal and frontal lobes highlights focal parietal atrophy (middle picture). Coronal view demonstrates normal volume of hippocampal formation bilaterally (right picture)

Full size image

Positron emission tomography (Siemens-ECAT scanner, Siemens Medical Systems, Erlangen, Germany) demonstrated severe bilateral hypometabolism in visual association cortices, which extended into parietal regions up to hemispheric convexities being more pronounced on the left. On both sides hypometabolic regions were well demarcated. The visual perception area and the entire frontal lobe were spared (Fig. 2).

Fig. 2
figure 2

18F-FDG positron emission tomography (PET) study demonstrating severe bilateral hypometabolism in visual association cortices (arrows) extending to parietal region, more pronounced on the left side; the sparing of visual perception areas and relative perseveration of metabolism in primary motor and sensory areas and basal ganglia

Full size image

Based on the assumption that PCA commonly represents a focal variant of AD, treatment attempt was made with a cholinesterase inhibitor in combination with memantine but there was no treatment effect. BP’s symptoms and signs progressed gradually, and due to severe functional impairment requiring intense assistance in activities of daily living, he was admitted to a nursing home 1 year after first diagnosis of PCA was made. BP died 6½ years after presumed disease onset from bronchopneumonia and right heart failure.

Neuropathology

Neuropathological assessment was performed according to standard protocols. Formalin-fixed, paraffin-embedded blocks from multiple brain regions (frontal, temporal, parietal, occipital cortex, hippocampus, basal ganglia, amygdala, brainstem, and cerebellum) were examined using routine stains, Bodian’s and Gallyas silver impregnations, and immunohistochemistry for tau (monoclonal AT8; 1:200, Innogenetics, Ghent, Belgium), monoclonal antibodies against 4R-tau (RD4; clone 1E/A5, dilution 1:400) and against 3R-tau (RD3, clone 8E5/C11, dilution 1:1,000; Upstate, Charlottesville, VA, USA), with additional pretreatment with citrate buffer (2 × 5 min) and microwaving, anti-β-amyloid (Aβ) (monoclonal antibody 4G8, Signet Labs, Dedham, MA, USA), α-synuclein (monoclonal and polyclonal rabbit antibodies, Chemicon, Hofheim, Germany and Signet Labs), ubiquitin (Dako, Glastrup, Denmark; dilution 1:800), anti-TDP-43 (dilution 1:2,000, Abnova Corp., Taipei, Taiwan), anti-glial fibrillary acidic protein (dilution 1:3,000, Dako). Sections were deparafinized and incubated with primary antibodies overnight at 5°C, then visualized by the avitin–biotin–peroxidase complex using fast red salt as chromogen. The spinal cord was not available for examination. Examination of genomic DNA extracted from blood was performed, whereas frozen tissue samples for biochemical examination were not available.

Genetics

Eleven MAPT exons (exon 1–5, 7, and 9–13), including exon/intron boundaries, were amplified from genomic DNA using the following PCR conditions: 95°C for 15 min; 10 cycles of 94°C for 30 s, 60°C for 30 s (decrease 1°C each cycle), and 72°C for 90 s; 25 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 90 s; 72°C for 6 min. Primer sequences are available upon request. All primer pairs were designed with alternate forward MTR and reverse M13 tails to facilitate dye-primer sequencing of both DNA strands. Each exon was individually analyzed by direct sequencing of each PCR product on an automated DNA sequencer (Long-Read Tower, Visible Genetics) using the dye-primer cycle sequencing chemistry (GE Healthcare).

Macroscopic findings

At naked eye, the brain (weight unfixed 1,200 g) showed severe diffuse bilateral cortical atrophy in the parietal and parieto-occipital areas with relative preservation of the occipital poles and frontal areas (Fig. 3). Coronal brain sections revealed almost symmetrical atrophy of the parietal convolutions, much less of the frontal and mediobasal temporal convolutions, with sparing of hippocampus, frontobasal regions, and corpus callosum. The cerebral ventricles were not enlarged; substantia nigra, brainstem, and cerebellum were macroscopically unremarkable. There was no considerable atherosclerosis of the basal cerebral arteries.

Fig. 3
figure 3

Macroscopic view of the brain demonstrating severe parietal atrophy slightly more pronounced on the left side

Full size image

Histopathology and immunohistochemistry

Common histopathological features comprised mild to moderate diffuse neuronal cell loss with mild superficial spongiosis in the frontal cortex and, less severe, in the frontobasal cortex, in temporal, transentorhinal and entorhinal cortex, pre- and prosubiculum and subareas CA 1-3 of the hippocampus, with occasional ballooned neurons showing diffuse α-B-crystallin immunoreactivity (not shown). Argyrophilic tau-positive neuropil threads without NFTs were seen in the olfactory tract and bulb (not shown). There was considerable bilateral thinning of the parietal convolutions with diffuse neuronal loss and superficial spongiosis in the parietal cortex, less severe in the occipital cortex, associated with considerable diffuse reduction of myelin and, less extensively, of axons, along with microglial activation, correlating with the severity of myelin loss and of white matter tau pathology. Early stage neurofibrillary degeneration with “pre-tangles” immunoreactive for 3R-tau corresponding to Braak stage II was observed (not shown). Amyloid pathology of brain parenchyma and vasculature as well as TDP-43 and α-synuclein inclusions were totally absent.

The most significant alteration was the presence of abundant tau-immunoreactive inclusions in both cortical and subcortical neurons, usually presenting as classical NFTs, neuropil threads in the neuropil, and multiple tau-positive inclusions in both astroglia and oligodendroglia (coiled bodies), involving both the frontal cortex and white matter (Fig. 4a, b), the parieto-occipital cortex and white matter (Fig. 4c, d), and sector CA 1 of hippocampus (Fig. 4e, f). There were multiple tau-positive threads and astroglial inclusions in putamen, claustrum, thalamus (Fig. 4g), and subthalamic nucleus. Substantia nigra showed mild neuronal loss without anatomical predominance and severe tau pathology (Fig. 4h–j). Similar involvement was seen in the pontine nuclei, the medullary reticular formation and hypoglossal nucleus (Fig. 4h–j).

Fig. 4
figure 4

a Frontal cortex with neurofibrillary tangles and neuropil threads (scale bar 20 μm). b Frontal white matter showing comma-like tau-positive inclusions (coiled bodies) in oligodendroglia and neuropil threads (scale bar 50 μm). c Occipital cortex showing tau-positive neuronal spherical cytoplasmic inclusions (scale bar 20 μm). d Occipital white matter with coiled bodies and neuropil threads (scale bar 50 μm. e Hippocampal CA1 subregion showing 4R-tau-positive NFTs (scale bar 100 μm). f Hippocampal CA1 subregion showing 4R-tau-positive spherical cytoplasmic inclusions in neurons (scale bar 50 μm) g Medial thalamus showing neuronal loss and multiple 4R-tau-positive neuronal inclusions (scale bar 200 μm). h Substantia nigra compacta with mild neuronal loss and multiple tau-positive neuronal inclusions (scale bar 100 μm). i Neuronal inclusions and neuropil threads in substantia nigra are 4R-tau-immunoreactive (scale bar 50 μm). j Pontine base nuclei with multiple tau-positive neuronal inclusions (scale bar 200 μm). k Medulla oblongata, magnocellular reticular formation with RD4-immunoreactive neuronal inclusions and neuropil threads (scale bar 50 μm). l Medulla oblongata, hypoglossal nucleus with tau-positive NFTs, globose neuronal inclusions and neuropil threads (scale bar 50 μm). m Frontal cortex with many Gallyas-positive neuropil threads and one only slightly Gallyas-positive neuron (scale bar 50 μm). a–d, e, g, j, l AT8 antibody; f, h, i, k RD4 antibody; m Gallyas

Full size image

All neuronal and glial inclusions were uniformly immunoreactive for the antibody RD4 with consistently negative results for the antibody RD3. Typical argyrophilic, tau-positive tufted astrocytes as seen in PSP were rarely observed, while oligodendroglial coiled bodies were frequent (Fig. 4b, c). Double-labeling revealed occasional co-expression of tau and GFAP in the astrocytes (not shown). The astrocytic tau lesions, but only very few tau-positive neurons were ubiquitinated (not shown). Occasional spheric cytoplasmatic inclusions immunoreactive for RD4 but not RD3 were observed in neurons (Fig. 4c, d). Anti-RD3 immunolabeled NFTs and neuropil threads were occasionally seen in hippocampus, while 3R-tau-positive Pick bodies, globose NFTs, argyrophilic grains and astrocytic plaques were not present. Glial and neuronal tau lesions, including the NFT-like neuronal inclusions, were well recognized by Bodian staining. Neuropil threads and only occasional neuronal inclusions were Gallyas-positive (Fig. 4m).

Anatomical distribution

Neuronal loss was observed in frontal, temporal, anterior cingulate cortices, more severe in the parietal cortex, CA 1-3 subareas of hippocampus, subiculum, and nucleus accumbens. The basal nucleus of Meynert and the basal ganglia were only mildly affected. There was some nerve cell loss in the substantia nigra and less so in other brainstem nuclei. Tau pathology was more widespread than nerve cell loss in both cortical and subcortical gray matter involving both tau-positive NFTs and neuropil threads as well as astrocytic inclusions (Table 1). Spherical tau-positive cytoplasmic inclusions were restricted to the frontal and temporal cortices (distributed almost randomly throughout cortical layers) and the hippocampal CA 1 region. Tau-positive tangles were observed in most regions investigated, including the dentate gyrus, substantia nigra, amygdala, basal ganglia, thalamus, and many brainstem nuclei (Table 1). Oligodendroglial coiled bodies and astrocytic tau-immunoreactive inclusions were observed in most cortical regions as well as in hemispheral white matter and subcortical myelinated areas. Glial inclusions were infrequent in the transverse pontine and pyramidal tracts, while the inferior olives, cerebellar cortex and white matter, and optic chiasm were preserved (Table 2).

Table 1 Distribution of neuronal loss and tau immunoreactivity in gray matter
Full size table
Table 2 Distribution of tau immunoreactivity in white matter
Full size table

Genetics

The patient was analyzed for mutations in the coding exons and exon–intron boundaries 1–5, 7, and 9–13 of the microtubule-associated protein tau (MAPT) gene. We could not detect any pathogenic changes in MAPT.

Summary of neuropathologic findings

The following anatomical features were characteristic: (1) neuronal loss predominantly in parietal and parieto-occipital cortices, less severe in frontal, temporal, visual occipital cortices and hippocampus, striatum and substantia nigra. Severe 4R-tau pathology in cortical and subcortical neurons (ballooned neurons with tau-positive NFTs or round cytoplasmic inclusions), usually visible with Bodian staining, and ubiquitinous occurrence of neuropil threads in the gray nuclei. 3R-tau-positive NFTs in hippocampus were rare, and generally absent in cerebral cortices. Preservation of the cerebellum and only mild involvement of inferior olives. (2) Prominent involvement of the hemispheral white matter with abundant neuropil threads, globular, glia-associated 4R-tau-immunoreactive inclusions in astroglia and oligodendroglia (coiled bodies). Less severe but consistent involvement of the interconnecting (anterior commissure), frontopontine and corticospinal (pyramidal) tracts. Correlation of the severity of white matter tau pathology with myelin density (not shown), accompanied by 4R immunoreactivity in astrocytic processes (tufted astrocytes) and in oligodendroglia (coiled bodies). Absence of astrocytic plaques, amyloid pathology in cerebral parenchyma and vasculature, 3R-tau-positive Pick bodies, and TDP-43 and α-synuclein-positive inclusions. Tau-positive neuropil threads were usually both Bodian- and Gallyas-positive, whereas only very few affected neurons were ubiquitinated and showed only mild Gallyas positivity.

Discussion

The reported patient, with negative family history (which does not entirely exclude a mutation in MAPT), presenting the full-blown clinical picture of PCA of rather long duration (6½ years) [34, 45, 51, 66], showed severe symmetrical parieto-occipital atrophy and severe bilateral hypometabolism in visual association cortices (see [71]). MRI-spectroscopy revealed increased levels of the gliosis-related marker myo-inositol in parietal areas with most pronounced atrophy and hypometabolism. Both neuroimaging and CSF biomarker findings (increase of both total tau protein and β-amyloid) showed differences from typical AD findings [7, 24, 39, 60].

Neuropathologic examination revealed a unique phenotype of 4R-tauopathy with involvement of both the gray and white matter, diffuse neuronal loss, NFTs, NTs and tau deposition in astroglia and oligodendroglia in parietal, less temporal, occipital, frontal cortex and hippocampus, and severe involvement of the white matter with tau-positive NTs, comma-like inclusions in oligodendroglia (coiled bodies), and astroglia, with various involvement of the basal ganglia, substantia nigra and many brainstem nuclei but preservation of the cerebellum. The distribution pattern of neuronal loss and tau pathology was compatible with the suggested clinical diagnosis of PCA and correlating neuroimaging findings, while despite considerable involvement of substantia nigra and other brainstem nuclei no extrapyramidal or motor neuron symptoms had been reported clinically. Despite involvement of large parts of cerebral cortex and white matter overt dementia only developed in the final stages of the disease. There was no amyloid pathology. The lack of globose NFTs, astrocytic plaques or argyrophilic grains did not fit into current morphologic criteria for most of the hitherto reported sporadic 4R-tauopathies such as AGD, PSP [14], FTLD, parkinsonism and motor neuron disease [18, 56], PDC of Guam [72], from tauopathies with massive involvement of white matter [6, 20, 57] and white matter tauopathy with globular inclusions [36]. The distribution pattern of 4R-tau inclusion pathology also differed from most hitherto known 4R-tauopathies [41].

Both the quality and distribution of the tau-positive neuronal and glial pathologies appear to be consistent with the diagnosis of possible corticobasal degeneration (CBD) [13, 14], the presenting clinical symptoms of which only rarely resembled those of PCA [65]. However, the present case showed some morphological differences from “classical” CBD: (1) in contrast to Gallyas positivity of neuropil threads, only a minority of tau-positive neuronal and glial inclusions was Gallyas-positive. (2) Astrocytic plaques—not requested for the diagnosis of CBD, but when present having diagnostic significance—were completely absent. (3) While the cerebellum is usually unaffected in CBD, the dentate nucleus may be involved [63], and there may be scattered Purkinje axonal torpedoes and Bergman gliosis [13], all of which were absent in our case. Despite some histologic similarities, i.e. the presence of coiled bodies and argyrophilic threads in a diffuse distribution, the present case differs from an atypical late-onset dementia, in which these lesions predominantly involved the limbic areas [55].

Lack of TDP-43 immunoreactivity of the neuronal and glial inclusions exclude any form of sporadic or familial FTD or FTLD [33, 42, 43]. Lack of 3R-immunoreactivity in the neuronal and glial inclusions and of typical Pick bodies distinguishes our case also from Pick disease, a representative sporadic 3R-tauopathy [12, 27]. In contrast, PHFs and NTs in both AD and TDD are made of both 3R- and 4R-tau [10, 11, 29, 30]. Lack of α-synuclein inclusions [classical glial cytoplasmic inclusions (GCI or Papp-Lantos bodies)] excluded a diagnosis of multiple system atrophy (MSA) [31, 37].

In order to determine whether a pathogenic mutation in MAPT was the cause of the reported disorder, the coding region of the brain-specific isoform was sequenced. No base pair changes in MAPT could be detected. These findings indicate a sporadic 4R-tauopathy.

In summary, this is the first case with the fully developed clinical and neuroimaging picture of PCA, morphologically presenting as a distinct phenotype of 4R-tauopathies that morphologically closely resembles (atypical) CBD.