Alzheimer’s Disease May Follow Four Distinct Tau Pathways, Including a Newly Identified Motor-Cortex Pattern
Alzheimer’s disease may be far less uniform than its traditional staging system suggests. A post-mortem study of 144 people with advanced Alzheimer’s-related brain changes has identified four distinct patterns in the distribution of neurofibrillary tangles, including a previously described but under-recognized subtype in which tau pathology is unusually concentrated in the brain’s primary motor and sensory regions. The pattern, called “idiotypic-susceptible” Alzheimer’s disease by the researchers, was strongly associated with corticobasal syndrome, a neurological condition involving movement problems, impaired coordination and asymmetric motor symptoms. The findings suggest that the location of tau damage—not simply its overall severity—could help explain why Alzheimer’s disease sometimes appears primarily as memory loss, but in other people emerges as language impairment, visual dysfunction, behavioral change or movement disorder.
The study, published in Acta Neuropathologica, expands on the framework developed by German neuropathologists Braak and Braak, whose staging system has shaped Alzheimer’s research for decades. In the conventional model, tau-positive neurofibrillary tangles appear first in areas near the medial temporal lobe, then spread through the hippocampus and into association cortices involved in complex cognition. At advanced stages, the pathology reaches “idiotypic” or primary cortices, including the motor, somatosensory and visual areas. A typical end-stage Braak pattern therefore contains the greatest tangle burden in the hippocampus, a moderate burden in association regions and the least in primary cortices. The new results indicate that this sequence is not an inevitable template. In a substantial minority of cases, tau accumulates disproportionately outside the hippocampus, potentially disrupting brain systems that govern movement, language, executive function or visual processing.
John L. Robinson of the University of Pennsylvania and colleagues examined brains from individuals who had received a high level of Alzheimer’s disease neuropathologic change according to National Institute on Aging–Alzheimer’s Association criteria. The cohort included people diagnosed during life with late-onset Alzheimer’s disease, early-onset Alzheimer’s disease, behavioral-variant frontotemporal dementia, corticobasal syndrome, logopenic primary progressive aphasia and posterior cortical atrophy. These clinical syndromes differ sharply in their dominant symptoms. Memory impairment is typical in conventional Alzheimer’s disease, whereas logopenic aphasia causes word-finding and repetition problems, posterior cortical atrophy affects visual perception and spatial processing, and corticobasal syndrome can produce stiffness, clumsiness and difficulty controlling one side of the body. Behavioral-variant frontotemporal dementia is characterized by changes in personality, judgment and executive function, although Alzheimer’s pathology is an uncommon cause of that syndrome.
To measure tau distribution, the researchers used immunohistochemistry, a tissue-staining technique that attaches antibodies to phosphorylated tau deposits so they can be seen under a microscope. They sampled eight brain regions: the CA1 sector and subiculum of the hippocampus; the middle frontal, superior temporal and inferior parietal association cortices; and the primary motor, primary somatosensory and primary visual cortices. Digital images of stained tissue were scanned and analyzed with QuPath software. Within manually outlined regions of interest, the investigators divided tissue into small 175-square-micrometer tiles and counted neurofibrillary tangles by hand, converting the results into standardized densities per square millimeter. The analysis produced more than 1,100 usable tangle-density measurements, with repeated counts showing strong agreement between assessments.
The investigators then compared average tau burdens across the three broad anatomical zones—hippocampal, association and idiotypic cortex—and calculated ratios between them. These ratios were used to assign each case to one of four mutually exclusive patterns. The typical Braak subtype, found in 87 of the 144 cases, showed heavy hippocampal involvement, intermediate association-cortex pathology and relatively light primary-cortex pathology. The idiotypic-susceptible subtype, identified in 23 cases, displayed the opposite emphasis: relatively low tau burden in the hippocampus and association cortex but high tangle densities in primary regions, especially the motor and somatosensory cortices. Twenty-four cases were classified as associative-predominant, with a high burden in association cortices and comparatively little hippocampal involvement. Ten were limbic-predominant, with especially heavy hippocampal pathology and relatively limited association-cortex involvement.
The anatomical patterns mapped onto clinical symptoms with striking regularity. More than half—56 percent—of the corticobasal syndrome cases belonged to the idiotypic-susceptible group, while the remaining cases followed the typical Braak pattern; neither the associative-predominant nor limbic-predominant subtype appeared in that clinical group. The researchers say this is biologically plausible because the primary motor and somatosensory cortices are central to movement planning, body sensation and coordination. The associative-predominant pattern was most common in behavioral-variant frontotemporal dementia, accounting for 53 percent of cases, and was also frequent in logopenic primary progressive aphasia. In posterior cortical atrophy, about half of the cases showed the typical pattern, while idiotypic-susceptible and associative-predominant patterns together accounted for much of the remainder. By contrast, 74 percent of late-onset and 76 percent of early-onset amnestic Alzheimer’s cases had the typical Braak distribution.
The results also revealed demographic and genetic distinctions between the groups. Individuals with idiotypic-susceptible or associative-predominant pathology tended to die at younger ages and had shorter disease durations than those with the typical Braak pattern. Limbic-predominant cases, in contrast, were older at death. The researchers also found that the MAPT H1H1 haplotype—a common genetic configuration affecting the gene that encodes the tau protein—was most prevalent in the limbic-predominant subtype and least common in the idiotypic-susceptible and associative-predominant subtypes. The frequency of the APOE ε4 allele did not differ significantly among the groups. The study further found that female sex and longer disease duration were associated with greater hippocampal tangle burdens, while younger age at death and longer disease duration predicted higher tau densities in association and idiotypic cortices. These findings indicate that subtype alone does not determine pathology; age, sex, disease length and genetic background also shape the final pattern.
Importantly, the distribution of tau did not appear to be explained simply by the amount of amyloid plaque in each region. In a subset of 20 brains, the researchers quantified the percentage of tissue occupied by β-amyloid deposits in representative hippocampal, association and primary cortical areas. Amyloid burden was highest in association cortices, averaging 8.6 percent of the measured area, followed by idiotypic cortices at 5.4 percent and limbic regions at 3.1 percent. Yet regional amyloid levels did not correlate with regional neurofibrillary-tangle density, and plaque burden was generally similar across tau subtypes. This dissociation supports the idea that the cellular processes through which amyloid pathology promotes tau aggregation may vary by brain region. Differences in neuronal vulnerability, network connectivity, tau-seed biology or immune responses could help determine why one person develops motor-cortex tau while another accumulates it predominantly in the hippocampus.
To test whether their categories were artifacts of the chosen statistical cutoffs, the researchers applied an independent machine-learning method called k-means clustering. This approach groups cases according to similarities in their measured features without assigning them in advance to named categories. After standardizing regional tau measurements and accounting for a small number of missing values, the analysis also produced four clusters. The cluster dominated by high overall hippocampal and cortical burden corresponded mainly to typical Braak cases, while the idiotypic and associative clusters closely matched the idiotypic-susceptible and associative-predominant subtypes. Seven of the ten limbic-predominant cases fell into a low-association, high-hippocampal cluster. The agreement between the two approaches strengthens the evidence that the patterns reflect genuine biological variation rather than arbitrary divisions, although the authors emphasize that their findings require replication in larger, independent cohorts.
The study has limitations that temper its immediate clinical implications. It examined brains after death, so it cannot yet establish when each tau pattern emerges or whether the subtypes can be reliably detected during life. The investigators measured neurofibrillary tangles, but not the much more abundant tau-containing dystrophic neurites that extend through the cortex and may influence symptoms. Tissue-processing methods, antibody selection and image-analysis protocols can also differ between laboratories. In addition, atypical Alzheimer’s presentations are relatively rare, leaving some clinical groups small, and the selected cases came largely from a specialized research brain bank. Even so, the findings challenge the idea that a single Braak stage captures the full biology of advanced Alzheimer’s disease. As tau-PET imaging and blood-based biomarkers become more precise, recognizing regionally distinct pathological trajectories could improve diagnosis, clarify why patients develop different symptoms and help determine whether future tau-directed treatments work differently across Alzheimer’s subtypes.
Subject of Research: Regional neurofibrillary-tangle patterns and clinically relevant tau subtypes in Alzheimer’s disease
Article Title: Idiotypic-susceptible Alzheimer’s disease: a clinically relevant, neurofibrillary tangle subtype
Article References: Robinson JL, Cai H, Loh NJ, et al. “Idiotypic-susceptible Alzheimer’s disease: a clinically relevant, neurofibrillary tangle subtype.” Acta Neuropathologica 151, article 51 (2026). Original research article
Image Credits: AI Generated
DOI: 10.1007/s00401-026-03013-6
Keywords: Alzheimer’s disease, tau pathology, neurofibrillary tangles, Alzheimer’s subtypes, corticobasal syndrome, atypical dementia, hippocampal-sparing Alzheimer’s, primary cortex
Tags: advances in Alzheimer’s disease neuropathologyAlzheimer’s disease clinical heterogeneityAlzheimer’s disease post-mortem brain analysisAlzheimer’s disease tau pathologybrain region-specific tau distributioncorticobasal syndrome neurodegenerationdistinct Alzheimer’s tau pathwaysmotor-cortex tau patternneurodegenerative movement disordersneurofibrillary tangle subtypestau pathology staging and classificationtau-driven Alzheimer’s clinical phenotypes
