Howard Rosen
Professor
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (3)
Frontotemporal lobar degeneration with tau inclusions (FTLD-tau) comprise a class of fatal heterogeneous neurodegenerative diseases. Approximately 10% arise from pathogenic MAPT mutations and often cause severe, early-onset disease with pathology that is distinct yet partially overlapping with sporadic cases. Here, we evaluated post-mortem tissue from a patient with FTLD-tau due to MAPT S305I showing neuropathology most consistent with argyrophilic grain disease (AGD), a prevalent limbic tauopathy of aging. Structures determined by cryo-electron microscopy reveal tau filament folds that differ from those found in sporadic AGD or other tauopathies and feature a 4-layer architecture stabilized by the Ile substitution within its core. Comparative structural analysis reveals conserved motifs are shared among AGD, corticobasal degeneration, and MAPT P301T. A well-defined density stacks along a cationic cleft, indicative of a bound RNA-like polyanion or small-molecule. In vitro analysis shows the S305I mutation promotes fibrilization relative to normal tau. These results demonstrate that MAPT S305I stabilizes a distinct aggregation-prone tau fold that likely contributes to disease pathology and heterogeneity beyond its known splicing defects, and underscore potential limitations of using the most pathologically similar genetic form as a model for sporadic FTLD-tau.
Year:
2026
Importance Accurate clinical and pathological diagnoses are essential in neurodegenerative diseases, especially given the emergence of pathology-specific disease-modifying therapies. However, diagnostic accuracy remains challenging due to heterogeneous clinical presentations, complexity of integrating multimodal data, and limited access to multidisciplinary expertise. Primary Progressive Aphasia (PPA) exemplifies these challenges, requiring specialized clinical, neuropsychological, and imaging evaluations. Generative artificial intelligence (AI), powered by large language models, may offer scalable diagnostic support in this context. Objective To evaluate the diagnostic performance of an agentic generative AI system in classifying prototypical PPA cases by clinical syndrome and underlying pathology. Design Retrospective diagnostic validation study using a multi-agent generative AI architecture simulating expert-level reasoning. Setting Single tertiary academic referral center (University of California San Francisco, Memory and Aging Center). Participants Fifty-four individuals with a definite diagnosis of PPA and post-mortem confirmation (18 semantic [svPPA], 17 logopenic [lvPPA], 19 nonfluent [nfvPPA]), selected as prototypical cases with congruent clinical, imaging, and pathological profiles. Exposure Multimodal input data, including clinical notes, neuropsychological and language assessments, and MRI brain images, were processed through a multi-agent architecture. The system generated diagnostic predictions under two conditions: (1) open-ended diagnosis from a set of 15 neurodegenerative clinical syndromes; (2) constrained classification of PPA variant and underlying neuropathology. Main Outcomes and Measures Generative AI system diagnostic accuracy for clinical syndrome and pathology, based on expert clinical diagnoses and post-mortem confirmations as gold standard. Results In the open-ended setting, the system correctly identified PPA in 49 of 54 cases (90.7%, chance level=6.7%). When constrained to PPA, it achieved 100% accuracy for svPPA and nfvPPA, and 94.1% for lvPPA as primary prediction. Neuropathological predictions were most accurate for FTLD-TDP type C (100%) and FTLD-4R tau (100%), and high for Alzheimer’s disease (94.4%). The full diagnostic pipeline of all 54 cases was completed in under 10 minutes. Conclusions and Relevance The AI system demonstrated expert-level performance in classifying prototypical PPA cases, integrating multimodal data and mirroring specialist reasoning. Its speed and accuracy support its potential role in extending access to specialized diagnostic expertise, particularly in non-tertiary settings. Further validation in larger and more heterogeneous populations is warranted.
Year:
2025
Background and Objectives Converging evidence hints at neurodevelopmental effects in people at risk of genetic frontotemporal dementia (FTD), including associations between FTD-causing mutations and neurodevelopmental disorders, and differences in young adult mutation carriers compared to familial non-mutation carriers in total intracranial volume (TIV) and cognition. We aimed to investigate TIV and educational attainment differences between adult mutation carriers and familial non-mutation carriers, as measures of the structural and functional neurodevelopmental effects of the FTD-causing genetic mutations. Methods This cross-sectional cohort study was facilitated through the FTD Prevention Initiative (FPI). Participants, aged 18 to 86 years, were pathogenic mutation carriers of GRN , MAPT , or C9orf72 , or familial non-carriers. ANCOVAs were computed per gene to compare outcome means for the main effect of group by carrier status, while controlling for birth decade, sex, and visit site (to account for unique scanners and educational systems). Pearson’s correlations were used to examine associations between TIV and education. Results Nine-hundred two mutation carriers (mean±SD; age=50.0±13.2 years, sex=55% female, n ( GRN )=298, n( MAPT )=187, n ( C9orf72 )=417) were compared to 532 familial non-carriers (age=48.0±12.9 years, sex=58% female, n ( GRN )=201, n( MAPT )=114), n ( C9orf72 )=217). Consistent with prior findings in young adults, GRN carriers showed larger TIV compared to familial non-carriers (95% confidence interval [CI]=1431994-1457123, p =0.049, η 2 p =0.008). Larger TIV correlated with higher years of education in GRN carriers (95% CI=0.01-0.24, r (295)=0.12, p =0.03) and GRN non-carriers (95% CI=0.08-0.34, r(198)=0.21, p =0.002). MAPT carriers demonstrated smaller TIV than non-carriers (95% CI=1417819-1450628, p =0.039, η 2 p =0.02). Models with C9orf72 and education as outcome variables did not reveal significant differences. Discussion In support of the neurodevelopmental hypothesis of FTD, GRN and MAPT mutations are linked to likely structural neurodevelopmental changes in TIV, some of which correlate to years of education. These findings motivate further research to identify mechanisms by which FTD mutations influence neurodevelopment and ascertain their suitability as targets for interventions.
Year:
2025
Collaborators (31)
Arabella Bouzigues
University College London
John C. van Swieten
-
Lize Jiskoot
Assistant professor
Erasmus University Rotterdam
Harro Seelaar
Neurologist
Erasmus MC
Jonathan Rohrer
University College London
Maria Luisa Mandelli
Associate Professor
University of California, San Francisco
Maria Carmela Tartaglia
University of Toronto
Rik Vandenberghe
Katholieke Universiteit Leuven
Pedro Pinheiro-Chagas
Assistant Professor, Director of NeuroAI Lab
University of California, San Francisco
Zachary Miller
Professor
-
Caroline Graff
Professor,MD PhD
Karolinska Institutet
James B. Rowe
Associate Director
-
William F. DeGrado
Professor
University of California
Pietro Tiraboschi
Fondazione IRCCS Istituto Neurologico Carlo Besta
Markus Otto
-
Bruce Miller
A.W. and Mary Margaret Clausen Distinguished Professor
University of California, San Francisco
Matthis Synofzik
Professor
Hertie Institute for clinical brain research
Maria Luisa Gorno Tempini
Professor of Neurology and Psychiatry
University of California, San Francisco
sandro sorbi
Full Professor of Neurology
University of Florence
Gregory Merz
Adjunct Assistant Professor
University of California, San Francisco
Johannes Levin
Professor for Clinical Neurodegeneration
Ludwig-Maximilians-Universität München
Robert Jr Laforce
Professeur agrégé de clinique
Laval University
Hyunil Jo
Specialist
University of California, San Francisco
Daniela Galimberti
Associate Professor
Università degli Studi di Torino
Lea T. Grinberg
-
Barbara Borroni
Docenti di ruolo di IIa fascia
University of Brescia
Florence Pasquier
Professor
Université de Lille
Salvatore Spina
Associate Professor
University of California, San Francisco
William Seeley
University of California, San Francisco
Mario Masellis
University of Toronto
Chiara Gallingani
University of Modena and Reggio Emilia
Make your next move.
Prepare your application
Review drafts before you send them.