Frontotemporal Dementia is a progressive neurodegenerative disorder characterized by the gradual loss of neuronal function. This page provides comprehensive information about the disease, including its pathophysiology, clinical presentation, diagnosis, and current therapeutic approaches.
Frontotemporal Dementia (FTD) is a group of neurodegenerative disorders characterized by progressive atrophy of the frontal and anterior temporal lobes. It is the second most common cause of young-onset dementia after Alzheimer's Disease, accounting for 10-20% of all dementia cases and up to 50% of cases with onset before age 651.
FTD encompasses a spectrum of clinical syndromes with distinct phenotypes:
The pathology involves selective degeneration of the frontal and temporal cortices, with approximately 40-50% of cases having tau-positive inclusions (FTD-tau) and 40-50% having TDP-43-positive inclusions (FTD-TDP)2.
Approximately 30-50% of FTD cases have an autosomal dominant pattern, with three major genes accounting for the majority of familial cases3:
| Gene | Protein | Inheritance | Typical Pathology | Frequency |
|---|---|---|---|---|
| MAPT | Tau | Autosomal dominant | Tau (3R/4R) | 5-10% |
| GRN | Progranulin | Autosomal dominant | TDP-43 Type A | 5-10% |
| C9orf72 | Dipeptide repeats | Autosomal dominant | TDP-43 Type B | 5-10% |
Mutations in the MAPT gene lead to abnormal tau protein accumulation:
Progranulin and C9orf72 mutations lead to TDP-43 dysfunction:
The GGGGCC repeat expansion in C9orf72 is the most common genetic cause of both FTD and ALS:
Core diagnostic features:
Supportive features:
Semantic variant (svPPA):
Nonfluent/agrammatic variant (nfvPPA):
Logopenic variant (lvPPA):
FTD diagnosis follows established clinical criteria (Neary et al., Rascovsky et al.)5:
bvFTD Diagnostic Requirements:
A practical imaging sequence in suspected Frontotemporal Dementia starts with structural MRI, then escalates to molecular imaging when diagnosis remains uncertain. MRI is used first to characterize syndrome-pattern atrophy (e.g., frontal/insular loss in bvFTD, left anterior temporal loss in svPPA, and left posterior fronto-insular patterns in nfvPPA) and to exclude major vascular, neoplastic, or hydrocephalus mimics13.
When MRI findings are subtle or mixed, FDG-PET improves syndrome-level discrimination by highlighting frontal and anterior temporal hypometabolism before advanced atrophy is evident12. Amyloid and tau PET should be interpreted as differential-diagnosis tools rather than standalone FTD confirmation tests, especially in older patients with mixed pathology risk.
Current biomarker use in FTD is best framed as a triage-and-stratification workflow:
NfL is currently the most mature cross-syndrome severity/prognostic marker for FTD-spectrum disease, but it remains non-specific for molecular subtype; it should be integrated with imaging and phenotype rather than used in isolation11. Biomarker limitations are most pronounced in mixed-pathology cases (e.g., coexisting AD pathology), where discordant fluid and imaging signals require multimodal adjudication.
For suspected Frontotemporal Dementia, diagnostic workups are most useful when sequencing modalities by the primary syndrome (behavioral,
language, or motor-neuron-associated presentation) rather than applying a single test stack to all cases81213.
A practical sequence used in specialist pathways is:
Important limitations remain: biomarker distributions overlap across phenotypes, mixed pathology is common in older adults, and most thresholds are not yet harmonized for universal primary-care deployment811.
Behavioral interventions:
Pharmacological approaches:
No disease-modifying therapy is currently approved for FTD. Evidence quality is heterogeneous and should be interpreted by development stage and endpoint rigor14:
For clinical counseling, symptomatic treatment remains standard of care while disease-modifying programs should be discussed as investigational and trial-dependent.
To avoid overstating maturity, disease-modifying claims in Frontotemporal Dementia should be interpreted by evidence grade:
| Program Class | Current Typical Evidence Grade | What Is Demonstrated | Key Gaps Before Confirmatory Use |
|---|---|---|---|
| Progranulin-restoration biologics (e.g., latozinemab class) | Phase 1-2 biomarker-supportive; Phase 3 pending | Target engagement and pharmacodynamic shifts are demonstrated in early studies10 | Robust clinical efficacy on functional/cognitive endpoints is still pending |
| Small-molecule modulators tested in GRN/other genetic subgroups | Early phase / mixed | Biological rationale and early safety are established1415 | Durable clinical benefit and subgroup reproducibility are not yet established |
| Gene or precision-targeted approaches in familial FTD | Early translational / exploratory clinical | Mechanistic plausibility and narrow-cohort feasibility | Small sample sizes and short follow-up limit confidence on progression endpoints |
Clinical care should therefore clearly separate approved symptomatic management from investigational disease-modifying strategies and present the latter as hypothesis-testing until phase-appropriate endpoint replication is available714.
FTD must be distinguished from:
Major unresolved questions in Frontotemporal Dementia now focus on stratified biology, trial-readiness biomarkers, and circuit-level
vulnerability in Frontotemporal Dementia and ALS Spectrum678.
These questions align with current cohort and trial infrastructure such as ALLFTD and genotype-enriched intervention programs910.
Proteins/CTSD (Cathepsin D) - Lysosomal dysfunction in FTD
Proteins/FBXO7 - F-box protein in FTD/PD spectrum
Proteins/ATP13A2 - Lysosomal ATPase in neurodegeneration
Genes/GSTP1 - Oxidative stress response in FTD
National Institute of Neurological Disorders and Stroke - FTD
Proteins/CTSD (Cathepsin D) - Lysosomal protease in FTD pathology
Proteins/FBXO7 - F-box protein in ubiquitin-proteasome system
Proteins/ATP13A2 - Lysosomal ATPase in neuronal degeneration
Genes/GSTP1 - Glutathione S-transferase in oxidative stress response
The study of Frontotemporal Dementia has evolved significantly over the past decades. Research in this area has revealed important insights into the underlying mechanisms of neurodegeneration and continues to drive therapeutic development.
Historical context and key discoveries in this field have shaped our current understanding and will continue to guide future research directions.
◎ Proposed
Abnormalities in the retina might alter susceptibility of occipital cortex to neurodegenerative processes associated with age-related dementia
Retina is composed of retinal ganglion cells whose axons form the optic nerve and travel to the lateral geniculate nucleus and primary visual cortex in the occipital cortex. The authors hypothesized that abnormalities in the retina might affect the occipital cortex through anterograde or retrograde degenerative mechanisms.
Source: Cecilia S. Lee et al., Ophthalmology-Based Neuropathology Risk Factors: Diabetic Retinopathy is Associated with Deep Microinfarcts in a Community-Based Autopsy Study (2019) DOI:10.3233/JAD-181087
◉ Supported
Diabetic retinopathy is associated with increased risk of deep cerebral microinfarcts, likely representing a common mechanism at the capillary or arteriolar level involving microvascular injury
OR = 1.91 (95% CI 1.11, 3.27, p = 0.02) for association between DR and deep microinfarcts. The relationship was specific to DR with diabetes, not diabetes alone, suggesting DR represents either more advanced systemic diabetes and/or specific microvascular injury in the retina that correlates with brain microvasculature. Loss of endothelial pericytes is believed to be the initiating pathology of DR, and similar loss of endothelial integrity of brain microvasculature may occur.
Source: Cecilia S. Lee et al., Ophthalmology-Based Neuropathology Risk Factors: Diabetic Retinopathy is Associated with Deep Microinfarcts in a Community-Based Autopsy Study (2019) DOI:10.3233/JAD-181087
○ Speculative
Eye diseases may increase susceptibility to pathological tau deposition in the occipital cortex, potentially through decreased afferent input from retinal ganglion cells
Trend for increased PHF-tau in occipital cortex among participants with glaucoma (OR = 1.36, 95% CI 0.91, 2.03, p = 0.13) was not statistically significant. The authors propose that retinal ganglion cell activity-dependent function may serve a protective role against tau pathology in the occipital cortex, and that loss of ganglion cells in glaucoma (approximately 50% lost even in mild cases) may decrease afferent input to lateral geniculate nucleus and eventually occipital cortex.
Source: Cecilia S. Lee et al., Ophthalmology-Based Neuropathology Risk Factors: Diabetic Retinopathy is Associated with Deep Microinfarcts in a Community-Based Autopsy Study (2019) DOI:10.3233/JAD-181087
◉ Supported
A novel event-based model incorporating kernel density estimation (KDE) mixture modeling can estimate fine-grained sequences of cognitive decline in neurodegenerative diseases from cross-sectional neuropsychological data
The authors developed a new event-based model using KDE components instead of Gaussian mixture models to handle non-Gaussian cognitive test score data. They validated it in simulation experiments before applying to patient data.
Source: Firth, Nicholas C. et al., Sequences of cognitive decline in typical Alzheimer's disease and posterior cortical atrophy estimated using a novel event-based model of disease progression (2020) DOI:10.1002/alz.12083
◉ Supported
Visual processing deficits precede memory deficits in posterior cortical atrophy, while episodic memory deficits precede visual deficits in typical Alzheimer's disease
In PCA, measures of visual processing (A Cancellation time, VOSP Fragmented Letters) became abnormal earlier, while episodic memory (SRMT, PAL) declined relatively late. In tAD, episodic memory (PAL, SRMT) deteriorated early while visual processing deficits appeared at the end of the sequence.
Source: Firth, Nicholas C. et al., Sequences of cognitive decline in typical Alzheimer's disease and posterior cortical atrophy estimated using a novel event-based model of disease progression (2020) DOI:10.1002/alz.12083
◎ Proposed
Within each syndrome (bvFTD and svPPA), individual patients harbor distinct epicenters that create subtly different atrophy patterns by templating distinct patterns of transneuronal spreading.
The main hypothesis of the study: individual patients show heterogeneous atrophy patterns, and the study aimed to identify patient-specific epicenters and use them to predict longitudinal atrophy spread.
Source: Brown, Jesse A. et al., Patient-Tailored, Connectivity-Based Forecasts of Spreading Brain Atrophy (2019) DOI:10.1016/j.neuron.2019.08.037
◉ Supported
Neurodegenerative diseases progress by spreading via brain connections (transneuronal degeneration hypothesis), where disease pathology spreads directly between anatomically connected brain regions.
Previous literature cited showing protein misfolding and aggregation progress through predictable anatomical patterns, potentially via transsynaptic spread. The study provides longitudinal evidence supporting this model.
Source: Brown, Jesse A. et al., Patient-Tailored, Connectivity-Based Forecasts of Spreading Brain Atrophy (2019) DOI:10.1016/j.neuron.2019.08.037
◉ Supported
Disease onset occurs within a focal brain region ('epicenter'), whose intrinsic functional connectivity pattern guides the spread of misfolded protein into new regions.
Epicenter identified by finding the brain region whose healthy functional connectivity map most strongly correlates with the patient's atrophy pattern. Epicenters were located in predicted regions (anterior cingulate/frontoinsular for bvFTD, anterior temporal lobe for svPPA).
Source: Brown, Jesse A. et al., Patient-Tailored, Connectivity-Based Forecasts of Spreading Brain Atrophy (2019) DOI:10.1016/j.neuron.2019.08.037
◉ Supported
Contrastive trajectory inference (cTI) can identify data-driven disease stages in genetic FTD using multi-modal MRI metrics without clinical information, providing a unified staging system across genetic and phenotypic variations.
cTI disease scores significantly correlated with MMSE (r=-0.273), CBI-R (r=0.516), neuropsychological tests (all r>0.276), and estimated years to symptom onset (r=0.334). Significant differences found between symptomatic, presymptomatic, and noncarrier groups.
Source: McCarthy, Jillian et al., Data-driven staging of genetic frontotemporal dementia using multi-modal MRI (2022) DOI:10.1002/hbm.25727
✓ Established
MAPT mutations are associated with tau pathology while GRN mutations and C9orf72 expansions are associated with TDP-43 pathology in genetic FTD.
Well-established pathological knowledge in the field; cited references on genetic FTD pathology.
Source: McCarthy, Jillian et al., Data-driven staging of genetic frontotemporal dementia using multi-modal MRI (2022) DOI:10.1002/hbm.25727
◎ Proposed
Abnormalities in the retina might alter susceptibility of occipital cortex to neurodegenerative processes associated with age-related dementia
Retina is composed of retinal ganglion cells whose axons form the optic nerve and travel to the lateral geniculate nucleus and primary visual cortex in the occipital cortex. The authors hypothesized that abnormalities in the retina might affect the occipital cortex through anterograde or retrograde degenerative mechanisms.
Source: Cecilia S. Lee et al., Ophthalmology-Based Neuropathology Risk Factors: Diabetic Retinopathy is Associated with Deep Microinfarcts in a Community-Based Autopsy Study (2019) DOI:10.3233/JAD-181087
◉ Supported
Diabetic retinopathy is associated with increased risk of deep cerebral microinfarcts, likely representing a common mechanism at the capillary or arteriolar level involving microvascular injury
OR = 1.91 (95% CI 1.11, 3.27, p = 0.02) for association between DR and deep microinfarcts. The relationship was specific to DR with diabetes, not diabetes alone, suggesting DR represents either more advanced systemic diabetes and/or specific microvascular injury in the retina that correlates with brain microvasculature. Loss of endothelial pericytes is believed to be the initiating pathology of DR, and similar loss of endothelial integrity of brain microvasculature may occur.
Source: Cecilia S. Lee et al., Ophthalmology-Based Neuropathology Risk Factors: Diabetic Retinopathy is Associated with Deep Microinfarcts in a Community-Based Autopsy Study (2019) DOI:10.3233/JAD-181087
○ Speculative
Eye diseases may increase susceptibility to pathological tau deposition in the occipital cortex, potentially through decreased afferent input from retinal ganglion cells
Trend for increased PHF-tau in occipital cortex among participants with glaucoma (OR = 1.36, 95% CI 0.91, 2.03, p = 0.13) was not statistically significant. The authors propose that retinal ganglion cell activity-dependent function may serve a protective role against tau pathology in the occipital cortex, and that loss of ganglion cells in glaucoma (approximately 50% lost even in mild cases) may decrease afferent input to lateral geniculate nucleus and eventually occipital cortex.
Source: Cecilia S. Lee et al., Ophthalmology-Based Neuropathology Risk Factors: Diabetic Retinopathy is Associated with Deep Microinfarcts in a Community-Based Autopsy Study (2019) DOI:10.3233/JAD-181087
◉ Supported
Contrastive trajectory inference (cTI) can identify data-driven disease stages in genetic FTD using multi-modal MRI metrics without clinical information, providing a unified staging system across genetic and phenotypic variations.
cTI disease scores significantly correlated with MMSE (r=-0.273), CBI-R (r=0.516), neuropsychological tests (all r>0.276), and estimated years to symptom onset (r=0.334). Significant differences found between symptomatic, presymptomatic, and noncarrier groups.
Source: McCarthy, Jillian et al., Data-driven staging of genetic frontotemporal dementia using multi-modal MRI (2022) DOI:10.1002/hbm.25727
✓ Established
MAPT mutations are associated with tau pathology while GRN mutations and C9orf72 expansions are associated with TDP-43 pathology in genetic FTD.
Well-established pathological knowledge in the field; cited references on genetic FTD pathology.
Source: McCarthy, Jillian et al., Data-driven staging of genetic frontotemporal dementia using multi-modal MRI (2022) DOI:10.1002/hbm.25727
◉ Supported
Parkinson's disease progression is non-deterministic and heterogeneous, with patients able to follow different progression pathways among disease states rather than sequential progression through fixed subtypes
The model discovered non-sequential, overlapping disease progression trajectories. The ranking of starting state did not match the ranking of reaching terminal state within 5 years. State transitions directly from 1 to 8, 2 to 3, and 4 to 7 were never observed, but many non-sequential transitions occurred. Patients starting in State 5 had shortest time to terminal state (median 2.75 years), contradicting the notion that lower-numbered states always progress slower.
Source: Severson, Kristen A et al., Discovery of Parkinson's disease states and disease progression modelling: a longitudinal data study using machine learning (2021) DOI:10.1016/S2589-7500(21)00101-1
◉ Supported
Static subtype assignment is ineffective at capturing the full spectrum of Parkinson's disease progression because disease manifestations and progression are heterogeneous
Previous approaches assumed deterministic progression within subtypes. The study found that non-sequential, overlapping progression trajectories exist, suggesting that defining subtypes based on baseline cross-sectional data cannot capture the dynamic nature of PD progression. The model allowed any progression path under progressive constraints, revealing that data-driven states better describe progression than fixed subtypes.
Source: Severson, Kristen A et al., Discovery of Parkinson's disease states and disease progression modelling: a longitudinal data study using machine learning (2021) DOI:10.1016/S2589-7500(21)00101-1
◎ Proposed
Dopaminergic medications affect observed symptoms but do not modify the underlying disease trajectory
The model assumed medications do not modify the underlying disease trajectory. Medication effects were incorporated into the observation model as a function of disease state and individualized response. Results showed bradykinesia and tremor measures improved with medication, whereas non-motor measures were worse with medication. The model with medication effects had improved performance (higher log-likelihood) compared to model without medication effects.
Source: Severson, Kristen A et al., Discovery of Parkinson's disease states and disease progression modelling: a longitudinal data study using machine learning (2021) DOI:10.1016/S2589-7500(21)00101-1
◉ Supported
Tau protein plays a central role in AD pathogenesis and serves as both a diagnostic and therapeutic target
The paper cites evidence supporting the central role of tau in AD pathogenesis and notes tau as both a diagnostic and therapeutic target (Medeiros et al. 2011)
Source: Hartmuth C. Kolb et al., Tau Positron Emission Tomography Imaging (2017) DOI:10.1101/cshperspect.a023721
✓ Established
The density and neocortical spread of NFTs correlate with progressive neuronal degeneration and cognitive impairment in AD
Multiple studies show correlation between NFT density and cognitive scores (Arriagada et al. 1992; Maccioni et al. 2010; Villemagne et al. 2012)
Source: Hartmuth C. Kolb et al., Tau Positron Emission Tomography Imaging (2017) DOI:10.1101/cshperspect.a023721
◉ Supported
Although amyloid and tau pathology may begin independently in separate brain regions, the presence of amyloid intensifies and accelerates an otherwise limited tauopathy
18F-T807 PET imaging showed amyloid burden often accompanied by increased tau pathology, with stronger correlation between cognitive decline and tau burden in people with amyloid
Source: Hartmuth C. Kolb et al., Tau Positron Emission Tomography Imaging (2017) DOI:10.1101/cshperspect.a023721
◉ Supported
Tau protein plays a central role in AD pathogenesis and serves as both a diagnostic and therapeutic target
The paper cites evidence supporting the central role of tau in AD pathogenesis and notes tau as both a diagnostic and therapeutic target (Medeiros et al. 2011)
Source: Hartmuth C. Kolb et al., Tau Positron Emission Tomography Imaging (2017) DOI:10.1101/cshperspect.a023721
✓ Established
The density and neocortical spread of NFTs correlate with progressive neuronal degeneration and cognitive impairment in AD
Multiple studies show correlation between NFT density and cognitive scores (Arriagada et al. 1992; Maccioni et al. 2010; Villemagne et al. 2012)
Source: Hartmuth C. Kolb et al., Tau Positron Emission Tomography Imaging (2017) DOI:10.1101/cshperspect.a023721
◉ Supported
Although amyloid and tau pathology may begin independently in separate brain regions, the presence of amyloid intensifies and accelerates an otherwise limited tauopathy
18F-T807 PET imaging showed amyloid burden often accompanied by increased tau pathology, with stronger correlation between cognitive decline and tau burden in people with amyloid
Source: Hartmuth C. Kolb et al., Tau Positron Emission Tomography Imaging (2017) DOI:10.1101/cshperspect.a023721