Frontotemporal dementia: latest evidence and clinical implications

Abstract

Background:

Frontotemporal dementia (FTD) describes a cluster of neurocognitive syndromes that present with impairment of executive functioning, changes in behavior, and a decrease in language proficiency. FTD is the second most common form of dementia in those younger than 65 years and is expected to increase in prevalence as the population ages. This goal in our review is to describe advances in the understanding of neurobiological pathology, classification, assessment, and treatment of FTD syndromes.

Methods:

PubMed was searched to obtain reviews and studies that pertain to advancements in genetics, neurobiology, neuroimaging, classification, and treatment of FTD syndromes. Articles were chosen with a predilection to more recent preclinical/clinical trials and systematic reviews.

Results:

Recent reviews and trials indicate a significant advancement in the understanding of molecular and neurobiological clinical correlates to variants of FTD. Genetic and histopathologic markers have only recently been discovered in the past decade. Current therapeutic modalities are limited, with most studies reporting improvement in symptoms with nonpharmacological interventions. However, a small number of studies have reported improvement of behavioral symptoms with selective serotonin reuptake inhibitor (SSRI) treatment. Stimulants may help with disinhibition, apathy, and risk-taking behavior. Memantine and cholinesterase inhibitors have not demonstrated efficacy in ameliorating FTD symptoms. Antipsychotics have been used to treat agitation and psychosis, but safety concerns and side effect profiles limit utilization in the general FTD population. Nevertheless, recent breakthroughs in the understanding of FTD pathology have led to developments in pharmacological interventions that focus on producing treatments with autoimmune, genetic, and molecular targets.

Conclusion:

FTD is an underdiagnosed group of neurological syndromes comprising multiple variants with distinct neurobiological profiles and presentations. Recent advances suggest there is an array of potential novel therapeutic targets, although data concerning their effectiveness are still preliminary or preclinical. Further studies are required to develop pharmacological interventions, as there are currently no US Food and Drug administration approved treatments to manage FTD syndromes.

Introduction

Frontotemporal dementia (FTD) is a term used to describe a group of neurocognitive disorders that encompass progressive dysfunction in executive functioning, behavior, and language. It is considered the third most common form of dementia following Alzheimer’s disease (AD) and dementia with Lewy bodies. ¹ As per its namesake, it is a cluster of syndromes that result from degeneration of the frontal and temporal lobes, and is subdivided into two categories that are unique in respect to their predominating presentations; namely, the behavioral subtype that accounts for about half of FTD cases, and the language subtype. ² The language subtype is further subdivided into nonfluent and semantic variants of primary progressive aphasia (PPA), which are characterized by diverging localizations and underlying cerebral dysfunction.^2,3 The semantic and nonfluent variants of PPA are both characterized by prominent language impairment, but also differ in neurological localization, severity of specific symptoms, and the overall course of their presentations. Illustrative of this, the semantic variant exhibits bilateral anterior temporal lobe atrophy and is associated with dysfunction of emotional processing, compulsions, and decline in language skills. In contrast, the nonfluent variant commonly presents with greater left hemisphere atrophy, and is associated with speech problems earlier while behavioral disturbances develop later in the disease course. ³ As the population of geriatric patients grow and neurocognitive disorders become more prevalent, there will be an increased need for physicians with an expert understanding of the diverse clinical findings that define the heterogeneous FTD subtypes. This review will focus on the most recent findings concerning FTD neurobiology, current classification and assessment systems, and the most up-to-date expert consensus on the treatment of this unique collection of syndromes.

Epidemiology

According to the World Health Organization (WHO), there are an estimated 47.5 million people who have dementia with another 7.7 million new cases every year. ⁴ Although AD is the most common form of dementia, contributing to 60–70% of cases, mixed types exist, and the overlap between different forms of dementia is sizeable. FTD is the second most common cause of dementia in patients aged < 65 years, ⁵ with a further 25% of dementia cases in patients older than 65 years also attributed to FTD disorders. ⁶ However, determining the total population with underlying FTD pathology is difficult due to the low disease frequency in a relatively large at-risk population. Additionally, studies estimating incidence and prevalence rates of FTD are limited by the inherent difficulty identifying FTD disorders.^6,7 In the USA, Knopman and Roberts ⁷ estimated that the prevalence of FTD between the ages of 45–65 years ranged from 15 to 22 per 100,000 people, with incidence estimates ranging from 2.7 to 4.1 per 100,000 members in the same age range. Overall, they estimated approximately 20,000 to 30,000 cases of FTD in the USA alone. Yet, these numbers may actually be underestimating the occurrence of FTD syndromes. This is especially evident, as older cases of FTD begin to echo symptoms of AD further confounding the delineation between the neurocognitive disorders. ⁶ Further emphasizing the limitations of FTD epidemiologic studies, Lambert and colleagues ⁸ conducted a systematic review with the goal of estimating prevalence rates of early-onset dementia. They found a significant disparity between studies estimating the prevalence of early-onset FTD with estimates ranging from 1.0 to 15.4 per 100,000 members of the population. Adding to this, it appears that the data presented was principally collected from populations studied in the UK, mainland Europe, and Japan, suggesting that the information gathered may not be generalizable to the world-wide population as a whole. For this reason, standardization of study designs will likely aid in more precisely defining FTD epidemiology globally, a focus that future studies should pursue in the future.

Classification and assessment systems

Researchers have developed multiple methodologies to classify FTD disorders using biological and molecular signatures. However, diagnosis is still obtained clinically due to the absence of definitive biomarkers. The international consensus criteria for the behavioral variant of FTD (bvFTD) was published by the International Behavioral Variant FTD Criteria Consortium ⁹ to develop a sensitive standard for early screening and management of bvFTD. To be diagnosed, a patient must show progressive deterioration of behavior or cognition by observation or history provided by a reliable caretaker or informant in order to be diagnosed. Possible bvFTD is diagnosed if three or more of the following are present: (a) early behavioral disinhibition described as socially inappropriate behavior, loss of manners/decorum, impulsivity and rash actions; (b) early apathy or inertia; (c) early loss of empathy or sympathy, including diminished response to other people’s need or sympathies, and diminished social interest; (d) early perseverative or compulsive/ritualistic behavior (i.e. simple repetitive movements, complex, compulsive, or ritualistic behavior, stereotypy of speech); (e) hyperorality and dietary changes, including altered food preferences, binge eating, increased consumption of alcohol and cigarettes, oral exploration or consumption of inedible objects; (f) neuropsychological profile demonstrates deficits in executive functioning and relative sparing of episodic memory and visuospatial functioning. Probable bvFTD diagnosis would need to meet criteria for possible bvFTD, but also includes imaging suggestive of frontal or anterior atrophy on magnetic resonance imaging (MRI) or computed tomography (CT). Alternatively, positron emission tomography (PET) or single-photon emission computed tomography (SPECT) imaging demonstrating hypoperfusion or hypometabolism in the frontal or anterior temporal regions could be used to diagnose probable bvFTD. Definitive bvFTD with definite FTD pathology is diagnosed when a patient meets criteria for possible bvFTD and has one or both of the following: histopathological evidence of FTD or if there is evidence of a known pathogenic mutation. bvFTD is excluded if a patient’s presentation is better accounted for by other nondegenerative nervous system or medical disorders, a psychiatric diagnosis, or biomarkers are strongly indicative of other neurocognitive disorders or neurodegenerative processes such as AD. ⁹

Gorno-Tempini and colleagues developed a common framework for the classification of PPA subtypes to provide consistency in clinical diagnosis and standardize PPA assessments for future clinical studies. ¹⁰ According to this framework, PPA is diagnosed when a patient exhibits prominent difficulty with language wherein the deficits are the principal cause of impairment of daily activities, with aphasia being the most prominent deficit at symptom onset and during the initial phase of the disease. PPA diagnosis is excluded if the presenting symptoms are more consistent with other neurocognitive disorders, medical conditions, neurodegenerative processes, or a psychiatric diagnosis. PPA is also excluded if there are prominent initial episodic memory, visual memory, visuospatial impairments, and initial behavioral disturbance. Patients may be diagnosed with the semantic variant of PPA (svPPA) if they exhibit impaired confrontation naming (i.e. difficulty naming or recognizing objects or drawings) and impaired single-word comprehension, and have at least three of the following: (a) impaired object knowledge; (b) surface dyslexia (i.e. inability to recognize words as a whole) or dysgraphia; (c) spared repetition; (d) spared speech production. For an imaging-supported diagnosis, there must either be predominant anterior temporal lobe atrophy or predominant anterior temporal hypoperfusion/hypometabolism on SPECT or PET. For a definite pathology diagnosis, there must be histopathologic evidence or the presence of a known pathogenic mutation.

The nonfluent variant of PPA (nfvPPA) diagnosis requires at least one of two core features (agrammatism in language production or effortful, halting speech not consistent with apraxia of speech), and two out of three of the following features: (a) impaired comprehension of complex sentences; (b) spared single-word comprehension; (c) spared object knowledge. Imaging-supported nfvPPA variant diagnosis is obtained when there is imaging that demonstrates predominant left posterior-fronto-insular atrophy on MRI or predominantly left posterior fronto-insular hypoperfusion or hypometabolism on SPECT or PET. Alternatively, nfvPPA with definite pathology diagnosis is present when there is histopathologic evidence or the presence of a known pathogenic mutation.

Like other PPA subtypes, individuals with the logopenic variant of PPA exhibit language impairment but with more specific difficulties in confrontation naming and syntax. These individuals are more likely to have speech interrupted with filling sounds, frequent pauses, phonological errors, and impaired sentence repetition. ¹¹ The diagnosis of logopenic variant PPA requires the following core features to be present: impaired single-word retrieval in spontaneous speech and naming, and impaired repetition of sentences and phrases. Additionally, three of four of the following features must also be present for a diagnosis: (a) speech (phonologic) errors in spontaneous speech and naming; (b) spared single-word comprehension and object knowledge; (c) spared motor speech; (d) absence of frank agrammatism. As with other PPA-variant diagnoses, imaging-supported diagnosis is made with corresponding abnormalities in imaging. In the logopenic variant, MRI, SPECT, or PET scans must demonstrate predominant left posterior perisylvian or parietal atrophy/hypoperfusion/hypometabolism. Definite histopathologic evidence or the presence of a known pathogenic mutation would allow for a definite pathology specifier.

Familial frontotemporal dementia caused by frontotemporal dementia-associated mutations

FTD is a highly heritable disorder despite varying heritability among different clinical syndromes and subtypes due to a range of gene mutations. ¹² Up to half of FTD cases with autosomal-dominant inheritance report a family history of FTD. ¹³ Familial FTD also accounts for approximately one third to half of all FTD cases and presents more commonly as bvFTD than other FTD subtypes.^6,14 Mutations in microtubule-associated protein tau (MAPT),^15–25 progranulin (PGRN),^26–38 and chromosome 9 open reading frame 72 (C9orf72) expansion mutations^32,39–50 have been found to be the most common causes of familial FTD.^51,52 Although FTD presentations are relatively homogenous early in the disease course, different biological correlates and varying genetic mutations ultimately result in diverging clinical courses. ⁵³ MAPT-associated familial FTD typically presents younger than FTD associated with other mutations, and presents as bvFTD with or without Parkinsonian symptoms, as well as with or without language decline. MAPT mutations occur on average at a frequency of 6–11% of all FTD subjects, is inherited in an autosomal-dominant manner, has high penetrance, and may have a shorter duration of illness relative to other mutations.^17,54,55 The most prominent behavioral features with MAPT mutations include disinhibition, stereotyped repetitive behaviors, and obsessions. Notably, apathy is less common than in PGRN and C9orf72 cases. Semantic impairment is more common later during the disease course. Episodic memory loss may occur and mimic early onset AD. ⁵⁶ MAPT mutations have been implicated in the formation of hyperphosphorylated tau in cortical and subcortical grey and white matter suggesting a neuroanatomical correlation in the development of clinical symptoms. ⁵⁵

PGRN mutations present with more variable clinical presentations, but most commonly present as bvFTD and less commonly as a nfvPPA presentation. ⁵⁷ Mutations occur at a frequency of 5–20% in the FTD population, are inherited in an autosomal-dominant manner, and have relatively low penetrance until the age of 70.^17,58,59 Apathy and social withdrawal are prominent in their presentation, and 10–30% of cases may present with episodic memory impairment. ⁶⁰ Neuropsychiatric symptoms are common and patients may report delusions, hallucinations, or exhibit ritualistic behaviors. There is also a predominance of early language involvement compared with MAPT or C9orf72 mutations. Extrapyramidal symptoms occur between 40% and 60% of PGRN-associated FTD cases.^57,60 Interestingly, PGRN mutations, though thought to be pathogenic, do not appear to increase the risk of developing FTLD. ⁵⁴

FTD patients with the C9orf72 expansion mutation present more commonly as bvFTD, motor neuron disease, or a combination of the two, although clinical presentations tend to be heterogeneous early in the disease process.^41,61 Apathy, disinhibition, and loss of empathy are prominent behavioral features in this type of familial FTD. ⁶² Complex stereotyped behaviors are common, but language decline is rarer compared with patients with MAPT or PGRN mutations. Patients typically have profound disruption of executive functioning, may have reduced spontaneous speech, echolalia, perseveration, impaired visual and verbal episodic memory, apraxia, anomia, and dyscalculia. ⁶³ C9orf72 expansion mutations exhibit an autosomal-dominant pattern of inheritance, is implicated in disease anticipation that affect successive generations, ⁴⁴ and is considered the most common cause of ALS-FTD (amyotrophic lateral sclerosis associated with frontotemporal dysfunction).^42,43,49 However, the minimum repeat length required to indicate increased risk of disease is unknown. ⁴⁷ These expansion mutations are highly associated with transactive response DNA-binding protein (TDP-43) pathology, and the presence of neuronal cytoplasmic and intranuclear inclusions that are immunoreactive to ubiquitin proteasome system markers. A confounding factor to consider when diagnosing this subtype is the possibility of AD-like biomarkers present in the cerebrospinal fluid. ⁶⁰ Psychiatric presentations, including delusions, auditory/visual/tactile hallucinations, somatoform symptoms, agitation, and anxiety may be observed. Parkinson-like symptoms are common and may present as gait disturbance, tremor, akinesis, and rigidity. Most patients do not respond to levodopa when administered.^60,62

Two other studied pathognomonic mutations found in a minority of cases include mutations in vasolin-containing protein (VCP) and transactive DNA-binding protein (TARDBP). ⁶⁴ VCP mutations have been implicated in ubiquitin–proteasome pathway dysfunction, disruption of TDP-43 localization, and inducing neuronal apoptosis. ⁶⁵ Individuals with VCP mutations historically present as the triad of autosomal-dominant hereditary inclusion body myopathy with Paget’s disease of the bone and frontotemporal dementia (IBMPFD).^66–69 FTD symptoms typically include behavioral changes, executive dysfunction, and aphasia. ⁷⁰ Consequently, a number of European and Asian cases detailing progressive myopathy with signs and symptoms consistent with FTD have been associated with VCP mutations.^71–77 Of particular note is one study ⁷⁸ that investigated how the location of VCP gene mutations in 27 families related to the progression and severity of symptoms. Critically, the investigators found that the onset of neurocognitive deterioration foreshadowed a rapid progression of the disease leading to an average life span of 6 years.

Alternatively, TARDBP was found to be a major component of ubiquitin-positive and tau-negative inclusions in patients that have been diagnosed with both amyotrophic lateral sclerosis (ALS) and FTD. ⁷⁹ TARDBP has also been implicated in both behavioral and semantic FTD cases with co-occurring motor neuron disease (MND). ⁸⁰ However, other studies have reported cases of individuals with TARDBP mutations that have developed FTD without the progression of ALS or MND symptoms.^81,82 Clinical presentations of patients with TARDBP mutations are highly heterogeneous and may present with apathy, poor executive functioning, poor language comprehension, or disinhibition. ⁸³ It has been suggested that mutations of TARDBP have led to increased translocation of TDP-43 causing dysregulation of neuronal endoplasmic reticulum calcium signaling resulting in an increase in B-cell lymphoma 2 (Bcl-2) mediated neuronal apoptosis. ⁸⁴

Neurobiology

FTD is classified into distinct, major subtypes based on their respective protein-based inclusions and underlying molecular pathologies. Namely, they are separated into the tau-based frontotemporal lobar degeneration (FTLD)-tau subtype, the FTLD-TDP subtype associated with TDP-43, the FTLD-FET subtype that is associated with the fused in sarcoma (FUS) proteinopathy and other FET [i.e. FUS, EWSR1 (EWS RNA-binding protein 1), TAF15 (TATA-box binding protein associated factor 15)] family proteins, and FTLD-ubiquitin–proteasome system (UPS).^59,85 (See Table 1 for a summary of FTD tauopathies) FTLD-tau accounts for approximately 40% of all FTD cases, and is characterized by misfolded tau tangles similar to those found in AD. ⁸⁶ The tauopathy present in this variant involves hyperphosphorylated tau protein that forms insoluble filamentous inclusions, which adversely affect axonal transport regulation, microtubule formation, and microtubule stabilization. These changes lead to a decrease in microtubule binding and an increase in tau aggregation, the latter of which induces neurotoxicity and nerve cell dysfunction. ⁸⁷ Tau aggregates in FTLD-tau subtypes vary in the degree of phosphorylation and the amount of tau isoforms they contain. This has prompted their subclassifications into predominantly 3R or 4R (e.g. corticobasal degeneration, progressive supranuclear palsy, argyrophilic grain disease, globular glial tauopathy) tau isoform disorders.^59,88 These subtypes of FTLD tauopathies also have distinct anatomical and histopathological correlates. In general, FTD is associated with severe neuronal pathology, and spherical argyrophilic neuronal cytoplasmic inclusions composed primarily of 3R tau isoforms. ⁶⁴ Progressive supranuclear palsy (PSP), a movement disorder that presents with rigidity, bradykinesia, and ophthalmoplegia, is characterized by significant degeneration of subcortical regions including the subthalamic nucleus, midbrain, substantia nigra, globus pallidus, and striatum. Postmortem examinations of PSP patients demonstrate spherical and flame-shaped neurofibrillary tangles, tufted astrocytes, and inclusions composed predominantly of the 4R tau isoform. ⁸⁹ FTD tauopathies may also exhibit the corticobasal syndrome, which may present as rigidity, bradykinesia, dystonia, apraxia, and alien limb phenomenon, is characterized by substantia nigra depigmentation, globus pallidus atrophy, and asymmetric and focal cortical atrophy. This syndrome exhibits similar traits compared with PSP, but develop more white matter involvement and has unique ring-shaped collections called astrocytic plaques caused by tau accumulation in astrocytes.^59,89

Table 1.

Frontotemporal dementia tauopathies. ⁵⁹

Tauopathies	Presentation	Histopathology	Tau isoform
Pick’s disease	bvFTD or nfvPPA	Pick bodies, argyrophilic neuronal cytoplasmic inclusions	3R
Progressive supranuclear palsy	Rigidity, ophthalmoplegia, bradykinesia	Spherical and flame-shaped neurofibrillary tangles, tufted astrocytes	4R
Corticobasal degeneration	Rigidity, bradykinesia, dystonia, apraxia, alien limb phenomenon	White matter degeneration with ring-shaped astrocytic plaques	–

FTD, frontotemporal dementia; bvFTD, behavioral variant FTD; nfvPPA, nonfluent variant primary progressive aphasia.

Alternatively, neuronal inclusions immunoreactive to ubiquitin are characteristic of the FTLD-TDP and FTLD-FET forms.^85,90 In FTLD-TDP, these neuronal inclusions are primarily composed of TDP-43, which has been associated with multiple genetic markers thought to be pathognomonic of FTD. These include mutations of PGRN, C9orf72, and VCP.^86,91 In mice, accumulation of the insoluble, phosphorylated cytoplasmic TDP-43 and loss of nuclear TDP-43 is associated with brain atrophy, muscle denervation, motor neuron loss, and progressive motor impairments that are commonly observed both in FTLD-TDP and ALS suggesting a common neuropathological pathway. ⁹² FTLD-TDP is further subdivided into Types A, B, C, and D that differ in the constitution of pathological inclusions and their distribution within the cortices. ⁹³ Patients with FTLD-TDP Type A have predominantly neuronal cytoplasmic inclusions and dystrophic neurites with TDP-43 immunoreactive lesions in the second layer of the cortices. FTLD-TDP type B features more neuronal cytoplasmic inclusions in the superficial and deeper cortical layers. FTLD-TDP type C has a significant amount of dystrophic neurites demonstrating a ‘corkscrew appearance’ through the cortical layers. FTLD-TDP type D is associated with multiple neuronal intranuclear inclusions and dystrophic neurites with fewer neuronal cytoplasmic inclusions.^85,93 Protein immunoreactivity to members of the FET protein family, with FUS being the more well-known FTD-associated proteinopathy, is the last major subgroup of FTD pathology, and is associated with co-aggregation of FET proteins into ubiquitinated inclusions found in severe cases of sporadic FTD.^59,86 Of note, FUS protein involvement in the pathology of certain FTD subtypes is suggestive of a common neuropathologic mechanism and close relationship between the FTD disorders and ALS, since FUS has been described as a common cause of autosomal-dominant ALS.^92,94 Interestingly, FTD-associated proteinopathies including tau, TDP-43, and FUS have demonstrated aggregation and seeding behavior in both in vivo and in vitro models, suggesting FTD pathologies have prion-like characteristics.^86,87 Lastly, the FTLD-UPS subtype is characterized by ubiquitin-immunoreactive inclusions that do not label for tau, TDP-43, or FET proteins.^12,95–97 This subtype is commonly associated with a mutation in the charged multivesicular body protein 2B (CHMP2B) gene that leads to dysfunction in the endosomal ESCRTIII complex that causes disruption of endosomal traffic.^98–101 A decline in levels of CHMP2B results in a decrease of hippocampal dendritic branching and reduced excitatory synapse activity effectively diminishing synaptic plasticity and causing neurodegenerative deficits. ¹⁰² This subtype typically presents with changes in behavior and personality (i.e. disinhibition, apathy, loss of emotion, dyscalculia, progressive aphasia that leads to mutism), as well extrapyramidal motor symptoms as the disease progresses. ⁹⁸ However, CHMP2B mutations are relatively rare, ¹⁰³ and do not appear to increase the risk of development of FTD. ¹⁰⁴

Neuroimaging

A recent review ¹⁰⁵ of neuroimaging studies have detailed advances in quantifying brain atrophy in FTD using volumetric MRI measurements and by observing hypometabolism in specific regions using [18F] fluorodeoxyglucose PET (18F-FDG PET). Volumetric MRI studies they reviewed focused mainly on cerebral loss of grey matter with the goal of improving accurate diagnosis and characterization of FTD variants. In these studies, MRI imaging demonstrated disruption of the salience network, default mode network, left hemisphere language and semantic networks, and executive control networks that imply pathoanatomical relationships in FTD syndromes. Notably, one study has demonstrated how baseline salience network resting state activity may predict behavioral changes in FTD. ¹⁰⁶ MRI studies of bvFTD patients in particular show losses in the prefrontal cortex, anterior temporal regions, insula, anterior cingulate, striatum, and thalamus, which differentiate bvFTD from AD with relatively high sensitivity and specificity. A recent retrospective observational study also suggests that there are specific anatomical phenotypes based on the salience and semantic appraisal networks that may contribute in classifying the otherwise heterogeneous bvFTD subgroups. ¹⁰⁷ The svPPA exhibits more asymmetric temporal lobe atrophy on imaging scans with primarily anterior and inferior regional atrophy more commonly seen in the left hemisphere. The earliest changes observed are grey matter losses in the inferior temporal and fusiform gyri, temporal pole, and parahippocampal entorhinal cortex. Notably, temporal pole atrophy could be seen on MRI imaging despite a patient’s continued ability to maintain activities of daily functioning. However, the severity of functional decline does later correlate with the extent of atrophy as it progresses to orbitofrontal, inferior frontal, insular, and anterior cingulate cortices anteriorly. While both variants of PPA share the trait of primarily left hemisphere atrophy, the nfvPPA differs from svPPA in distribution during the course of the disease. In nfvPPA, atrophic losses may begin in the inferior frontal gyrus (particularly in the pars opercularis), dorsolateral prefrontal cortex, superior temporal gyrus, and insula. Over time, these losses will include the prefrontal and temporal lobe structures in the right hemisphere. Ipsilateral anterior frontal, lateral temporal, and anterior parietal lobes are also affected. Lastly, the logopenic PPA subtype has a more posterior profile involving the left temporoparietal and posterior cingulate atrophy in the early stages of disease progression. In summary, bvFTD subtypes present with frontal and temporal lobe pathology that includes key tracts such as the corpus callosum and cingulum, while PPA subtypes are more focal and asymmetric presenting with white matter changes in specific networks associated with language processing. Other pathologically defined forms also exhibit specific atrophic patterns. PSP is associated with midbrain atrophy, and TDP-43 proteinopathies present with a variation of symmetric and asymmetric atrophic findings according to the underlying TDP type.

18F FDG-PET imaging studies also demonstrate patterns of hypometabolism that correlate with areas of atrophy, and are useful in differentiating AD from FTD. In FTD, PET studies typically demonstrate hypometabolism in the anterior frontotemporal regions including the cingulate gyri, uncus, insula, subcortical areas, basal ganglia, and medial thalamic regions. Hypometabolism is limited to frontal, parietal, and temporal cortices during the early stages of the disease but spreads outwards as the disorder progresses. Orbitofrontal, dorsolateral, medial prefrontal cortex, and anterior poles display hypometabolism in bvFTD patients’ PET scans. Likewise, svPPA patients exhibit more asymmetrical presentations with exclusively left temporal lobe hypometabolism, nfvPPA imaging shows hypometabolism in the left frontal and superior temporal regions, and logopenic variants demonstrate a left parietotemporal hypometabolism extending to anterior temporal and frontal regions.^105,108 However, it is important to note that there is considerable individual variability when using imaging to classify FTD disorders as a consequence of heterogeneity in genetic associations and the underlying pathological causes that are yet to be fully understood. Consequently, neurobiological and imaging markers remain ancillary modalities in the diagnosis and classification of FTD subtypes in contrast to the gold standard of clinical evaluations. Please see Table 2 for a summary of neuroimaging findings commonly found in FTD variants.

Table 2.

Frontotemporal dementia neuroimaging correlations. ¹⁰⁵

Variant	MRI findings	PET/SPECT findings
bvFTD	Atrophy in prefrontal cortex, anterior temporal regions, anterior cingulate, the striatum, thalamus, the insula	Decreased FA in uncinate fasciculus, genu of corpus callosum and cingulum, superior and inferior longitudinal fasciculi, inferior fronto-occipital fasciculus
nfvPPA	Predominant left hemisphere cortical loss; atrophy predominantly affecting pars opercularis in the inferior frontal gyrus extending to prefrontal and temporal cortices, caudate and putamen atrophy bilaterally	Decreased FA in left uncinate fasciculus, inferior longitudinal fasciculus tracts, corpus callosum, cingulum; notable sparing of occipital white matter, brainstem, cerebellum
svPPA	Predominant left hemisphere cortical loss; grey matter loss predominantly in inferior temporal and fusiform gyri, temporal pole, parahippocampal cortex, entorhinal cortex; losses extend to anterior cingulate, orbitofrontal, inferior frontal, and insular cortices	Decreased FA in left superior longitudinal fasciculus, corpus callosum, left cingulum, left orbitofrontal, inferior frontal, anterior temporal, inferior parietal white matter regions
Logopenic PPA	Atrophy of left temporoparietal and posterior cingulate	Notable white matter losses in left frontoparietal regions including left superior longitudinal fasciculus and arcuate fasciculus

FTD, frontotemporal dementia; bvFTD, behavioral variant FTD; FA, fractional anistotrophy; MRI, magnetic resonance imaging; nfvPPA, nonfluent variant primary progressive aphasia; PET, positron emission tomography; PPA, primary progressive aphasia; SPECT, single-photon emission computed tomography; svPPA, semantic variant primary progressive aphasia.

Management and treatment

There is a lack of quality studies and reports concerning the management of problematic behaviors in FTD with mainly sparse case reports and case series detailing their effectiveness. Nevertheless, nonpharmacological strategies are considered the preferred intervention before resorting to pharmacological therapies that may exacerbate medical comorbidities that affect more elderly patients. The main purpose of nonpharmacological interventions is to prevent disruptive behaviors, provide symptom relief, and lessen caregiver distress. Environmental approaches aim to decrease irritability, aggression, and anxiety that arise from patients’ difficulty processing information from the array of daily stimuli. These interventions include reducing noise, limiting stimulation, simplifying social parameters by limiting interaction to small groups of people, and facilitating the removal of complicated daily activities that may confuse, and therefore agitate, patients. There should be special care in implementing hearing aids, and optimizing sensory stimulation to limit patient discomfort. Introducing games or old hobbies may reduce disinhibition and inappropriate behavior. Exercise has been suggested to reduce behavioral symptoms as well. Confronting and challenging psychotic symptoms may lead to more agitation, so reassurance and distracting patients are considered to be more acceptable alternatives in decreasing behavioral symptoms. Clinicians have attempted to modify compulsions, apathy, and lack of motivation using reward systems, although there are limited data concerning this intervention. Cognitive behavioral therapy is currently being studied as a possible option to modulate behavior, but studies are still preliminary.^109,110

There are currently no US Food and Drug Administration (FDA) approved pharmacological therapies for FTD. ¹¹¹ There is limited evidence that antidepressants such as selective serotonin reuptake inhibitors (SSRIs) may be useful for impulsivity, irritability, eating behavior, and disinhibition. ¹¹² For example, one small, randomized-controlled, open-label 14-month trial ¹¹³ reported improvement of behavioral symptoms in subjects with ages ranging from 64 to 68 years after being prescribed paroxetine at dosages of up to 20 mg/day. Notably, subjects tolerated paroxetine well and exhibited improvement in Neuropsychiatric Inventory (NPI) scoring. However, a follow-up study ¹¹⁴ using placebo controls with a similarly small subject population determined that paroxetine administration at dosages up to 40 mg/day provided little objective improvement and may have caused impairment in several neuropsychological tests including visual discrimination tasks, errors in paired-associates learning task, and delayed pattern recognition memory accuracy. Discrepancies between study results were attributed to subjects responding more favorably to lower doses of paroxetine suggesting a higher rate of tolerability at 20 mg daily compared with higher dosages. Additionally, paroxetine may be more effective when prescribed earlier in the disease process, although it is difficult to determine without more in-depth and higher-powered studies.

Citalopram is another SSRI that has been recently shown to provide some improvement in FTD patient behavior. Herrmann and colleagues found improvement with a citalopram challenge at 40 mg daily dosages that led to a decrease in behavioral symptoms including irritability, depression, apathy, and disinhibition, while also improving overall NPI scores. ¹¹⁵ More recently, in a randomized double-blinded placebo-controlled crossover study, bvFTD patients who were prescribed 30 mg daily dosages of citalopram demonstrated improved disinhibition symptoms in addition to a partial restoration of serotonergic neurotransmission in dysfunctional prefrontal cortical systems as measured by magnetoencephalography and electroencephalography. ¹¹⁶

Studies of sertraline efficacy in treating FTD symptoms are limited to mainly observational studies. One study comparing stereotypical behavior in FTD and AD subjects suggested that stereotypical movements in FTD could be decreased with sertraline administration at a dosage range of 50–100 mg/day. ¹¹⁷ A case study of a 53-year-old ALS-FTD patient indicated that sertraline treatment could also decrease inappropriate sexual behavior in addition to decreasing aggressive behavior which was helpful in minimizing caregiver burden. ¹¹⁸

Alternatively, treatment with trazodone at dosages of at least 300 mg/day over 12 weeks has been reported to be helpful in decreasing symptoms of problematic eating, agitation, irritability, dysphoria, and depression, although adverse events were more prominent in subjects being prescribed the psychotropic. Typical emergent side effects included fatigue, dizziness, and hypotension, but the authors reported that these symptoms were mild in severity. ¹¹⁹ At least one open-label 12-week trial showed a treatment response with fluvoxamine administration from a dosage range of 50–150 mg/day (mean dose was 110 mg/day) improving stereotypic behavior, eating behavior, and roaming behavior. ¹²⁰

Stimulants have been proposed as alternative treatments due to their mechanism of action in elevating extracellular catecholamine concentrations to improve cognitive functioning and lessen frontostriatal and orbitofrontal dysfunction reported to affect risk-taking behavior in FTD patients.^121–123 Exemplifying this, one study ¹²³ found that eight patients who were prescribed methylphenidate at 40 mg daily dosages demonstrated decreased risk-taking behavior with overall improvement in general behavior. Additionally, a randomized double-blinded crossover trial comparing dextroamphetamine and quetiapine use in FTD patients found that dextroamphetamine administration decreased disinhibition and apathy symptoms, while being more tolerable than quetiapine, which caused sedation in their subject population. ¹²⁴ Despite these studies suggesting the safety and efficacy of stimulants in treating specific symptoms common in the FTD population, a clinician’s medical decision-making process should take into account the limited data provided by these stimulant studies due to their low power and questionable generalizability, as well as possible side effects and tolerability issues of stimulants in more elderly patients.

Memantine and cholinesterase inhibitors have not demonstrated clinically significant efficacy in treating FTD patients, and may also hasten cognitive decline or worsen behavioral symptoms.^125,126 This is likely due to a lack of consistent evidence of a cholinergic deficit in FTD without concomitant pathology of other major neurocognitive disorders. ¹²⁷ However, at least one study reported some efficacy in improving behavioral symptoms with galantamine treatment in an aphasic subgroup of FTD without a preponderance of serious adverse events, although no effect was found in overall treatment of the general FTD subject population. ¹²⁸

Complaints of insomnia may be due to behavioral dysfunction or neuronal degeneration causing impairment of the sleep/wake cycle. However, primary sleep disorders such as restless leg syndrome may also contribute to insomnia and can be treated with alpha-2-delta calcium channel ligands (e.g. gabapentin, pregabalin) or dopaminergic agents. Agitation may be controlled by atypical antipsychotics such as risperidone, olanzapine, quetiapine, and clozapine, but they are not FDA-approved and could be detrimental in a dementia-associated agitation due to increased risk of mortality and unwanted side effects.^111,129,130 Of note, antipsychotics have been suggested to be more helpful in treating psychotic symptoms in patients with the C9orf72 subtype of FTD, but once again, data is limited with increased risk of extrapyramidal side effects and morbidity in dementia patients preventing safe utilization of these neuroleptics. ¹³¹ Consequently, due to the limited data and power of pharmacologic treatment studies in FTD populations; nonpharmacological interventions and caregiver support are the preferred management options. ¹²⁵

Due to the limited contemporary psychopharmacologic interventions available to physicians, several novel approaches are being researched to develop more effective treatments for FTD symptoms. One such approach is the use of the neuropeptide, oxytocin, to aid in mediating behavior and improving apathy and empathy in svPPA patients. ¹³² Another study suggested agomelatine, a melatonin receptor agonist, may improve apathy in FTD patients. ¹³³ A preclinical trial reported active immunization against tau aggregates using anti-tau antibodies may improve cognition through decreasing tauopathy. ¹³⁴ A study investigating the effectiveness of inhibiting tau aggregation in bvFTD patients using leuco-methylthioninium was recently completed, but detailed results are pending release and publication. ¹³⁵ Other targets for therapy include disrupting the downstream effects of PGRN and C9orf72 mutations. ¹¹¹ For example, antisense oligonucleotides are being used to target the transcription elongation factor SUPT4H1/SUPT5H to bi-directionally suppress C9orf72 RNA transcription. ¹³⁶ Other proposed methods involve elevating PGRN levels through molecular manipulation of biological pathways using suberoylanilide hydroxamic acid, androgen treatment, and bafilomycin A1. ¹³⁵

Conclusion

The entirety of the FTD clinical syndrome is now known to be composed of complex and heterogeneous variants that vary in presentation, genetic alterations, and pathological processes. Consequently, there is a need for physicians to become well versed in the diagnostic criteria set forth by consortiums and expert consensus in order to properly assess and manage FTD patients. According to recent epidemiologic studies, FTD is likely to be underdiagnosed due to the inherent difficulty accurately diagnosing these disorders. These findings underline the need to standardize and update current criteria to increase diagnostic sensitivity, and allow for better management of these variants at an earlier stage of the disease course. There have been several breakthroughs in genetic and neuroimaging studies that have elucidated the pathophysiological characteristics of these FTD subtypes in the past decade. However, FTD pathology is still relatively unknown and requires further investigations to produce more effective diagnostic biomarkers and therapies, especially as later course presentations begin to mimic AD symptomatology. Current management guidelines recommend a focus on nonpharmacological interventions, as these have been found to be the most studied and effective practices at the disposal of clinicians. Contemporary pharmacological treatment options are still lacking due to a deficit of comprehensive randomized controlled trials to determine efficacy and safety, although there is limited evidence of SSRI efficacy in decreasing FTD-specific symptomatology and difficult behavior. Additionally, the extensive list of symptoms that are being targeted for intervention typically necessitates a variety of therapies for clinical improvement, which fundamentally complicates treatment planning. However, this renewed interest in FTD research gives hope for future advancements that aim to increase diagnostic capabilities, and expand treatment options through novel therapeutic targets.