Clinical manifestations

Presentation and course

Clinically, frontotemporal dementia is expressed as three variants: behavioral variant, nonfluent/agrammatic variant primary progressive aphasia, and semantic variant primary progressive aphasia (37). Consensus and nosology are consolidating around a fourth variant, the right temporal variant of frontotemporal dementia (55; 60). Each is characterized by slow, insidious changes in behavior and/or language. Many patients also go on to develop motor dysfunction.

(1) Behavioral variant frontotemporal dementia (bvFTD). The earliest signs of disease in behavioral variant frontotemporal dementia (bvFTD) are frequently subtle personality and behavioral changes that become increasingly pronounced. The six core clinical features include disinhibition, apathy, loss of empathy, compulsivity, hyperorality, and executive dysfunction (46). Childish behavior, rudeness, inappropriate sexual remarks or jokes, impatience, careless driving, excessive spending or hoarding of certain items, perseverative routines, compulsive roaming, insistence of certain foods or excessive food intake, neglect of personal hygiene, and disinterest in the immediate family are all common symptoms. The repetitive behaviors in frontotemporal dementia can take the form of compulsions, the absence of impulse control, stereotypy, hoarding, restrictive behaviors, and insistence on sameness (47). Moreover, these changes, at least initially, may stand in stark contrast to patients’ cognitive ability, which may remain intact for some time.

The diagnostic criteria for behavioral variant frontotemporal dementia were revised in 2011 and are called the FTD Criteria Consortium FTDC criteria (Table 1) (46). With these criteria, diagnosis of possible bvFTD is based solely on clinical presentation (three of six core criteria), whereas a diagnosis of probable bvFTD is based on a diagnosis of possible bvFTD with concomitant functional decline and evidence of frontal and/or temporal atrophy on neuroimaging. The diagnostic sensitivity and specificity of the FTDC criteria have been investigated via retrospective chart review of pathologically confirmed cases in two separate studies. In the first study, they found the FTDC criteria to have greater sensitivity to the diagnosis of bvFTD when compared to the previous (“Neary”) criteria (72% vs. 51%, respectively) (46).

Table 1. International Consensus Criteria for Behavioral Variant Frontotemporal Dementia (FTDC)

(2) Nonfluent/agrammatic variant primary progressive aphasia (nfvPPA). NfvPPA is also known as progressive nonfluent aphasia or primary progressive aphasia-agrammatic (PPA-G). NfvPPA is characterized by early disturbances in motor speech output and loss of syntax (ie, grammatical structure of language). Most patients with symptoms initially develop apraxic speech that results in articulatory difficulty and phonological paraphasias (19). These patients tend to show preservation of social graces for some time into the course of their illness. This variant accounts for approximately 20% of the phenotypic expression of the disease (21).

(3) Semantic variant primary progressive aphasia (svPPA). SvPPA is also referred to as semantic dementia or the temporal variant of frontotemporal dementia. Patients with svFTD display progressive loss regarding the meaning of words, but retain fluent speech. Articulation, phonology, and syntax also remain intact, but the patient does not comprehend others’ speech and has significant word-finding difficulty (19). Characteristically, patients will ask “what is ___?” questions regarding the meaning of common nouns (26). Behavioral changes, similar to those seen in bvFTD, also occur, including disinhibition, reduced empathy, compulsions, and altered food preferences (10). This variant accounts for approximately 12% of the phenotypic expression of the disease (10).

Similar to bvFTD, the criteria for diagnosis of nfvPPA and svPPA have been significantly revised to facilitate scientific exchange across centers. In the new International Consensus Criteria outlined for primary progressive aphasia (19) (see Table 2), establishing a diagnosis is based on a 2-step process. First, a patient must meet criteria for Mesulam’s guidelines for primary progressive aphasia (35). If the patient meets criteria for Step 1, then his or her language disturbance is further parcellated into one of three variants: nonfluent variant primary progressive aphasia, semantic variant primary progressive aphasia, and logopenic variant primary progressive aphasia. Because the majority of cases of logopenic primary progressive aphasia are pathologically linked to Alzheimer disease, it will not be reviewed here.

(4) Right temporal variant frontotemporal dementia (rtvFTD). Right anterior temporal lobe degeneration leads to neurobehavioral features, especially early loss of empathy, prosopagnosia, loss of place knowledge, cognitive rigidity, disinhibition, apathy, obsessive-compulsive symptoms, changes in eating habits, and depression with person-specific and socioemotional knowledge deficits (55; 09; 60). Patients are diagnosed later than the left temporal variant of frontotemporal dementia and have shorter mortality (14). Genetic risk can be familial or sporadic with multiple genes implicated (56).

Table 2. Classification of Primary Progressive Aphasia

Associated disorders. Our enhanced understanding of the genetics, pathology, and epidemiology of frontotemporal dementia has led to recognizing that the three clinical variants of frontotemporal dementia often occur within the context of progressive supranuclear palsy (PSP), corticobasal syndrome (CBS), and motor neuron disease (MND)/amyotrophic lateral sclerosis (ALS) (04).

Progressive supranuclear palsy. The motor syndrome of progressive supranuclear palsy, which includes postural instability, axial rigidity, bradykinesia, frequent falls, dysphagia, and vertical gaze palsy, was originally described by Steele and colleagues in 1964 (53). In addition to its motor features, some symptoms resemble bvFTD as well, including significant apathy and dysexecutive symptoms (04).

Corticobasal syndrome. Corticobasal syndrome has traditionally been characterized as a clinical syndrome consisting of asymmetric rigidity, prominent apraxia, cortical sensory loss, reflex myoclonus, bradykinesia, tremor, and dystonia without prominent early dementia (17). However, an international research committee has analyzed the clinical symptomatology of 267 pathologically confirmed cases of corticobasal degeneration (a term which refers specifically to pathologically confirmed corticobasal syndrome). Based on their analysis, they set forth guidelines that reflect the heterogeneity of corticobasal syndrome, which includes phenotypes similar to both bvFTD and nfvPPA (02).

Motor neuron disease/amyotrophic lateral sclerosis (MND/ALS). MND/ALS involves degeneration of the pyramidal tract or anterior horn cell disease. Heterogeneous cognitive and behavioral impairment occurs in amyotrophic lateral sclerosis, with neuropsychological deficits affecting over 50% of people with amyotrophic lateral sclerosis (54). The discovery of the C9ORF72 and FUS gene mutations in familial frontotemporal dementia and frontotemporal dementia-amyotrophic lateral sclerosis likely account for the high reported incidence of behavioral (and sometimes aphasia) symptoms in these patients (45).

Prognosis and complications

Across all clinical variants of frontotemporal dementia, survival from time of symptom onset has been estimated to range from 6 to 11 years (29; 10). As the disease progresses, immobility due to motor symptoms and difficulties with swallowing and choking may shorten survival.

Clinical vignette

A 68-year-old male shopkeeper became ill approximately 4 years prior to his first clinical examination. His initial symptoms included errors in judgment while driving (which resulted in rear-ending another driver) and carelessness while handling money. Two years later, he began to display socially inappropriate behavior, rudeness, and a short attention span. He would often stand up during the middle of a conversation and abruptly leave. He developed a preference for salami and would eat half a pound in one sitting. He would often eat an entire bunch of bananas.

He became apathetic. For example, he stopped doing yard work and did not shovel the snow during the winter. His hygiene declined, and he wore the same clothes for a week at a time. He only showered at the YMCA after exercising. When requested, he shaved with an electric shaver but while sitting and watching television. He was often fidgety and would manipulate objects for no reason. He became emotionally distant and disinterested and only talked about what he watched on television. Later, he stopped participating in conversations, had difficulty getting words out, and talked less. Customers at his shop, which had become quite messy, noted his lack of patience and easy frustration.

On initial examination, he was impulsive and interrupted conversations with unrelated thoughts and concerns. He had trouble with some of the sequential motor tests as well as a similarities test (considered to assess complex reasoning). His gait at that time seemed rushed; at other times he shuffled with reduced stride length, and he turned stiffly, resembling patients with Parkinson disease. His upward gaze appeared markedly reduced. He was diagnosed with early frontotemporal dementia, which led to subsequent referral for neuroimaging. His CT scan was notable for mild frontoparietal atrophy. (Author’s note: the characteristic atrophy pattern in frontotemporal dementia is frontotemporal atrophy, and an MRI is considered a better modality than CT to assess for atrophy).

Over time, he started falling, which led to his using a walker. A rapid tremor developed in the right hand, especially when he was agitated. He continued to be impulsive and restless. For example, when eating at restaurants with his family, he would finish his meal and then leave to stand beside the car until the family was ready to go home. He was emotional and tearful, at times with no provocation (pseudobulbar affect). His speech became perseverative, slurred, and less intelligible. He often repeated words and only spoke in short sentences.

On a subsequent evaluation, he was noted to have festinating gait, en bloc turns, and axial rigidity. He also displayed severe vertical gaze palsy and impaired lateral pursuit. To check his watch, he had to raise his arm up in front of his eyes (the “watch salute”). He also had several bursts of disinhibited crying and laughing with no provocation (pseudobulbar palsy), and his speech was slow and slurred, with a sing-song inflection. A rapid jerky tremor was evident in the right hand. His clock drawing was very small, and the numbers were crowded together and difficult to recognize. He failed to place the hands appropriately. Language testing demonstrated nonfluent aphasia in addition to his dysarthria. Eventually, he was admitted to a nursing home.

His case is an example of the behavioral variant of frontotemporal dementia with superimposed progressive supranuclear palsy and progressive aphasia, a convergence of the behavior, language, and extrapyramidal manifestations of the disease. This case is detailed in the case-based book on frontotemporal dementia in the chapter “The Hero of Bolero” (23).

Biological basis

Etiology and pathogenesis

Frontotemporal dementia is thought to be caused by the abnormal aggregation of proteins in the brain, which begin in selective, vulnerable neuroanatomical “hubs.” As the disease progresses, the proteins disseminate through specific neuroanatomical networks, conferring the unique clinical characteristics seen at each stage of the disease (52). The pathology is heterogeneous, though three major proteins have now been shown to cause the majority of cases of frontotemporal dementia. A large proportion of cases are genetic and caused by mutations in the tau and progranulin genes on chromosome 17 and the C9ORF72 gene on chromosome 9.

Gross pathology. The gross pathology of frontotemporal dementia is asymmetric atrophy of the frontal and anterior temporal lobes.

In bvFTD, disease begins primarily in paralimbic structures such as the anterior cingulate cortex, frontoinsular region, dorsal anterior insula, and lateral orbitofrontal cortex (51). As the disease progresses, it moves towards the dorsolateral prefrontal cortex and posteriorly towards the anterior and lateral temporal lobes (05). Nonfluent variant primary progressive aphasia typically involves the left inferior frontal gyrus and insula, whereas semantic variant primary progressive aphasia typically begins in the left anterior temporal lobe, though it eventually becomes bilateral, progressing to include degeneration of the posterior insula and ventromedial frontal lobe as well (18).

Histopathology. The first protein to be identified in the pathogenesis of frontotemporal dementia was tau (FTLD-tau). Normal tau proteins contribute to axonal transport by binding to microtubular proteins. However, abnormally phosphorylated and aggregated tau proteins are biochemical markers of various forms of degenerative dementia, collectively referred to as “tauopathies,” and include Alzheimer disease, Pick disease, corticobasal degeneration, progressive supranuclear palsy, and chronic traumatic encephalopathy (amongst others) (39).

The majority of cases of frontotemporal dementia, however, are not tau-reactive. Rather, they display ubiquitin-immunopositive histology. These cases were termed dementia lacking distinctive histology for many years (28). It was eventually found that these cases displayed ubiquitinated protein inclusions (FTLD-U); however, the exact protein was unknown. This changed in 2006, when it was determined that transactive response DNA-binding protein 43kDa (TDP-43) was the protein responsible for most ubiquitin-positive cases of frontotemporal dementia (38). It has now been shown that approximately 50% of all cases of frontotemporal dementia have TDP-43 pathology, as well as the vast majority of cases with amyotrophic lateral sclerosis (38). A harmonized scheme for the pathological classification of TDP-43 was proposed, which outlines four subtypes of FTLD-TDP: type A (TDP-A), found in the majority of cases with mutations in the GRN gene; type B (TDP-B), associated with FTD-MND; type C (TDP-C) associated predominantly with semantic variant PPA; and type D (TDP-D), associated with mutations in the valosin-containing protein (VCP) gene (34). Overall, in bvFTD, tau pathology (3R > 4R) and TDP-43 (type A > type B > type D) are equally as likely (32).

In 2009, ubiquitinated FUS protein inclusions were found to be responsible for the majority of the remaining tau/TDP-43 negative frontotemporal dementia cases, constituting the third major pathological subtype of frontotemporal dementia (30). FTLD-FUS includes three closely related entities: atypical frontotemporal dementia with ubiquitin-positive inclusions (aFTLD-U), neuronal intermediate filament inclusion disease (NIFID), and basophilic inclusion body disease (BIBD) (45). Underlying FUS pathology is the least common neuropathology in bvFTD (32).

Clinicopathological correlations. Prediction of the underlying molecular pathology of clinical frontotemporal dementia is tenuous; however, some patterns have been established and can be used as guidelines. For example, svPPA is typically sporadic and associated with FTLD-TDP type C pathology, (more likely than 3R tauopathy), whereas cases of sporadic nfvPPA are more likely to be FTLD-tau, (4-repeat (4R) tau—CBD protein and PSP protein > 3-repeat (3R) tau and TDP-43 type A), especially if they also display prominent parkinsonism. BvFTD is the most difficult variant in which to predict pathology; however, certain symptoms are more likely to present with particular pathological subtypes. For example, when bvFTD is also associated with motor neuron disease, the pathological substrate is usually FTLD-TDP; if parkinsonism is present (eg, corticobasal syndrome or progressive supranuclear palsy), then pathology will likely be FTLD-tau. Should onset be very early with significant psychiatric symptoms, pathology is likely to be FTLD-FUS (45). FTD-MND is most likely to be due to underlying TDP-43 type B pathology (with TDP-43 type A and FUS being less likely). CBS is most likely to be due to 4R tauopathy (CBD), although in rare cases, CBS may be caused by Alzheimer pathology, 4R tau (PSP protein), TDP-43, or 3R tau. PSP-S (syndrome) is most often caused by underlying 4R tau pathology (PSP> CBD) (32) (see Table 1 of their paper).

Genetics. In contrast to Alzheimer disease, genetic causes of frontotemporal dementia are common and up to 40% of cases have a family history of dementia, psychiatric disease, or motor symptoms. In addition, 10% of patients with frontotemporal dementia have an autosomal dominant pattern of inheritance (40). The first genetic mutation to be linked to familial frontotemporal dementia was discovered in 1998 and was found to cause mutations in the microtubule-associated protein (“MAPT”) gene, which is located on chromosome 17 and codes for the protein tau. More than 44 tau mutations on the MAPT gene have been identified, accounting for 5% to 20% of familial frontotemporal dementia. Clinical presentation is heterogeneous, though bvFTD and extrapyramidal motor syndromes such as corticobasal syndrome and progressive supranuclear palsy tend to predominate (49).

After the discovery of the MAPT gene mutations, there remained many chromosome 17-linked families with frontotemporal dementia that did not display mutations in MAPT. The second locus was eventually discovered in 2006, in extremely close proximity to the MAPT gene. The identified mutation was found to occur in the progranulin (GRN) gene. Since 2006, 69 pathogenic mutations have been reported worldwide, which are thought to account for approximately 5% to 20% of familial cases and approximately 1% to 5% of sporadic cases of frontotemporal dementia. Mutation carriers tend to develop symptoms characteristic of bvFTD or nfvPPA.

Since 2006, increasing evidence suggested the presence of a mutation on chromosome 9, responsible for familial cases of combined frontotemporal dementia-amyotrophic lateral sclerosis. In 2011, this gene was identified as a hexanucleotide expansion mutation in a noncoding region of C9ORF72. Furthermore, this analysis demonstrated the mutation in C9ORF72 to be the most common genetic abnormality in both familial frontotemporal dementia (11.7%) and familial amyotrophic lateral sclerosis (23.5%). The clinical presentation can include bvFTD, amyotrophic lateral sclerosis, or features of both, whereas other clinical variants of frontotemporal dementia and associated disorders are rare.

Six other genes have been identified as responsible for a minority of the familial frontotemporal dementia cases and include: valosin-containing protein (VCP), chromatin-modifying protein 2B (CHMP2B), transactive response DNA-binding protein 43kDa (TARDBP), fused in sarcoma (FUS), TBK1, and TIA1 (50).

Reidl and colleagues clearly outline the molecular classification of frontotemporal lobar degeneration with its genetic and clinical correlations in Table 2 of their paper (48).

Epidemiology

Frontotemporal dementia is considered the third most common cause of dementia, behind Alzheimer disease and dementia with Lewy bodies. Estimates of its prevalence vary considerably and likely relate to challenges in identification and accurate diagnosis of the disease. Using the 2011 revised diagnostic criteria, the incidence of frontotemporal dementia in two UK counties was demonstrated as 1.57 per 100,000 person-years (10). Age of onset can range from 40 to 90, and prevalence is equivalent between male and female gender (10).

Prevention

No strategies are known except genetic counseling of mutation carriers in familial frontotemporal dementia.

Differential diagnosis

Confusing conditions

BvFTD is commonly misdiagnosed as early-onset Alzheimer disease, given that Alzheimer disease is the most prevalent dementia syndrome, and many symptoms of the two diseases can overlap. Diagnostic studies, such as CSF biomarkers (eg, Aβ1-42:tau ratio) and molecular PET amyloid imaging, can be useful to distinguish between these diseases, if available. Evidence of vascular disease, endocrine or metabolic conditions, neoplasms or paraneoplastic syndromes, CNS infections, or dietary deficiencies should be ruled out.

Many patients with bvFTD are initially diagnosed with late-onset psychiatric disturbance. Symptoms of disinhibition, euphoria, and poor judgment can mimic those of mania, whereas profound apathy and eating disturbance might be misconstrued as depression. For example, in a retrospective chart review of 252 patients with neurodegenerative disease, Woolley and colleagues found that 51% of patients with bvFTD had received a prior diagnosis of a psychiatric disorder (eg, major depression, bipolar disorder, schizophrenia), compared to 23% of patients with Alzheimer disease (eg, major depression, anxiety) (59). Psychiatric symptoms, including delusions and hallucinations, appear to be a common symptom in individuals who carry the C9ORF72 gene mutation (03). In general, late age of onset of psychiatric symptoms should be a “red flag” when considering whether an individual might have a neurodegenerative disease.

Associated or underlying disorders

Our enhanced understanding of the genetics, pathology, and epidemiology of frontotemporal dementia has led to the recognition that the four clinical variants of frontotemporal dementia often occur within the context of progressive supranuclear palsy, corticobasal syndrome, and motor neuron disease/amyotrophic lateral sclerosis (04).

Diagnostic workup

A comprehensive diagnostic evaluation for frontotemporal dementia should include a clinical interview, neurologic exam, neuropsychological assessment, and review of neuroimaging.

The clinical interview should focus on onset and progression (eg, has it been slow and insidious or abrupt and explicit?), the nature of the change, which is typically in the realm of personality, emotionality, language, and social behavior (use of the FTDC criteria as a way to frame questions is helpful), and family history (which should be probed for family members who exhibited significant changes in personality and social behavior after the fifth decade or who also may have had symptoms of a motor syndrome). It is important to remember that due to the lack of awareness and insight that typically occurs in this disease, it is unlikely that the patient will be able to give an accurate account of his or her history. Therefore, it is crucial that an informant be present to provide details of the history.

Use of global assessment measures such as the Mini-Mental State Exam (13) are typically insensitive to the earliest changes in cognition that occur in frontotemporal dementia. Patients frequently score highly, despite obvious impairment in behavior and judgment. Neuropsychological assessment should examine the relative test score pattern between measures of executive function versus episodic memory and visuospatial function (executive dysfunction > memory/visuospatial dysfunction), in conjunction with behavioral observations such as perseverative behavior, inappropriate remarks, lack of social engagement, and apathy.

Asking informants to fill out questionnaires that address changes in behavior or psychiatric symptoms, such as the Neuropsychiatric Inventory (11) or the Frontal Systems Behavior Scale (20), can also be extremely helpful. If the patient simply displays a “frontal” or “dysexecutive” syndrome without significant behavioral problems or a history of personality change, the diagnosis is more likely to be early-onset Alzheimer disease. Other useful functional and social cognition assessment tools in frontotemporal dementia include Frontal Behavioral Inventory (24), FTLD-CDR (27), Frontotemporal Dementia Rating Scale (FRS; 36), Cambridge Behavioral Inventory (57; 58), Functional Activities Questionnaire (FAQ) (43), and the Revised Self-Monitoring Scale (31).

Clinically, structural magnetic resonance imaging (MRI) is best for reviewing findings. Atrophy is often asymmetric and in the early-middle stages, confined to the medial frontal and anterior temporal lobes. Clinicians should review the MRI readings themselves because radiologists often fail to comment on atrophy patterns.

FDG-PET can be used to reveal a specific pattern of hypometabolism in the frontal and temporal lobes consistent with the synaptic dysfunction present in that region. It has been FDA approved to distinguish between frontotemporal dementia and Alzheimer disease and is particularly beneficial in the early stages of disease. FDG-PET is highly specific for neurodegeneration, even in the preclinical and prodromal phase (08). A tau-PET ligand has been FDA-approved for use in identifying the neurofibrillary tangles of Alzheimer disease, but ligands for frontotemporal dementia-associated tau and TDP-43 aggregates are still in development.

CSF analysis can exclude Alzheimer disease pathology and distinguish frontotemporal dementia from non-neurodegenerative disorders. Amyloid PET and emerging plasma biomarkers (αβ42/490, p-tau217) can also exclude Alzheimer disease. There is no current fluid biomarker that confirms frontotemporal dementia pathology, but CSF and blood-based neurofilament light chair (NfL) can differentiate frontotemporal dementia from primary psychiatric disorders (12). pNfL may be able to track progression of disease as well as predict onset in familial mutation carriers (16).

Management

Frontotemporal dementia is managed through both pharmacological and behavioral interventions. There are currently six medications approved for treatment of amyotrophic lateral sclerosis, including tofersen, edaravone, riluzole, thickened riluzole, riluzole oral film, and dextromethorphan-quinidine sulfate—the last of these is used for pseudobulbar affect, which is characterized by involuntary crying or laughing without a clear trigger (22). However, there are no FDA-approved medications for treating symptoms in frontotemporal dementia. Selective serotonin reuptake inhibitors (SSRIs) have proven most effective in treating behavioral symptoms such as apathy, depression, anxiety, agitation, delusions, self-harm, and compulsive behaviors (33). Despite widespread use, the evidence for using acetylcholinesterase inhibitors, NMDA receptor antagonists, antipsychotics, and anticonvulsants in patients with frontotemporal dementia has not been consistently demonstrated (33). Antipsychotics are often used when SSRIs are insufficient, but they have an FDA black box warning of increased mortality in patients with dementia.

Education for caregivers regarding behavioral management techniques can be extremely useful. For example, learning the “ABC”s of problem behavior (antecedent – behavior – consequence) can allow the caregiver to pinpoint what situations give rise to problem behaviors so they can anticipate and change the situation before a behavior occurs. Physical therapy for gait and balance training, home safety evaluations by occupational therapists, speech therapy for dysarthria and compensatory techniques, cognitive rehab conducted by speech therapists for patients with insight into their impairments, and swallow evaluations due to dysphagia are also important management tools. Counseling, education, and support groups are also important given the dramatic changes that occur in patients with frontotemporal dementia. These outlets provide important information regarding coping strategies (such as taking respite), resources (day programs), and information about long-term care and advanced care directives.

Efforts to develop specific, disease-modifying therapies for frontotemporal dementia are advancing rapidly (04). For example, research in tau-related therapies has led to a number of treatment strategies that have already or will soon be in the clinical trial phase. In addition, clinical trials are currently under way with agents presumed to normalize progranulin levels in individuals who bear this genetic mutation. Finally, strategies to block the expression of the C9ORF72 hexanucleotide repeat expansions are being developed for the most common genetic form of the disease. There are currently six clinical trials targeting FTD-GRN and four trials targeting FTD-C9ORF72.

Special considerations

Anesthesia

General anesthesia for elective surgeries should be avoided due to the risk of delirium leading to rapid progression of deficits.