Frontotemporal dementia (FTD) refers to a group of neurodegenerative brain disorders characterized by atrophy of the frontal and anterior temporal lobes (
1) and is one of the most common forms of early-onset dementia (
2). Clinically, there are three main syndromes of FTD that are generally recognized on the basis of their clinical presentations: a behavioral variant FTD (bvFTD) characterized by a progressive deterioration of personality, social comportment and cognition (
3); and two language presentations, classified under primary progressive aphasia (PPA), in which an insidious decline in language skills is the primary feature (
1). These PPAs are divided on the basis of the pattern of language breakdown into a nonfluent variant of aphasia (nfvPPA) and a semantic variant (svPPA). A third form of PPA, the logopenic variant (lvPPA), usually occurs in association with the pathology of Alzheimer’s disease but can also be found in relation to FTD (
4). There are forms of the disease that escape the descriptions of the main syndromes in FTD; an example of this is the right temporal variant, which is often associated with semantic memory impairment, prosopagnosia, and behavioral symptoms often associated with a socioemotional deficit (
5). These syndromes have specific clinical symptoms and neuroimaging and pathological characteristics, although considerable heterogeneity and overlap exist in clinical practice, particularly as the disease progresses (
1). In this article, pharmacological and nonpharmacological treatments for the neuropsychiatric aspects of FTD are reviewed. Promising advances in molecule-based therapies for the genetic forms are highlighted.
NEUROPATHOLOGY
Frontotemporal lobar degeneration (FTLD) is the term used to refer to a group of progressive brain diseases that predominantly affect the frontal and anterior temporal lobes. Thus, FTLDs include the clinical syndromes that are part of FTD: bvFTD, nfvPPA, and svPPA and also include progressive supranuclear palsy (PSP), corticobasal degeneration (CBD), and bvFTD with motor neuron disease (FTD-MND). These diseases, although sharing similar anatomy, have diverse etiologic causes at the neuropathological and genetic levels. Pathologically, FTLD is subdivided according to the composition of the abnormal inclusions of misfolded proteins. There are three main subgroups: FTLD-tau, FTLD-TDP, and FTLD-FET. The first two, characterized by the accumulation of tau protein and TDP-43, account for 90%−95% of FTLD cases; the third, a consequence of FET protein accumulation, is related to the remaining 5%−10% of cases (
6).
Tau is a microtubule-associated protein that has been linked to multiple molecular processes, including synaptic plasticity, cell signaling, and regulation of axonal stability (
7). There are six isoforms of tau expressed in the brain from alternative mRNA splicing of a single gene,
MAPT. What separates the longer forms of tau from the shorter forms is the inclusion (or exclusion) of the 31 amino acids encoded in exon 10 at the carboxy-terminal end into three isoforms with four repeats (4R forms) and their exclusion into three isoforms with three repeats (3R forms) (
8). The expression of tau is regulated throughout development, and in the adult brain, all six tau isoforms are present, with equal numbers of 3R and 4R forms. In Pick’s disease, we mainly find 3R forms, whereas 4R forms can be related to PSP, CBD, nfvPPA, and other pathologies such as globular glial tauopathy (GGT); mixed 3R-4R forms can be found in Alzheimer’s disease (AD) and chronic traumatic encephalopathy (
9). In addition, the tau protein undergoes posttranslational changes, the most commonly described being phosphorylation, which can modify the affinity of tau for microtubules and lead to self-aggregation. There are 85 known putative phosphorylation sites (
7), and tau posttranslational modifications characterize disease heterogeneity and stage in dementia (
10).
FTLD-TDP is characterized by the accumulation of 43-kDa transactive response DNA binding protein (TDP-43), a multifunctional nucleic-acid-binding protein related to RNA metabolism to which other functions such as neurite outgrowth and axonal repair after injury have been attributed (
11). TDP-43 loss from the nucleus leads to the up- and downregulation of more than 100 different proteins, including stathmin-2 (
12). Upregulation of stathmin-2 is a proposed therapy for FTLD-TDP. In FTLD-TDP, abnormally phosphorylated and ubiquitinated versions of TDP-43 manifest with different morphology and are grouped according to the number of neuronal cytoplasmic inclusions, dystrophic neurites, and glial cytoplasmic inclusions into subtypes A, B, C, and D (
13). Approximately 90% of svPPAs present TDP-43 type C, FTD-MND is almost exclusively related to type B, amyotrophic lateral sclerosis (ALS) is related to types B and D, and bvFTD and nfvPPA are related to types A and B (
6,
14).
The FUS protein (fused in sarcoma) is one of several FET proteins (also including EWS and TAF15 proteins) associated with FTLD and ALS. FUS, like TDP-43, regulates mRNA and, in addition to being associated with bvFTD and PPA, is also a cause of ALS and neuronal intermediate filament inclusion disease (
6). Functions linked to alternative splicing, transcription, and RNA transport are attributed to FUS. By affecting splicing, FUS dysfunction could affect normal
MAPT expression, leading to an increased 4R-Tau–3R-Tau ratio (
15).
GENETICS
FTD is strongly heritable. A positive family history is found in 30%−50% of cases, and in 10%−27%, the inheritance is autosomal dominant (
2,
16–
18). In comparison, in less than 1% of AD cases, the inheritance is autosomal dominant (
19). Additionally, genetic causes are found in 1%−10% of sporadic bvFTD cases (
20). Thus, genetics is of fundamental importance in the assessment of patients with FTD (
21). Numerous genes are associated with FTD; however, the most commonly implicated genes are
MAPT,
GRN, and
C9orf72.
MAPT, located on chromosome 17, is the gene encoding for the tau protein, whose function has been described previously. There are more than 50 known mutations for this gene. These mutations account for 5%−20% of familial FTD cases but are rarely found in sporadic forms of FTD (0%−2%) (
21). Disease onset with these
MAPT mutations varies but is often before 60 years of age. These mutations usually cause a bvFTD phenotype, but parkinsonism is often prominent.
Also located on chromosome 17 adjacent to
MAPT is the
GRN gene encoding for progranulin. This protein functions as a multifunctional growth factor in development, wound repair, neuroinflammation, autophagy, and lysosomal function (
21,
22). More than 70 known
GRN mutations lead to the generation of nonsense mRNA, which is subsequently eliminated by physiological surveillance mechanisms, leading to haploinsufficiency of the progranulin protein, which thus leads to FTLD-TDP pathology by an unclear mechanism (
23,
24).
In 2011, it was described for the first time that the six-nucleotide noncoding repeat (G4C2) in the first intron of the
C9orf72 gene, located on the short arm of chromosome 9, could lead to FTD, ALS, and FTD-MND (
25,
26). Up to 24 G4C2 repeats have been described in healthy control groups. Although there is no universally established cutoff point, it is suspected that more than 30 expansions increase susceptibility to neurodegeneration (
26,
27). Although the mechanism of pathogenicity by which this nucleotide expansion leads to the development of FTD is not well defined, multiple hypotheses have been formulated, including haploinsufficiency of the homonymous protein; toxicity from the transcribed, expanded-repeat-containing RNA; up- and downregulation of numerous proteins, including stathmin-2; and toxic dipeptide repeat (DPR) proteins (
21). Regardless of the pathological mechanism by which it produces the disease, the
C9orf72 mutation is the leading cause of familial FTD (20%−30% of cases) and the leading known genetic cause of sporadic FTD (6%) (
21,
28–
30). In addition, this expansion leads to the accumulation of DPRs that aggregate in the cerebellum and hippocampus (
31) and TDP-43 type A and type B pathology (
14). Clinically, this mutation can present as bvFTD, ALS, or both and is characterized by a shorter disease duration (6.4±4.9 years) relative to other genes such as
MAPT or
GRN (
32). However, a group of carriers of this mutation may have extremely slow-evolving forms of the disease syndromically indistinguishable from bvFTD, categorized as “FTD phenocopies” (
33). Moreover, between 10% and 50% of patients with this mutation may manifest psychotic symptoms (hallucinations, delusions, or both), which may lead to confusing this disease with psychiatric conditions such as schizophrenia, bipolar disorder, or obsessive-compulsive disorder (
2,
27,
34).
There are recommendations for genetic testing for the three main bvFTD-related genes (
MAPT,
GRN, and
C9orf72) in patients with at least one affected first-degree relative. This recommendation extends to FTD or early-onset dementia relatives, but a history of ALS, Parkinson’s disease, or unexplained late-onset psychiatric disorders should also be considered (
2). Also, because of its association with sporadic cases of FTD, one should look for
C9orf72 mutations in cases of late-onset behavioral symptoms (even if they do not meet all criteria for bvFTD), and there are no neuroimaging abnormalities, as a diagnostic element (
2). With a significant proportion of apparent sporadic cases that are due to unexpected mutations, several groups are moving toward genetic testing in all FTD cases, even without family history. This approach will become more routine when therapies become available.
Commonly, when mental health professionals explain the diagnosis of FTD to patients and their family members, they are often concerned about the heritability of the disease. Before obtaining genetic testing, the implications for the individual and their family unit should be discussed with a genetic counselor. Family members may also be interested in genetic testing. Genetic counseling has a cost and can have legal and, sometimes, ethical implications. Before testing a family member (or members), the ideal scenario is to have the affected gene identified and search for the mutation in those concerned. This is not always possible, for various reasons. For example, affected relatives may be deceased; the afflicted patient may refuse to be tested, and so forth. When the test for a single gene is negative, another gene may be responsible for the clinical syndrome (
35). Even when the presence of a mutated gene is demonstrated, it is not possible to predict the exact age of onset, severity or type of symptoms, or the course of the disease (
35). Also, there are numerous factors to consider when evaluating whether to perform genetic testing. For example, does the person understand what having the mutation implies, or will the genetic result affect the person’s life decisions? Additionally, it is important to consider what psychological impact this information can have and whether the patient is prepared for this information (
36). These decisions also have legal and economic consequences, including the possibility that the result affects the patient’s health insurance coverage or the possibility that this information can be used by an employer to decide whether to hire the person. In the United States, in 2008, the Genetic Information Nondiscrimination Act, or GINA, was passed at the national level to prohibit information such as this from being used in the context of administrative decisions such as health insurance or employment decisions (
37). However, many countries do not have similar legislation on this type of information, which could potentially expose mutation carriers to stigma and marginalization.
LEGAL ASPECTS
Brain disorders have long been considered as a cause of criminal behavior (
45). This seems to be particularly true in the case of FTD, where such behaviors can be found in up to 50% of cases (
46,
47), up to five times more frequent than in patients with AD (
48). One of the most prominent symptoms in FTD is behavioral disinhibition (
3). These behaviors are often labeled as disinhibited because they break with social norms and frequently transgress legal boundaries (
49,
50). There are at least two forms of disinhibition (
51): impulsivity, acts involving general rule violations that are related to an impairment of cognitive control mechanisms (
49); and person-based disinhibition, where the behavior is more related to disturbed interpersonal interactions violating social tact and personal boundaries (
52), in which case behaviors may emerge because of the compromise of cognitive systems related to semantic knowledge (
53) or personal salience (
54). This type of behavior has been labeled as “acquired sociopathy” (
55). Because antisocial behavior in patients with FTD arises as a result of compromised functioning of brain structures responsible for directly or indirectly modulating behavior, these cases present a challenge in defining the degree of autonomy in their actions, this being particularly true for the early stages of the disease (
56). Thus, this disease presents a challenge for the criminal justice system.
CURRENT CLINICAL TRIALS AND BIOMARKERS
Molecule-based therapies are being considered for the genetic forms of FTD and to treat the symptoms of the disease (
Table 2). For each of the major genetic subtypes,
MAPT,
GRN, and
C9orf72, different approaches will be needed. Several advances have made it possible to consider such efforts. First, the discovery of powerful biomarkers such as the neurofilament light-chain protein (NfL) will make it possible to follow progression in a clinical trial, because NfL begins to rise during the transition from asymptomatic to mildly symptomatic FTD (
130). Similarly, structural imaging can detect significant changes in atrophy over 6 months, making it likely that the magnetic resonance imaging (MRI) can also be used as a surrogate marker (
131).
For
MAPT carriers, the major emphasis of clinical trials will be to lower tau either by decreasing its production or by increasing its clearance. There is extensive evidence that lowering tau will ameliorate symptoms in animal models of AD and FTD (
132,
133). Further, in humans, antibodies against tau have been demonstrated to reach the brain and bring tau into the plasma (
134), but clinical trials using antibodies for both AD and FTD have been disappointing. These failures have probably occurred because of relatively low levels of the antibody that cross the blood-brain barrier. Therefore, technologies such as antisense oligonucleotides and CRISPR could lower tau in a highly effective manner. If
MAPT carriers treated with effective tau-lowering therapies show slowing or progression or even halting of the disease, these approaches will next be used to treat other tau-related forms of FTD. Other efforts are focused on increasing the degradation of tau in the lysosome or the proteosome (
135).
With
GRN, different mechanisms and different approaches are being considered.
GRN mutation carriers show markedly reduced brain and blood levels of progranulin, suggesting a haploinsufficiency mechanism with the deficiency of progranulin production on one chromosome sufficient to cause FTD. Many strategies are being considered to increase brain progranulin, with many focused on better ways to deliver progranulin into the brain. Arrant and colleagues found, when using an AAV vector (AAV-
Grn) to deliver progranulin in
Grn−/− mice, that lysosomal dysfunction and microglial pathology were both ameliorated (
136). It is likely that, in the coming year, AAV transplantation studies will begin with
GRN gene carriers, and others delivery systems are also being considered.
Finally, C9orf72 mutations produce a long hexanucleotide repeat that is already the target for gene carriers with ALS, and therapies for FTD are being considered. As with all these gene-related therapies, multiple questions will need to be answered regarding delivery of the drug to the brain; timing of the therapy (presymptomatic versus symptomatic); the reliability of biomarkers; and, most importantly, efficacy. A new chapter in therapy for FTD and related conditions is beginning. Once the genetic forms of the disease have been effectively treated, new approaches to the sporadic form of the disease are likely to emerge.
CONCLUSIONS
FTD is frequently misdiagnosed, and when the diagnosis occurs, it often comes late in the course of the illness or is missed. Recognition that behavioral changes represent a neurodegenerative condition is difficult, leading clinicians to diagnose a primary psychiatric disorder. Also, diagnostic tools such as blood biomarkers or neuroimaging can be difficult to access, particularly in low- and middle-income communities. Another barrier to its identification is the lack of knowledge and training for health care providers about FTD.
FTD treatment has been limited to the management of neuropsychiatric symptoms, these being the most prominent feature of the disease. Therapeutic strategies have focused on nonpharmacological interventions such as behavioral and environmental manipulation, caregiver interventions, and speech therapy for the language variants of FTD. Also, pharmacological treatment has also been used to treat these symptoms, with variable but sometimes positive results. With the advance of knowledge regarding the pathophysiology of FTD, pharmacological interventions such as the use of SSRIs, trazodone, or second-generation antipsychotics have a solid scientific basis for the treatment of FTD.
In the past 10 years, thanks to new techniques in neuroimaging, genetics, and biomarker analysis, much has been discovered about the phenomena underlying frontotemporal lobar degeneration. This has allowed the design of new molecule-based therapies that are still in the early stages of research but show promising results.