Clinical manifestations

Presentation and course

Cognitive and behavioral impairment associated with amyotrophic lateral sclerosis. The diagnosis of a dementia syndrome requires acquired, progressive impairment in at least 2 cognitive domains that is severe enough to cause social disability. Some patients with amyotrophic lateral sclerosis have cognitive impairment but do not progress to meet criteria for dementia. In particular, frontal lobe dysexecutive syndrome and attentional deficits may accompany the motor signs of amyotrophic lateral sclerosis in patients who never develop a memory disturbance or who become dependent on others because of their relentless motor disorder rather than their cognitive or behavioral changes. Impaired performance on the Wisconsin Card Sorting Test, Reitan's Trail Making Tests A and B, and the Stroop Interference Test may be indicative of a frontal dysexecutive syndrome. Word generation tests can be a useful screening tool for cognitive decline in amyotrophic lateral sclerosis patients (100) and are recommended for use in either the oral or written formats (04; 182; 183) by the current consensus criteria. Some patients may not show impairment on such tasks but, nevertheless, manifest personality changes, disinhibition, and obsessive-compulsive behaviors, which are behavioral manifestations of frontal and temporal lobe dysfunction. Informant measures such as the Neuropsychiatric Inventory (41) or the Frontal Behavioral Inventory (91) may be useful to elicit these symptoms in a structured interview. Language abnormalities in amyotrophic lateral sclerosis include both reduced word fluency and impaired confrontational naming, and both may be detected before dementia is diagnosed (03).

According to guidelines (162), patients with amyotrophic lateral sclerosis and associated cognitive or behavioral impairment should be classified according to the following schema:

(1) Patients meeting at least 2 non-overlapping supportive diagnostic features from either the Neary criteria (118) or Hodge’s criteria (76) for frontotemporal dementia shall be diagnosed ALSbi (amyotrophic lateral sclerosis behavioral impairment).

(2) Patients with evidence of cognitive impairment at or below the 5th percentile on at least 2 distinct tests of cognition deemed sensitive to executive functioning shall be diagnosed ALSci (amyotrophic lateral sclerosis cognitive impairment).

Population data are lacking. In 2012 Phukan and colleagues performed a prospective study of cognitive function in 160 Irish motor neuron disease patients residing in the community and found that 13.8% had frontotemporal dementia and 34.12% had cognitive impairment without dementia – these patients had a high frequency of language disability and memory dysfunction in comparison to controls; these abnormalities existing in patients with executive dysfunction (137). Cognitive impairment without executive dysfunction was seen in 14%. Almost 50% had no cognitive dysfunction.

A Korean study made inquiries of 318 motor neuron disease patients who were studied prospectively for 4 years from the time of diagnosis (124). About 50% were cognitively or behaviorally impaired and had more executive dysfunction. Patients with cognitive impairment or frontotemporal dementia had a poorer prognosis than motor neuron disease patients with normal cognition.

Little is known about cognitive impairment in Chinese patients with amyotrophic lateral sclerosis. One hundred and six Chinese patients from Beijing with sporadic amyotrophic lateral sclerosis were studied neuropsychologically and were classified into 4 groups: amyotrophic lateral sclerosis with normal cognition (approximately 80%), amyotrophic lateral sclerosis with executive cognitive impairment (approximately 11%), amyotrophic lateral sclerosis with nonexecutive cognitive impairment (approximately 5%), and amyotrophic lateral sclerosis with frontotemporal lobe degeneration (approximately 5%) (40). Executive cognitive impairment was greater in the nondemented amyotrophic lateral sclerosis groups than in healthy controls. ALS-FTD had more severe bulbar dysfunction.

A systematic review and metaanalysis suggested that cognitive impairment was found in about 30% of people with amyotrophic lateral sclerosis (20) – when present adversely affecting prognosis. Forty-four studies including 1287 patients and 1130 controls were studied and revealed that all cognitive domains except visuospatial functions showed impairment without motor impairment bias: fluency, language, social cognition, delayed verbal memory, and executive functions. In this study, diverging effect sizes could be explained with impairment bias. In this study, bulbar disease and mood state did not affect the results. The cognitive profile, on the basis of this analysis, suggests fluency, language, social cognition, executive function, and verbal memory compose the variable cognitive profile in nondemented amyotrophic lateral sclerosis subjects. The findings in relation to social cognition accentuate the relationship between frontal lobe dysfunction, frontotemporal lobe degeneration, and amyotrophic lateral sclerosis. This study emphasizes the importance of correcting for motor impairment when performing neuropsychological tests on patients with amyotrophic lateral sclerosis.

Dementia associated with amyotrophic lateral sclerosis. The pattern of dementia in cases of sporadic amyotrophic lateral sclerosis resembles frontotemporal dementia (165). In many cases, the behavioral alterations dominate over motor signs as a clinical feature. Dementia does not necessarily predate the onset of motor signs, but dementia sometimes precedes motor signs by several months to years (33; 119). Some patients develop dysarthria and dysphagia prior to the onset of dementia, and the incidence of dementia in the subgroup of patients with bulbar onset is particularly high (141).

Changes in personality and comportment are seen. Patients may display emotional disinhibition, lack of concern, apathy, and impulsiveness (170; 119; 76). Emotional lability was reported even in early cases (186), although this may in some instances be referable to a pseudobulbar state due to bilateral corticobulbar lesions rather than psychological dysfunction (114). Pseudobulbar affect with pathological laughter and crying may be seen. This phenomenon is most likely due to pathology involving the cerebro-ponto-cerebellar pathways and is a disinhibition or release effect. This distressing syndrome might respond to a combination of dextromethorphan/quinidine (Nuedexta) and selective serotonin reuptake inhibitors. Depression is common and is typically more severe than in age-matched controls (94). Neuropsychological testing demonstrates severe attention deficits and impaired judgment and insight (114; 134). Language presentations are variable (170; 172), and current consensus guidelines (162) suggest categorization of language impairment according to the Neary classification of progressive non-fluent aphasia and semantic dementia (118).

Progression of cognitive deficits does not correlate with the motor function decline (94). End-stage patients are mute and in a locked-in-like state, and it is difficult to determine their cognitive status; studies using event-related brain potentials demonstrated a variety of responses, from probably normal responses to no responses, consistent with severe cortical dysfunction (98). Visuospatial function and calculation are relatively preserved until late in the course of the illness but may be difficult to test if the patient's frontal dysexecutive function is prohibitive. Strong and colleagues prospectively studied cognitive decline in 8 amyotrophic lateral sclerosis patients (164); after 6 months, patients developed at least mild new deficits in oral and written word fluency, recognition memory for faces but not verbal material, and visual perception. Patients with bulbar amyotrophic lateral sclerosis declined more quickly with severe deficits in working and episodic memory, cognitive flexibility, and visuospatial processing within the 6-month period, paralleling the more rapid physical decline seen in these patients relative to non-bulbar amyotrophic lateral sclerosis. The cognitive changes could not be linked to depression in this study, although the patients developed other neuropsychiatric symptoms, such as agitation, anxiety, delusions, disinhibition, apathy, and irritability during the 6-month interval.

According to guidelines (162), patients with amyotrophic lateral sclerosis and associated dementia should be classified according to the following schema:

(1) Patients meeting either the Neary criteria (118) or Hodge’s criteria (76) for frontotemporal dementia shall be diagnosed ALS-FTD.

(2) Patients meeting criteria for dementia not typical of a frontotemporal lobar degeneration shall be diagnosed with ALS-dementia (ie, ALS-Alzheimer disease, ALS-vascular dementia, ALS-mixed dementia).

Amyotrophic lateral sclerosis is a clinical syndrome with a diverse phenotype involving motor and frontal neocortical neurons and implies molecular mechanisms propagating through the motor-frontal network (181; 69).

In 2017 the Strong diagnostic criteria were revised on the basis of an international workshop held in London, Canada in 2015. On the basis of increasing delineation of the neuropsychology of amyotrophic lateral sclerosis, it has become clear that there exists a panorama of neuropsychological observations with combinations of findings that might be discerned in about 50% of people with amyotrophic lateral sclerosis – this evidence of cognitive involvement deleteriously affecting survival. This new consensus proposes the notion of frontotemporal spectrum disorder of amyotrophic lateral sclerosis (ALS-FTLD) (161). This clinical heterogeneity is consistent with intricate molecular and stochastic processes affecting prion-like propagation, autophagy, vesicle trafficking, and RNA metabolism within the frontal-motor network (179; 58; 128; 129).

Cognitive differences have been emphasized in cognition and behavior between patients with behavioral variant frontotemporal dementia, frontotemporal amyotrophic lateral sclerosis patients, and those with C9orf72 (151).

CSF neurofilament light (NfL) may predict disease progression in amyotrophic lateral sclerosis and, therefore, in amyotrophic lateral sclerosis frontotemporal dementia (48). Motor cortical excitability might predict cognitive phenotypes in amyotrophic lateral sclerosis (06). NfL, a marker of axonal pathology, shows that NfL is increased in amyotrophic lateral sclerosis and most of the other neurodegenerative diseases, elevation in neurofilament light correlated with disease progression, and a poorer prognosis, and studies suggest that long-term kinetics in blood might help in monitoring the gene mutations and the natural history of ALS-FTD (176).

In 3-Tesla MRI scanning, changes in cortical thickness and increased cortical diffusivity may be a biomarker of extra-motor cortical neurodegeneration in ALS-FTD (84). The studies of Hakkinen and colleagues support selective patterns of atrophy in genetic frontotemporal dementia and amyotrophic lateral sclerosis (74).

A metaanalysis showed that frontal, medial, and caudate circuits in behavioral and frontotemporal dementia are also found in amyotrophic lateral sclerosis patients without dementia, the grey matter particularly involving the frontal limbic circuitry, the anterior insula, and anterior cingulate regions (104). Intrusion errors may occur during verbal fluency tasks in patients with amyotrophic lateral sclerosis, suggesting this may be a marker of cognitive change (136).

In patients with amyotrophic lateral sclerosis, 7 Tesla fMRI reveals changes in long-range functional connectivity between the superior sensory and motor cortex, the precentral gyrus, and the bilateral cerebellar lobule VI. This is of interest as cerebellar lobule VI is a transition zone between anterior motor networks and the posterior nonmotor networks in the cerebellum, cerebellar lobule VI having function in complex motor and cognitive processing tasks (15).

Familial amyotrophic lateral sclerosis. Familial cases of amyotrophic lateral sclerosis demonstrate a similar clinical picture when dementia is involved, except that the dementia most often follows the onset of amyotrophic lateral sclerosis symptoms. Dementia has been reported in cases of juvenile familial amyotrophic lateral sclerosis where patients have survived into their third decade of life (126). However, it is unknown whether these patients have mutations in alsin or represent a separate genetic type of juvenile amyotrophic lateral sclerosis (73; 184).

Previous guidelines suggest that patients with familial amyotrophic lateral sclerosis and associated cognitive or behavioral changes or dementia should be classified according to the previously detailed schema, with the proviso that the familial component of the amyotrophic lateral sclerosis has confirmed genetic linkage or clinical evidence of autosomal dominant, autosomal recessive, or X-linked dominant inheritance pattern (162).

Motor neuron disease-like variant of frontotemporal dementia. Along with clinical features of frontotemporal dementia (disinhibition, change in mood or personality, loss of personal and social awareness, mental rigidity, hyperorality, stereotyped behaviors, impulsivity, reduction and stereotypy of speech, echolalia, late mutism, etc.), patients with associated motor neuron disease have earlier age at onset than is seen in other frontotemporal dementias and may exhibit both upper and lower motor neuron signs (77; 119).

According to guidelines (162), the entity FTD-MND-like will remain a neuropathological diagnosis in which the characteristic findings are those of frontotemporal lobar degeneration with associated evidence of motor neuron degeneration insufficient to be classified as amyotrophic lateral sclerosis.

ALS-FTD may represent a disease spectrum. By utilizing mathematical modelling of MRI connectomic information, disordered frontotemporal and parietal networks, and focal structural damage within the sensorimotor-basal ganglia may be seen (36).

The cognitive profiles of amyotrophic lateral sclerosis patients differ in functional connectivity measured in the resting state, depending on the clinical presentation of predominantly motor involvement versus cognitive shifting (171).

Prognosis and complications

All cases will end fatally, usually due to the natural course of the motor neuron disorder. Infectious complications are always likely. Supportive care is in order but differs according to the wishes of the patient and family. Recognizing the presence of a frontotemporal cognitive or behavioral syndrome in amyotrophic lateral sclerosis is prognostically important as it is a predictor of shorter survival time (125; 146; 50). Patients presenting with language variants rather than behavioral features may be more likely to have bulbar-onset, which portends a poorer prognosis (37).

Retrospective analysis of 625 patients revealed that a poor prognosis was associated with amyotrophic lateral sclerosis and frontotemporal dementia versus amyotrophic lateral sclerosis, shorter time to diagnosis and higher age at symptom onset with rapid decline for a person with a high amyotrophic lateral sclerosis functional rating scale sum score, and if the individual has chronic obstructive pulmonary disease (144).

The Edinburgh Cognitive and Behavioral ALS Screen was found to be useful to determine if patients with amyotrophic lateral sclerosis and behavioral variant frontotemporal dementia were portending a poorer prognosis (19; 185).

Studies of the clinical heterogeneity of ALS-FTD indicate those with amyotrophic lateral sclerosis have predominant frontoparietal involvement, whereas those with ALS-FTD behavioral variant frontotemporal dementia have more frontal, cingulate, amygdala, insula, thalamic, hippocampal, and temporal lobe involvement, suggesting there are distinct atrophic profiles across the ALS-FTD spectrum with more involvement of the motor cortex and insula in those with ALS-FTD. Cortical involvement might explain the origin of some of the behavioral difficulties (07). Pathological expansion in the C9orf72 gene is associated with accelerated decline of respiratory function and reduced survival in amyotrophic lateral sclerosis (112).

Clinical vignette

A 74-year-old, right-handed woman with 7 years of education had initial symptoms of dysphonia, dysphagia, increasing respiratory difficulties, and fatigue. On neuropsychological evaluation 5 months into her illness, her affect was apathetic and restricted. She had little insight into her cognitive impairments. Speech was hypophonic, dysarthric, and non-fluent. Her responses were bradyphrenic and bradykinetic. Her assessment revealed mild to moderate impairment in most cognitive areas including:

The patient died 1 year into the course of her illness. Autopsy of the brain and spinal cord confirmed the clinical diagnosis of amyotrophic lateral sclerosis. There was anterior horn motor neuron loss, accompanied by astrocytic proliferation and axonal swelling. Ubiquitin stains revealed rare dense cytoplasmic rounded granular inclusions. Extensive spongiosis and widespread tau immunoreactive glial cells were observed in the frontal lobes. Further neuronal loss appeared in the substantia nigra, periaqueductal grey, mamillary bodies, midline thalamic nuclei, and the hippocampal subiculum. Only rare neurofibrillary tangles and neuritic plaques were located. [Case courtesy of Strong and colleagues (163).]

Clinical diagnosis and management remains challenging (109).

Biological basis

Etiology and pathogenesis

The cause of cognitive and behavioral changes as well as dementia associated with amyotrophic lateral sclerosis is variable and related to genetic predispositions and the identified neuropathological substrates to be discussed further in the pathogenesis and pathophysiology section of this article. Although amyotrophic lateral sclerosis may co-exist with well-defined dementing illnesses such as Alzheimer disease, it does not in most instances appear to have linkage. At present, the majority of the cognitive and behavioral changes demonstrated by patients with amyotrophic lateral sclerosis are considered to be referable to the accumulation of TDP-43 neuropathology (62) or the abnormal accumulation of tau in a smaller but well-defined subset of patients (180).

Genetics. Genetic studies of patients with amyotrophic lateral sclerosis and dementia may provide an insight into the pathogenesis and pathophysiology of both the cognitive deficits and motor neuron degeneration. Amyotrophic lateral sclerosis is genetically heterogeneous, and several loci for autosomal dominant and recessive amyotrophic lateral sclerosis have been identified (105). Previously reported as the most common type of familial amyotrophic lateral sclerosis, mutations in the copper-zinc superoxide dismutase (SOD1) gene on chromosome 21 may be associated with a cognitive and behavioral syndrome characterized by apathy, anxiety, inattention, reduced verbal fluency, and hypersexuality (150; 108; 16).

The linkage to a locus on chromosome 17q21, which was later shown to be associated with frontotemporal dementia, was discovered in a pedigree whose phenotype was classified as disinhibition-dementia-parkinsonism-amyotrophy complex (180). Mutations in the microtubule associated protein tau gene have been later discovered as a cause of this type of dementia-amyotrophic lateral sclerosis complex (83; 140). Single cases of dementia in families with amyotrophic lateral sclerosis and mutations in ANG/VEGF on chromosome 14q11 (162) and VAPB on chromosome 20q13.33 have also been reported and require further study (121). A mutation in CHMP2B was initially described in a Danish cohort with autosomal dominant frontotemporal dementia and has subsequently been identified in patients with amyotrophic lateral sclerosis and amyotrophic lateral sclerosis with frontotemporal dementia; the mutation may be associated with as much as 10% of lower motor neuron-predominant amyotrophic lateral sclerosis (38). Mutations in UBQLN2, which encodes the ubiquitin-like protein ubiquilin 2, have also been reported to cause dominantly inherited, chromosome-X-linked amyotrophic lateral sclerosis and amyotrophic lateral sclerosis/dementia. Novel ubiquilin 2 pathology in the spinal cords of amyotrophic lateral sclerosis cases and in the brains of amyotrophic lateral sclerosis dementia cases with or without UBQLN2 mutations has also been reported in association with this genetic discovery. Ubiquilin 2 is known to regulate the degradation of ubiquitinated proteins and preliminary work suggests that mutations in UBQLN2 lead to an impairment of protein degradation (43).

Additional loci on chromosome 9q21-q22 (79) and 9p13.2-21.3 (116) were initially identified in families with overlapping motor neuron disease and frontotemporal dementia, and the chromosome 9p linkage was confirmed by a genome-wide association replication study of patients with frontotemporal lobar degeneration and frontotemporal lobar degeneration and amyotrophic lateral sclerosis in 2 British cohorts (149). The discovery of the critical gene change as an expanded sequence of hexanucleotide repeats in C9orf72 was subsequently reported independently by 2 international research groups (42; 143). The genetic change affects a region outside of the normal protein coding portion of the gene and affects the non-coding RNA. Unaffected individuals may carry up to 30 DNA repeats in the gene, whereas affected patients with motor neuron disease or frontotemporal dementia may carry hundreds of repeats. Scientists have discovered that this genetic mutation may account for up to 12% of familial frontotemporal dementia, and as much as 22% of familial motor neuron disease, rendering the C9orf72 gene the most common genetic cause of both frontotemporal dementia and motor neuron disease identified to date. Clinically, motor neuron disease progression may be slower (154) and psychiatric symptoms may be more common in this genetic cohort of families presenting with frontotemporal dementia and amyotrophic lateral sclerosis (157). The role of the protein C9orf72 within neurons of the brain and spinal cord is not currently known and the mechanism of pathogenesis remains to be determined. Preliminary work with induced pluripotent stem cells suggests that compromised autophagy function may be involved (10).

Remarkably, although mutations in the TARDBP gene on chromosome 1p36.2 have been identified in patients with amyotrophic lateral sclerosis and the characteristically associated neuropathological inclusions representing the abnormal accumulation of TDP-43 have been identified in both amyotrophic lateral sclerosis and frontotemporal lobar degeneration, to date very few cognitive or behavioral phenotypes have emerged in patients with this genetic mutation (35). Conversely, although mutations in the PRGN gene on chromosome 17q.21.23 have been observed to account for as many as 5% of familial cases of frontotemporal dementia resulting in abnormal TDP-43 neuropathological inclusions, only a few pedigrees reported to date have overlapping motor neuron disease phenotypes (153; 159). Mutations in the FUS gene, an RNA-processing gene linked to chromosome 16q12 and functionally related to TDP-43, have been reported in approximately 3% of familial amyotrophic lateral sclerosis patients, with at least 2 cases presenting with frontotemporal dementia features (24; 29). Further highlighting the evolving, complicated pathological and genetic relationship between frontotemporal lobar degeneration and amyotrophic lateral sclerosis, FUS pathological inclusions have been reported in frontotemporal lobar degeneration to account for many of cases that were previously considered ubiquitin positive, TDP-43 and tau negative, but such cases have not been associated with FUS gene mutations (158).

There is a clinical and genetic spectrum of amyotrophic lateral sclerosis-frontotemporal dementia (14; 44; 31), mutations associated with amyotrophic lateral sclerosis-frontotemporal dementia (TBK1, C9orf72, TARDBP, FUS, CCNF, VCP, UBQLN2, CHCHDIO, SQSTM2), mutations with frontotemporal dementia (CHM2B, PGRN, MAPT), and those with amyotrophic lateral sclerosis only (SOD1, OPTN, PFN1) (17; 39; 86; 63). This array of genetic mutations implicates molecular mechanism involving protein homeostasis, the functions of RNA binding proteins, cytoskeletal kinetics, and the autophagy/lysomal machinery (75; 63).

The known genes causing frontotemporal dementia/motor neurone disease are listed in Table 1 along with their functions: autophagy, vesicular trafficking, RNA metabolism, mitochondrial function, apoptosis, ubiquitination, mitophagy, innate immunity, transcription, chaperone protein, microtubule function, cytoskeletal integrity, axonal transport, and stress granule formation. The frequencies of these gene mutations in populations of ALS-FTD patients is unknown, but C9orf72, TD4-43, and FUS are the most common, and other gene mutations are not (05; 70).

The C9orf72 hexanucleotide repeat expansion in Exon 1 is the most common cause of ALS-FTD and is not detected by Sanger sequencing. This DNA expansion is first searched for; if not detected, next generation sequencing on ALS-FTD panel is performed.

Table 1. Genes causing ALS + FTD and their functions

The C9orf72 repeat expansion is the most frequent genetic cause of familial amyotrophic lateral sclerosis-frontotemporal dementia and a number of families have been reported in which mutations in other genes coexist in the same family, resulting in the concept of the oligogenic hypothesis where mutated alleles of causative, at-risk, and modifier genes give rise to disease (99). Giannoccaro and colleagues identified the p.V47A variant in the TYROBP gene in amyotrophic lateral sclerosis-frontotemporal dementia as well as other genetic variants: in their study if patients carried a double mutation there was an earlier age of onset, more likely to be a family history, and parkinsonism was more prevalent (65). In another study these authors described a pathogenic variant in the ITM2B gene associated with amyotrophic lateral sclerosis-frontotemporal dementia plus movement disorder, psychiatric problems, cognitive dysfunction, deafness, and optic atrophy (66). An intermediate length expansion has been identified in the ataxin 2 gene (ATXN2 polyQ) and an optineurin point mutation (OPTN p.Met 468 Arg) in 2 different families with amyotrophic lateral sclerosis-frontotemporal dementia and the C9orf72 hexanucleotide repeat expansion, providing further evidence that oligogenic inheritance may influence the phenotypic expression of C9orf72 (54).

A large Australian family with frontotemporal dementia-motor neurone disease was shown to have the cytochrome lysine 63 deubiquitinase (CYLD) gene mutation as the cause (47); there was increased glial cytochrome CYLD immunoreactivity. Mice expressing this mutation also increased cytoplasmic localization of TDP-43 and shortened axons. In vitro studies reveal inhibition of NF-kB cell single transition molecule resulting in reduced phagosome fusion to lysosomes, a key process in autophagy. That is, this mutation leads to a loss of autophagosome function resulting in protein accumulation and cellular dysfunction. CYLD interacts with 3 other genes and proteins important in autophagosome function TBK1, optineurin, and sequestosome-1. This contributes to impaired autophagy and neurodegeneration. CYLD is a protease that removes ubiquitin from proteins and, therefore, regulates protein activity and functions as a tumor suppressor gene. CYLD mutation has also been identified in Chinese populations (71). A rare mutation in the cytoplasmic dynein 1 heavy chain gene has been associated with ALS-FTD (111) and is an endoplasmic reticulum ATPase—an ATPase associated with diverse cellular activities, with functions in ubiquitination and protein degradation, autophagy, lysosome clearance, and mitochondrial quality control. Mutations in this gene lead to amyotrophic lateral sclerosis with TDP-43 mislocalization from the nucleus into the cytoplasm (152). Patients with SOD1 mutations have been shown to have increased metabolism in the motor cortex compared to sporadic amyotrophic lateral sclerosis, suggesting that there may be different mechanisms (32). Studies suggest that 30% of patients with ALS-FTD have genetic pathogenic mutations or variants (147).

Machine learning studies suggest polygenic risks factors contribute to cognitive function in amyotrophic lateral sclerosis and might be useful in frontotemporal dementia, in that cognitive decline and cortical thinning from motor neuronal loss might be related to polygenic risk score (138). The genetics of ALS-FTD show significant overlap (05; 142).

Neurophysiology. Physiological mechanisms operate in the pathophysiology of ALS-FTD. Studies of C9orf72 repeat expansion mutations in ALS-FTD indicate hyperexcitability, both increase in excitatory signals and reduced inhibition, synaptic vesicle disruption, impaired plasticity, morphological deficits, glutamate excitotoxicity, and reduced synaptic transition with general increased excitability of surrounding neuronal structures (132). Transmagnetic stimulation is the best technique to explore cortical hyperexcitability in ALS-FTD (61; 160).

Neuropathology. Aside from cases revealing pathological markers for Alzheimer disease, Creutzfeldt-Jakob disease, or Pick disease, patients with dementia associated with amyotrophic lateral sclerosis show gross frontal and temporal lobe atrophy. On closer inspection, neuronal cell loss in frontal and temporal cortex is more extensive than the usual depletion of Betz cells observed in amyotrophic lateral sclerosis without dementia (78; 55; 106; 135). In autopsied cases of familial amyotrophic lateral sclerosis, a pattern of predominantly frontotemporal neuronal degeneration, status spongiosis, and gliosis similar to dementia associated with sporadic amyotrophic lateral sclerosis is observed (81). Familial cases also show ubiquitin-positive inclusions in the temporal and frontal lobes (117; 90).

Affected areas show dropout of interneurons that immunolabel with calbindin D-28k but not those that label with parvalbumin (56); calbindin D-28k interneurons predominate in upper cortical layers, whereas parvalbumin interneurons are featured in deeper cortical layers. Because these studies did not compare demented to nondemented amyotrophic lateral sclerosis patients, it is not clear how much of a role the interneurons play in the dementing aspect of the syndrome. Their position in the superficial layers II and III of cortex may reflect damage to neuronal dendritic processes (135). Furthermore, cortical lesions are more prominent at the crests of gyri than in sulci (88). When compared with nondemented patients, the loss of the Betz cells in the primary motor cortex appears to be more severe in amyotrophic lateral sclerosis patients who are demented (173).

In 2006, TDP-43 (TAR DNA Binding Protein 43), a nuclear protein encoded by the TARDBP gene on chromosome 1, was identified as a major disease protein in amyotrophic lateral sclerosis as well as frontotemporal lobar degeneration with ubiquitinated inclusions and frontotemporal dementia with motor neuron disease (120), highlighting the potential for a common pathogenesis for these disorders, which may have considerable clinical overlap. In the majority of cases the TDP-43 protein co-localizes to inclusions previously identified with ubiquitin immunoreactivity (62). Evidence suggests that TDP-43 plays an essential role in dendritic branching, such that diminished neuronal connectivity may precede neuronal loss in TDP-43-associated amyotrophic lateral sclerosis and frontotemporal dementia (103).

The emotional disturbances seen in the dementia syndrome associated with amyotrophic lateral sclerosis focus interest on pathological changes in the limbic system. The basolateral nucleus of the amygdala; subiculum; nucleus accumbens; dorsomedial cortex of the anterior temporal horns; anterior cingulate; and orbital and insular gyri show degenerative changes in demented amyotrophic lateral sclerosis patients (88).

Degeneration of the substantia nigra is common in frontal lobe dementia cases and may be a marker for dementing cases of amyotrophic lateral sclerosis (78; 67; 106). The nucleus basalis of Meynert, locus ceruleus, and dorsal raphe appear unaffected (78; 106). Neurochemical investigations have disclosed no significant alterations in the major brain neurotransmitters, except for reductions in dopamine in the corpus striatum and substantia nigra in a case of amyotrophy-dementia with parkinsonism (67). Cholinergic activity, especially in the nucleus basalis, has been normal in several studies (78; 67).

MR spectroscopy has shown reduction of the N-acetylaspartate and creatine ratio in nondominant precentral motor gyrus in patients with amyotrophic lateral sclerosis, but there has been no clear correlation between this change and the degree of cognitive decline observed clinically (164).

More recent studies have emphasized the importance of clinical and pathological correlations of the dementias, including amyotrophic lateral sclerosis-frontotemporal dementia and the significance of proteomics, protein misfolding, and aggregation both in histopathological definition and molecular pathophysiology (49; 174). In patients with the linguistic presentation of nonfluent primary progressive aphasia and the semantic variant, 12% of these 139 primary progressive aphasia patients had amyotrophic lateral sclerosis – emphasizing that evidence of amyotrophic lateral sclerosis should be screened for in primary progressive aphasia; all of the primary progressive aphasia-amyotrophic lateral sclerosis patients had TDP pathology and half had mutations in recognized frontotemporal dementia/amyotrophic lateral sclerosis genes (168).

A fundamental problem remains that the majority of amyotrophic lateral sclerosis-frontotemporal dementia patients are sporadic and nongenetic, suggesting a role of stochastic processes involving DNA, RNA, and proteins in their pathogenesis (128; 129).

Chemical biology. Amyotrophic lateral sclerosis-frontotemporal dementia in many situations is unified by the loss of TDP-43 from the nucleus of neurons in the brain and spinal cord and its translocation into the cytoplasm. TDP-43 functions to repress inclusion of cryptic exons during RNA splicing. When TDP-43 is depleted, this function is lost. Cryptic splicing targets include STM2, UNC13A, and others (08). UNC13A is important in ALS-FTD as a risk factor. Gene variants make this gene more susceptible to cryptic exon inclusion. Cryptic splicing events suggest novel therapeutic targets for ALS-FTD. The study of ALS-FTD emphasized the importance of autophagy and RNA homeostasis (80). Autophagy is the multi-target removal of malfunctioning cellular components with many layers of regulation including protein degradation and RNA catabolism. ALS-FTD proteins are of importance in the clearance of stress granules. Genes such as TDP-43, sequestosome-1, optineurin, TBK-1, C9orf72, SMN, VCP, VAP, SPG11, CHMP2B, ALS2, Sigma-1, CCNF, GRN, and ubiquitin 2 are involved in the regulation of these mechanisms. DNA damage response and effective DNA protein repair is also considered one of the important mechanisms in frontotemporal dementia amyotrophic lateral sclerosis with the proteins involved in DNA damage and repair being affected in ALS-FTD such as TAR TDP-43, FUS, C9orf72, SOD1, SETX, VCP, CCNF, and NEK1 (96). The 4C2G expansion found in C9orf72 mutations gives rise to dipeptide repeat polypeptides (DPRs), which are neurotoxic. The dipeptide repeat polypeptides interfere with tethering of the endoplasmic reticulum to the mitochondria. This tethering is mediated by the VAPB-PTPIP51 tethering proteins, which the dipeptide repeat polypeptides disrupt. This disruption of the tethering gives rise to an increase in the glycogen synthase kinase beta gene, a negative regulator of VAPB-PTPIP51, which further impairs the endoplasmic reticulum and mitochondria tethering leading to neurodegeneration (68). Cyclophilin A known as a peptidylprolyl isomerase A (PPIA) is a foldase and molecular chaperone. PPIA knockout mice have neurodegeneration like frontotemporal dementia with TDP motor neurone changes. Reduction in PPIA leads to an increase in the GTP binding nucleoprotein RAN, and this causes mislocation of TDP-43. Therefore, reductions in PPIA affect TDP-43 autoregulation, decrease TDP-43 synaptic function, and impair synaptic plasticity; PPIA is reduced in patients with ALS-FTD (131). Inflammation is thought important in ALS-FTD as experiments with induced pluripotent stem cells (iPS) reveal that cytokines, chemokines, growth factors, and extracellular matrix molecules interfere with TDP-43 and other neuropathological processes in ALS-FTD (102). ALS-FTD both have evidence of chronic inflammation mediated by microglia and astrocytes. Immune signaling kinases have been shown to be abnormal in ALS-FTD including TBK-1, RIPK-1, 3, RAK1, and EPHA4; this led to clinical trials for immune kinase inhibitors in ALS-FTD (60). Protein synthesis modulation is a possible therapeutic intervention in ALS-FTD; therapies that restore protein imbalance, but not affect translation activity of cells is a desired goal. The mechanisms of protein folding and unfolding, accumulation, and mislocalization are shared between amyotrophic lateral sclerosis and frontotemporal dementia, with a complex interaction between the endoplasmic reticulum and elongation factors, which increase expression of chaperone proteins which, in turn, repress translation, to restore protein homeostasis, that is proteostasis. In ALS-FTD, and all neurodegenerative disorders, there is unbalanced protein overload; therefore, modulation of protein translation without affecting the normal translational activity of cells is a potential therapeutic approach. This therapy approach includes regulation of activation initiation of elongation factors, inhibition of unfolded protein response activation, and induced chaperone expression (see Table 2) (34).

Table 2. Potential molecular mechanisms ALS-FTD

MicroRNAs (miRNAs) are single-stranded noncoding RNA less than 22 nucleotides in length. They recognize sequences in 3 prime untranslated regions of target mRNA and induce mRNA degradation or inhibit translation. miRNA dysregulation in neurodegenerative disease is known in Alzheimer disease, amyotrophic lateral sclerosis, Parkinson disease, Huntington disease, frontotemporal dementia, macular degeneration, multiple sclerosis, and ALS-FTD. Expression of miRNAs depends on the diagnosis and the state of the disease. The most robust and microRNA markers in serum and CSF have been increased miR-233-3P and reductions in miR-15A-5P. These markers help to distinguish frontotemporal dementia from normal subjects. Cortical tissue mi132-3P in the frontal and temporal regions is helpful in distinguishing frontotemporal dementia from normal. Strong miRNA in behavioral variant frontotemporal dementia is found in blood serum and CSF including mi233-3P, mi22-3P, miR15A-5P, miR124, and CSF. These micro-MRI are thought important in neurodegeneration, and changes in TDP-43 and other mechanisms of the molecular cascade in amyotrophic lateral sclerosis/frontotemporal dementia interact through miRNAs to affect translation (107). Experimental models of C9orf72 mutations in Drosophila revealed reduced transcription, decreased function, and also gain of function with repetitive sense and antisense RNA, with the reduction of DPR polypeptides, which cause toxicity, with the gain-of-function affecting RNA molecules, and the DPR toxicity compromising endoplasmic reticulum stability leading to neurodegeneration (155). In ALS-FTD, TDP-43 mislocalizes to the cytoplasm from the nucleus. Mutant FUS reduces protein transport in the cell and fragments the Golgi apparatus causing endoplasmic reticulum stress with reduced Golgi trafficking contributing to neurotoxicity, along with the FUS mutant protein impeding DNA repair. Disulfide isomerase (PDI) is protective against FUS in vitro models, suggesting a therapeutic target (130). Phenols such as flavanone and non-flavanone molecules have been successfully used in experimental models of ALS-FTD and might be a future therapy (123). Heterogeneous nuclear ribonucleoproteins (hnRNPs) are important in ALS-FTD; hnRNP are family of RNA-binding proteins. In frontotemporal dementia-amyotrophic lateral sclerosis, pre-mRNA splicing regulation, cryptic exon expression, stress expansion assembly and DNA damage response, functions mediated by hnRNPs, are disrupted, especially hnpA1 (13). TDP43 is a member of this hnRNP family and is mislocated into the cytoplasm and has an effect on ubiquitination of protein aggregates. TDP also has immune and inflammatory functions. TDP-43 can increase inflammation, which can increase neurodegeneration. TDP-associated genes are linked with immunity and inflammation; TDP proteins initiate immune inflammation pathways; TDP is involved in immune pathway; and TDP is implicated in acute and chronic inflammatory processes (13; 28). Amyotrophic lateral sclerosis patients have increased intrathecal T cell activation (148). Mislocalization of FUS and TDP-43 into cytoplasm are prominent pathophysiological mechanisms, and a rapid in vitro test to quantify mislocalization has been developed (127). Localization of TDP from the nucleus to the cytoplasm is associated with cell stress signaling abnormalities and stress granule dynamics; stress granule assembly is also affected, causing increased cellular vulnerability to death. The mRNA G3BP1 molecule is a RAS-GAPSH3 domain binding protein I and is critical in stress granule assembly. TDP-43 stabilizes G3BP1 through mRNA transcription. When TDP-43 is reduced, G3PB1 concentration is low, contributing to neurodegeneration; in ALS-FTD, G3PB1 transcripts are low and in experimental models of TDP-43 mutations. Therefore, mutations that cause disruption of TDP-43 can destabilize G3BP1, leading to impaired stress granule formation and neurodegeneration (156). G3BP1 MRNA stability is affected by TDP-43 depletion, leading to loss of function and disease. The 4G2C noncoding mutation, C9orf72, gives rise to dipeptide repeat polymers of protein and arginine. These derail protein folding by sequestering molecular chaperones leading to neurodegeneration and inhibit folding catalyzed by PPIA (12). The TDP-43 “knock in” mouse shows parvalbumin interneuron loss and impaired neurogenesis, which might be relevant to the pathophysiology of ALS-FTD (101). Glial cell dysfunction has been identified in C9orf72 mutations in microglia and astrocytes, with loss and gain of function. The C9orf72 mutation is associated with gliosis, TDP-43 aggregation, and toxic gain of function mRNA, leading to dipeptide release and non-ATG RAN translation (64). Four micro mRNAs in the plasma represent a molecular signature for ALS-FTD, particularly mi 349-5p, miR 345-5p, mi 200C-3p, and mi 10a-3p (95). TDP-43 oligomers and their interactions may be important in ALS-FTD (113). Measurement of serum protein levels suggests impaired calcium and immune dysregulation in frontotemporal dementia-amyotrophic lateral sclerosis (89). Findings of differential gene expression data from a number of different studies in the human central nervous system show there are shared gene expressions between ALS-FTD, Lewy Body disease, and Alzheimer disease, with these gene expressions driving the pathophysiology of these diverse proteinopathies. The principal genes identified in 2600 studies are innate immunity genes, cytoskeletal genes, and RNA processing genes with downregulation of mitochondrial electron transfer genes. Also, genes involving neural inflammation, phagocytosis, are increased along with mitochondrial oxidative phosphorylation, lysosomal function, and ubiquitin-proteasome pathways. This suggests that in the pathophysiology of neurodegenerative disorders there are shared pathophysiological pathways (122). Interestingly, reduced telomere length has not been related to amyotrophic lateral sclerosis and might not be relevant to ALS-FTD, but further studies are required (59). Pathogenic Huntington repeat expansions have been identified in frontotemporal dementia and amyotrophic lateral sclerosis; their significance is unclear, but they might reflect spontaneous mutations (45). Studies of the UNC13A polymorphism suggests vesicle maturation during exocytosis, and neurotransmitter release might be affected in ALS-FTD (167). Immune activation in C9orf72 expansions is associated with increased cytokine activation, stimulation of innate inflammatory mechanisms, and intracellular receptors that elicit inflammation response to cellular stress, indicating the importance that immunology might have in ALS-FTD (132). Stathmin 2 (STMN2) is a microtubular associated protein that is important in axonal development and repair. STMN2 promotes microtubular stability necessary for axonal length and regeneration. STMN2 is regulated by TDP-43. Evidence reveals that reduced nuclear TDP-43 function results in impaired splicing of the gene encoding STMN2, leading to reduced production of a truncated protein resulting in axonal degeneration. Reduction in TDP-43 in the nucleus and its translocation to the cytoplasm caused impaired splicing of STMN2 and truncated STMN2 proteins causing axonal degeneration, a finding linking changes in TDP to axonal pathology (21).

There has been increasing interest in the liquid-liquid phase separation (LLPS) of protein and nucleic acid scaffolds in the biogenesis of membraneless organelles. These organelles marshal biochemical reactions and allow the expression of genotype-->phenotype. Liquid-liquid phase separation are important physiologically and pathologically. Liquid-liquid phase separation of important protein and nucleic acids can result in disease. Liquid-liquid phase separation can condense RNA-binding proteins like TDP-43, FUS, hnRNP A1, tau, HTT-PolyQ C9orf72, and others, which have prion-like domains that form self-replicating fibrils that result in disease, promising novel therapeutic interventions involving phase boundary interactions, material properties of condensed phases such as viscoelasticity reversibility of exchange within the dynamics of molecular movement, and fiber formation (51).

Epidemiology

Estimates of the prevalence of dementia in amyotrophic lateral sclerosis range from 15% to 40% (100; 145), with a wider range of estimates, from 10% to 75%, for cognitive or behavioral impairments not meeting criteria for the diagnosis of dementia (139; 57). Historical estimates placed the frequency of dementia at 1% to 2% in sporadic amyotrophic lateral sclerosis (114; 134) and higher (15%) for familial amyotrophic lateral sclerosis (81).

Investigators are still debating whether dementia associated with amyotrophic lateral sclerosis is the equivalent of the motor neuron disease variant of frontotemporal dementia (118; 119). However, there is emerging evidence for the overlap of these 2 entities (23) based on convergence of clinical, neuropathological, and genetic evidence suggesting a continuum model of the diseases (62). Most investigators agree that the order of appearance (neurologic changes before cognitive impairment or vice versa) do not preclude classification of the dementia associated with amyotrophic lateral sclerosis as a frontotemporal dementia syndrome.

Prevention

Until the pathogenesis and pathophysiology of these disorders are fully ascertained, there can be no definitive comments about prevention.

Once the molecular mechanisms have been elucidated it may be possible to use new technologies such as CRISPR and preimplantation genetic diagnoses to eradicate amyotrophic lateral sclerosis-frontotemporal dementia and other neurodegenerative conditions (178).

Now that many of the genetic and molecular mechanisms are understood, prevention of ALS-FTD is possible, especially in the situation where genetic mutations are recognized, but more difficult in the more common sporadic cases of frontotemporal dementia-amyotrophic lateral sclerosis (22). Investigations with induced pluripotent stem cells derived motor neurons carrying the hexanucleotide repeat mutation induced by CRISPR show correction of the amyotrophic lateral sclerosis phenotype, suggesting a possible role of gene editing in prevention ALS-FTD (01).

Differential diagnosis

The differential diagnosis may include a dementia specifically resulting from neuropathological changes associated with amyotrophic lateral sclerosis or the co-occurrence of amyotrophic lateral sclerosis and another dementing illness. It may be difficult to distinguish the etiology of the dementia until one can make a pathologic diagnosis from brain biopsy or autopsy. Postmortem studies on patients with dementia associated with amyotrophic lateral sclerosis commonly reveal concomitant frontotemporal lobar degeneration, and less commonly Alzheimer disease, Creutzfeldt-Jakob disease, or Pick disease. Dementia has been reported rarely in spinal muscular atrophy (53).

Similar clinical dementia syndromes can be found with non-amyotrophic lateral sclerosis motor neuron disorders. Motor neuron involvement has been recognized in postencephalitic states and in forms of Creutzfeldt-Jakob disease. An association of these symptoms with pantothenate kinase-associated neurodegeneration has also been recognized (25; 133).

Diagnostic workup

In obvious cases where there is ample evidence of both motor neuron involvement and cognitive or behavioral changes, both EMG and neurobehavioral screening evaluations will serve to confirm the joint presence of these deficits. Neurobehavioral screening should be performed on all patients diagnosed with amyotrophic lateral sclerosis and should always include tests of executive functioning. Current consensus recommendations (162) suggest that a screening assessment be completed that includes a measure of verbal fluency and a neurobehavioral assessment measure. The MMSE, a standard cognitive screening instrument, is not recommended. Diagnoses of ALSbi and ALS-FTD are permitted without a formal neuropsychological testing battery; however, more detailed neuropsychological testing is recommended in all cases with an abnormal brief screening exam and required when ALSci or ALS-dementia are diagnostic considerations. Premorbid level of function and deficits referable to motor symptomatology should be considered when interpreting neuropsychological test scores. In cases of frontotemporal dementia without evidence of motor neuron dysfunction (fasciculations, altered deep tendon reflexes, etc.), it may be useful to perform EMG testing for evidence of subclinical disease.

Although neuroimaging in both structural and functional modalities is promising with regard to reflection of early cognitive and behavioral abnormalities in patients with amyotrophic lateral sclerosis, they are currently deemed not yet suitable for clinical use as predictive tools (162). Structural neuroimaging studies on patients with amyotrophic lateral sclerosis reveal frontotemporal atrophy in patients with and without dementia (87; 93; 02). Unfortunately, in clinical practice, the rapidity of the progression of the illness may preclude identification of specific structural neuroimaging changes. In a more recent longitudinal study of 4 patients with ALS-FTD, there was a 1% annual rate of atrophy in the premotor, motor, and anterior parietal cortices compared to a 0.25% annual atrophy rate in the same regions in age-matched controls, suggesting that normalized methods of regional comparison may prove useful in detecting subtle atrophy previously attributed to individual variability (11).

In the realm of functional neuroimaging, PET scans from demented patients with amyotrophic lateral sclerosis identify a significant worsening of regional cerebral blood flow in bilateral frontal cortices, right temporal cortex, and bilateral cerebellar hemispheres, exceeding the reductions detected in nondemented patients with amyotrophic lateral sclerosis (170; 18). SPECT studies tend to correlate well with this pattern of dementia. Reduced isotope uptake in the frontal areas is consistently seen in those amyotrophic lateral sclerosis patients with dementia, whereas nondemented patients had only reduced uptake in the motor cortices (02; 169). Similar patterns of reduced uptake are demonstrated in patients with ALS-FTD and FTD compared to controls, involving the frontal, temporal, cingulate, and insular cortices, and supporting the continuum model of the 2 illnesses (72).

In the appropriate clinical or research setting, genetic counseling and commercially available genetic testing may be offered if there is a compelling family history of either amyotrophic lateral sclerosis or frontotemporal dementias, using World Federation of Neurology guidelines (178).

Management

The following have all been approved by the FDA for the treatment of amyotrophic lateral sclerosis: riluzole, a glutamate antagonist; Edaravone, an antioxidant; a combination of phenylbutyrate/taurursodiol (Relyvrio) thought to prevent neuronal death by stabilizing mitochondria; and the endoplasmic reticulum. Nuedexta is an approved treatment for pseudobulbar affect.

Other agents approved or under investigation for treating amyotrophic lateral sclerosis include glutamate antagonists, neurotrophic factors, protease inhibitors, and antioxidants.

In patients with frontotemporal dementia, selective serotonergic reuptake inhibitors such as sertraline HCl and paroxetine have shown efficacy for controlling agitation, mood disorders, and obsessive-compulsive behaviors (166; 09). Open-label trials of the NMDA-receptor antagonist memantine have demonstrated variable results in treating the behavioral symptoms of frontotemporal dementia (26; 46). Randomized, controlled clinical trials have failed to demonstrate a benefit of memantine in frontotemporal dementia (175; 27). Similar treatment strategies might be useful for dementia associated with amyotrophic lateral sclerosis, but the heterogeneous etiologies of this syndrome should dictate pharmacological trials. Therapies targeting the accumulation of TDP-43 are in the preclinical phase along with the development of appropriate animal models (177).

A patient with concomitant Alzheimer disease and amyotrophic lateral sclerosis might respond better to a cholinesterase inhibitor such as donepezil than to a selective serotonergic reuptake inhibitor, although cholinesterase inhibition has not been successful in treating patients with frontotemporal dementia (92). The cholinesterase inhibitor donepezil increased frontal behaviors in a small group of patients (110) whereas rivastigmine showed some benefit (115). However, larger more rigorous studies are required to investigate the role of neurotransmitter modulation in frontotemporal dementia (82). In general, treatment for both the motor deterioration and the dementia associated with amyotrophic lateral sclerosis is symptomatic.

Experimental models suggest that antisense oligonucleotides silencing FUS expression may be a therapeutic approach in amyotrophic lateral sclerosis and also ALS-FTD (97). Medications targeting proteostasis might be useful in ALS-FTD, and one example is trehalose, which has been shown to inhibit protein misfolding (52). Monoaminergic dysfunction might suggest treatment possibilities for frontotemporal dementia-amyotrophic lateral sclerosis, but not kynurenine genetic pathways (85).