Neuropathologic diagnostic and nosologic criteria for frontotemporal lobar degeneration: consensus of the Consortium for Frontotemporal Lobar Degeneration
The aim of this study was to improve the neuropathologic recognition and provide criteria for the pathological diagnosis in the neurodegenerative diseases grouped as frontotemporal lobar degeneration (FTLD); revised criteria are proposed. Recent advances in molecular genetics, biochemistry, and neuropathology of FTLD prompted the Midwest Consortium for Frontotemporal Lobar Degeneration and experts at other centers to review and revise the existing neuropathologic diagnostic criteria for FTLD. The proposed criteria for FTLD are based on existing criteria, which include the tauopathies [FTLD with Pick bodies, corticobasal degeneration, progressive supranuclear palsy, sporadic multiple system tauopathy with dementia, argyrophilic grain disease, neurofibrillary tangle dementia, and FTD with microtubule-associated tau (MAPT) gene mutation, also called FTD with parkinsonism linked to chromosome 17 (FTDP-17)]. The proposed criteria take into account new disease entities and include the novel molecular pathology, TDP-43 proteinopathy, now recognized to be the most frequent histological finding in FTLD. TDP-43 is a major component of the pathologic inclusions of most sporadic and familial cases of FTLD with ubiquitin-positive, tau-negative inclusions (FTLD-U) with or without motor neuron disease (MND). Molecular genetic studies of familial cases of FTLD-U have shown that mutations in the progranulin (PGRN) gene are a major genetic cause of FTLD-U. Mutations in valosin-containing protein (VCP) gene are present in rare familial forms of FTD, and some families with FTD and/or MND have been linked to chromosome 9p, and both are types of FTLD-U. Thus, familial TDP-43 proteinopathy is associated with defects in multiple genes, and molecular genetics is required in these cases to correctly identify the causative gene defect. In addition to genetic heterogeneity amongst the TDP-43 proteinopathies, there is also neuropathologic heterogeneity and there is a close relationship between genotype and FTLD-U subtype. In addition to these recent significant advances in the neuropathology of FTLD-U, novel FTLD entities have been further characterized, including neuronal intermediate filament inclusion disease. The proposed criteria incorporate up-to-date neuropathology of FTLD in the light of recent immunohistochemical, biochemical, and genetic advances. These criteria will be of value to the practicing neuropathologist and provide a foundation for clinical, clinico-pathologic, mechanistic studies and in vivo models of pathogenesis of FTLD.
KeywordsFrontotemporal dementiaSemantic dementiaProgressive non-fluent aphasiaFrontotemporal lobar degenerationMotor neuron diseaseTauopathyUbiquitinTDP-43 proteinopathyProgranulinValosin-containing proteinCharged multivesicular body protein 2BNeuronal intermediate filament inclusion diseaseNeuropathologic diagnosis
In this paper, we follow the convention that FTLD is an umbrella term that groups several different neurodegenerative diseases characterized by predominant destruction of the frontal and temporal lobes. After Alzheimer disease (AD) and dementia with Lewy bodies (DLB), frontotemporal lobar degeneration (FTLD) is the third most common neurodegenerative cause of dementia in industrialized countries [59, 60, 69]. Most commonly, patients with FTLD present with frontotemporal dementia (FTD), a change in personal and social conduct, often associated with disinhibition, with gradual and progressive changes in language . Other patients falling under the diagnostic umbrella of FTLD may present with early and progressive changes in language function, and two syndromes have been recognized: semantic dementia (SD) and primary progressive non-fluent aphasia (PNFA) [40, 43, 53, 65, 67, 78]. In later stages of these particular syndromes, both behavioral and language dysfunction may be present. A proportion of patients with FTLD present with or develop parkinsonism as part of their disease process. Clinical amyotrophic lateral sclerosis/motor neuron disease (ALS/MND) may also be found in a proportion of patients with FTLD, especially those with FTD, indicating a spectrum of clinical phenotypes that relate to common neuropathologic lesions [3, 61, 69].
FTD, SD, or PNFA refer to the main clinical syndromes linked to the FTLD group. Typically, at least in the early course of the disease, patients with FTD do not have an amnestic syndrome, which distinguishes them clinically from AD [46, 53], but there are exceptions . Although no pre-symptomatic biomarkers have been identified, at least in sporadic cases, clinical assessment, neuropsychology, and neuroimaging may help to distinguish FTD and the related disorders of SD and PNFA from other neurodegenerative causes of dementia [12, 43, 46]. The diagnosis of FTD, SD, or PNFA may only be considered when other potential causes of dementia including other nervous system diseases (e.g., small and large vessel disease), systemic conditions (e.g., hypothyroidism), tumors, and substance abuse have been excluded.
The apolipoprotein E (APOE) gene ε4 allele is a major risk factor for AD, though this is not the case in most association studies of FTLD (but see ref. ), and none of the autosomal dominant mutations in genes associated with some familial cases of AD [amyloid precursor protein (APP), presenilin 1 (PS1) and presenilin 2 (PS2)] acts as a risk factor for FTLD.
Recent developments in the molecular pathology and genetics of FTLD now dictate that a minimal panel of pathological investigations is required for correct diagnosis in this group of diseases. Standardization of nomenclature and approach will facilitate better understanding of clinico-pathologic correlations, provide insights into pathogenesis, and guide the construction and validation of in vivo models.
With the exception of those cases in which a gene defect has been identified, examination of the brain and neuropathology are essential in order to determine the disease entity underlying FTLD. Even in those cases that have been genetically characterized, it is not uncommon to find coexisting neurodegenerative disease and other pathology, which may have contributed to the clinical picture to a varying degree. The neuropathology of the brain, either on autopsy or, rarely, on biopsy, remains the “gold standard” for determining the neuropathologic diagnosis. Although most cases seen by a neuropathologist are likely to be cases of advanced disease, there is an increasing awareness that the molecular pathology of all neurodegenerative disease is present often several years prior to the onset of clinical symptoms, and this knowledge will inform the neuropathologist of pre-clinical FTLD in an otherwise cognitively and behaviorally normal subject.
Examination of the brain of a patient with FTLD typically shows symmetrical focal atrophy of the frontal or temporal lobes, or both. In some patients there is asymmetry of atrophy, typically reflected in perisylvian loss on one side of the brain. Macroscopic atrophy of the basal ganglia and loss of pigmentation from the substantia nigra are seen in a proportion of cases. This focal atrophy is, not infrequently, the most dramatic in all of neuropathology. Conversely, in some individuals, for example, those who die at an earlier stage, the brain is unremarkable. The pattern of atrophy may assist in staging disease severity [11, 39, 41].
In most forms of FTLD, examination of the cerebral cortex with H&E staining shows microvacuolation and neuronal loss. In many cases, this is most evident around layer II of the affected cortical regions. In advanced cases, there is transcortical microvacuolation and neuronal loss. Swollen cortical neurons may be seen and highlighted with immunostaining for alpha B-crystallin; however, they are not specific for any disease subtype. White matter myelin loss and astrocytic gliosis may be seen. There may be significant neuronal loss from the basal ganglia and substantia nigra in some cases.
Specific diagnosis of disease within the broad group of FTLD now requires immunohistochemistry (IHC) to determine the molecular pathology, morphology, and distribution of lesions in the neuraxis, and thereby identify the neurodegenerative disease. In the routine microscopic evaluation of the brain of a patient with FTLD, other neurodegenerative diseases may be identified, most commonly AD , DLB , and, rarely, prion disease  and hereditary diffuse leukoencephalopathy with axonal spheroids .
Although some neurodegenerative diseases can be readily identified using conventional staining techniques (e.g., modified Bielschowsky and Gallyas silver impregnations and thioflavine-S for AD pathology [8, 18, 54]), more sensitive and reliable IHC techniques are now preferred. IHC methods are more consistent and dependable than are silver impregnation techniques, they have greater inter-rater reliability, as shown by the BrainNet Europe Consortium study , and IHC results can suggest or identify underlying molecular pathology. For example, antibodies raised against epitopes of tau readily label the neurofibrillary tangles, neuritic plaques, and neuropil threads of AD; anti-β-amyloid antibodies detect diffuse and compact β-amyloid deposits and cerebral amyloid angiopathy ; while anti-α-synuclein antibodies label Lewy bodies and Lewy neurites, the signature lesions of DLB [10, 52]. Neuropathologic staging schemes have been developed using tau, β-amyloid and α-synuclein IHC, and IHC is now replacing conventional stains in the neuropathologic diagnostic criteria for AD and DLB [10, 18, 52]. Prion IHC may be used reliably to detect or exclude prion disease in most cases . Proteins targeted for degradation are ubiquitinated and several hallmark inclusions in neurodegenerative disease either in neurons or glia or both are detected by ubiquitin IHC [17, 19, 47, 48]. There is also age-related accumulation of ubiquitinated material in the brain , which can make detection of certain ubiquitin-related pathologies difficult. Until recently, ubiquitin IHC was the only marker for certain neuronal inclusions seen in FTLD and ALS/MND that contained neither tau nor α-synuclein epitopes. P62 (sequestosome-1) IHC has recently been highlighted as an alternative method to detect a range of ubiquitin-immunoreactive structures in neurodegenerative diseases including ALS/MND, and FTLD. Like ubiquitin IHC, a range of pathological and age-related abnormalities are detected, but an advantage over anti-ubiquitin IHC appears to be that there is better contrast in the detection of intracellular pathology. More recently, TDP-43 has been identified as a major component of the inclusions of FTLD with ubiquitin-positive, tau- and α-synuclein-negative inclusions (FTLD-U) [3, 61], formerly called FTLD with MND-type inclusions, but without MND . This protein now defines a novel class of neurodegenerative diseases collectively called TDP-43 proteinopathies , and TDP-43 IHC may be used to characterize a majority of FTLD-Us, but not all .
Although IHC is essential for determining the underlying molecular pathology of the majority of neurodegenerative diseases, other techniques may be available in dementia research centers and complement the routine neuropathologic diagnosis. The reliable and robust detection of abnormally aggregated proteins either within neurons or glia or both, or in the neuropil, is necessary for neuropathologic diagnosis. However, the density and distribution of abnormal protein aggregates, as identified by IHC, do not always correlate well with clinical symptoms. Other markers, such as synaptic and neuronal loss in affected brain areas, may correlate better with cognitive impairment and motor dysfunction. Thus, stereologic methods that assess synaptic and neuronal loss in an unbiased manner may be useful in clinico-pathologic studies in the dementia research center, but are not necessary, or usually feasible, for routine neuropathologic diagnosis.
Biochemistry is also useful, but not essential for diagnosis. Methods of fractionating brain homogenates may be used to rationally classify the tauopathies in a research setting . In the adult brain, there are normally six isoforms of the microtubule-associated protein tau (MAPT): three isoforms with 0, 1, or 2 inserts contain three microtubule-binding repeats (3R tau) and three isoforms, also with 0, 1, or 2 inserts, contain four microtubule-binding repeats (4R tau) . The tauopathies have a biochemical signature: tau protein in these disorders is relatively insoluble and these insoluble species can be detected by biochemical fractionation methods. The insoluble fractions may be further characterized according to the pattern of tau isoforms. For example, in AD, all six isoforms are abnormally hyperphosphorylated and migrate as three major bands and one minor band when visualized by immunoblotting. Treatment with the enzyme alkaline phosphatase removes phosphate groups, and the tau isoforms appear as six bands (3R and 4R tau). This biochemical signature may be used to distinguish AD from the FTLD tauopathies . Thus, brain tissue from patients with FTLD where Pick bodies are present is characterized biochemically by predominantly 3R tau, while CBD, PSP, argyrophilic grain disease (AGD), and sporadic multiple system tauopathy with dementia (MSTD) are predominantly 4R tauopathies [13, 74], and neurofibrillary tangle dementia (NTD), also called tangle predominant form of senile dementia, has inclusions containing a mixture of 3R and 4R tau [34, 35]. FTLD with MAPT mutation, of which more than 40 have been described, is biochemically heterogeneous with different mutations being associated with 3R, 4R, or 3R and 4R tauopathy . Monoclonal antibodies, which discriminate between 3R and 4R tau  are now commercially available; so, the molecular classification of tauopathies by isoform type may be easily undertaken in the histology laboratory that does not have access to biochemistry.
FTLD with ubiquitin-positive, tau-negative inclusions (FTLD-U), also known as FTLD with MND-type inclusions or MND inclusion dementia, is the most common underlying pathology in FTLD with and without clinical MND [45, 73]. TAR DNA-binding protein 43 (TDP-43), a nuclear protein implicated in exon skipping and transcription regulation, was recently identified as a major protein component of the ubiquitin-immunoreactive inclusions characteristic of sporadic and familial FTLD-U, with and without clinical MND, as well as in sporadic ALS [3, 16, 21, 61]. Biochemistry in these disorders shows TDP-43 to be abnormally phosphorylated, ubiquitinated and cleaved to generate C-terminal fragments, and is recovered only from areas with ubiquitin-immunoreactive inclusions including hippocampus, neocortex, and spinal cord . The neuropathology of these conditions is characterized by ubiquitin- and TDP-43-positive neuronal cytoplasmic inclusions (NCIs), neuronal intranuclear inclusions (NIIs), dystrophic neurites (DNs), and glial cytoplasmic inclusions (GCIs) that are negative for tau, α-synuclein, β-amyloid, neuronal intermediate filaments, and expanded polyglutamines [21, 61]. The variability in the morphologic types of neuronal inclusions, their distribution, density, and immunohistochemical profile has led to the proposed classification of FTLD-U into four pathologic subtypes [16, 63, 66]. Recently, mutations in the progranulin (PGRN) gene [4, 20, 57], the molecular genetic basis of non-tau familial FTD linked to chromosome 17, were discovered. The neuropathology in these cases is FTLD-U with ubiquitin-positive neurites, NCIs and, most characteristically, NIIs [4, 50] (but see ref. ). However, NIIs can be seen in other FTLD-U cases where PGRN mutations are not found  and therefore such NIIs cannot be considered pathognomonic for PGRN or other (i.e., valosin-containing proteinVCP) mutations associated with FTLD. As demonstrated by IHC and biochemistry, the ubiquitinated pathologic protein in these cases is not progranulin, but TDP-43 [4, 57]. Pathologic TDP-43 is detected biochemically in both affected gray and white matter, suggesting that both glial and neuronal pathology may contribute to the pathogenesis of FTLD-U caused by PGRN mutations .
Frontotemporal lobar degeneration with VCP gene mutation, also called inclusion body myopathy associated with Paget’s disease of bone and frontotemporal dementia (IBMPFD), is a rare autosomal dominant disorder caused by mutations in the VCP gene . VCP, a member of the AAA-ATPase gene super family (ATPase associated with diverse cellular activities), has multiple cellular functions including acting as a molecular chaperone in endoplasmic reticulum-associated protein degradation, stress response, programmed cell death, and interactions with the ubiquitin-proteasome system. The neuropathology in FTLD with VCP mutation is a unique subtype of FTLD-U and is characterized by numerous NIIs and relatively few NCIs and DNs ; the ubiquitinated inclusions are not primarily composed of the mutated protein (VCP), but rather TDP-43 [26, 63].
Frontotemporal lobar degeneration with charged multivesicular body protein 2B (CHMP2B) gene mutation is the cause of FTD linked to chromosome 3 in a large Danish pedigree . Human CHMP2B is a component of the endosomal secretory complex, which becomes dysregulated by the gene defects. Recent studies have revealed ubiquitin-positive, but TDP-43 negative, NCIs in the frontal neocortex and hippocampus, so that this disease is an FTLD-U, but not a TDP-43, proteinopathy .
A genetic locus on chromosome 9p for familial FTD-MND has been described . In one family, candidate gene sequencing revealed the presence of a putative disease segregating stop codon mutation (Q342X) in the intraflagellar transport protein 74 (IFT74) gene . IFT74 is a protein that localizes to the intracellular vesicle compartment and is a component of the intraflagellar transport system responsible for vesicular transport of material synthesized within the cell body into and along dendrites and axons. Neuropathology in a single case with the IFT74 mutation was reported as showing all the stigmata of FTLD-U (ubiquitinated NCIs, DNs, and NIIs) and TDP-43 proteinopathy similar to that seen in other reported families with FTD, with or without MND linked to chromosome 9p . Nonetheless, it remains to be established in other families and patients that IFT74 is indeed a true locus for FTLD.
Neuropathologic classification of FTLDs
Following the principles of the previous consensus criteria for the neuropathologic diagnosis of FTLD , and the consensus criteria for the postmortem diagnosis of AD  and DLB , we acknowledge that only probabilistic statements can be made as to the causal relationship between the neuropathology and the clinical phenotype. Just as the constellation of clinical symptoms associated with FTD, SD, or PNFA do not predict reliably the underlying causative neurodegenerative disease, the presence of the neuropathology of FTLD does not predict with certainty one or other of the clinical phenotypes associated with FTLD, or even if the subject was demented. Small series of cases are inadequate to reliably and robustly determine clinico-pathologic correlations with any one form of FTLD. Multi-center collaborations are beginning to address this challenge , and it is only by pooling relatively rare cases from several research centers that reliable clinico-pathologic correlations are likely to emerge.
Comparison between the present proposed criteria and McKhann et al.  neuropathologic diagnostic criteria for FTLD
McKhann et al. criteria
1. Tauopathy (with associated neuron loss and gliosis) and insoluble tau with a predominance of 3R tau, the most likely diagnoses are:
1. When the predominant neuropathological abnormalities are tau-positive inclusions (with associated neuron loss and gliosis) and insoluble tau has a predominance of tau with three microtubule-binding repeats, the most likely diagnoses are:
FTLD with Pick bodies
(a) Pick disease
FTLD with MAPT mutation
(b) Frontotemporal dementia with parkinsonism linked to chromosome 17
(c) Other as yet unidentified familial and sporadic frontotemporal disorders
2. Tauopathy (with associated neuron loss and gliosis) and insoluble tau with a predominance of 4R tau, the most likely diagnoses are:
2. When the predominant neuropathological abnormalities are tau-positive inclusions (with associated neuron loss and gliosis) and insoluble tau has a predominance of four microtubule-binding repeats, the most likely diagnoses are:
(a) Corticobasal degeneration
Progressive supranuclear palsy
(b) Progressive supranuclear palsy
Argyrophilic grain disease
(c) Frontotemporal dementia with parkinsonism linked to chromosome 17
Sporadic multiple system tauopathy with dementia
(d) Other as yet unidentified familial and sporadic frontotemporal disorders
FTLD with MAPT mutation
3. Tauopathy (with associated neuron loss and gliosis) and insoluble tau, with a predominance of 3R and 4R tau, the most likely diagnoses are:
3. When the predominant neuropathological abnormalities are tau-positive inclusions (with associated neuron loss and gliosis) and insoluble tau has a predominance of three and four microtubule-binding repeats, the most likely diagnoses are:
Neurofibrillary tangle dementia
(a) Neurofibrillary tangle dementia
FTLD with MAPT mutation
(b) Frontotemporal dementia with parkinsonism linked to chromosome 17
(c) Other as yet unidentified familial and sporadic frontotemporal disorders
4. Frontotemporal neuronal loss and gliosis without tau- or ubiquitin/P62-positive inclusions, the most likely diagnosis is:
4. When the predominant neuropathological abnormalities are frontotemporal neuronal loss and gliosis without tau- or ubiquitin-positive inclusions and without detectable amounts of insoluble tau, the most likely diagnoses are:
FTLD (also known as dementia lacking distinctive histologic features)
(a) Frontotemporal lobar degeneration (also known as dementia lacking distinct histopathological features)
(b) Other as yet unidentified familial and sporadic frontotemporal disorders
5. TDP-43 proteinopathy with associated neuronal loss and ubiquitin-positive/P62-positive, tau-negative inclusions, with MND or without MND but with MND-type inclusions, the most likely diagnoses are:
5. When the predominant neuropathological abnormalities are frontotemporal neuronal loss and gliosis with ubiquitin-positive, tau-negative inclusions and without detectable amounts of insoluble tau, with MND or without MND but with MND-type inclusions, the most likely diagnoses are:
FTLD-U with MND (FTLD-U types 1–3)
(a) Frontotemporal lobar degeneration with MND
FTLD-U but without MND (FTLD-U types 1–3)
(b) Frontotemporal lobar degeneration with MND-type inclusions but without MND, or
FTLD-U with PGRN mutation (FTLD-U type 3)
(c) Other as yet unidentified familial and sporadic frontotemporal disorders.
FTLD-U with VCP mutation (FTLD-U type 4)
FTLD-U linked to chromosome 9p (FTLD-U type 2)
Other as yet unidentified TDP-43 proteinopathies
6. Frontotemporal neuronal loss and gliosis with ubiquitin-positive/P62-positive, TDP-43- and tau-negative inclusions, the most likely diagnoses are:
FTLD-U with CHMP2B mutation
Basophilic inclusion body disease (BIBD)
Other as yet unidentified FTLD-U, non-TDP-43 proteinopathies
7. Frontotemporal neuronal loss and gliosis with ubiquitin/P62 and α-internexin-positive inclusions, the most likely diagnosis is:
Neuronal intermediate filament inclusion disease (NIFID)
Algorithm for the neuropathologic diagnosis of FTLD
Recently, staging schemes have been developed that no longer rely on capricious silver impregnation methods, but employ instead sensitive monoclonal and polyclonal antibodies that detect, by IHC, the pathologic proteins of the major neurodegenerative diseases [9, 10, 52]. Multi-center studies have confirmed the reproducibility and reliability of IHC over traditional staining methods, and IHC is recommended for the detection of the signature lesions of FTLD [1, 53], when appropriate. Thus, neurodegenerative diseases with α-synuclein pathology, with or without Aβ plaques and tau-positive neurofibrillary tangle (AD) pathology (i.e., DLB, Parkinson’s disease, and multiple system atrophy), or those with plaque and tangle pathology (i.e. Alzheimer disease) are excluded using established sampling schemes and diagnostic criteria for these diseases [18, 52].
Where neurofibrillary tangles alone are present, in the absence of Aβ plaques, in the context of neuronal loss and gliosis, NTD, also called tangle predominant form of senile dementia, which, like AD tangles, contain tau composed of all six isoforms, is a diagnostic possibility. Where the distribution of neuronal and glial tau pathology is more widespread and includes frontal, temporal, and parietal neocortex, basal ganglia, and brainstem nuclei, then sporadic MSTD may be indicated . Neurofibrillary tangles in more subcortical regions including the basal ganglia, subthalamic nucleus, midbrain, and pontine nuclei indicate progressive supranuclear palsy [31, 72]. Distinguishing lesions in PSP are tau-positive tufted astrocytes and are found in affected neocortical and subcortical regions. Corticobasal degeneration is characterized by frontal and temporal atrophy that is not infrequently asymmetric, neuronal loss, gliosis, swollen achromatic neurons that are faintly tau-positive, and tau-positive neurofibrillary tangles in the neocortex, basal ganglia, and brainstem nuclei [24, 27]. Distinguishing lesions in CBD are tau-positive astrocytic plaques and threads found in the affected neocortex and subcortical white matter and in the basal ganglia. In both PSP and CBD, tau-positive oligodendroglial inclusions called coiled bodies are seen, but these are generally at a lower density than inclusions in astrocytes. Tau-positive ovoid structures (glial processes), astrocytes, and oligodendroglial inclusions (coiled bodies), when confined to the medial temporal lobe and limbic structures, indicate another tauopathy, AGD . If globose tau-positive NCIs, called Pick bodies, are present in the non-pyramidal (dentate gyrus granule cells) and pyramidal neurons of area CA1 of the hippocampus, and pyramidal neurons of the temporal and frontal lobes, then FTLD with Pick bodies may be present . Pick bodies are largely or wholly composed of 3R tau, which can be demonstrated by IHC or immunoblotting, while the tau-positive inclusions of PSP, CBD, MSTD, and AGD all contain 4R tau indicating that these latter disorders may represent a spectrum of 4R tauopathies [13, 74]. Finally, FTLD with MAPT mutation, also called FTD with parkinsonism linked to chromosome 17 (FTDP-17), is not only clinically and genetically heterogeneous (more than 40 mutations have been reported in the MAPT gene), but is also neuropathologically heterogeneous. The spectrum of neuronal and glial pathology seen in 3R, 4R, and combined 3R and 4R tauopathies is also found in such cases of familial tauopathy [15, 44, 70]. For the practicing neuropathologist, the presence of this spectrum of pathology in a case warrants further genetic investigation particularly if there is an autosomal dominant pattern of inheritance. Similar 3R and 4R tau heterogeneity is seen in some tauopathies that cannot be categorized as FTLD with Pick bodies, PSP, CBD, or AGD, and in individuals without MAPT mutations. Such a case may have tau pathology in the distribution described in sporadic MSTD, or may have to be categorized as “unclassifiable sporadic tauopathy.”
Tau-negative, ubiquitin-positive inclusions
Immunohistochemistry for ubiquitin, P62, and TDP-43 in cases with FTLD generally reveals similar pathology that includes a spectrum of neuronal (NCIs, DNs, and NIIs) and glial, predominantly oligodendroglial, cytoplasmic inclusions (GCIs) [62, 61]. However, there are significant differences in the immunohistochemical findings with ubiquitin, P62, and TDP-43. Ubiquitin immunoreactivity is present in extensive age-related pathology in gray and white matter, e.g., dot-like bodies and granular degeneration of myelin, which can mask subtle neuronal and glial pathology and can be difficult to interpret. P62 immunostaining detects the same range of pathological structures as anti-ubiquitin, but highlights less age-related pathology, making interpretation somewhat less demanding. TDP-43 immunoreactivity is present in nuclei of most cells types and changes in distribution within affected neurons in disease. Screening cases using TDP-43 immunostaining as a primary diagnostic tool, as might be used for tau and α-synuclein accumulations, is more demanding. Four histologic subtypes of FTLD-U have been proposed, based on the predominant type of inclusion present as detected using anti-ubiquitin, its distribution in the cortex, and density [16, 63, 66]. Other types have been proposed as well and take into account involvement of other brain regions (e.g., hippocampus or corpus striatum) and the morphology of the lesions [2, 38]. Patterns of FTLD-U histology based solely on cortical pathology include a system proposed by Sampathu et al. , and Neumann et al. ; while Mackenzie et al.  have proposed a system that includes cortical and dentate fascia inclusions. In all these schemes the same descriptors were employed to differentiate subtypes, though the numbering of each subtype differed among schemes. In the present composite scheme, type 1 cases (as in , but known as type 2 in Mackenzie et al. ) are characterized by an abundance of long DNs, predominantly in the superficial cortical laminae, with few or no NCIs or NIIs. Type 2 cases (as in , but known as type 3 in Mackenzie et al. ) are characterized by numerous NCIs in both superficial and deep cortical laminae as well as infrequent DNs and sparse or no NIIs. Type 3 cases (as in , but known as type 1 in Mackenzie et al. ) are characterized by pathology predominantly in the superficial cortical layers with numerous NCIs, DNs, and variable numbers of NIIs. Type 4 cases [16, 63] are distinguished by numerous NIIs and infrequent NCIs and DNs in the neocortical areas with relative sparing of the hippocampus, consistent with the pathology described in cases of FTLD with VCP mutations (but see ref. ). Consensus on the validity, reliability, and reproducibility of the various proposed typing schemes, as well as their clinical significance, remain to be determined.
If TDP-43- and ubiquitin- or P62-positive, tau- and α-synuclein-negative inclusions are found in the superficial laminae of the frontal and temporal neocortex and neurons of the dentate gyrus, the most likely diagnosis is FTLD-U [21, 36, 45]. If in addition, there is TDP-43 proteinopathy and ubiquitin-positive inclusions in the lower motor neurons and evidence of motor neuron loss, gliosis, Bunina bodies, and corticospinal tract degeneration, FTLD with MND is the most likely diagnosis [53, 64, 79].
There is heuristic value in the sub-classification of FTLD-U beyond pathologic description alone, with clinical and genetic correlates of histologic patterns emerging. Cases with type 1 histology are associated with SD, whereas cases of FTLD with MND frequently show type 2 histology . In genetic terms, cases with PGRN or VCP gene mutations, and in cases linked to chromosome 9, distinct patterns of ubiquitin- and TDP-43-positive inclusions are also seen. FTLD with PGRN mutation cases often display a PNFA clinical phenotype and exclusively show type 3 histology , whereas those with VCP mutations have type 4 histology. Cases linked to chromosome 9 have type 2 histology  and the majority of such cases also have ubiquitin- and TDP-positive inclusions in the upper and lower motor neurons, identical to those encountered in sporadic MND  or sporadic FTLD with MND where a similar type 2 histology is often present . These latter data indicate that the pathology of FTLD linked to chromosome 9p is a specific subtype of FTLD-U (type 2) and that TDP-43 is the disease-associated protein. Biochemistry of sporadic cases of FTLD with and without MND, cases of sporadic ALS, and familial cases of FTLD-U with PGRN and VCP mutations, and those linked to chromosome 9p, all have a characteristic biochemical signature: TDP-43 is detected in the detergent-insoluble urea fractions from affected regions and is abnormally phosphorylated, with additional protein bands of ∼25 and 45 kDa, as well as a high molecular smear, and is ubiquitinated [16, 61, 62, 63]. The quantity of these modified TDP-43 species detected by biochemistry [62, 63] may be variable, but correlates with the amount of pathology detected by IHC. Additionally, the 45 kDa species collapse into a 43 kDa band upon dephosphorylation with alkaline phosphatase, indicating that TDP-43 is abnormally phosphorylated, with features paralleling the biochemical changes seen in the tauopathies. Hippocampal sclerosis (HS) may be found as a coexisting pathology in FTLD-U with or without MND, and small numbers of ubiquitin- and TDP-43-positive inclusions may be seen exclusively in the dentate granule cells. Preliminary data indicate that some, or perhaps most, cases of HS are TDP-43 proteinopathies , but further studies on larger samples of “pure HS” and biochemical studies are required to determine the nosologic status of HS.
In those cases that have ubiquitin- or P62-positive, TDP-43-, tau-, and α-internexin-negative NCIs in the frontal and temporal lobes and dentate gyrus, the most likely diagnosis is FTLD with CHMP2B mutation . Mutations in the CHMP2B gene are the cause of FTD-linked to chromosome 3 in a large Danish pedigree . Human CHMP2B is a component of the endosomal secretory complex, which becomes dysregulated by the gene defect. However, the absence of DN and the presence of granular, ubiquitin-positive structures within the neocortex of these cases distinguish this FTLD-U subtype from the types 1–4 described above. Thus, based on the small number of cases studied to date, familial FTLD with CHMP2B mutation appears to be a distinctive pathologic subtype of FTLD-U and is not a TDP-43 proteinopathy on the basis of IHC.
In cases of FTLD where there is frontotemporal neuron loss and gliosis, α-internexin- or neurofilament-positive, TDP-43-, α-synuclein-, and tau-negative, and variably ubiquitinated but P62-positive neuronal inclusions, the most likely diagnosis is neuronal intermediate filament inclusion disease . Where there is FTLD and P62-positive, variably ubiquitin-positive, TDP-43-, α-internexin-, α-synuclein-, and tau-negative basophilic inclusions, the most likely neuropathologic diagnosis is basophilic inclusion body disease (BIBD) . If all of the antibodies listed above and routine histological stains such as hematoxylin and eosin fail to reveal signature lesions, and if prion disease has been excluded by IHC or molecular genetics, the remaining FTLD diagnosis is dementia lacking distinctive histologic features (DLDH) . However, it should be borne in mind that in many of the earlier histopathologic surveys, a relatively high proportion of cases of DLDH were encountered [32, 67]. Re-evaluation of such cases using either more sensitive ubiquitin IHC methodologies [25, 45, 51] or TDP-43 IHC  shows DLDH to be an uncommon cause of FTLD and, indeed, it still remains to be proven as a discrete entity with diagnostic criteria other than default characteristics being employed when all IHC and other methods have failed to reveal signature lesions.
The proposed criteria for the neuropathologic diagnosis of FTLD described here are an evolution of the previous criteria described by McKhann et al. . When formulating these proposed criteria, we acknowledged revised staging schemes for other disorders (e.g., AD and DLB), which recommend the use of IHC for the diagnosis of neurodegenerative diseases, replacing silver impregnation methods and averting the lack of reproducibility between centers when these stains are used. We have incorporated the recent identification of new entities into the nosology of FTLD. Most significantly, the discovery of TDP-43 as the major pathologic protein of most forms of FTLD identifies a novel molecular pathology, TDP-43 proteinopathy, and this is now included in the revised criteria. We have also considered the great progress in determining the molecular genetics of FTLD. In addition to FTLD with MAPT mutation, other familial subtypes are now recognized on the basis of the neuropathology of ubiquitin and/or TDP-43 IHC, biochemistry, and molecular genetics (FTLD with PGRN, CHMP2B, and VCP mutations, and cases linked to chromosome 9p), which reveal a strong correlation between genotype and neuropathologic phenotype. To facilitate neuropathologic diagnoses of FTLD at research and other centers, commercially available antibodies are now readily available for the identification of the underlying molecular pathologies (e.g., TDP-43 proteinopathy and tauopathy) and the rational diagnosis and nosology of a case of FLTD that comes to autopsy. The neuropathologic diagnostic algorithm described here is based on the current level of knowledge, but the consortium members appreciate that further study of TDP-43 proteinopathy may reveal new subtypes and that other FTLD entities may yet be identified. The proposed neuropathologic algorithm will facilitate efforts to improve the diagnosis of FTLD and encourage multi-center clinico-pathologic studies. Together, these efforts should improve the early and reliable neuropathologic diagnosis of the FTLD, raise the awareness of possible coexisting pathologies, and facilitate research efforts into pathogenesis and potential treatments where none currently exists.
We thank the clinical, genetic, pathology, and technical staff of the collaborating centers for making information and tissue samples available for this study and we thank the families of patients whose generosity made this research possible. Support for this work was provided by grants from the National Institute on Aging of the National Institutes of Health (P50-AG05681, P01-AG03991, U01-AG16976, P30-AG13854, P30-NS057105, AG10124 and AG17586), the Buchanan Fund, the Winspear Family Center for Research on the Neuropathology of Alzheimer Disease, and the Charles and Joanne Knight Alzheimer Research Initiative.