The criteria ought to provide the foundation for research work into the neuropsychology, neuropathology, genetics, molecular biology, and epidemiology of these important clinical disorders that account for a substantial proportion of cases of primary degenerative dementia occurring before the age of 65 years.

Based on a combination of the clinical and pathological data, we suggest that microinfarcts in the cerebral cortex associated with severe CAA may be the primary pathological substrate in a significant proportion of VaD cases. Future studies should be undertaken to confirm or dismiss the hypothesis that these cases exhibit a different symptom profile than VaD cases without CAA.

We investigated the distribution of neuropathologically defined dementia subtypes among individuals with dementia disorder. The neuropathological reports were studied on all patients (n=524; 55.3% females; median age 80, range 39-102 years) with clinically diagnosed dementia disorder who underwent complete autopsy including neuropathological examination within the Department of Pathology at the University Hospital in Lund, Sweden, during the years 1974-2004. The neuropathological diagnosis was Alzheimer's disease (AD) in 42.0% of the cases, vascular dementia (VaD) in 23.7%, dementia of combined Alzheimer and vascular pathology in 21.6%, and frontotemporal dementia in 4.0% of the patients. The remaining 8.8% of the patients had other dementia disorders, including combinations other than combined Alzheimer and vascular pathology. The registered prevalence of dementia subtypes depends on many variables, including referral habits, clinical and neuropathological judgments and diagnostic traditions, all of these variables potentially changing over time. This, however, does not seem to obscure the delineation of the major dementia subgroups. In this material of 30 years from Lund in the south of Sweden, AD by far dominated among dementia subtypes, while cerebrovascular pathology corresponded with the dementia disorder, either entirely or partly, in almost half of the demented patients.

BackgroundFrontotemporal dementia (FTD) is recognised as a clinically and morphologically heterogeneous group of interrelated neurodegenerative conditions. One of the subtypes within this disease spectrum is the behavioural variant FTD (bvFTD). This is known to be a varied disorder with a mixture of tau-positive and tau-negative underlying pathologies. The other subtypes include semantic dementia (SD), which generally exhibits tau-negative pathology, and progressive non-fluent aphasia (PNFA), which is usually tau-positive. As the clinical presentation of these subtypes may overlap, a specific diagnosis can be difficult to attain and today no specific biomarker can predict the underlying pathology. Neurofilament light chain protein (NFL), a cytoskeletal constituent of intermediate filaments, is thought to reflect neuronal and axonal death when appearing in the cerebrospinal fluid (CSF). NFL has been shown to be elevated in CSF in patients with FTD compared with AD and controls. Our hypothesis was that the levels of NFL also differ between the subtypes of FTD and may indicate the underlying pathological subtype.MethodsWe retrospectively analysed data from previous CSF analyses in 34 FTD cases (23 bvFTD, seven SD, four PNFA), 20 AD cases, and 26 healthy controls. A separate group of 10 neuropathologically verified and subtyped FTD cases (seven tau-negative, three tau-positive) were also analysed.ResultNFL levels were significantly higher in FTD compared with both AD (p<0.001) and controls (p<0.001). The NFL levels of SD and bvFTD were significantly higher (p<0.001) compared with AD. The biomarker profiles of PNFA and AD were similar. In the neuropathologically verified FTD cases, NFL was higher in the tau-negative than in the tau-positive cases (exact p=0.017).ConclusionsThe marked NFL elevation in some but not all FTD cases is likely to reflect the different underlying pathologies. The highest NFL values found in the SD group as well as in the neuropathologically verified tau-negative cases may be of subtype diagnostic value, if corroborated in larger patient cohorts. In bvFTD, a mixture of tau-positive and tau-negative underlying pathologies could possibly explain the intermediate NFL values.

More knowledge about early presenting cognitive and behavioural signs of FTD is needed in both primary and secondary health care to reduce the time period needed to establish a clinical diagnosis of FTD.