Abstract
This article examines the evolution in understanding of frontotemporal dementia (FTD) during the last four decades. A central theme is the recognition of heterogeneity. Originally construed as a disorder of behaviour and executive impairment, FTD is now known also to be associated with alterations in language, conceptual knowledge and praxis. An absence of neurological signs is the hallmark of many FTD patients, but there is also an established association with motor neurone disease (MND), progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD). FTD is commonly defined as an early onset dementia, yet about a quarter of patients present after the age of 65. The underlying pathological protein is tau, TDP-43 or more rarely fused-in-sarcoma (FUS). Distinct genetic mutations have been identified in familial FTD. There are predictable relationships between clinical phenotype, pathological substrate and genetic mutation. For example, a circumscribed semantic disorder predicts TDP-43 pathology, and speech or limb apraxia tau pathology. The co-occurrence of MND predicts TDP-43 pathology, and PSP and CBD tau pathology. FUS pathology is associated with very youthful onset, stereotyped behaviours and caudate atrophy. Non-fluent aphasia is linked to progranulin (GRN) mutations and MND and psychosis to repeat expansions in the C9orf72 gene. Despite striking worldwide consensus in findings there remain some issues of contention, largely related to the classification of FTD and its sub-variants. Understanding the diverse nature of FTD is crucial for effective diagnosis, management and the development of targeted therapies.