Introduction

Over the two past decades there have been considerable advances in our understanding of the major neurodegenerative diseases producing focal cognitive deficits, most commonly referred to collectively as either Pick’s disease or, more recently, frontotemporal dementia (FTD). These advances have come from the fields of neuropsychology, neuropsychiatry, neuroimaging, neuropathology, and molecular genetics. Unfortunately, most non-experts’ ability to follow these developments has been hindered by the confusing plethora of terms which have been used. Central to the problem is a lack of clarity concerning the level of description (clinical syndrome versus clinicopathologic entity versus specific histologic diagnosis) and the poor concordance between these levels. In other words, while some labels denote a clinical syndrome without specific histologic implications (e.g., progressive aphasia, semantic dementia, or dementia of frontal type), others denote specific neuropathologic entities (e.g., Pick’s disease, familial tauopathy, ubiquitin-inclusion disease), hybrid clinicopathologic entities (frontotemporal dementia), or even specific genetic disorders (e.g., chromosome 17-linked frontotemporal dementia with parkinsonism [FTDP-17]). The resurgence of interest in these disorders and the differences in opinion over terminology are well illustrated by the titles of the previous books published on the topic: Pick’s Disease and Pick’s Complex (Kertesz and Munoz, 1998), Frontotemporal Dementia (Pasquier et al., 1996), and Frontotemporal Lobar Degeneration: Frontotemporal Dementia, Progressive Aphasia, Semantic Dementia (Snowden et al., 1996b).

The aims of this introductory chapter are to review the evolution of the terms used to describe this spectrum of disorders, to highlight recent advances and areas of continuing controversy, and to set the scene for the rest of the book. While my own preference has always been to use the term Pick’s disease for this group of disorders – partly because this term is more readily understood by carers and parallels our use of the label Alzheimer’s disease – the tide of medical opinion turned in favor of FTD in the late 1990s. We have, therefore, adopted this general label within which we distinguish two main clinical variants: behavioral variant FTD (bvFTD) and primary progressive aphasia (PPA) with further subclassification of the aphasic cases. The sections that follow describe the meandering route that led to the adoption of these terms. The chapter draws heavily upon my own experience of more than 500 patients assessed in Cambridge and in Sydney over the past 20 years and has been updated since the first edition of the book to reflect areas of evolution and change.

What did Arnold Pick actually describe?

In 1892 Arnold Pick (Girling and Berrios, 1994), working in Prague, reported a 71-year-old man with progressive mental deterioration and unusually severe aphasia who at post-mortem had marked atrophy of the left temporal lobe. Twelve years later in 1904 he published his landmark paper (“On the symptomatology of left-sided temporal lobe atrophy”) in which he described three further cases (Girling and Berrios, 1997). The first, a 58-year-old woman (Josephina) had a two-year history characterized by a striking loss of memory for names (amnesic aphasia) culminating in almost complete loss of speech and accompanied by changes in personality. She deteriorated rapidly and at post-mortem, two years after presentation, Pick observed asymmetric temporal lobe atrophy involving particularly the inferior and middle gyri (i.e., not Wernicke’s area). Methods of staining brain sections were not available at that time and Pick was able to make observations on the macroscopic pathology only. The other two cases were clinically similar, except that case three had the complication of cerebral syphilis, preventing firm conclusions about the cause of the focal brain atrophy. Pick wanted to draw attention to the fact that progressive brain atrophy may lead to symptoms of focal disturbance (in this instance aphasia) through local accentuation of the disease process. He also made specific and, as we will see below, highly perceptive predictions regarding the role of the mid temporal region of the left hemisphere in the representation of word meaning. It was only in his later publication that Pick turned his attention to bilateral frontal atrophy with resultant behavioral disturbance.

Pick made major contributions which have sadly been rather overlooked in recent years. Current classifications have also relegated him to a minor role but several points should be emphasized: (1) Pick’s primary interest was the language and behavioral disorder, particularly the clinico-anatomical correlates of aphasia; (2) he did not claim to have discovered a new disease, merely novel phenomena arising from asymmetric degeneration; (3) two of the major syndromes now included under the rubric of FTD (bvFTD and semantic dementia) were clearly described by Pick; (4) he did not describe distinct histopathologic changes in his patients with focal atrophy.

The histologic abnormalities associated with Pick’s disease were, in fact, described a few years later by Alzheimer (1911) who recognized changes distinct from those found in the form of cerebral degeneration later associated with his name. Alzheimer recognized both argyrophilic intracytoplasmic inclusions (Pick bodies), and diffusely staining ballooned neurons (Pick cells) in association with focal lobar atrophy. It is interesting to note that a comprehensive review of 20 patients from the literature with aphasia due to focal lobar atrophy written soon after Alzheimer’s description (Mingazzini, 1913) did not use the label Pick’s disease. Onari and Spatz (1926) were among the first to use the eponym Pick’s disease but Carl Schneider (1927, 1929) is probably most responsible for its introduction. Unfortunately, however, he concentrated on the frontal lobe component of the syndrome and began the neglect of the temporal lobe syndromes associated with focal atrophy that continued for at least half a century. He distinguished three clinical phases – the first characterized by impaired judgment and behavior, the second by focal symptoms, and the third by generalized dementia. Many papers describing similar cases appeared in the 1930s and 1940s (e.g., Ferraro and Jervis, 1940; Lӧwenberg and Arbor, 1936; Lӧwenberg et al., 1939; Neumann, 1949; Nichols and Weigner 1938) which mainly focused on the frontal lobe aspects of the disorder. Given the more recent genetic discoveries related to the gene for tau protein, special mention should be made of the large Dutch family first reported by Sanders et al. (1939) and then again by Schenk (1951). These families were central to developments in the 1990s when linkage to the tau gene region on chromosome 17 was established by workers in the USA (Wilhelmsen et al., 1994) and Europe (Heutink et al., 1997).

With the general waning of interest in the cognitive aspects of neurology in the English-speaking world, interest in focal dementia syndromes faded, as reflected by the dearth of clinical papers in the neurologic literature after the Second World War. Indeed, many authors went as far as to claim that Alzheimer’s and Pick’s disease were clinically indistinguishable in life (Kamo et al., 1987; Katzman, 1986). The focus of interest in English language publications became the neuropathology, and latterly the genetics, of these conditions. This resulted in a gradual change in the criteria for Pick’s disease, which evolved to include the necessity for specific pathologic changes (i.e., focal atrophy with Pick cells and/or Pick bodies). In continental Europe, however, there remained a strong interest in the clinical phenomena of the dementias; Pick’s remained an in vivo diagnosis based on a combination of clinical features suggestive of frontal and/or temporal lobe dysfunction and focal lobar atrophy (e.g., Mansvelt, 1954; Tissot et al., 1975, 1985).

This controversy continues and has contributed to the adoption of the many labels to describe patients with the clinical syndrome of progressive frontal or temporal lobe degeneration.

Rediscovering Pick’s disease: from dementia of the frontal type and progressive aphasia to frontotemporal dementia

A renaissance of interest in the focal dementias began in the 1980s. Workers from Lund, Sweden (Brun, 1987; Gustafson, 1987) reported on a large series of patients with dementia and found that of 158 patients studied prospectively who came to post-mortem, 26 had evidence of frontal lobe degeneration. Since only a small proportion had Pick cells and Pick bodies – the remainder had very similar findings but without specific inclusions (i.e., focal lobar atrophy with severe neuronal loss and spongiosis) – the Lund group preferred to adopt the term “frontal degeneration of non-Alzheimer type.” At approximately the same time, Neary and co-workers in Manchester (Neary et al., 1986) began a series of important clinicopathologic studies of patients with presenile dementia. They, likewise, found a high proportion of cases with a progressive frontal lobe syndrome who had neither specific changes of Alzheimer’s disease (plaques and tangles) nor specific inclusion pathology. They introduced the term “dementia of frontal type.” Over the next few years other groups described very similar cases under the labels “frontal lobe degeneration” (Miller et al., 1991) and “dementia lacking distinct histologic features” (Knopman et al., 1990). These papers were important in defining the key clinical features associated with progressive frontal lobe degeneration, notably: alterations in social conduct, inhibitory control, sexual behavior, appetite; ritualized and stereotypic behaviors; reduced empathy; and apathy.

In more recent classifications these patients have been given either the general label of FTD (as distinct from progressive aphasia and semantic dementia) or alternatively frontal variant FTD and more recently bvFTD. Key advances have been the development of carer-based interview schedules or questionnaires such as the Neuropsychiatric Inventory (NPI; Cummings et al., 1994), the Frontal Behavioral Inventory (FBI; Kertesz et al., 2000), and the Cambridge Behavioural Inventory (CBI; Bozeat et al., 2000). It has become apparent that conventional frontal lobe tests based largely on executive abilities (planning, set-shifting, problem-solving) are not very sensitive to the beginnings of this behavioral form of FTD. A range of exciting recent research has focused on ways of measuring the alterations in social conduct, theory of mind, emotion processing, and complex decision-making (Bertoux et al., 2012; Gregory et al., 2002; Keane et al., 2002; Kumfor and Piguet, 2012; Kumfor et al., 2013; Lough et al., 2006; Rahman et al., 1999; Rankin et al., 2003; Torralva et al., 2007). It had been long assumed that the orbital cortex bears the brunt, particularly in the early stages of the disease, but recent quantitative imaging work has emphasized rather the role of the mesial surface. Moreover, some of the symptoms typically regarded as “frontal” in nature may, in fact, be secondary to amygdala or insula damage. Studies attempting to relate individual clinical features to site(s) of brain dysfunction using structural or functional imaging are in their infancy (Hornberger et al., 2011; Kloeters et al., 2013; Rankin et al., 2003; Rosen et al., 2002a, 2005; Williams et al., 2005) and it is certain that there will be considerable advances over the next few years.

Of relevance to the story of bvFTD was the realization a few years ago that a proportion of patients with this clinical label, bvFTD, failed to progress even over many years of follow-up. Such patients typically lacked atrophy on MRI (Davies et al., 2006). Subsequent work showed that these non-progressors or “phenocopy cases,” who were virtually all men, could be identified by their lack of executive (Hornberger et al., 2008) or memory deficits (Hornberger et al., 2010) and preservation of activities of daily living. (Piguet et al., 2011). The etiology of the phenocopy syndrome remains unclear. A proportion of cases may have the C9orf72 (chromosome 9 open reading frame 72) mutation (discussed below). Others have long-standing personality disorders and decompensate in later life. These findings contributed to the revision of criteria for bvFTD with much more clearly defined symptoms and the need for brain imaging changes, plus evidence of progression, to qualify for a diagnosis of probable, rather than possible, bvFTD (Rascovsky et al., 2011).

Familial chromosome 17-linked frontotemporal dementia and the discovery of unique tau pathology

As described in more detail in Chapter 14, linkage was established in a number of families in which FTD is inherited as an autosomal dominant trait to the region of chromosome 17 (q21–22) containing the gene for the microtubule-associated protein tau (Wilhelmsen, 1997). The story of the chromosome 17 linkage is extraordinary in a number of ways. Families around the world with what has become known as FTD with parkinsonism linked to chromosome 17 or FTDP-17 (Spillantini et al., 1998a) had originally been reported under a range of headings including: disinhibition–dementia–parkinsonism–amyotrophy complex (Wilhelmsen et al., 1994), rapidly progressive autosomal dominant parkinsonism and dementia with pallido-ponto-nigral degeneration (Wszolek et al., 1992), familial progressive subclinical gliosis (Petersen et al., 1995), hereditary dysphasia and dementia (Morris et al., 1984), hereditary frontotemporal dementia (Heutink et al., 1997), familial multiple system tauopathy with presenile dementia (Spillantini et al., 1997), familial presenile dementia with psychosis (Sumi et al., 1992), and Pick’s disease (Schenk, 1951). In 1996 a meeting of representatives from all of the groups identifying linkage to chromosome 17 was held in Ann Arbor, Michigan (Foster et al., 1997). Comparison of clinical and pathologic data revealed a great deal of similarity between the families who all shared the characteristics of predominantly frontotemporal distribution of pathology with marked behavioral changes. Extrapyramidal dysfunction was present in most. In some families psychotic symptoms were a major feature and a number had amyotrophy. It was recognized at that time that some of the families shared the common pathology with microtubule-associated protein tau-positive inclusions. Progress in the field was then rapid. It was soon discovered that most, if not all, families had diffuse neuronal and glial tau inclusions with a distinctive morphologic pattern, leading to the coining of the term “familial tauopathy” and the suggestion that the disease might reflect a mutation in the gene for tau protein known to be located in the 17q21–22 region (Spillantini et al., 1998a). Within two years of the Ann Arbor meeting, several groups had identified the genetic mutation which, as predicted, was in the gene for tau protein (Dumanchin et al., 1998; Hutton et al., 1998; Poorkaj et al., 1998; Spillantini et al., 1998b).

Since 1998, more than 30 different mutations of the gene for tau protein have been identified, largely involving the coding regions, particularly the so-called microtubule-binding domains (exons 9–12) of the gene for tau. There has been an explosion of interest in the molecular pathology of tau. Although the histopathologic appearances in cases with mutations of the gene for tau are consistent, the clinical phenotypes across and even within families have varied considerably, suggesting that other factors influence the distribution of pathologic changes within the brain. It is also clear that much remains to be learnt.

Very recently, interest has shifted to cases with ubiquitin-positive pathology, particularly the growing number of familial cases which have all been linked to chromosome 17 (Mackenzie et al., 2006; Van der Zee et al., 2006). Just as the first edition of this book was nearing completion, two groups reported mutations in the gene encoding progranulin, close to but apparently independently of the microtubule-associated protein tau (MAPT) gene (Baker et al., 2006; Cruts et al., 2006). Moreover, it seems that progranulin mutations are relatively common as over 30 families were discovered within months of the original discovery. Another recent genetic breakthrough involves the well-known Danish kindred from Jutland. Linkage to chromosome 3 was established in 1993 (Brown et al., 1993) but in 2005 a mutation in the endosomal sorting complex required for transport III (ESCRT-III) complex subunit CHMP2B gene was described in affected members of the family and in one unrelated sporadic Cambridge patient (Skibinski et al., 2005). A recent large-scale screen of 141 familial probands from the USA and UK suggests that in contrast to MAPT and progranulin this mutation is extremely rare (Cannon et al., 2006).

The even more recent discovery of the C9orf72 gene expansion is best considered after discussing the overlap between FTD and MND.

Frontotemporal dementia with motor neuron disease

Although MND has traditionally been regarded as a disorder which spares higher cognitive abilities, it has become clear since early reports from Japan (Mitsuyama and Takamiya, 1979) that the rate of dementia in MND is significantly greater than expected, and conversely a significant minority of patients with FTD develop features of MND (for reviews see Bak and Hodges, 1999; Burrell et al., 2011; Caselli et al., 1993; Lillo et al., 2010, 2011; Neary et al., 1990; Rakowicz and Hodges, 1998). Many patients with the overlap syndrome present with behavioral changes and/or progressive aphasia, which then progresses rapidly, followed by the emergence of bulbar features and mild limb amyotrophy, although the reverse sequence can be seen. Such patients were noted to have prominent neuropsychiatric features including psychosis (Lillo and Hodges, 2010), which is interesting in the context of the C9orf72 mutation cases.

There is also evidence that patients with the MND-dementia/aphasia complex have disproportionate impairment of verb, compared with noun, knowledge (Bak and Hodges, 1997; Bak et al., 2001), which is pertinent to the hypothesis that some components of the widespread semantic network in the brain are located in or near corresponding sensory/motor areas.

The topic of the degree of overlap between MND and FTD has become one of active investigation. Our studies have suggested that subtle behavioral changes, particularly apathy, are very common in MND and often precede classic motor symptoms (Lillo et al., 2011; Mioshi et al., 2014), and that such symptoms impact significantly on caregiver burden (Lillo et al., 2012). Viewed from the opposite perspective, it appears that perhaps 10–20% of patients with FTD will develop frank MND although a much higher proportion show subtle signs of motor neuron dysfunction; however, the long-term implications of the latter finding remains unclear (Burrell et al., 2011). Chapter 6 provides more detail on this dimension of overlap within the FTD spectrum.

Discovery of the C9orf72 mutation

The latest piece of the genetics puzzle links FTD to MND. It had been clear that certain families included members suffering from both of these disorders and that the responsible gene was C9orf72. In 2011 an exciting discovery was the expansion of a hexanucleotide repeat in the gene C9orf72, soon established to be the commonest genetic cause of both FTD and MND (DeJesus-Hernandez et al., 2011; Renton et al., 2011). As well as accounting for a high proportion of familial cases, this mutation appears to be a relatively common cause of cases (perhaps around 5–10%) of apparently sporadic bvFTD (Hodges, 2012). From a clinical perspective, such cases appear atypical in that they may have a slowly progressive or indolent course with prominent psychiatric features and relatively little in the way of brain atrophy (Devenney et al., 2014). These findings beg the question of how many cases with the non-progressive or phenocopy syndrome may turn out to have this expansion.

Corticobasal syndrome

The other clinical syndrome which overlaps considerably with FTD is CBS, originally described as a movement disorder characterized by an asymmetric akinetic-rigid syndrome with prominent apraxia culminating, in some instances, in the striking feature of alien or anarchic hand (Gibb et al., 1989) and associated with a characteristic pattern of tau-positive neuropathology involving basal ganglia and parietal and frontal cortices (Dickson et al., 2002; Feany and Dickson, 1996). It is now clear that cognitive deficits are virtually universal in CBS (Graham et al., 2003a, 2003b). The pattern of dementia fits most closely with PNFA, although frontal-executive deficits are also common (Graham et al., 2003a). One relatively unique feature is the prominence of visuospatial and perceptual deficits not seen in other forms of FTD (Bak et al., 2005, 2006). Another facet of the overlap is that the typical tau-positive CBD neuropathology may be found in patients presenting with FTD syndromes without motor features, at least in the earlier stages of their illness (Mathuranath et al., 2000). To add further to the confusion, it is emerging that some patients with an in vivo diagnosis of CBS may have Alzheimer’s disease neuropathology at autopsy (Boeve et al., 1999; Doran et al., 2003; Schneider et al., 1997; Shelley et al., 2009). It remains unclear whether such patients can be distinguished in life from those without Alzheimer’s pathology (Alexander et al. 2014; Burrell et al., 2013). Further details on this domain of overlap within the FTD spectrum can be found in Chapter 7.