Frontotemporal lobar degeneration and dementia with Lewy bodies: Clinicopathological issues associated with antemortem diagnosis

Authors


  • This review article was presented by the author in Symposium of the 23rd annual meeting of Japanese Psychogeriatric Society in Kobe, 27–28 June 2008.

Dr Osamu Yokota MD, PhD, Department of Neuropsychiatry, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, 2-5-1 Shikata-cho, Okayama 700-8558, Japan. Email: oyokota1@yahoo.co.jp

Abstract

Currently, the clinical diagnostic criteria of frontotemporal lobar degeneration (FTLD) and dementia with Lewy bodies (DLB) are well known to neurologists and psychiatrists. However, the accuracy of the clinical diagnosis of these diseases in autopsy series is not always adequate. For example, FTLD is a syndrome rather than a clinicopathological disease entity that is comprised of various pathological substrates, including Pick's disease, FTLD with microtubule-associated protein tau gene mutation, FTLD with tau-negative ubiquitin-positive inclusions (FTLD-U), FTLD-U with progranulin gene mutation, corticobasal degeneration, basophilic inclusion body disease, and neuronal intermediate filament inclusion disease. Whether these underlying pathologies can be identified clinically is one of the greatest interests in neuropathological research. The pathophysiological relationship between Lewy pathology and Alzheimer pathology in DLB is explored with interest because it may be associated with the accuracy of clinical diagnoses. For example, although Lewy pathology may progress from the brain stem nuclei to the cerebral cortex in Parkinson's disease, recent studies have demonstrated that the progression pattern in DLB is not always identical to that in Parkinson's disease. It is also considered that the progression pattern of Lewy pathology correlates with the evolution of clinical symptoms and that the progression pattern of Lewy pathology may be altered when Alzheimer pathology coexists. In the present paper, the clinicopathological features of two demented cases are presented, and some pathological issues associated with the clinical diagnosis of FTLD and DLB are discussed.

INTRODUCTION

In clinical practice, various dementias are diagnosed based on operational criteria, as well as on a careful clinical history and neurological and psychiatric examinations. We clinicians tend to believe that making diagnoses became easier after operational criteria were proposed. However, the diagnostic power of the representative criteria is not always adequate. For example, the sensitivity and specificity of the representative operational clinical criteria for Alzheimer's disease (AD) range from 70% to 90%,1–4 and the sensitivity of the criteria for dementia with Lewy bodies (DLB) is often lower.5,6 In a common psychiatric hospital outpatient setting, the frequency of AD tends to be rather high: in our experience, the frequency of AD with and without cerebrovascular disease reached approximately 80%, whereas that of DLB was approximately 4%.7 In general, to improve diagnostic accuracy, clinical diagnoses need to be confirmed by pathological examinations. However, the autopsy rate in Japan is gradually decreasing.

The accuracy of a clinical diagnosis of dementia directly affects the accuracy of all clinical studies, including epidemiological, neuroradiological, and therapeutic intervention studies. Recent quantitative analyses using magnetic resonance imaging (MRI) and single photon emission computed tomography (SPECT) are attempts to improve the accuracy of the diagnosis of dementia. However, it has not yet been confirmed whether various quantitative MRI and SPECT findings (e.g. hypoperfusion in the posterior portion of the cingulate cortex in AD patients) actually correlate with a specific underlying pathology. One of the causes of the dissociation between clinical and pathological diagnoses (misdiagnosis) is that patients with different pathological disease entities often show indistinguishable clinical presentations. For example, at present neurofibrillary tangle-predominant dementia (NFTPD;8 note, the terms ‘senile dementia of the neurofibrillary tangle type’9 (SD-NFT) and ‘limbic neurofibrillary tangle dementia’10 (LNTD) are often used in Japan) cannot be distinguished clinically from AD. Similarly, it is difficult to clinically distinguish Pick's disease (this term now refers only to cases having tau-positive Pick bodies) from frontotemporal lobar degeneration with ubiquitin-positive tau-negative inclusions (FTLD-U), the most frequent pathological substrate of frontotemporal lobar degeneration (FTLD). Another cause of misdiagnosis is that a substantial number of patients have several distinct pathologies, such as AD and dementia with Lewy bodies (DLB). For example, although clinicians usually consider the clinical diagnosis of dementia on the assumption that a single patient has a single disorder, 60% of pathologically confirmed AD cases have α-synuclein pathology including Lewy bodies with a variable distribution. Given these findings, it may be difficult, or even impossible, to sharply demarcate the boundary between AD and DLB.

In the present paper, two autopsy cases are first described that have significant implications regarding the clinical diagnosis of FTLD and DLB. Then, some pathological issues associated with the accuracy of the clinical diagnosis of these diseases are discussed. Detailed clinical, genetic, and pathological findings in the present cases have been reported previously.11,12 In the present paper, the term DLB is used to refer to cases that fit the third consensus guideline of DLB,13 including brain stem-predominant type, limbic type, and diffuse neocortical type, whereas the term Lewy body disease (LBD) is used to refer to all cases having Lewy pathology, including cases that do not fit the criteria of DLB.

CASE 1

Clinical course

The patient was a right-handed man who was 74 years old at the time of death. He had no family or past history of neurological or psychiatric disorders. At the age of 55 years, one of his colleagues, a neurologist, became aware of language disturbances, including difficulty finding words, an excessive use of pronouns, and semantic paraphasia in daily conversation. The patient's spontaneous speech was fluent, his auditory comprehension appeared to be preserved, and there were no problems with his academic work as a chief editor and reviewer of a scientific journal. The patient's neurologist colleague suspected that he was suffering from slowly progressive aphasia. By the age of 60 years, the patient's wife was aware of his difficulty in naming common objects, such as vegetables, yet he could recall day-to-day events and appointments perfectly. Spatial function and face recognition were normal. At age 63 years, the patient presented with disturbances in reading and writing Kanji (Chinese characters). Motor disturbances, including parkinsonism and motor neuron signs, were not found. At 66 years of age the patient became aware of his difficulty in recognizing faces. Thereafter, fixed routines, stereotypic behaviors, and compulsive behaviors gradually appeared. At 71 years of age, he began shuffling his right leg. In addition, swallowing disturbances, changes in eating behaviors, pica, evident euphoria, total aphasia, forced crying, and rigidity in the right lower extremity and, to a lesser degree, in the right upper extremities were observed. Computed tomography (CT) and MRI revealed severe lobar atrophy in the bilateral temporal tips, especially on the left side, and a remarkably dilated Sylvian fissure (Fig. 1a–d). At 72 years of age, increased deep tendon reflexes and bilateral ankle clonus first occurred. Babinski reflex was not found. Rigidity was evident in all four extremities, especially on the right side. The patient finally became bedridden and he died of pneumonia at 74 years of age, 19 years after disease onset. His clinical diagnosis was semantic dementia (SD), a clinical subtype of FTLD.14

Figure 1.

Brain computed tomography (CT) and magnetic resonance images (MRI) in Case 1. (a,b) The CT images were obtained when the patient was 71 years of age. Severe temporal atrophy predominant in the left side is noted. The frontal cortex is relatively spared. The caudate nucleus is slightly atrophic, but the structure of this site is not flattened. The bilateral Sylvian fissures, especially on the left side, are dilated. (c,d) Coronal MRI in the patient at 73 years of age. The bilateral temporal poles are severely atrophic. The anterior portion of the caudate nucleus exhibits evident atrophy and the ventricle is bilaterally dilated and concave in the left side (c). Conversely, the posterior portion of the caudate nucleus is not flattened (d). Reprinted with the permission of Springer from Yokota et al.11

Pathological findings

After fixation, brain weight was 905 g (the mean brain weight of men ranges between 1350 and 1400 g). Macroscopically, severe cortical atrophy was observed in the tips of the bilateral temporal lobes, especially on the left side (Fig. 2a). The gyri in these regions showed knife-blade atrophy (Fig. 2b). Atrophy was also evident in the left pars opercularis and precentral gyrus (Fig. 2a). The frontal cortex showed moderate atrophy, whereas atrophy in the amygdala severe (Fig. 2c). Histopathologically, neuronal loss with glial proliferation was most remarkable in the temporal cortex, which was more evident in the anterior rather than the posterior portion. In the frontal lobes, the pars opercularis and inferior frontal gyrus were affected by neuronal loss. The primary motor cortex demonstrated loss of Betz cells. The bilateral pyramidal tract was evidently degenerated at the levels of the medulla oblongata and spinal cord (Fig. 3a,b). Lower motor neurons in the spinal anterior horn and hypoglossal nuclei were spared in number (Fig. 3b). No Bunina bodies were found. Although the hippocampal region was macroscopically unremarkable, mild to moderate neuronal loss in the hippocampus, subiculum, and parahippocampal gyrus was observed. The amygdala was severely affected by neuronal loss with tissue rarefaction. The caudate nucleus and putamen showed moderate to severe neuronal loss with gliosis. The number of pigmented neurons in the substantia nigra was mildly reduced and free melanin was found. Neurons in the other regions were spared in number. A few neurofibrillary tangles (NFTs) were found in the parahippocampal gyrus and CA1 region of the hippocampus on Gallyas–Braak-stained sections, and a few senile plaques (SPs) were observed in the frontal cortex on methenamine silver-stained sections. Neither Pick bodies, argyrophilic grains, nor Lewy bodies were noted in any region.

Figure 2.

Macroscopic pathological findings in Case 1. (a) The most severely affected site is the temporal tip. In addition, the precentral gyrus (long arrow) and adjacent pars opercularis (short arrow) also show moderate atrophy. Thus, the atrophy in the frontal cortex is posterior gradient. The evident atrophy in these sites results in remarkable dilation of the Sylvian fissure. (b) Coronal section at the level of the head of the caudate nucleus. The temporal lobes show knife edge-like atrophy. The insular cortex, especially on the left side, is also involved. The caudate nucleus appears to be mildly reduced in volume, but the structure is not flattened. (c) Coronal section at the level of the amygdala. The inferior horns are markedly dilated. The volume of the left frontal lobe appears to be smaller than that of the right. Reprinted with the permission of Springer from Yokota et al.11

Figure 3.

Microscopic pathological findings in Case 1. (a) The bilateral pyramidal tracts at the level of the medulla oblongata show myelin pallor, suggesting the degeneration (arrows). Klüver-Barrera stain. (b) The lateral tract at the level of the lumbar cord also shows evident myelin pallor (long arrow). However, the anterior horn cells (lower motor neurons) are well spared in number (short arrow). Klüver-Barrera stain. (c) Ubiquitin-positive inclusions in the granular cells in the hippocampal dentate gyrus. Ubiquitin immunohistochemistry. Bars, 1 mm (a,b); 10 µm (c). Reprinted with the permission of Springer from Yokota et al.11

Ubiquitin immunohistochemistry demonstrated ubiquitin-positive inclusions and neurites in the rostral portion of the hippocampal dentate gyrus (Fig. 3c), parahippocampal gyrus, amygdala, caudate nucleus, putamen, and frontal, temporal, parietal, and occipital cortices. No ubiquitin-positive lesions were found in the hypoglossal nuclei and anterior horn in the spinal cord. These ubiquitin-positive lesions were negative for both tau and α-synuclein. Thus, the pathological diagnosis in this case was FTLD-U and the ubiquitin pathology was morphologically classified as Type 1 because many ubiquitin-positive dystrophic neurites with few neuronal cytoplasmic inclusions and no neuronal intranuclear inclusions were present.15,16

CASE 2

Clinical course

A 49-year-old Japanese man was first noticed a decline in his efficiency at work, and his personality became mildly childish. At the 51 years of age, anxiety and irritability became obvious. His concentration was gradually impaired and his appetite was reduced. At his first neurological assessment at 51 years of age, his speech was incoherent and his understanding was very poor. He had no family history of neurological or psychiatric diseases. He showed mild dysarthria. Gaze palsy could not be evaluated precisely owing to his poor understanding. Neither parkinsonism nor pathological reflex were noted. Muscle weakness and atrophy were not observed. Baseline blood examination, urinalysis, and cerebrospinal fluid investigations were normal. Electroencephalography, head CT, and head MRI were within normal limits. At 53 years of age, the patient was admitted to the psychiatric ward of a general hospital. His score on the Hasegawa dementia scale revised (HDS-R) was 7 points (full score 30 points; cutoff 19/20). Magnetic resonance imaging revealed cerebral atrophy in the temporal and frontal lobes and hippocampal region with dilatation of the third ventricle. The patient died at 56 years of age after a clinical course of 8 years. Parkinsonism, visual hallucinations, delirium, or cognitive fluctuation were not seen during the clinical course. The patient was diagnosed clinically with AD.

Pathological findings in central nervous system

Before fixation, brain weight was 1300 g (cf. a mean brain weight in men of 1350–1400 g). Macroscopic examination revealed no significant cerebral atrophy. Coronal sections of the substantia nigra did not show depigmentation (Fig. 4a), but mild pallor was noted in the locus coeruleus. Histopathologically, the most severely affected region was the amygdala. In this region, severe neuronal loss associated with tissue rarefaction and abundant Lewy bodies, NFTs, and SPs were found. Moderate neuronal loss associated with many NFTs and a few SPs was also noted in the hippocampus. Many cortical Lewy bodies were observed in the other regions of the cerebrum and, to lesser degree, in the frontal cortex. Many NFTs and SPs were found throughout the cerebral cortex; however, neuronal loss in the frontal and temporal cortices was slight. In the brain stem nuclei, the substantia nigra, locus coeruleus, and dorsal vagal nucleus were relatively well preserved. Hematoxylin and eosin-stained sections showed a few classic Lewy bodies in the substantia nigra, locus coeruleus, and dorsal vagal nucleus. A small number of NFTs was also present in the substantia nigra and locus coeruleus. Phosphorylated α-synuclein immunostaining revealed many Lewy bodies and Lewy neurites in the neocortex, limbic system, and brain stem nuclei (Fig. 4b). The Braak stages of neurofibrillary changes and senile plaques in this case were Stage V and Stage C, respectively (Fig. 4c,d).17 Thus, consensus recommendations for a diagnosis of AD according to the National Institute on Aging and the Reagan Institute working group indicated a high likelihood that AD was the cause of dementia.18 This case also corresponded to the diffuse neocortical type of DLB according to the pathological diagnostic criteria of DLB.13 Thus, the DLB criteria also indicated an intermediate likelihood that the Lewy-related pathology was associated with the clinical dementia. Genomic DNA was extracted from paraffin-embedded brain tissue blocks, but there was no mutation in exons 3–12 and the exon–intron boundary regions of the presenilin-1 gene. The Apo E genotype was ε3/ε3.

Figure 4.

Pathological findings in Case 2. (a) The substantia nigra is well pigmented (arrows), suggesting that the pigmented neurons may not be reduced in number. (b) Although neurons in the substantia nigra were relatively spared in number, many Lewy bodies (brown) are encountered at this site. Phosphorylated α-synuclein immunohistochemistry. (c) Numerous senile plaques in the temporal cortex. Methenamine silver stain. (d) The primary motor cortex with Betz cells (arrow). Senile plaques are noted in this site. Thus, senile plaques are classified as Braak Stage C (severe). Methenamine silver stain. Bars, 100 µm (b); 500 µm (c); 200 µm (d). Reprinted with the permission of Wiley-Blackwell from Yokota et al.12

Immunohistochemical findings in cardiac sympathetic nerves

Neurofilament immunostaining revealed that the nerve fibers in the nerve fascicles of the cardiac sympathetic nerve were reduced in number in this case (Fig. 5a), as well as in another DLB case examined as a disease control (Fig. 5d), compared with normal controls (Fig. 5b). However, the degree of the degeneration was slighter than that in Parkinson's disease (PD) cases, the fibers of which had almost completely disappeared (Fig. 5c). Phosphorylated α-synuclein immunostaining revealed many α-synuclein-positive Lewy neurites in the nerve fascicles of the epicardium and myocardium in the present case, as well as in another DLB case and a PD case (Fig. 5e,f). Double-labeling immunofluorescence examination using anti-neurofilament and anti-tyrosine hydroxylase (TH) also disclosed a reduction of fibers in the cardiac sympathetic nerve, as well as thick neurofilament-positive TH-positive fibers that appeared to correspond to Lewy neurites (Fig. 5g,h).

Figure 5.

(a) Neurofilament-positive fibers in the cardiac sympathetic nerve from a patient with dementia with Lewy bodies (DLB) plus Alzheimer's disease (AD). The number of labeled fibers is reduced compared with the normal control (b). However, the fibers in the DLB plus AD case (a) are rather spared compared with those in a Parkinson's disease case (c). A mild reduction of labeled fibers in the cardiac sympathetic nerve was also found in another DLB case (d). (e,f) Phosphorylated α-synuclein-positive thick fibers in a DLB plus AD case (e) and Parkinson's disease case (f). (g) Double immunofluorescence labeling for neurofilament and tyrosine hydroxylase (TH) showing a reduction in the number of neurofilament-positive fibers (green) and TH-positive fibers (red) in a DLB plus AD case. This section is adjacent that shown in (f). Neurofilament- and TH-positive thick fibers (yellow) appear to be phosphorylated α-synuclein-positive fibers shown in (f). (h) Neurofilament- and TH-positive fibers in the cardiac sympathetic nerve in a normal control case. Bars, 30 µm. Reprinted with the permission of Wiley-Blackwell from Yokota et al.12

DISCUSSION

Frontotemporal lobar degeneration

The clinical diagnosis of Case 1 was SD, which is characterized by fluent speech output and loss of conceptual knowledge, resulting in loss of understanding of nominal terms, impaired recognition of faces and common objects, and impaired comprehension. This clinical phenotype can develop when the temporal poles are severely involved but other regions are relatively spared.

In a classification of FTD by the Lund and Manchester groups,19 a prototype of the consensus criteria of FTLD,14 three pathological bases of FTLD were divided into the frontal lobe degeneration type, Pick type, and motor neuron disease type. However, this pathological classification is not used now in neuropathological research because many novel disease entities defined by accumulated disease-specific proteins have been proposed. Consequently, the most recent pathological classification proposed in 200715 includes Pick's disease (defined by tau-positive Pick bodies), FTLD-U (ubiquitin-positive tau-negative inclusions, most of which are TDP-43 positive),20 neuronal intermediate filament inclusion disease (neurofilament-positive, α-internexin-positive, and tau-negative inclusions),21–23 and basophilic inclusion body disease (p62-positive, tau-negative, and TDP-43-negative spherical basophilic inclusions).23,24 Severe, circumscribed frontotemporal atrophy, which was regarded as a hallmark of the ‘Pick type’ in the Lund–Manchester criteria, usually develops in all these diseases. Of FTLD-U cases, 5–10% have mutations in the progranulin (PGRN) gene. In addition, several tauopathies in addition to Pick's disease, including corticobasal degeneration (CBD), progressive supranuclear palsy (PSP), argyrophilic grain disease,25 and FTLD with microtubule-associated protein tau gene mutation (FTLD with MAPT mutation), have been included in this classification. Although argyrophilic grain disease usually causes only mild temporal atrophy, some cases exhibit severe atrophy at the site.26–28 The category of dementia lacking distinctive histologic features (DLDH)29 has remained in this classification; however, because a large proportion of DLDH (approximately 90%) is actually FTLD-U,30 it is considered that the frequency DLDH is far lower than believed previously. In fact, Munoz et al.31 reported that the frequency of DLDH was only 6% of clinical FTLD. No Japanese case of FTLD-U with mild cortical atrophy has been reported to date.

Several epidemiological studies have demonstrated that FTLD-U is the most common pathology among FTLD cases, and the frequency of FTLD-U ranges from 30% to 40%.31,32 The second most common pathology may be Pick's disease or CBD, although the results are inconsistent among the series examined.32,33 The clinical characteristics of FTLD-U have been examined mainly in familial cases with PGRN mutations. The most common clinical presentation in FTLD-U with PGRN mutations is FTD.34–36 In addition, several features that are thought to be uncommon in FTLD were noted in FTLD with PGRN mutations, such as forgetfulness or memory loss,36,37 visuospatial dysfunction,37–39 asymmetric paraparesis,37 apraxia,36,38 and visual hallucination,36 which are more suggestive of underlying AD, CBD, or DLB. The CBD-like symptoms (e.g. asymmetric parkinsonism, pyramidal signs, and/or apraxia)40 and PSP-like symptoms41 have also been observed in sporadic FTLD-U. Severe amnesia has also been noted in sporadic FTLD-U.42 Conversely, whether the clinical features of sporadic FTLD-U are similar to those of familial FTLD-U remains unclear. The CBD-like symptoms, including alien-hand sign and early memory impairment, are also encountered in other rare pathological substrates of FTLD, such as basophilic inclusion body disease and neuronal intermediate filament inclusion disease.23

It is of note that although some FTLD-U cases present clinically with SD,11,43 the most common clinical phenotype in FTLD-U with PGRN mutations was reported to be FTD.35 Beck et al.44 also noted that language output impairment, but not impaired semantic memory or impaired comprehension, was common in FTLD-U with PGRN mutations. Like FTLD-U with PGRN mutations, it was reported that Pick's disease patients frequently showed speech output impairment.45 In contrast, to our knowledge there is no previous case of CBD that exhibited SD. Because the cerebral atrophy is usually more pronounced in the convexity of the frontoparietal region in CBD, the low prevalence of SD seems to be reasonable. Considering the possibility that therapies targeting specific histopathologies will be developed in the future, it may be increasingly important to understand the clinical and histopathological characteristics and their relationship in each pathological disease entity of FTLD.

Dementia with Lewy bodies

Approximately 50 years after the first description of Lewy bodies in the nucleus basalis of Meynert and dorsal vagal nucleus by Lewy,46 in 1961 Okazaki et al.47 described two atypical demented cases. As noted in the title ‘Diffuse intracytoplasmic ganglionic inclusions (Lewy type) associated with progressive dementia’, the outstanding pathological feature in these cases was the occurrence of numerous Lewy bodies in the cerebral cortex in addition to the brain stem nuclei. To our knowledge, this was the first report of DLB characterized by many Lewy bodies in the cerebral cortex. Fifteen years later, in 1976, Kosaka et al.48,49 again noticed the association between cortical Lewy bodies and the development of dementia. However, although over 40 years have passed since the first cases were reported, the clinical and pathological features in DLB have not been fully clarified.

The most important issue associated with the accuracy of the clinical diagnosis of DLB may be the fact that, as observed in Case 2 in this paper, Lewy pathology frequently overlaps Alzheimer pathology in the same patient. For example, Uchikado et al.50 demonstrated that, in a large series of pathologically confirmed AD cases (n = 347), only approximately 40% had pure AD, whereas 15% had AD with Lewy pathology consistent with diffuse neocortical-type DLB, 9% had AD with transitional-type DLB, and 0.9% had AD with brain stem-type DLB (i.e. PD). Further, the proportion of AD patients having Lewy bodies in the amygdala was 18%, whereas that of AD with Lewy pathology in the brain stem and/or limbic system but not in the amygdala was 18%. Now, when a case with extensive Alzheimer pathology (i.e. NFT Braak Stage V or VI and probable or definite AD according to the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) criteria) fits the pathological criteria of DLB, the case is usually called ‘AD plus DLB’ or ‘Lewy body variant AD (LBV/AD)’.51,52 Marui et al.53,54 and Iseki55 have proposed an original system for staging Lewy pathology and the classification of LBD. According to their classification system, our Case 2 can be classified as ‘AD form of DLB’.

In clinical practice, it is a critical issue that the clinical presentation of DLB is affected by concomitant Alzheimer pathology and that the clinical presentation of DLB may not be as uniform as believed previously.56 For example, concurrent Alzheimer pathology in DLB cases correlates with less frequent parkinsonism,57,58 less frequent visual hallucinations,59–61 less frequent delusions,59 and less frequent cognitive fluctuation.59 The sensitivity of the diagnostic accuracy of DLB with severe Alzheimer pathology is lower than that of DLB with mild Alzheimer pathology (22% vs 70%).62 Indeed, our Case 2, which had severe Alzheimer pathology, did not show cognitive fluctuation, visual hallucinations, or parkinsonism.

Why the clinical presentation of DLB cases with and without severe Alzheimer pathology is different remains unclear. However, it is considered that the progression pattern of Lewy pathology as well as neuronal loss in at least some DLB cases may be different from that in PD cases, and the pattern may be influenced by Alzheimer pathology. For example, according to the staging system for Lewy pathology by Braak et al.,63–66 Lewy pathology progresses from the medulla oblongata to the cerebral cortex (a bottom-up or caudorostral progression pattern) and the stages are divided into: (i) preclinical (Stages 1–2); (ii) early (Stages 3–4; 35% with clinical PD); and (iii) late (Stages 5–6; 86% with clinical PD).67 However, Gomez-Tortosa et al.68 noticed that DLB does not occur as a severe or long-lasting PD because paralimbic Lewy pathology was significantly correlated with that in the neocortex, but neither was correlated with Lewy pathology in the substantia nigra. Also in a series of LBD from a brain bank in the UK, only approximately 53% of 71 cases fit the caudorostral progression pattern.69 Jellinger57 noticed that among AD cases having Lewy pathology, approximately 70% did not show lesions in the medulla oblongata. Saito et al.70 also noticed the possibility that α-synuclein pathology begins in the amygdala if associated with Alzheimer pathology. Likewise, Yamamoto et al.71 noted that Lewy pathology was distributed predominantly in the cerebrum rather than in the brain stem in DLB with AD. Because α-synuclein pathology also frequently develops in the amygdala in several tauopathies, it is considered that the accumulation of α-synuclein at this site is associated with abnormal tau accumulation.72,73 Conversely, Tsuboi and Dickson74 reported frequent involvement of the amygdala by Lewy pathology and a relatively spared neuronal population in the substantia nigra in DLB cases with mild Alzheimer pathology (mean Braak stage 2.8 ± 0.6) compared with PD cases. These findings suggest that ‘the top-down progression pattern’ of degeneration (i.e. from the cerebrum to the brain stem) is not always associated with severe Alzheimer pathology. Zaccai et al.75 also noted that the progression patterns of Lewy pathology were not associated with the severity of Alzheimer pathology. Interestingly, the first two cases of DLB reported by Okazaki et al.47 clinically exhibited progressive dementia but lacked any parkinsonism stigmata. Further, these cases had numerous Lewy bodies in the cerebral cortex and brain stem nuclei but lacked depigmentation in the substantia nigra. That is, the clinical and pathological features in these cases may be consistent with those in the ‘cerebral-type DLB’ proposed by Kosaka.76 Unfortunately, the presence or absence of neurofibrillary changes in these cases was not described. Although cerebral-type DLB was regarded as very rare, unexpectedly, Zaccai et al.75 recently reported that six of 76 LBD cases (7.9%) showed neocortex-predominant Lewy pathology with very limited involvement of the amygdala and substantia nigra (isolated Lewy bodies or a few Lewy neurites). They called this pathology the ‘cortical form of LBD’.

Somewhat surprisingly, although several studies have demonstrated a correlation between the stage of Lewy pathology and the development of dementia and parkinsonism,77,78 the impact of Lewy pathology on the development of clinical symptoms remains to be elucidated. For example, parkinsonism is correlated with neuronal loss in the substantia nigra, but not with Lewy pathology at this site.79 The fact that some of the cognitively spared individuals have many cortical Lewy bodies has also been noticed repeatedly.64,80,81 Recently, Parkkinen et al.82 reported that 56% of cases in a Finnish autopsy series having widespread cortical Lewy bodies in the limbic system or neocortex did not exhibit either cognitive impairment or parkinsonism. Furthermore, in that series, only 48% of cases of diffuse neocortical type DLB with only mild Alzheimer pathology had dementia and only 54% displayed parkinsonism. Thus, the authors raised the question of whether assessment of Lewy pathology is an assessment of the actual disease process.82

The development of Lewy pathology in the autonomic nervous system was first described in 1961.83 In 1989, Wakabayashi84 first described Lewy pathology in the cardiac sympathetic nerve. Recently, Orimo et al.85–87 reported that all cases of DLB, LBV/AD, and PD showed almost complete loss of nerve fibers in the cardiac sympathetic nerve. In addition, Fujishiro et al.88 disclosed that in PD cases, the cardiac sympathetic nerve denervation correlated with the Braak PD stage as well as with the Hoehn and Yahr clinical stage.89 It is considered that the reduction in the number of these fibers can be evaluated by the decreased cardiac uptake on [123I]-metaiodobenzylguanidine (MIBG) myocardial scintigraphy in the early stage of DLB and PD.90,91 However, as demonstrated in the present paper, whether the cardiac sympathetic nerve in LBV/AD always shows severe degeneration in the early stage remains unclear.12

At present, whether the progression pattern of Lewy pathology is identical among DLB, LBV/AD, and PD remains unclear and the clinical presentation of DLB may be different from that in LBV/AD. Further, it is plausible that the clinical difference between DLB and LBV/AD results in differences in their reported frequencies between neurological departments and psychiatric departments. Thus, I believe that clinicopathological examinations in case series in various clinical settings, including psychiatric hospitals, are necessary for a better understanding of the clinical features in LBD.

ACKNOWLEDGMENTS

The author thanks Dr Koshiro Fujisawa (Department of Psychiatry, Yokohama Hoyu Hospital) for insightful comments on cases, Dr Kuniaki Tsuchiya (Department of Laboratory Medicine and Pathology, Tokyo Metropolitan Matsuzawa Hospital) for an insightful review of the manuscript, Dr Hiroshi Ujike (Department of Neuropsychiatry, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences) for the genetic analysis, Dr Toshiki Uchihara (Department of Neurology, Tokyo Metropolitan Institute for Neuroscience) for the immunohistochemical examination of Case 2, Ms Hiromi Kondo (Department of Neuropathology, Tokyo Institute of Psychiatry) for her excellent technical assistance, and Mr Atsushi Sasaki for help with the production of the manuscript. The author's work reported herein was supported by a Grant-in-Aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology (14570957) and a research grant from the Zikei Institute of Psychiatry.

Ancillary