Acta Neuropathologica

, Volume 106, Issue 2, pp 143–149

Distribution of astrocytic plaques in the corticobasal degeneration brain and comparison with tuft-shaped astrocytes in the progressive supranuclear palsy brain

Authors

    • Department of Neurology and NeuroscienceNagoya City University Graduate School of Medical Science
    • Department of Internal Medicine and Molecular ScienceNagoya City University Graduate School of Medical Science
  • Yoshio Hashizume
    • Department of Neuropathology, Institute for Medical Sciences of AgingAichi Medical University
  • Mari Yoshida
    • Department of Neuropathology, Institute for Medical Sciences of AgingAichi Medical University
  • Yasushi Iwasaki
    • Department of NeurologyNagoya University Graduate School of Medicine
  • Nozomi Hishikawa
    • Department of NeurologyNagoya University Graduate School of Medicine
  • Ryuzo Ueda
    • Department of Internal Medicine and Molecular ScienceNagoya City University Graduate School of Medical Science
  • Kosei Ojika
    • Department of Neurology and NeuroscienceNagoya City University Graduate School of Medical Science
Regular Paper

DOI: 10.1007/s00401-003-0711-4

Cite this article as:
Hattori, M., Hashizume, Y., Yoshida, M. et al. Acta Neuropathol (2003) 106: 143. doi:10.1007/s00401-003-0711-4

Abstract

Corticobasal degeneration (CBD) and progressive supranuclear palsy (PSP) have some clinical and pathological features in common. Each, however, has been shown to have specific astrocytic inclusions. CBD is characterized by astrocytic plaques, and PSP is characterized by tuft-shaped astrocytes. To clarify differences between these inclusions, we investigated intracerebral distribution of astrocytic plaques and tuft-shaped astrocytes in autopsied brains of patients with either CBD or PSP. Specimens from ten patients with CBD and five patients with PSP were stained by the Gallyas-Braak method. Densities of the astrocytic plaques and tuft-shaped astrocytes were determined for 11 cerebral cortical regions, 6 subcortical nuclei, 5 brain stem regions, the cerebellar cortex and the dentate nucleus. Astrocytic plaques were most abundant in the prefrontal and premotor areas in the cerebral cortex of CBD brains, whereas tuft-shaped astrocytes were most prominent in the precentral gyrus and premotor area of PSP brains. Many astrocytic plaques were observed in the caudate nucleus of CBD brains, whereas tuft-shaped astrocytes were abundant in both the caudate and putamen and moderate in number in the globus pallidus, subthalamic nucleus and thalamus in PSP brains. Very slight accumulation of astrocytic plaques was seen in the brain stem of CBD brains, whereas numerous tuft-shaped astrocytes were found in the red nucleus and superior colliculus of PSP brains. Distribution of the astrocytic plaques and tuft-shaped astrocytes differed greatly. Thus, CBD and PSP can be considered different entities.

Keywords

Corticobasal degenerationProgressive supranuclear palsyTauAstrocytic plaqueTuft-shaped astrocyte

Introduction

Corticobasal degeneration (CBD) and progressive supranuclear palsy (PSP) are rare, sporadic neurodegenerative disorders, and both are characterized by neuronal and glial cytoplasmic inclusions caused by accumulation of the highly phosphorylated microtubule-associated protein tau [5, 9]. These two neurodegenerative diseases, along with Pick's disease and Alzheimer's disease, have been grouped under the term tauopathies.

Since CBD was originally reported by Rebeiz et al. in 1967 [25], it has been considered the most important disease in the differential diagnosis of PSP. CBD is characterized by asymmetric rigid-akinetic parkinsonism and a variety of cerebral cortical symptoms including apraxia, aphasia and alien hand syndrome [12, 19, 20]. The major postmortem feature of CBD is degeneration of the cerebral cortex and subcortical nuclei, especially of the substantia nigra. The diagnostic feature most emphasized is the presence of ballooned achromatic neurons [5, 12, 18, 25]. PSP, as described by Steele et al. in 1964 [28], is characterized by supranuclear gaze palsy, muscle rigidity, truncal dystonia and instability upon standing and walking [13, 20]. The neuropathology of PSP includes degeneration of the substantia nigra, globus pallidus, subthalamic nucleus and the dentate nucleus [5, 18]. Although the diseases have been considered discrete clinical entities, the accumulation of atypical cases with overlapping features has led to controversy over the distinctiveness of CBD versus PSP [3, 8, 11, 14, 23, 24]. Furthermore, advances in immunohistochemistry and Gallyas-Braak staining have revealed similar tau-containing inclusion bodies in glial cells and neurons in association with both CBD and PSP [6, 29, 32].

Komori et al. [15], however, showed that astrocytic plaques and tuft-shaped astrocytes, the two morphologically different types of astroglial inclusions found in CBD and PSP brains, never coexist in a single individual. They found that all patients with astrocytic plaques also had many ballooned neurons; these patients were diagnosed pathologically with CBD. Komori et al. [15] concluded that astrocytic plaques are well correlated with neuronal degeneration associated with ballooned neurons and that they are more specific and sensitive than neuronal pathology for diagnosis of CBD. The current hypothesis is that the astrocytic plaque is a disease-specific hallmark of CBD and that the tuft-shaped astrocyte is the hallmark of PSP [5, 6, 8, 11, 15]. Neuropathological studies of diagnostically difficult cases have revealed that CBD can show extremely diverse symptoms, more diverse than has been thought in the past [3, 8, 14, 31].

Despite pathological similarities between CBD and PSP such as tau deposits in the neurons and glial cells, the morphology of the astroglial inclusions of each disease differs clearly. However, the difference in localization of astrocytic inclusions between CBD and PSP has not been well described. Thus, to clarify differences between CBD and PSP, we investigated regional distribution of astroglial inclusions quantitatively.

Materials and methods

Cases

We studied the brains of ten consecutive patients (seven men, three women; mean age at death, 65.4 years; range, 54–83 years; mean duration of illness, 3.9 years) who were diagnosed with CBD according to the presence of astrocytic plaques and the brains of five patients (two men, three women; mean age at death, 76.6 years; range, 72–84 years; mean duration of illness, 5.2 years) who were diagnosed with PSP according to the neuronal pathology and presence of tuft-shaped astrocytes. All brains were provided by hospitals in Aichi prefecture, Japan. Cases are summarized in Table 1.
Table 1.

Summary of cases (CBD corticobasal degeneration, PSP progressive supranuclear palsy, ALS-D amyotrophic lateral sclerosis with dementia, DLBD diffuse Lewy body disease, a asymmetrical, AH alien hand, Ag agnosia, Ap apraxia, Ak akinesia, Dem dementia, Dep depression, Dys dystonia, I instability on standing, M memory disturbance, P personality change, R rigidity, T tremor, V vertical gaze palsy, W weakness, NA not available)

Disease

Case

Age/sex

Duration (years)

Clinical diagnosis

Brain weight (g)

Chief symptoms

CBD

1

83/F

6

CBD

955

AH, V, a(Dys, R, W, T)

2

65/M

4

PSP

1,315

V, I, Ak, R

3

67/F

3

PSP

1,080

V, I, Ak, R

4

55/M

4

Parkinsonism

1,215

Dep, truncal Dys, Ak

5

61/M

3

CBD

1,250

AH, Ag, Ap, a(W, R, Ak, Dys)

6

69/M

4

PSP

NA

P, V, M, Ak

7

72/M

3

Parkinsonism

1,220

M, R

8

69/M

4

PSP

1,110

P, Ak, I, R

9

59/M

5

ALS-D

1,260

aW, Dem

10

54/F

3

DLBD

1,400

V, I, P

PSP

11

73/M

5

PSP

1,210

I, V, Ak, R, Dem, neck Dys

12

84/F

4

PSP

1,090

P, I, V, R, Ak, neck Dys

13

81/M

6

PSP

1,150

V, Ak, I, neck R

14

72/F

6

PSP

1,080

I, Ak, T, neck R

15

73/F

5

PSP

1,130

V, Ak, I, neck R

Analysis of specimens

Specimens were fixed in 20% formaldehyde and embedded in paraffin. They were cut into 9-μm slices and stained by the Gallyas-Braak method [4]. Astroglial inclusions were identified by their shape. Astrocytic plaques were identified as fuzzy short argyrophilic processes arranged annularly with fine collaterals at vertical or sharp angles [5, 8, 15]. Tuft-shaped astrocytes were observed as conglomerated processes arranged radially [5, 15]. Typical appearances of both structures are shown in Fig. 1. The astrocytic plaques in CBD specimens and tuft-shaped astrocytes in PSP specimens were observed by the same examiner: moderate magnification (×100 or ×200) was used, and inclusions were marked with a pen on the glass slide. Structures that were cut at their edge and lacking a center were excluded. The astrocytic plaques and tuft-shaped astrocytes were counted, and the densities of each type of inclusion were determined in the following 11 regions of the cerebral cortex: the precentral gyrus, premotor area, convexity and orbital surfaces of the prefrontal area, postcentral gyrus, parietal association area, hippocampus/parahippocampus, temporal lobe, visual area, cingulate gyrus and insula. The densities were calculated as "line densities" by dividing the number of astrocytic inclusions by the length of curved lines tracing the mean depth of the cerebral cortex. Area densities were calculated by dividing the number of inclusions by the area of each region in 6 subcortical nuclei (putamen, caudate, external and internal segments of the globus pallidus, subthalamic nucleus and thalamus), 5 regions of the brain stem (substantia nigra, red nucleus, superior colliculus, pontine nucleus and inferior olivary nucleus), the cerebellar cortex and the dentate nucleus. Images of the areas were viewed on a Macintosh computer. An Epson color image scanner and NIH-Image software were used. The length or area of each region was determined. Average densities were calculated from as many slices as possible for each brain area; at least two or three slices were used. Cerebral cortices, which were cut tangentially to the brain surface, were excluded from measurement for the sake of accuracy.
Fig. 1.

Typical appearances of astrocytic plaques (arrow) (a) and tuft-shaped astrocytes (arrows) (b) stained by the Gallyas-Braak method. Bars 50 μm

To minimize the influence on distribution data of large individual dispersion in absolute density, the density of each region relative to that of the base regions consistently showing astrocytic inclusions in both diseases, i.e., the premotor area in the cerebral cortex or the caudate in the subcortex, the brain stem and the cerebellum, was calculated.

Distribution patterns especially in the frontal lobe and striatum were compared statistically between CBD and PSP by densities of the precentral gyrus and convexity of the prefrontal area relative to density of the premotor area and by density of the putamen relative to that of the caudate. Statistical analysis was performed by Mann-Whitney U test.

Results

The densities of astrocytic inclusions in each region of CBD brains and PSP brains are shown in Fig. 2a–d. The original density data and relative densities are shown in Table 2. Distribution patterns expressed by mean densities relative to densities of the base regions are shown in Fig. 2e, f.
Fig. 2.

Densities and distributions of astrocytic inclusions. Individual cases are shown by dotted lines, and averages are shown by solid lines (ad). Distributional patterns of astrocytic plaques and tuft-shaped astrocytes (e, f). Filled bars (astrocytic plaques) and empty bars (tuft-shaped astrocytes) indicate mean density of each region relative to that of the premotor area or caudate in which these inclusions are observed most consistently in both diseases

Table 2.

Densities of the astrocytic inclusions in each region. Underlines indicate relative densities. Densities are expressed per cm in the cerebral cortex, and per cm2 in the subcortical nuclei, brain stem or cerebellum (PreC precentral gyrus, PM premotor area, PF-C prefrontal convexity, PF-O orbital surface of prefrontal area, PostC postcentral gyrus, PA parietal association area, Hip hippocampus/parahippocampal gyrus, Temp temporal lobe, VA visual area, Cing cingulate gyrus, Ins insula, Pt putamen, Ca caudate, GPe external segment of globus pallidus, GPi internal segment of globus pallidus, ST subthalamic nucleus, Th thalamus, SN substantia nigra, RN red nucleus, SC superior colliculus, Po pontine nucleus, IO inferior olive, CC cerebellar cortex, De dentate nucleus, AP astrocytic plaque, Tu-SA tuft-shaped astrocyte, NA not available)

Disease

Case

Line-densities of astrocytic plaques (/cm)

Area-densities of astrocytic plaques (/cm2)

PreC

PM

PF-C

PF-O

PostC

PA

Hip

Tmp

VA

Cin

Ins

Pt

Ca

GPe

GPi

ST

Th

SN

RN

SC

Po

IO

CC

De

CBD

1

1.01

20.56

15.97

8.86

4.27

19.76

0.00

5.10

0.67

16.15

17.08

0.00

56.25

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.05

1.00

0.78

0.43

0.21

0.96

0.00

0.25

0.03

0.79

0.83

0.00

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

2

0.00

0.80

1.61

2.04

0.00

0.11

0.21

0.15

0.00

0.26

0.98

22.68

139.56

0.00

0.00

0.00

0.00

0.00

0.00

2.13

0.00

0.00

0.00

0.00

0.00

1.00

2.00

2.54

0.00

0.14

0.26

0.18

0.00

0.32

1.22

0.16

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.02

0.00

0.00

0.00

0.00

3

0.17

8.35

12.98

4.76

0.34

0.08

0.29

0.72

0.00

4.09

0.76

3.04

22.17

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.02

1.00

1.56

0.57

0.04

0.01

0.03

0.09

0.00

0.49

0.09

0.14

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

4

0.00

1.44

1.93

1.89

NA

0.38

0.00

0.15

0.00

1.19

0.46

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

1.00

1.34

1.31

0.26

0.00

0.11

0.00

0.83

0.32

5

0.06

0.58

0.66

0.51

0.41

0.00

0.00

0.00

0.00

0.12

0.67

0.48

5.68

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.10

1.00

1.13

0.87

0.71

0.00

0.00

0.00

0.00

0.21

1.15

0.08

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

6

0.11

0.89

1.03

0.38

0.05

0.00

0.00

0.06

0.00

0.31

0.00

0.19

7.69

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.13

1.00

1.16

0.43

0.05

0.00

0.00

0.07

0.00

0.35

0.00

0.03

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

7

0.83

13.59

3.00

4.17

1.03

1.61

0.25

0.59

NA

3.58

1.29

5.14

101.18

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.06

1.00

0.22

0.31

0.08

0.12

0.02

0.04

0.26

0.10

0.05

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

8

0.59

3.46

NA

NA

0.00

NA

0.00

0.13

NA

0.94

1.12

2.01

38.55

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.17

1.00

0.00

0.00

0.04

0.27

0.32

0.05

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

9

1.81

5.44

4.98

5.19

0.17

0.14

0.20

0.06

0.29

0.76

0.15

2.97

17.91

0.00

0.00

0.00

1.61

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.33

1.00

0.92

0.95

0.03

0.03

0.04

0.01

0.05

0.14

0.03

0.17

1.00

0.00

0.00

0.00

0.09

0.00

0.00

0.00

0.00

0.00

0.00

0.00

10

13.53

19.06

20.83

9.22

1.73

7.62

0.41

5.14

NA

7.24

3.36

2.37

13.07

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.71

1.00

1.09

0.48

0.09

0.40

0.02

0.27

0.38

0.18

0.18

1.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Average

1.81

7.42

7.00

4.11

0.89

3.30

0.14

1.21

0.14

3.46

2.59

3.89

40.21

0.00

0.00

0.00

0.16

0.00

0.00

0.21

0.00

0.00

0.00

0.00

0.16

1.00

1.13

0.88

0.13

0.21

0.04

0.11

0.01

0.40

0.42

0.10

1.00

0.00

0.00

0.00

0.01

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Disease

Case

Line-densities of tuft-shaped astrocytes (/cm)

Area-densities of tuft-shaped astrocytes (/cm2)

PreC

PM

PF-C

PF-O

PostC

PA

Hip

Tmp

VA

Cin

Ins

Pt

Ca

GPe

GPi

ST

Th

SN

RN

SC

Po

IO

CC

De

PSP

11

14.83

20.38

9.35

1.95

10.78

2.23

0.00

0.20

0.48

0.62

0.30

101.88

141.18

66.00

11.11

80.00

101.23

2.78

245.16

248.00

0.00

20.00

0.00

4.26

0.73

1.00

0.46

0.10

0.53

0.11

0.00

0.01

0.02

0.03

0.01

0.72

1.00

0.47

0.08

0.57

0.72

0.02

1.74

1.76

0.00

0.14

0.00

0.03

12

59.19

24.45

7.94

1.58

4.63

1.91

0.15

1.45

2.32

2.04

0.47

144.23

175.45

23.19

20.83

50.00

37.80

9.26

23.08

411.11

0.00

13.04

0.00

6.33

2.42

1.00

0.32

0.06

0.19

0.08

0.01

0.06

0.09

0.08

0.02

0.82

1.00

0.13

0.12

0.28

0.22

0.05

0.13

2.34

0.00

0.07

0.00

0.04

13

62.39

59.87

4.42

1.42

3.27

3.07

0.40

0.39

0.00

1.66

2.41

131.64

82.72

32.00

26.67

16.67

76.82

15.15

144.44

150.00

0.00

14.71

0.00

9.43

1.04

1.00

0.07

0.02

0.05

0.05

0.01

0.01

0.00

0.03

0.04

1.59

1.00

0.39

0.32

0.20

0.93

0.18

1.75

1.81

0.00

0.18

0.00

0.11

14

13.41

10.88

2.98

1.10

5.83

1.58

0.00

0.13

0.00

1.19

2.34

348.15

282.26

253.85

83.33

50.00

84.81

12.50

323.68

272.09

0.00

4.00

0.00

12.50

1.23

1.00

0.27

0.10

0.54

0.15

0.00

0.01

0.00

0.11

0.21

1.23

1.00

0.90

0.30

0.18

0.30

0.04

1.15

0.96

0.00

0.01

0.00

0.04

15

7.57

22.84

2.40

1.99

3.75

2.40

0.00

0.66

0.00

1.43

1.49

80.13

101.56

23.53

17.65

200.00

27.75

23.53

93.33

100.00

0.00

23.08

0.00

33.33

0.33

1.00

0.11

0.09

0.16

0.11

0.00

0.03

0.00

0.06

0.07

0.79

1.00

0.23

0.17

1.97

0.27

0.23

0.92

0.98

0.00

0.23

0.00

0.33

Average

31.48

27.69

5.42

1.61

5.65

2.24

0.11

0.57

0.56

1.39

1.40

161.21

156.63

79.71

31.92

79.33

65.68

12.64

165.94

236.24

0.00

14.97

0.00

13.17

1.15

1.00

0.25

0.07

0.29

0.10

0.00

0.02

0.02

0.06

0.07

1.03

1.00

0.42

0.20

0.64

0.49

0.11

1.14

1.57

0.00

0.13

0.00

0.11

Cerebral cortex

Astrocytic plaques were most frequent in the prefrontal areas [average relative density, 1.13 (average density, 7.00/cm) on the convexity; 0.88 (4.11/cm) on the orbital surface] and premotor areas [1.00 (7.42/cm)] in the CBD brain (Fig. 2a, e). Density of the astrocytic plaques was much lower in the precentral gyrus [0.16 (1.81/cm)] than in the prefrontal and premotor areas. Densities were moderate in the cingulate gyrus [0.40 (3.46/cm)] and the insula [0.42 (2.59/cm)]. Densities were low in the postcentral gyrus [0.13 (0.89/cm)], parietal association area [0.21 (3.30/cm)] and temporal lobe [0.11 (1.21/cm)]. The average density in the parietal association area was influenced by the particularly high number of astrocytic plaques in case 1. Astrocytic plaques were very rare in the hippocampus [0.04 (0.14/cm)] and the visual cortical area [0.01 (0.14/cm)].

Tuft-shaped astrocytes were most abundant in the precentral gyrus [1.15 (31.48/cm)] and premotor area [1.00 (27.69/cm)] in PSP brains (Fig. 2c, e). Tuft-shaped astrocytes were much less frequent in the prefrontal area [0.25 (5.42/cm) on the convexity, 0.07 (1.61/cm) on the orbital surface] than in the posterior part of the frontal lobe. There was a small number of tuft-shaped astrocytes in the postcentral gyrus [0.29 (5.65/cm)]. Only a few tuft-shaped astrocytes were found in the parietal association area [0.10 (2.24/cm)], hippocampus [0.00 (0.11/cm)], temporal lobe [0.02 (0.57/cm)], visual area [0.02 (0.56/cm)], cingulate [0.06 ([1.39/cm)] and insula [0.07 (1.40/cm)].

The difference in distribution in the frontal lobe between astrocytic plaques and tuft-shaped astrocytes is shown in Fig. 3a, b. The precentral/premotor distribution ratios were significantly larger in PSP brains than in CBD brains (P<0.01) (Fig. 3a), and the prefrontal/premotor distribution ratios were significantly larger in CBD brains than in PSP brains (P<0.01) (Fig. 3b).
Fig. 3.

Comparison of distribution patterns in the frontal lobe and striatum between CBD and PSP. The ratios of the density in the precentral gyrus to that in the premotor area (a), of the prefrontal area to that of the premotor area (b) and of the putamen to that of the caudate (c). *p<0.01. (CBD corticobasal degeneration, PSP progressive supranuclear palsy, PreC precentral gyrus, PM premotor area, PF-C prefrontal convexity, Pt putamen, Ca caudate)

Subcortical nuclei, brain stem and cerebellum

The CBD brain had many astrocytic plaques in the caudate [1.00 (40.21/cm2)] (Fig. 2b, f). There were fewer astrocytic plaques in the putamen [0.10 (3.89/cm2)] than in the caudate, and astrocytic plaques were nearly absent in other subcortical and infra-tentorial regions. Tuft-shaped astrocytes in PSP brains were abundant in the caudate [1.00 (156.63/cm2)], putamen [1.03 (161.21/cm2)], red nucleus [1.14 (165.94/cm2)] and superior colliculus [1.57 (236.24/cm2)] and moderate in number in the globus pallidus [external segment, 0.42 (79.71/cm2); internal segment, 0.20 (31.92/cm2)], subthalamic nucleus [0.64 (79.33/cm2)] and thalamus [0.49 (65.68/cm2)] (Fig. 2d, f). A small number of tuft-shaped astrocytes was also observed in the substantia nigra [0.11 (12.64/cm2)], inferior olive [0.13 (14.97/cm2)] and dentate nucleus [0.11 (13.17/cm2)]. The putamen/caudate distribution ratio of astroglial inclusions was significantly lower in CBD brains than in PSP brains (P<0.01) (Fig. 3c). Neither CBD nor PSP brains had astroglial inclusions in the cerebellar cortex. The distribution of astrocytic plaques is shown schematically in Fig. 4.
Fig. 4.

Schematic illustration of the distribution of astrocytic plaques. Dots: astrocytic plaques, asterisk: precentral gyrus, double asterisk: postcentral gyrus

Discussion

We performed quantitative analysis of the density of astrocytic plaques in CBD brains and of tuft-shaped astrocytes in PSP brains and showed distinct distribution patterns characterizing each disease. Distribution of tuft-shaped astrocytes in PSP brains was previously reported [22, 30], and our findings accord with the reported findings. We believe the present study is the first such investigation of astrocytic plaques in CBD brains.

Although astroglial inclusions were most abundant in the frontal lobe in both CBD and PSP brains, the patterns of distribution differed (Fig. 2a, c, e). Astrocytic plaques were more prominent in the anterior than in the posterior part of the frontal lobe in CBD brains. Tuft-shaped astrocytes were more frequent in the posterior than in the anterior part of frontal lobe in PSP brains. CBD brains had small to moderate number of astrocytic plaques in the temporal neocortex, cingulate gyrus and insula, whereas PSP brains had only a few tuft-shaped astrocytes in these cortical regions. This wider distribution of astrocytic pathology in the cerebral cortex distinguishes CBD from PSP if the symptoms are mainly due to neuronal pathology (Fig. 2a, c).

In contrast to the astrocytic plaques, the tuft-shaped astrocytes of PSP showed widespread distribution in the subcortex and brain stem. Characteristic differences in the distribution patterns between CBD and PSP were also apparent, especially in the striatum (Fig. 2b, d, f). Although the putamen and caudate are generally considered to have the same cellular structure and function, densities of astrocytic plaques in the putamen and caudate were quite different in CBD brains. The reason for this difference is unclear, but different severities of degeneration in other cerebral regions connecting to the striatum might reflect the heterogeneity in density of astrocytic plaques in the striatum. Topographical observation on projection systems may provide further insight into this issue.

Another notable difference is that CBD brains showed marked variation in actual densities between individuals (Fig. 2a, b) but all PSP brains were relatively stereotypical (Fig. 2c, d). More than a few patients with CBD were clinically diagnosed with PSP, parkinsonism or another disorder, whereas all patients with PSP were diagnosed correctly with PSP. This clinical heterogeneity of CBD has often been described [8, 14, 15, 31]. Only two cases with astrocytic plaques (cases 1 and 5) showed typical clinical symptoms of CBD such as alien hand, agnosia or apraxia. Case 1 had the largest number of astrocytic plaques in the cerebral cortex, whereas case 5 had very few (Fig. 2a). Therefore, amounts of astrocytic plaque are considered less responsible for clinical profiles than differences in neuronal degeneration.

A diagnosis of CBD or PSP depends on the findings emphasized, clinical features [17, 19], conventional pathology based on neuronal loss or atrophy [18] or cytoskeletal abnormalities visualized by immunohistochemistry or Gallyas-Braak staining. Previous reports describing pathologically atypical PSP with astrocytic plaques or without tuft-shaped astrocytes [16, 22, 23] resulted from diagnoses that were not based on cytoskeletal pathology. Li et al. [16] described a clinically typical case of PSP with dementia and diagnosed the case pathologically as PSP despite the presence of ballooned neurons, astrocytic plaques and marked fronto-temporal cortical atrophy. By contrast, Shiozawa et al. diagnosed a case of clinical PSP as CBD, emphasizing the existence of ballooned neurons and astrocytic plaques to support their diagnosis [27].

These contrasting studies reflect the inconsistent diagnostic criteria present in the literature. This combined use of clinical and pathological criteria complicates any discussion about the distinctiveness of CBD and PSP. Furthermore, the sensitivity of the clinical diagnosis of CBD is at present considered extremely low [19]. In this study, we diagnosed CBD simply by the presence of astrocytic plaques. This criterion itself may be a subject of argument, but it seems clear that there are two distinct groups of tauopathies, one with astrocytic plaques and the other with tuft-shaped astrocytes, and that the astrocytic plaque is highly sensitive and specific for CBD [15]. Hence, we support the notion that CBD can be defined as a discrete tauopathy characterized by the existence of astrocytic plaques and that it can show a variety of symptoms and often mimics PSP or other neurodegenerative diseases.

In addition to clinical and pathological similarities, recent studies showed the biochemical similarity of CBD and PSP by showing that the tau isoform with four microtubule-binding repeats (4Rtau) accumulates in both CBD and PSP brains [1, 21, 26]. Analysis of the tau gene haplotype revealed that CBD shares a common genetic background with PSP [7]. The biochemical and genetic similarities between CBD and PSP indicate that the disorders share a common tau-aggregation mechanism. However, the final discrepancies in shape and distribution of astrocytic plaques and tuft-shaped astrocytes point to other unknown factors modulating the glial pathology. There is a morphological difference in the localization of abnormal tau. Astrocytic plaques involve the distal part of the astroglial processes, and the tuft-shaped astrocytes involve the perikarya and proximal processes. This difference implies a difference in expression, processing or degradation of the tau molecule. Some differences in the intracellular processing of tau have been shown immunohistochemically and biochemically [2, 10]. In addition, the different intracerebral distributions may indicate that different neuronal projection systems are involved. We did not clarify the relation between neurodegeneration and glial pathology in the present study. According to a previous report of the distribution of cortical lesions in CBD [31], neuronal and glial pathology do not seem parallel. Further research on these factors might provide further insight into the pathogenesis of CBD and PSP.

In conclusion, the distributions of astroglial inclusions in CBD and PSP brains were analyzed quantitatively in this study. Despite some clinical and pathological similarities between the two diseases, the morphology and distribution of astrocytic inclusions are quite distinct. These apparent differences lead us to believe that CBD and PSP are distinct clinical entities that differ in pathogenesis.

Copyright information

© Springer-Verlag 2003