Voxel-based comparison of regional cerebral glucose metabolism between PSP and corticobasal degeneration

Journal of the Neurological Sciences 199 (2002) 67 – 71 www.elsevier.com/locate/jns

Voxel-based comparison of regional cerebral glucose metabolism between PSP and corticobasal degeneration Kayo Hosaka a,b, Kazunari Ishii a,*, Setsu Sakamoto a, Tetsuya Mori a, Masahiro Sasaki a, Nobutsugu Hirono c, Etsuro Mori c a

Division of Imaging Research, Hyogo Institute for Aging Brain and Cognitive Disorders, 520 Saisho-Ko, Himeji, Hyogo 670-0981, Japan b Department of Radiology, Kobe University School of Medicine, Kobe, Japan c Division of Neurosciences, Hyogo Institute for Aging Brain and Cognitive Disorders, Himeji, Japan Received 27 February 2002; accepted 9 April 2002

Abstract Objectives: Progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD) are neurodegenerative disorders that may be accompanied by dementia and parkinsonism as clinical symptoms. The purpose of this study was to elucidate cerebral metabolic differences of these two diseases with cognitive impairments by [18F] fluorodeoxyglucose (FDG) and positron emission tomography (PET). Methods: A total of 12 patients with PSP (age: 62.8 F 6.0 years old, m: 7, f: 5, Mini-Mental State Examination (MMSE): 23.4 F 2.6), 12 patients with CBD (age: 64.8 F 6.3 years old, m: 6, f: 6, MMSE: 22.9 F 4.5), and age-matched healthy subjects (normal control (NC)) (age: 63.8 F 7.7 years old, m: 7, f: 5) were subjected to FDG-PET to obtain glucose metabolic images. We compared regional cerebral metabolic images by a voxel-by-voxel analysis with statistical parametric mapping (SPM) among PSP, CBD, and NC subjects, and evaluated differences of hypometabolic regions. Results: The patients with PSP showed reduced cerebral glucose metabolism in the medial and lateral frontal gyri, basal ganglia, and midbrain compared with NC, whereas the patients with CBD showed significant reduction in the parietal lobes ( p < 0.001). SPM also revealed parietal hypometabolism in CBD patients compared with PSP patients ( p < 0.001). Conclusions: The predominant parietal glucose metabolic reduction in CBD patients was different from previously reported findings. This finding would be the characteristic substance of patients with CBD accompanying cognitive impairments. Our findings suggest that measurement of glucose metabolism by PET and a voxel-based analysis is useful to understand the pathophysiology of these two diseases with cognitive impairments. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Progressive supranuclear palsy; Corticobasal degeneration; Positron emission tomography; Glucose metabolism; Voxel-by-voxel analysis

1. Introduction Progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD) are neurodegenerative disorders of middle and late age presenting with dementia and parkinsonism. The neurological features of PSP included impaired ocular motility, pseudobulbar palsy, and axial dystonia, while those of CBD include asymmetric rigidity of the limbs, myoclonus, alien limb sign, and localized cortical signs such as apraxia or cortical sensory loss. The clinical syndromes of typical cases of PSP and CBD are distinct, although atypical cases are described that have overlapping clinical and pathologic features. Several pathologic features, such as astrocytic

*

Corresponding author. Tel.: +81-792-95-5511; fax: +81-792-95-8199. E-mail address: [email protected] (K. Ishii).

lesions and ballooned neurons, differentiate PSP from CBD, although both diseases share several neuropathological features including filamentous tau inclusions in neurons and glia and biochemical alterations in the tau protein. Several clinical, pathological, and molecular features rationalize the concept of sporadic form of tauopathy. Nevertheless, there are sufficient differences to continue the present-day trend to consider PSP and CBD as separate disorders [1,2]. These two diseases are sometimes difficult to distinguish on clinical grounds. Functional neuroimagings such as positron emission tomography (PET) and single photon emission computed tomography (SPECT) may help to differentiate CBD from PSP. A number of studies with PET and SPECT have demonstrated cerebral metabolism or perfusion abnormalities in PSP and CBD, separately. CBD patients are characterized by an asymmetric hypometabolism mainly in the sensorimotor and parietal cortices and the thalamus,

0022-510X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 2 2 - 5 1 0 X ( 0 2 ) 0 0 1 0 2 - 8

68

K. Hosaka et al. / Journal of the Neurological Sciences 199 (2002) 67–71

while PSP patients are characterized by a global metabolic reduction mainly in the frontal lobes and the basal ganglia. However, only a few studies have made direct comparisons of the two diseases [3 –6], and there are some discrepancies in the results of these studies. The disagreement among the studies may be due to the use of groups of patients that were not matched in terms of age, cognitive impairment, or hemisphere that was most affected. The purpose of the present study was to delineate the pathophysiological difference between PSP and CBD by using fluorodeoxyglucose (FDG)-PET. By a voxel-by-voxel analysis, we compared two groups equivalent for age, gender, cognitive impairment, and hemisphere that was most affected.

Table 1 Clinical features of the patients with CBD and PSP

2. Methods

CBD: corticobasal degeneration; PSP: Progressive supranuclear palsy; MMSE: Mini-Mental State Examination; ADAS: Alzheimer’s Disease Assessment Scale.

2.1. Subjects A total of 12 patients with probable PSP diagnosed by NINDS-SPSP International Workshop [7] and 12 patients with probable CBD that fulfilled the criteria of the CBD Multicenter Case-Control Study (CBDMCCS) [8] were selected from patients who were admitted to our hospital for examination of cognitive impairment. The two groups were matched for age and scores on the Mini-Mental State Examination (MMSE) and Alzheimer’s Disease Assessment Scale (ADAS). The PSP patients consisted of seven males and five females with a mean age F standard deviation of 62.8 F 6.0 years. Their MMSE and ADAS scores were 23.4 F 2.6 and 18.6 F 7.4, respectively. The CBD patients consisted of six males and six females with a mean age of 64.8 F 6.3 years. Their MMSE and ADAS scores were 22.9 F 4.5 and 19.1 F 8.9, respectively. A group of 12 agematched healthy volunteers (mean age, 63.8 F 7.7 years) was used for a normal control (NC). All the subjects in this study were right-handed. Probable PSP requires the presence of a gradually progressive disorder with onset at age 40 or later, and with both vertical supranuclear gaze palsy and prominent postural instability occurring in the first year of onset, as well as no evidence of other diseases that could explain these features. The criteria for probable PSP are reportedly highly specific, making them suitable for therapeutic, analytic epidemiological, and biologic studies, but they are not very sensitive [8]. Although clinical diagnostic criteria for CBD have not been established, Litvan et al. [9] reported that the clinical diagnosis of CBD using characteristic clinical features had good specificity. The CBDMCCS criteria require the presence of rigidity plus one of apraxia, cortical sensory loss, or the alien limb phenomena at some time in the course of the disease. Alternatively, the presence of moderate-to-marked limb rigidity with a fixed dystonic posture and spontaneous and reflex myoclonus in asymmetrical fashion can fulfill the inclusion criteria. The clinical features of the patients with PSP and CBD are summarized in Table 1. The NC subjects showed no clinical evidence of

MMSE score ADAS score Limb rigidity Axial dystonia or rigidity Postural instability Apraxia Cortical sensory loss Alien limb Myoclonus Postural tremor Hyperreflexia Babinski’s sign Supranuclear ophthalmoplegia Pseudobulbar palsy

CBD (n = 12)

PSP (n = 12)

23.4 F 2.6 18.6 F 7.4 11 (92%) 2 (17%) 4 (33%) 6 (50%) 9 (75%) 7 (58%) 8 (67%) 5 (42%) 5 (42%) 3 (25%) 0 (0%) 0 (0%)

22.9 F 4.5 19.1 F 8.9 6 (50%) 12 (100%) 12 (100%) 0 (0%) 0 (0%) 0 (0%) 1 (8%) 0 (0%) 6 (50%) 3 (25%) 12 (100%) 10 (83%)

cognitive deficits or neurological disease and were taking no acute or chronic medications at the time of the scan. They had no abnormal findings on magnetic resonance (MR) images. Written informed consent was obtained for all of the patients and from all of the controls. The study protocol was approved by our institution’s ethical committee and the radiation protection authorities. Detailed PET and MRI procedures are described elsewhere [10,11]. In brief, before the PET scans, all the subjects received MR imaging for anatomical reference, for PET positioning, and to confirm that they had no abnormal conditions for each selection criteria. PET images were obtained using a scanner Headtome IV (Shimadzu, Kyoto, Japan) [12]. All subjects had fasted for at least 4 h before PET scanning. A transmission scan was performed using a 68Ga/68Ge pin source for attenuation correction. PET studies were performed with the subjects under resting conditions with eyes closed and ears unplugged, in a supine position, in a darkened and quiet room. A total of 60 min after injection of 185– 346 MBq of FDG, emission scanning was started and emission data were collected for 12 min. 2.2. Data analysis Asymmetries in glucose metabolism are common in CBD [3,5,13,14]. Therefore, the more affected (hypometabolic) hemisphere was fixed to the left side in order to lessen the variance due to asymmetry. Several circular regions of interest (ROIs) of 10-mm diameter were placed on the lateral frontal and lateral parietal cortices in each hemisphere, and the measurements of ROIs in each lobe were averaged [13]. The frontal and parietal metabolic asymmetry index was computed as 2
j(L  R)j/(L + R)
100 (%), where L and R refer to left- and right sided regional glucose metabolism, respectively. When either the frontal or parietal symmetry index exceeded the mean + 2 standard deviations of healthy subjects obtained in our previous study [13]

K. Hosaka et al. / Journal of the Neurological Sciences 199 (2002) 67–71

69

Table 2 Comparisons of PSP, CBD, and NC groups showing locations in which reductions in glucose metabolism are greater in one group than the other

NC > PSP

NC>CBD

CBD>PSP PSP>CBD

rt. cingulate gyrus rt. inferior frontal gyrus rt. caudate nucleus lt. inferior frontal gyrus lt. caudate nucleus midbrain lt. cuneus lt. inferior parietal lobule lt. middle frontal gyrus lt. post central gyrus rt. inferior parietal lobule midbrain rt. cingulate gyrus lt. middle occipital gyrus lt. inferior parietal lobule lt. precuneus

t

x

y

z

5.52 5.51 3.50 4.52 4.00 5.24 5.31 5.07 4.58 4.26 3.39 4.81 4.14 4.66 4.61 4.60

2  41  11 47 15 1 11 38 45 41  31 2 7 50 41 11

21 14 6 14 15  23  74  54 14  27  61  25 29  88  32  70

32 32 2 34 2 5 16 38 34 36 38 7 32 5 29 29

NC: normal control; CBD: corticobasal degeneration; PSP: progressive supranuclear palsy. Note: lt. indicates the affected side in CBD group.

Fig. 2. Specific voxels that have significantly lower metabolism in the CBD group than in the healthy control group at the threshold of p < 0.001, uncorrected.

(2.0 + 2
1.8 = 5.6% for the frontal lobe and 4.3 + 2
3.4 = 11.2% for the parietal lobe), we regarded the asymmetry to be positive, and made right – left mirror PET images to make the left hemisphere the more hypometabolic side if opposite. In addition, the mean of eight 10-mm diameter ROIs placed on the medial occipital cortices of the less affected side was obtained, and the medial occipital cortices were chosen for the reference region for normalization, as the occipital cortices are the less affected region both in PSP and CBD [5,13,14]. We used NEUROSTAT (University of Michigan, Ann Arbor, MI) for anatomical

standardization of PET images, as brain atrophy is often severe in CBD [15]. In our previous study [16], we found that NEUROSTAT was more suitable than Statistical Parametric Mapping 99 (SPM99, The Wellcome Department of Neurology, London, UK) for anatomical standardization of atrophied brains. The image sets were transformed to a standard stereotactic space [17] on a Power Mac G4 computer (Apple, California, USA) using the part of the program NEUROSTAT that generates standardized three-dimensional stereotactic surface projections (3D-SSP) data sets for individual subjects [18 –20]. Then, regional cerebral metabolic

Fig. 1. Specific voxels that have significantly lower metabolism in the PSP group than in the healthy control group at the threshold of p < 0.001, uncorrected.

Fig. 3. Specific voxels that have significantly lower metabolism in the PSP group than in the CBD group at the threshold of p < 0.001, uncorrected.

70

K. Hosaka et al. / Journal of the Neurological Sciences 199 (2002) 67–71

images were compared among PSP, CBD, and control subjects by using SPM99 [21]. The PET images were smoothed using an isotropic Gaussian filter of 12-mm diameter to compensate for intersubject gyral variability and to reduce high-frequency noise. The anatomically standardized images were then normalized to the less affected side of the occipital cortices. The three groups were compared with one-way ANOVA. Significance was accepted if the voxels survived an uncorrected threshold of p < 0.001.

3. Results Asymmetry was positive in 10 of the 12 patients with CBD, and mirror PET images were made for 5 CBD patients with a right predominance. Of the 12 patients with PSP, 1 had a left predominance and, as a result, there was no need for preparing mirror images. None of the healthy controls had hemispheric asymmetry. In the PSP group, cerebral glucose metabolism was significantly decreased in the anterior cingulate gyrus, inferior frontal gyrus, caudate nucleus, and midbrain compared with the control group ( p < 0.001, uncorrected) (Table 2, Fig. 1), whereas in the CBD group, significant hypometabolism was noted in the cuneus, inferior parietal lobule, middle frontal gyrus, and post central gyrus on the more affected side, and in a smaller part of the inferior parietal lobule on the less affected side ( p < 0.001, uncorrected) (Table 2, Fig. 2). A comparison between the CBD and PSP groups revealed a lower metabolism in the inferior parietal lobule, precuneus, and lateral occipital cortex of the more affected hemisphere in the CBD group and a lower metabolism in the anterior cingulate and medial frontal gyri of both

Fig. 4. Specific voxels that have significantly lower metabolism in the CBD group than in the PSP group at the threshold of p < 0.001, uncorrected.

hemispheres and the midbrain in the PSP group ( p < 0.001, uncorrected) (Table 2, Figs. 3 and 4).

4. Discussion In patients with PSP, glucose metabolism was decreased in the lateral and medial frontal lobes, caudate nucleus, and midbrain as compared with age-matched healthy controls. These findings were consistent with those of previous ROIbased studies and a voxel-based study. In PSP, the basal ganglia and brain stem are the main loci of pathological changes, while the cortical regions have only a slight pathological involvement [22]. However, there seems to be agreement among previous functional neuroimaging studies [23,24] as well as the present study that frontal dysfunction occurs in PSP. The frontal lobe, as a central constituent of the fronto-subcortical network, is closely connected with subcortical structures. Damage in the subcortical structures results in functional deprivation of the frontal lobe [25]. Behavioral derangement and cognitive deficits which are characteristic of PSP and which epitomize subcortical dementia, such as disinhibited and stereotyped behaviors, apathy, mental slowness, and attentional and executive dysfunctions, are likely due to the frontal hypofunction. On the other hand, in CBD patients, glucose metabolism was decreased in the sensorimotor cortices and the parieto-occipital regions. Mild reduction of metabolism was also noted even in the parietal lobe of the less affected side. The extent of hypometabolic regions demonstrated in this study was somewhat different from that demonstrated in previous studies; the parieto-occipital regions were more hypometabolic in this study than in previous studies. This appears to be caused by our patient cohort, most of whom had cognitive deficits, and our data analysis process, which was modified to give more accurate results. Modifications included standardization of the more affected hemisphere, normalization of measurements relative to the occipital lobe, and a voxel-by-voxel comparison with a new, more appropriate anatomical standardization technique. Cognitive deficits characteristic of CBD, such as cortical sensory loss, ideomotor and limb kinetic apraxias, aphasia or unilateral spatial neglect, and constructive impairment, were mainly attributed to the unilateral parietal dysfunction. Although previous studies with ROI analyses have demonstrated metabolic asymmetry or absolute hypometabolism in the thalamus [12,14,26], we failed to find a significant decrease in the thalamus as they did in a previous study with a voxel-by-voxel analysis. This failure is likely attributable to the normalization process. We normalized the glucose metabolism of the patients to the glucose metabolism in the seemingly least affected region in CBD brain, the medial occipital lobe of the less affected side. However, this normalization may still underestimate the degree of damage. Interestingly, there was little overlap of hypometabolic topography between PSP and CBD, despite the considerable

K. Hosaka et al. / Journal of the Neurological Sciences 199 (2002) 67–71

similarity of symptoms of the two disorders. A direct comparison between PSP and CBD demonstrated a remarkable contrast: a relative hypometabolism in the medial frontal cortices and the midbrain in PSP and a relative hypometabolism in the parietal cortices in CBD. These findings are consistent with those of previous PET studies that compared PSP and CBD [3,5]. The clear distinction between the two diseases may be caused by our criteria-based subject selection process. Since we used stringent clinical diagnostic criteria for each disorder, which have a high specificity and a low sensitivity, we probably included only patients that were typical of each disease. Nevertheless, these findings suggest that an accentuated medial frontal hypometabolism specifically occurs in PSP and an accentuated parietal hypometabolism specifically occurs in CBD. The involvement of the paracentral region and inferior parietal lobules is specific for CBD, and the fronto-subcortical pathways mediating cognition, emotion, and motor function are more affected in PSP than in CBD. These findings are compatible with the findings in previous behavioral and cognitive studies. Although patients with PSP and CBD have overlapping clinical manifestations, they express distinctive symptom profiles, which may arise from distinctive dysfunction topography profiles.

[6] [7]

[8]

[9] [10]

[11]

[12]

[13]

[14]

[15]

5. Conclusions In PSP patients, the metabolic reduction is dominant in the frontal lobe, while in CBD patients, the metabolic reduction is dominant in the parietal lobe. These differences in glucose metabolism appear to reflect the pathological and clinical differences between these two diseases that are especially characterized by cognitive impairments. These findings suggest that measurement of glucose metabolism by PET followed by a voxel-based analysis is useful to understand pathophysiology of these two diseases.

[16]

[17] [18]

[19]

[20]

[21]

References [1] Dickson DW. Neuropathologic differentiation of progressive supranuclear palsy and corticobasal degeneration. J Neurol 1999;246 (Suppl. 2):II6 – 15. [2] Di Maria E, Tabaton M, Vigo T, et al. Corticobasal degeneration shares a common genetic background with progressive supranuclear palsy. Ann Neurol 2000;47:374 – 7. [3] Nagahama Y, Fukuyama H, Turjanski N, et al. Cerebral glucose metabolism in corticobasal degeneration: comparison with progressive supranuclear palsy and normal controls. Mov Disord 1997;12:691 – 6. [4] Okuda B, Tachibana H, Kawabata K, Takeda M, Sugita M. Cerebral blood flow in corticobasal degeneration and progressive supranuclear palsy. Alzheimer Dis Assoc Disord 2000;14:46 – 52. [5] Garraux G, Salmon E, Peigneux P, et al. Voxel-based distribution of

[22] [23]

[24]

[25]

[26]

71

metabolic impairment in corticobasal degeneration. Mov Disord 2000; 5:894 – 904. Turjanski N, Brooks DJ. PET and the investigation of dementia in the parkinsonian patient. J Neural Transm, Suppl 1997;51:37 – 48. Litvan I, Agid Y, Calne D, et al. Clinical research criteria for the diagnosis of progressive supranuclear palsy (Steele – Richardson – Olszewski syndrome): report of the NINDS-SPSP international workshop. Neurology 1996;47:1 – 9. Lang AE, Riley DE, Bergeron C. Cortical-basal ganglionic degeneration. In: Calne DB, editor. Neurodegenerative Disease. Philadelphia: Saunders, 1994. pp. 877 – 94. Litvan I, Agid Y, Goetz C, et al. Accuracy of the clinical diagnosis of corticobasal degeneration. Neurology 1997;48:119 – 25. Ishii K, Sasaki M, Kitagaki H, Sakamoto S, Yamaji S, Maeda K. Regional difference of cerebral blood flow and oxidative metabolism in human cortex. J Nucl Med 1996;37:1086 – 8. Ishii K, Sasaki M, Kitagaki H, et al. Reduction of cerebellar glucose metabolism in advanced Alzheimer’s disease. J Nucl Med 1997;38: 925 – 8. Iida H, Miura S, Kanno I, et al. Design of evaluation of Headtome IV: a whole body positron emission tomograph. IEEE Trans Nucl Sci 1989;NS-37:1006 – 10. Hirono N, Ishii K, Sasaki M, et al. Features of regional cerebral glucose metabolism abnormality in corticobasal degeneration. Dement Geriatr Cogn Disord 2000;11:139 – 46. Eidelberg D, Dhawan V, Moeller JR, et al. The metabolic landscape of cortico-basal ganglionic degeneration: regional asymmetries studied with positron emission tomography. J Neurol Neurosurg Psychiatry 1991;54:856 – 62. Kitagaki H, Hirono N, Ishii K, Mori E. Corticobasal degeneration: evaluation of cortical atrophy by means of hemispheric surface display generated with MR images. Radiology 2000;216:31 – 8. Ishii K, Willoch F, Minoshima S, et al. Statistical brain mapping of FDG PET in Alzheimer’s disease: validation of anatomic standardization for atrophied brains. J Nucl Med 2001;42:548 – 57. Talairach J, Tournoux P. Co-Planar Stereotaxic Atlas of the Human Brain. Stuttgart: Thieme Verlag, 1988. Minoshima S, Berger KL, Lee KS, Mintun MA. An automated method for rotational correction and centering of three-dimensional functional brain images. J Nucl Med 1992;33:1579 – 85. Minoshima S, Koeppe RA, Mintun MA, et al. Automated detection of the intercommissural line for stereotactic localization of functional brain images. J Nucl Med 1993;34:322 – 9. Minoshima S, Koeppe RA, Frey KA, Kuhl DE. Anatomical standardization: linear scaling and nonlinear warping of functional brain images. J Nucl Med 1994;35:1528 – 37. Friston KJ, Holmes AP, Worsely KJ, Poline JB, Frith CD, Frackowiak RSJ. Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp 1995;2:189 – 210. Steele JC. Progressive supranuclear palsy. Brain 1972;95:693 – 704. Blin J, Baron JC, Dubois B, et al. Positron emission tomography study in progressive supranuclear palsy. Brain hypometabolic pattern and clinicometabolic correlations. Arch Neurol 1990;47:747 – 52. Otsuka M, Ichiya Y, Kuwabara Y, et al. Cerebral blood flow, oxygen and glucose metabolism with PET in progressive supranuclear palsy. Ann Nucl Med 1989;3:111 – 8. Mori E, Ishii K, Hashimoto M, Imamura T, Hirono N, Kitagaki H. The role of functional brain imaging in evaluation of vascular dementia. Alzheimer Dis Assoc Disord 1999;(Suppl. 13):1 – 10. Taniwaki T, Yamada T, Yoshida T, et al. Heterogeneity of glucose metabolism in corticobasal degeneration. J Neurol Sci 1998;161: 70 – 6.