Tomography Study in Progressive Supranuclear Palsy
Positron Emission Brain
Hypometabolic
Pattern and Clinicometabolic Correlations
Jer\l=o^\meBlin, MD; Jean Claude Baron, MD; Bruno Dubois, MD; Bernard Pillon, PhD; Henri Cambon, MD; Jean Cambier, MD; Yves
Agid, MD, PhD
\s=b\ In 41 patients with progressive supranuclear palsy (PSP) that was diagnosed on the basis of eight clinical criteria (25 patients with all eight criteria [probable PSP] and 16 with six or seven criteria [possible PSP]), we studied cerebral energy metabolism by using positron emission tomography and the fludeoxyglucose F 18 or the oxygen 15 method. Compared with age\x=req-\ matched controls, each of the cortical and subcortical metabolic values was significantly reduced, with a predominance in the frontal cortex, in both groups of patients with probable and possible PSP, without a difference between these two groups, suggesting similar underlying disease. The frontal metabolic value decreased with disease duration, but the relative frontal hypometabolism (expressed as the fronto-occipital metabolic ratio) was apparently already present in the early stages of the disease. The parkinsonian motor score was correlated with the caudate and thalamic metabolic values. The intellectual deterioration index was significantly correlated with both the frontal and the nonfrontal metabolic values. Finally, the frontal neuropsychological score was significantly correlated with only the fronto-occipital metabolic ratio. Hence, in PSP, a degenerative brain disease with subcortical lesions, the prominent frontal lobe-like syndrome essentially depends on the relative hypometabolism of the frontal cortex. (Arch Neurol. 1990;47:747-752)
regressive supranuclear palsy " (PSP) has recently been the matter
Accepted for publication November 29, 1989. From the Service Hospitalier Frederic Joliot, CEA, Departement de Biologie, Hopital d'Orsay (France) (Drs Blin, Baron, and Cambon); the Clinique de Neurologie et Neuropsychologie and Institut National de la Sant\l=e'\et de la Recherche Medicale U289, Hopital de la Salp\l=e^\tri\l=e`\re,Paris, France (Drs Dubois, Pillon, and Agid); the Service de Neurologie, Hopital Beaujon, Clichy (France) (Dr Cambier); and the Institut National de la Sant\l=e'\et de la Recherche Medicale, U320, Caen, France (Dr Baron). Reprint requests to INSERM U320, CYCERON, BP 5027, 14021 Caen Cedex, France (Dr Baron).
of
extensive clinical,14 neuropsychological,58 neurochemical,9 " neuropathological,1012 and neurometabolic13" investigations. This focus of in¬ terest on PSP essentially results from the quasi-exclusively subcortical local¬ ization of lesions in this neurodegen¬ erative process, making it a potential model of progressive cortical deafferentation to explain intellectual and behavioral impairment1819 ("subcortical dementia"), with attending thera¬ peutic implications.20 By using positron emission tomog¬ raphy (PET) in limited samples of in¬ tellectually impaired patients with PSP, D'Antona et al14 and, more re¬ cently, Leenders et al15 and Foster et
al16 have demonstrated the presence of a preferential depression of energy metabolism in the whole frontal cor¬ tex, suggesting that a neuronal dys¬ function underlies the frontal lobe¬ like syndrome typical of PSP and providing, at the same time, a pathophysiological basis to the concept of subcortical dementia. Goffinet et al,17 however, found the frontal hypometabolism to affect mainly the motor/ premotor cortical fields, relatively sparing the prefrontal lobe, and they found no correlations between frontal metabolic rates and neuropsychological measures. In the present PET study of a large series of patients with clin¬ ical diagnoses, we wished to establish further the pattern of brain hypometabolism in PSP and to investigate, in detail, the relationships between regional metabolic rates and the fol¬ lowing clinical data: disease duration, parkinsonian score, global intellectual deterioration, and frontal lobe-like
symptoms.
PATIENTS AND METHODS Patients With PSP and Controls
Cerebral energy metabolism ated in 41 patients with either
was
evalu¬
"probable"
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/15/2015
"possible" PSP on the basis of the fol¬ lowing clinical criteria8: (1) progressive course of the disease, (2) disease duration less than 10 years, (3) parkinsonism (hy¬ pertonia and akinesia), (4) no significant improvement with levodopa, (5) vertical voluntary gaze palsy, (6) falls, (7) pseudobulbar palsy or dysarthria, and (8) no evi¬ and
dence of a focal lesion on a sensorimotor examination or computed tomographic scan. Cognitive impairment and frontal lobe-like symptoms were not included as inclusion criteria to avoid a systematic bias in the selection of patients that could am¬ plify the metabolic disturbances. The con¬ ditions of patients with all criteria were di¬ agnosed as probable PSP (n 25). When one or two criteria were not available or missing (criterion 4 in four cases, criterion 5 in nine cases, criterion 6 in four cases, and criterion 7 in two cases) the diagnosis of possible PSP (n 16) was made. Further clinical information was obtained around the time (within 3 months) of the PET studies, including the motor parkinsonian score assessed after cessation of levodopa treatment (using the modified Columbia rating scale21) and neuropsychological per¬ formances. Of the 41 patients, 13 were treated with levodopa or dopamine agonists at the time of evaluation. Intellectual and memory function were evaluated with the following tests: verbal tests from the Ver¬ bal scale (digit span, similarities, and arith¬ metic tests) of the Wechsler Adult Intelli¬ gence Scale; a visuospatial task (Raven 47colored Progressive Matrices—PM47); and the Weschler Memory Scale. An intellec¬ tual deterioration index' was calculated to compare the neuropsychological scores ob¬ tained with the expected performance of the patient according to his or her age and educational level (optimal score 0). The following tests that were sensitive to fron¬ tal lobe disorders were added to these stan¬ dard tests: revised version of the Wisconsin card sorting test,22 tests of verbal fluency (word-list generation tests),23 and graphic series.24 Prehension, imitation, and utiliza¬ tion behaviors, which have been shown to be associated with frontal lobe dysfunction,25 were analyzed. A "frontal" score7 was de¬ fined as the sum of the scores for each cog¬ nitive and behavioral test believed to be sensitive to frontal dysfunction (worst score 0; optimal score 60). The frontal =
=
=
=
=
and intellectual deterioration index also assessed in 20 age-matched con¬ trol subjects (mean ± SEM, 63 ± 1.3 years). Specific clinical scores were occa¬ sionally missing for some patients (Table 1 ). The groups of patients with probable and possible PSP (Table 1) had a similar age at onset of the disease, disease duration, parkinsonian motor score, and intellectual de¬ terioration index. Both patient groups showed a significant difference in the fron¬ tal score and intellectual deterioration in¬ dex compared with controls (Table 1). The frontal score, but not the intellectual dete¬ rioration index, was significantly lower in the group with probable PSP when com¬ pared with the group with possible PSP (Table 1). Twenty-seven unmedicated sub¬ jects without any history of central nervous system disease or major psychiatric disor¬ der served as controls for the PET study after informed consent was obtained. score
Table 1.—Characteristics of Patients With PSP*
were
PET
Patients With
Characteristic Age at onset, y Disease duration, y
Age,
PSP,
Mean ± SEM
Probable
Possible
(n
(n
25)
=
=
All Patients
(n
16)
3.1 ± 0.3
65.3 33.2 ± 4.3
y
32.4
Motor scoret
Neuropsychological frontal scoret
Intellectual deterioration indexll
=
41)
62.8 ± 1.2 3.1 ± 0.3
62.4 ± 2.0
63.5
31.7 ± 3.0
10.8 ±
2.1§
24.9 ±
5.2§||
15.4 ±
27.2 ±
3.7§
16.4 ±
5.1§
23.3 ± 3.1 §
2.5§
*PSP progressive supranuclear palsy tMeasured in 23 patients. ¡ 20 patients with probable PSP and in 9 patients with possible PSP (age-matched controls ÍMeasured 53.7 ± 1.8, 20). §P < .05, patients vs controls. ||P< .05, patients with probable PSP vs patients with possible PSP. HMeasured in 20 patients with probable PSP and in 8 patients with possible PSP (age-matched controls value 2.2 ± 0.9, 20). indicates
=
=
Methodology
The PET studies of cerebral energy me¬ tabolism were performed in a dimly lit room and in the eyes-closed, ears-un¬ plugged situation, by using the fludeoxyglucose F18 method with a lumped constant of 0.42 and fludeoxyglucose F 18 rate con¬ stants detailed elsewhere,2'' or the steadystate oxygen 15 method (015)2' for measur¬ ing the regional cerebral metabolic rate of
glucose (CMRGlu)
or
oxygen
(CMR02,
un-
corrected for cerebral blood volume), re¬ spectively. These measurements were per¬ formed on a single-slice camera (ECAT II Model, ORTEC Ine, Oak Ridge, Tenn) until 1985 (lateral and axial resolution: 16 and 19 mm, respectively (used only for fludeoxy¬ glucose F 18), and on the four-ring time-offlight (model TTVOl, Laboratoire d'Elec¬ tronique et de Technologie de l'Informa¬
tique [LETI], Grenoble, France) (lateral and axial resolution: 13 mm for both by us¬ ing fludeoxyglucose F 18 and 015) PET camera, thereafter forming three method¬ ological patient groups: (fludeoxyglucose F 18-ECAT, fludeoxyglucose F 18-LETI, and 015-LETI (n 11 [including the 6 patients previously described by D'Antona et al14], 12, and 18 patients, respectively) (Table 2). In all studies, attenuation correction was carried out by using germanium 68-gallium 68 transmission scans. Imaging was per¬ formed along a plane parallel to the orbito=
meatal line and included two to seven cuts, depending on the camera used and the sub¬
ject's cooperation (see below). (All treat¬ ment was stopped about 16 hours before the PET study.) Circular regions of interest (ROIs) were positioned on the PET scans by using a standardized method that allowed metabolic data to be obtained essentially free of subjective judgment28 by a physician blind to the subject's condition (patients or controls). The cortical ROIs were assigned anatomical localization according to seven cortical areas (prefrontal superior, an
prefrontal inferior, temporal, temporoparietal, sensory-motor, parietal associative, and occipital) defined according to methods published previously,28·25 except that medial frontal and occipital ROIs were not used. The number of ROIs assigned to each of
Number and
Table 2.
Age of Patients With PSP and Controls According
to PET
Method Used*
—
Controls
Patients With PSP
For PET Method
Fludeoxyglucose F 18 with ECAT
camera
Fludeoxyglucose with LETI
(66.1 ± 2.5)
Oxygen 15 with LETI camera
4 ±
2.7) (65.1
(66.0
4 ±
4.9) (66.1
10
(67.8 ± 2.2)
±
25
±
16
2.3)
(53.7
±
2.2)
(52.6
11 ±
1.8)
(52.3
±
12 ±
9
18
8
(64.9
7
11
(65.5
8
F 18
camera
Possible
Probable 7
(64.9 ± 3.4)
For PET Method
2.7)
(66.3
±
41
27
Patients With PSP 6
4.5) (64.3 ± 2.1) 4
4.4)
(59.3 ± 2.3) 5
4.3) (60.6 ± 4.9) 15
(66.6 ± 1.5) (65.3 ± 1.8) (65.9 ± 1.2) (53.0 ± 2.5) (61.7 ± 1.9) PSP indicates progressive supranuclear palsy; PET, positron emission tomography; ECAT, data obtained with the ECAT II model camera, ORTEC Ine, Oak Ridge, Tenn; and LETI, Laboratoire d'Electronique et de Tech¬ nologie de l'Informatique, Grenoble, France. Data are given as the number and mean ± SEM age (years) of pa¬ tients and controls.
these seven cortical areas, and the number of PET slices over which the average met¬ abolic rate of a given area was obtained, varied according to the subject studied and the camera used: three brain levels were studied with the ECAT camera (except in poorly cooperative patients in whom the study was limited to two levels), and five levels were studied with the LETI camera (orbitomeatal, +20, +40, and +60 mm; orbitomeatal, +10, +25, +40, +55, and +70 mm, respectively; the "superior" prefrontal cor¬ tex was sampled on the +55-, +60-, and +70-mm cuts; and the "inferior" prefrontal cortex was sampled on the +25- and +40mm
cuts). Cerebellar, caudate, lenticular,
and thalamic ROIs were defined by using only one circular ROI per structure, cen¬ tered on the peak activity of the given structure, and per side.28·29 This procedure yielded for each of the above described cor¬ tical and subcortical areas, except in the few patients in whom one brain cut was missing, a single mean metabolic value per side. The metabolic activity of each area was calculated as the average of the left and right values. Data
Analysis
To normalize the regional metabolic val¬ ues obtained in the form of CMRGlu or
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/15/2015
CMR02, and with
two different PET
cam¬
regional metabolic rate from each patient was expressed as the fraction of the corresponding mean control value obtained by using the same method and the same camera. Therefore, each of the three "methodologically defined" patient groups (ie, fludeoxyglucose F 18-ECAT, fludeoxy¬ glucose F 18-LETI, and 015-LETI) was matched to one homologous control group of 7, 11, and 9 subjects of similar age, respectively (Table 2). It was justified to use eras, each
the data derived from this normalization
procedure as CMR02 and CMRGlu appears similarly affected in degenerative brain disease in general and especially in PSP.1415 In addition, comparability of the three methodological patient groups, as well as validity of the normalization procedure, were demonstrated by a one-factor analysis of variance that showed no significant dif¬
ference among these three groups either for clinical items or for any normalized re¬ gional metabolic values (data not shown, but available on request). This procedure, which provided what will be referred to as "normalized absolute metabolic values" in the following text, therefore allows data analysis within the patient sample (eg, comparison of groups with probable and
ery
Significancet Mean ± SEM
Patients
Patients
Normalized Absolute Metabolic Values
With
With
Probable PSP
Possible PSP
Age-Matched
Frontal superior cortex Frontal inferior cortex
0.741 ± 0.037 0.756 ± 0.036
0.748 ± 0.045 0.771 ± 0.047
Temporal cortex Temporoparietal
0.780 ± 0.041
0.754 ± 0.047 0.815 ± 0.041
Sensory-motor
cortex
cortex
Controls 0.982 ± 0.055 0.994 ± 0.034
Possible
Probable PSP vs Controls
PSP vs Controls
Positron Emission Brain
Hypometabolic
Pattern and Clinicometabolic Correlations
Jer\l=o^\meBlin, MD; Jean Claude Baron, MD; Bruno Dubois, MD; Bernard Pillon, PhD; Henri Cambon, MD; Jean Cambier, MD; Yves
Agid, MD, PhD
\s=b\ In 41 patients with progressive supranuclear palsy (PSP) that was diagnosed on the basis of eight clinical criteria (25 patients with all eight criteria [probable PSP] and 16 with six or seven criteria [possible PSP]), we studied cerebral energy metabolism by using positron emission tomography and the fludeoxyglucose F 18 or the oxygen 15 method. Compared with age\x=req-\ matched controls, each of the cortical and subcortical metabolic values was significantly reduced, with a predominance in the frontal cortex, in both groups of patients with probable and possible PSP, without a difference between these two groups, suggesting similar underlying disease. The frontal metabolic value decreased with disease duration, but the relative frontal hypometabolism (expressed as the fronto-occipital metabolic ratio) was apparently already present in the early stages of the disease. The parkinsonian motor score was correlated with the caudate and thalamic metabolic values. The intellectual deterioration index was significantly correlated with both the frontal and the nonfrontal metabolic values. Finally, the frontal neuropsychological score was significantly correlated with only the fronto-occipital metabolic ratio. Hence, in PSP, a degenerative brain disease with subcortical lesions, the prominent frontal lobe-like syndrome essentially depends on the relative hypometabolism of the frontal cortex. (Arch Neurol. 1990;47:747-752)
regressive supranuclear palsy " (PSP) has recently been the matter
Accepted for publication November 29, 1989. From the Service Hospitalier Frederic Joliot, CEA, Departement de Biologie, Hopital d'Orsay (France) (Drs Blin, Baron, and Cambon); the Clinique de Neurologie et Neuropsychologie and Institut National de la Sant\l=e'\et de la Recherche Medicale U289, Hopital de la Salp\l=e^\tri\l=e`\re,Paris, France (Drs Dubois, Pillon, and Agid); the Service de Neurologie, Hopital Beaujon, Clichy (France) (Dr Cambier); and the Institut National de la Sant\l=e'\et de la Recherche Medicale, U320, Caen, France (Dr Baron). Reprint requests to INSERM U320, CYCERON, BP 5027, 14021 Caen Cedex, France (Dr Baron).
of
extensive clinical,14 neuropsychological,58 neurochemical,9 " neuropathological,1012 and neurometabolic13" investigations. This focus of in¬ terest on PSP essentially results from the quasi-exclusively subcortical local¬ ization of lesions in this neurodegen¬ erative process, making it a potential model of progressive cortical deafferentation to explain intellectual and behavioral impairment1819 ("subcortical dementia"), with attending thera¬ peutic implications.20 By using positron emission tomog¬ raphy (PET) in limited samples of in¬ tellectually impaired patients with PSP, D'Antona et al14 and, more re¬ cently, Leenders et al15 and Foster et
al16 have demonstrated the presence of a preferential depression of energy metabolism in the whole frontal cor¬ tex, suggesting that a neuronal dys¬ function underlies the frontal lobe¬ like syndrome typical of PSP and providing, at the same time, a pathophysiological basis to the concept of subcortical dementia. Goffinet et al,17 however, found the frontal hypometabolism to affect mainly the motor/ premotor cortical fields, relatively sparing the prefrontal lobe, and they found no correlations between frontal metabolic rates and neuropsychological measures. In the present PET study of a large series of patients with clin¬ ical diagnoses, we wished to establish further the pattern of brain hypometabolism in PSP and to investigate, in detail, the relationships between regional metabolic rates and the fol¬ lowing clinical data: disease duration, parkinsonian score, global intellectual deterioration, and frontal lobe-like
symptoms.
PATIENTS AND METHODS Patients With PSP and Controls
Cerebral energy metabolism ated in 41 patients with either
was
evalu¬
"probable"
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/15/2015
"possible" PSP on the basis of the fol¬ lowing clinical criteria8: (1) progressive course of the disease, (2) disease duration less than 10 years, (3) parkinsonism (hy¬ pertonia and akinesia), (4) no significant improvement with levodopa, (5) vertical voluntary gaze palsy, (6) falls, (7) pseudobulbar palsy or dysarthria, and (8) no evi¬ and
dence of a focal lesion on a sensorimotor examination or computed tomographic scan. Cognitive impairment and frontal lobe-like symptoms were not included as inclusion criteria to avoid a systematic bias in the selection of patients that could am¬ plify the metabolic disturbances. The con¬ ditions of patients with all criteria were di¬ agnosed as probable PSP (n 25). When one or two criteria were not available or missing (criterion 4 in four cases, criterion 5 in nine cases, criterion 6 in four cases, and criterion 7 in two cases) the diagnosis of possible PSP (n 16) was made. Further clinical information was obtained around the time (within 3 months) of the PET studies, including the motor parkinsonian score assessed after cessation of levodopa treatment (using the modified Columbia rating scale21) and neuropsychological per¬ formances. Of the 41 patients, 13 were treated with levodopa or dopamine agonists at the time of evaluation. Intellectual and memory function were evaluated with the following tests: verbal tests from the Ver¬ bal scale (digit span, similarities, and arith¬ metic tests) of the Wechsler Adult Intelli¬ gence Scale; a visuospatial task (Raven 47colored Progressive Matrices—PM47); and the Weschler Memory Scale. An intellec¬ tual deterioration index' was calculated to compare the neuropsychological scores ob¬ tained with the expected performance of the patient according to his or her age and educational level (optimal score 0). The following tests that were sensitive to fron¬ tal lobe disorders were added to these stan¬ dard tests: revised version of the Wisconsin card sorting test,22 tests of verbal fluency (word-list generation tests),23 and graphic series.24 Prehension, imitation, and utiliza¬ tion behaviors, which have been shown to be associated with frontal lobe dysfunction,25 were analyzed. A "frontal" score7 was de¬ fined as the sum of the scores for each cog¬ nitive and behavioral test believed to be sensitive to frontal dysfunction (worst score 0; optimal score 60). The frontal =
=
=
=
=
and intellectual deterioration index also assessed in 20 age-matched con¬ trol subjects (mean ± SEM, 63 ± 1.3 years). Specific clinical scores were occa¬ sionally missing for some patients (Table 1 ). The groups of patients with probable and possible PSP (Table 1) had a similar age at onset of the disease, disease duration, parkinsonian motor score, and intellectual de¬ terioration index. Both patient groups showed a significant difference in the fron¬ tal score and intellectual deterioration in¬ dex compared with controls (Table 1). The frontal score, but not the intellectual dete¬ rioration index, was significantly lower in the group with probable PSP when com¬ pared with the group with possible PSP (Table 1). Twenty-seven unmedicated sub¬ jects without any history of central nervous system disease or major psychiatric disor¬ der served as controls for the PET study after informed consent was obtained. score
Table 1.—Characteristics of Patients With PSP*
were
PET
Patients With
Characteristic Age at onset, y Disease duration, y
Age,
PSP,
Mean ± SEM
Probable
Possible
(n
(n
25)
=
=
All Patients
(n
16)
3.1 ± 0.3
65.3 33.2 ± 4.3
y
32.4
Motor scoret
Neuropsychological frontal scoret
Intellectual deterioration indexll
=
41)
62.8 ± 1.2 3.1 ± 0.3
62.4 ± 2.0
63.5
31.7 ± 3.0
10.8 ±
2.1§
24.9 ±
5.2§||
15.4 ±
27.2 ±
3.7§
16.4 ±
5.1§
23.3 ± 3.1 §
2.5§
*PSP progressive supranuclear palsy tMeasured in 23 patients. ¡ 20 patients with probable PSP and in 9 patients with possible PSP (age-matched controls ÍMeasured 53.7 ± 1.8, 20). §P < .05, patients vs controls. ||P< .05, patients with probable PSP vs patients with possible PSP. HMeasured in 20 patients with probable PSP and in 8 patients with possible PSP (age-matched controls value 2.2 ± 0.9, 20). indicates
=
=
Methodology
The PET studies of cerebral energy me¬ tabolism were performed in a dimly lit room and in the eyes-closed, ears-un¬ plugged situation, by using the fludeoxyglucose F18 method with a lumped constant of 0.42 and fludeoxyglucose F 18 rate con¬ stants detailed elsewhere,2'' or the steadystate oxygen 15 method (015)2' for measur¬ ing the regional cerebral metabolic rate of
glucose (CMRGlu)
or
oxygen
(CMR02,
un-
corrected for cerebral blood volume), re¬ spectively. These measurements were per¬ formed on a single-slice camera (ECAT II Model, ORTEC Ine, Oak Ridge, Tenn) until 1985 (lateral and axial resolution: 16 and 19 mm, respectively (used only for fludeoxy¬ glucose F 18), and on the four-ring time-offlight (model TTVOl, Laboratoire d'Elec¬ tronique et de Technologie de l'Informa¬
tique [LETI], Grenoble, France) (lateral and axial resolution: 13 mm for both by us¬ ing fludeoxyglucose F 18 and 015) PET camera, thereafter forming three method¬ ological patient groups: (fludeoxyglucose F 18-ECAT, fludeoxyglucose F 18-LETI, and 015-LETI (n 11 [including the 6 patients previously described by D'Antona et al14], 12, and 18 patients, respectively) (Table 2). In all studies, attenuation correction was carried out by using germanium 68-gallium 68 transmission scans. Imaging was per¬ formed along a plane parallel to the orbito=
meatal line and included two to seven cuts, depending on the camera used and the sub¬
ject's cooperation (see below). (All treat¬ ment was stopped about 16 hours before the PET study.) Circular regions of interest (ROIs) were positioned on the PET scans by using a standardized method that allowed metabolic data to be obtained essentially free of subjective judgment28 by a physician blind to the subject's condition (patients or controls). The cortical ROIs were assigned anatomical localization according to seven cortical areas (prefrontal superior, an
prefrontal inferior, temporal, temporoparietal, sensory-motor, parietal associative, and occipital) defined according to methods published previously,28·25 except that medial frontal and occipital ROIs were not used. The number of ROIs assigned to each of
Number and
Table 2.
Age of Patients With PSP and Controls According
to PET
Method Used*
—
Controls
Patients With PSP
For PET Method
Fludeoxyglucose F 18 with ECAT
camera
Fludeoxyglucose with LETI
(66.1 ± 2.5)
Oxygen 15 with LETI camera
4 ±
2.7) (65.1
(66.0
4 ±
4.9) (66.1
10
(67.8 ± 2.2)
±
25
±
16
2.3)
(53.7
±
2.2)
(52.6
11 ±
1.8)
(52.3
±
12 ±
9
18
8
(64.9
7
11
(65.5
8
F 18
camera
Possible
Probable 7
(64.9 ± 3.4)
For PET Method
2.7)
(66.3
±
41
27
Patients With PSP 6
4.5) (64.3 ± 2.1) 4
4.4)
(59.3 ± 2.3) 5
4.3) (60.6 ± 4.9) 15
(66.6 ± 1.5) (65.3 ± 1.8) (65.9 ± 1.2) (53.0 ± 2.5) (61.7 ± 1.9) PSP indicates progressive supranuclear palsy; PET, positron emission tomography; ECAT, data obtained with the ECAT II model camera, ORTEC Ine, Oak Ridge, Tenn; and LETI, Laboratoire d'Electronique et de Tech¬ nologie de l'Informatique, Grenoble, France. Data are given as the number and mean ± SEM age (years) of pa¬ tients and controls.
these seven cortical areas, and the number of PET slices over which the average met¬ abolic rate of a given area was obtained, varied according to the subject studied and the camera used: three brain levels were studied with the ECAT camera (except in poorly cooperative patients in whom the study was limited to two levels), and five levels were studied with the LETI camera (orbitomeatal, +20, +40, and +60 mm; orbitomeatal, +10, +25, +40, +55, and +70 mm, respectively; the "superior" prefrontal cor¬ tex was sampled on the +55-, +60-, and +70-mm cuts; and the "inferior" prefrontal cortex was sampled on the +25- and +40mm
cuts). Cerebellar, caudate, lenticular,
and thalamic ROIs were defined by using only one circular ROI per structure, cen¬ tered on the peak activity of the given structure, and per side.28·29 This procedure yielded for each of the above described cor¬ tical and subcortical areas, except in the few patients in whom one brain cut was missing, a single mean metabolic value per side. The metabolic activity of each area was calculated as the average of the left and right values. Data
Analysis
To normalize the regional metabolic val¬ ues obtained in the form of CMRGlu or
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/15/2015
CMR02, and with
two different PET
cam¬
regional metabolic rate from each patient was expressed as the fraction of the corresponding mean control value obtained by using the same method and the same camera. Therefore, each of the three "methodologically defined" patient groups (ie, fludeoxyglucose F 18-ECAT, fludeoxy¬ glucose F 18-LETI, and 015-LETI) was matched to one homologous control group of 7, 11, and 9 subjects of similar age, respectively (Table 2). It was justified to use eras, each
the data derived from this normalization
procedure as CMR02 and CMRGlu appears similarly affected in degenerative brain disease in general and especially in PSP.1415 In addition, comparability of the three methodological patient groups, as well as validity of the normalization procedure, were demonstrated by a one-factor analysis of variance that showed no significant dif¬
ference among these three groups either for clinical items or for any normalized re¬ gional metabolic values (data not shown, but available on request). This procedure, which provided what will be referred to as "normalized absolute metabolic values" in the following text, therefore allows data analysis within the patient sample (eg, comparison of groups with probable and
ery
Significancet Mean ± SEM
Patients
Patients
Normalized Absolute Metabolic Values
With
With
Probable PSP
Possible PSP
Age-Matched
Frontal superior cortex Frontal inferior cortex
0.741 ± 0.037 0.756 ± 0.036
0.748 ± 0.045 0.771 ± 0.047
Temporal cortex Temporoparietal
0.780 ± 0.041
0.754 ± 0.047 0.815 ± 0.041
Sensory-motor
cortex
cortex
Controls 0.982 ± 0.055 0.994 ± 0.034
Possible
Probable PSP vs Controls
PSP vs Controls
