Neurobiologyof Aging, Vol. 13, pp. 93-98. ©Pergamon Press plc, 1991. Printed in the U.S.A.
0197-4580/92 $5.00 + .00
Differences of Regional Cerebral Glucose Metabolism Between Presenile and Senile Dementia of Alzheimer Type R. MIELKE, K. HERHOLZ, M. GROND, J. K E S S L E R A N D W . - D . H E I S S l Max-Planck-Institut fiir Neurologische Forschung, Gleueler Str. 50, D-5000 KOln 41, Germany R e c e i v e d 15 N o v e m b e r 1990; A c c e p t e d 15 A u g u s t 1991 MIELKE, R., K. HERHOLZ, M. GROND, J. KESSLER AND W.-D. HEISS. Differences of regional cerebral glucose metabolism between presenile and senile dementia of Alzheimer type. NEUROBIOL AGING 13(1) 93-98, 1992.--The effect of age on regional cerebral metabolic rate of glucose (rCMRG1) was studied in 14 patients with presenile dementia of Alzheimer type (DAT) and 24 patients suffering from senile DAT in comparison to 20 age-matched normal subjects by positron emission tomography (PET) of 2-(18F)-fluoro-2-deoxy-D-glucose (FDG). The metabolic pattern was condensed to a single metabolic ratio. It was calculated as the quotient of rCMRG1 in regions typically affected by AD (frontal and temporoparietal cortex) divided by that in regions typically not affected. In normals this ratio was 1.05---0.04 and did not depend on age. In patients, the metabolic ratio was generally smaller and there was a significant difference between presenile (0.82---0.1) and senile DAT (0.90 + 0.1). This was due to a different metabolic pattern in the two age groups: metabolic impairment was focused on frontal and temporo-parietal cortex in presenile DAT, whereas more global rCMRG1 reductions were present in senile DAT. The results suggest a more generalized disorder in senile dementia impairing metabolism globally in addition to the more localized changes that are typical for DAT. Alzheimer's disease
Aging
Cerebral metabolism
Positron emission tomography
SINCE the introduction of the term "Alzheimer's Disease" (AD) by Kraepelin (25) there have always been controversies, whether this disease may be considered an homogenous entity or whether it should be classified into a presenile and a senile type. Even Alzheimer waived his conception of a presenile disease (3) and applied this term to all kinds of presenile and senile types showing the typical macroscopic and microscopic pattern (2), although very common scientific views and neuropsychiatric nosology regarded the syndrome as an exclusively presenile disorder. A widely accepted pragmatic approach is to use the terms dementia of Alzheimer type (DAT) and senile dementia of Alzheimer type (SDAT) (24) with an arbitrary cutoff age of 65 years. Many parameters have been analysed previously with respect to the influence of age. Apart from psychopathometric and other clinical results, epidemiological (36), neuropathological (31,39), biochemical (38), and genetic (7, 21, 41) parameters have been inspected with contradictory results concerning pathophysiological or etiological inhomogenity between presenile and senile disease. Nowadays, measurement of regional cerebral glucose metabolism by positron emission tomography (PET) is a reliable method to analyse metabolic alterations in dementia of Alzheimer type (DAT) and its subtypes. It has consistently been shown in several studies that reductions of glucose metabolism are most prominent in the temporo-parietal and frontal association cortex, whereas metabolism of primary sensorimotor and visual cortex, basal ganglia and cerebellum is relatively well preserved (5, 8,
10, 13, 15, 29). Recently, we suggested to use this pattern, that appears to be relatively specific for AD, as a diagnostic marker (18). Grady et al. (16) and Small et al. (40) reported that temporo-parietal hypometabolism is not as prominent in senile than in presenile DAT. We therefore examined the dependency of the typical metabolic pattern on age, and compared it with healthy subjects of similar age. METHOD
AD Patients Thirty-eight patients, 13 male and 25 female, were studied. Average age was 66.5_+ 7.6 years (range 52-81 years). Fourteen belonged to the presenile group (5 male and 9 female, average age 57.9---2.9 years, range 52-61 years) and 24 to the senile group (8 male and 16 female, average age 71.5+__4.2 years, range 66-81 years). The diagnosis of probable Alzheimer's disease according to NINCDS-ADRDA criteria (30) was based on general medical, neurological and psychiatric examination and neuropsychological tests. Additionally the patients fulfilled the DSM-III-R criteria (4) of dementia. Laboratory tests included erythrocyte sedimentation rate, complete blood count, serum levels of glucose, bilirubin, urea, creatinine, sodium, potassium, calcium, total protein, lipids, liver enzymes, LDH, thyroid function and TPHA. Ultrasound examination of the neck vessels and CT or MRI of the brain were obtained in each case to rule out
1Requests for reprints should be addressed to W.-D. Heiss.
93
94
MIELKE ~/t ~:~i
hemodynamically significant stenoses of the carotid arteries, ischemic infarcts, hemorrhage and brain tumors. The modified Hachinski score according to Rosen et al. (37) was 3 or less. Patients did not receive any drug like minor or major tranquilizers, antidepressants, nootropics or psycbostimulants. History of intellectual deficits such as memory loss and impairment of abstractive ability lasted from 6 months until 6 years with an average duration of disease of 2 . 5 - 1.3 years. The degree of mental deterioration was quantified based on the Mini-Mental-State Test (MMSE) of Folstein et al. (12) and on the Global-DeteriorationScale (GDS) of Reisberg et al. (34). Average ~cores were: MMSE 16.2_+6.0 (range 3-27). GDS 4.2_+ 1.0 {range 3-6t. Controls
The 20 normal subjects had no clinical evidence of cognitive deficits or neurological disesase and were completely normal in clinical and neuropsychological examination. Average age was 66.3--_8.2 years (range 51-77 years), the group comprised 8 men and 12 women. They were divided into a presenile (4 men and 4 women) and senile (4 men and 8 women) group with average ages 57.4-+3.6 years ( n = 8 ; range 51-63 years) and 72.3 _+3.6 years (n = 12; range 66-77 years), respectively, PET
PET studies were performed under resting conditions with eyes closed and ears unplugged in a room with dim light and low ambient noise. One hundred and eighty-five MBq ~SF-2-fluoro-2-deoxyglucose (FDG) synthesized according to the procedure of Ehrenkaufer et al. (11) or Ido et al. (22) were injected intravenously. Details of the measurement protocol have been described by Heiss et al. (17). Regional cerebral glucose metabolic rates (rCMRGI) were calculated from tomograms recorded 30 to 50 min after tracer injection using an operational equation with adjustment of the local rate constants to measured tissue activity according to the procedure described by Wienhard et al. (42). A total of fourteen slices parallel to the canthomeatal line with a slice thickness of 11 mm and center to center distance of 6.85 mm comprising virtually the whole brain including the cerebellum were analysed. Regions of interest were adapted to major functional-anatomical structures according to a standard scheme using a semiautomatic computer assisted mapping procedure (20). The analysis was focused on regions typically affected by AD (temporo-parietal and frontal association cortex) in contrast to regions typically not affected by AD (cerebellum, brainstem, lentiform nucleus, visual and sensorimotor cortex). In a previous paper (18) we had demonstrated that a ratio of mean rCMRGI in typically affected regions divided by mean rCMRG1 in unaffected regions provides a measure that can discriminate between normals and DAT with high accuracy. It expresses the contrast of the metabolic pattern in DAT in a single quantitative variable. We therefore used this ratio now to analyse the effects of age on the pattern of glucose metabolism in normals and DAT patients. Data were evaluated by analysis of variance (ANOVA) and linear regression on age, treating the various regions as repeated measures within subjects with appropriate Huyn-Feldt corrections of significance probabilities. Average values are given as means _+standard deviation. All statistical procedures were performed using the SAS software package (SAS Institute, Cary, NC). RESULTS
In CT or MRI no signs of ischemic stroke were observed in patients and normals. Small (
0197-4580/92 $5.00 + .00
Differences of Regional Cerebral Glucose Metabolism Between Presenile and Senile Dementia of Alzheimer Type R. MIELKE, K. HERHOLZ, M. GROND, J. K E S S L E R A N D W . - D . H E I S S l Max-Planck-Institut fiir Neurologische Forschung, Gleueler Str. 50, D-5000 KOln 41, Germany R e c e i v e d 15 N o v e m b e r 1990; A c c e p t e d 15 A u g u s t 1991 MIELKE, R., K. HERHOLZ, M. GROND, J. KESSLER AND W.-D. HEISS. Differences of regional cerebral glucose metabolism between presenile and senile dementia of Alzheimer type. NEUROBIOL AGING 13(1) 93-98, 1992.--The effect of age on regional cerebral metabolic rate of glucose (rCMRG1) was studied in 14 patients with presenile dementia of Alzheimer type (DAT) and 24 patients suffering from senile DAT in comparison to 20 age-matched normal subjects by positron emission tomography (PET) of 2-(18F)-fluoro-2-deoxy-D-glucose (FDG). The metabolic pattern was condensed to a single metabolic ratio. It was calculated as the quotient of rCMRG1 in regions typically affected by AD (frontal and temporoparietal cortex) divided by that in regions typically not affected. In normals this ratio was 1.05---0.04 and did not depend on age. In patients, the metabolic ratio was generally smaller and there was a significant difference between presenile (0.82---0.1) and senile DAT (0.90 + 0.1). This was due to a different metabolic pattern in the two age groups: metabolic impairment was focused on frontal and temporo-parietal cortex in presenile DAT, whereas more global rCMRG1 reductions were present in senile DAT. The results suggest a more generalized disorder in senile dementia impairing metabolism globally in addition to the more localized changes that are typical for DAT. Alzheimer's disease
Aging
Cerebral metabolism
Positron emission tomography
SINCE the introduction of the term "Alzheimer's Disease" (AD) by Kraepelin (25) there have always been controversies, whether this disease may be considered an homogenous entity or whether it should be classified into a presenile and a senile type. Even Alzheimer waived his conception of a presenile disease (3) and applied this term to all kinds of presenile and senile types showing the typical macroscopic and microscopic pattern (2), although very common scientific views and neuropsychiatric nosology regarded the syndrome as an exclusively presenile disorder. A widely accepted pragmatic approach is to use the terms dementia of Alzheimer type (DAT) and senile dementia of Alzheimer type (SDAT) (24) with an arbitrary cutoff age of 65 years. Many parameters have been analysed previously with respect to the influence of age. Apart from psychopathometric and other clinical results, epidemiological (36), neuropathological (31,39), biochemical (38), and genetic (7, 21, 41) parameters have been inspected with contradictory results concerning pathophysiological or etiological inhomogenity between presenile and senile disease. Nowadays, measurement of regional cerebral glucose metabolism by positron emission tomography (PET) is a reliable method to analyse metabolic alterations in dementia of Alzheimer type (DAT) and its subtypes. It has consistently been shown in several studies that reductions of glucose metabolism are most prominent in the temporo-parietal and frontal association cortex, whereas metabolism of primary sensorimotor and visual cortex, basal ganglia and cerebellum is relatively well preserved (5, 8,
10, 13, 15, 29). Recently, we suggested to use this pattern, that appears to be relatively specific for AD, as a diagnostic marker (18). Grady et al. (16) and Small et al. (40) reported that temporo-parietal hypometabolism is not as prominent in senile than in presenile DAT. We therefore examined the dependency of the typical metabolic pattern on age, and compared it with healthy subjects of similar age. METHOD
AD Patients Thirty-eight patients, 13 male and 25 female, were studied. Average age was 66.5_+ 7.6 years (range 52-81 years). Fourteen belonged to the presenile group (5 male and 9 female, average age 57.9---2.9 years, range 52-61 years) and 24 to the senile group (8 male and 16 female, average age 71.5+__4.2 years, range 66-81 years). The diagnosis of probable Alzheimer's disease according to NINCDS-ADRDA criteria (30) was based on general medical, neurological and psychiatric examination and neuropsychological tests. Additionally the patients fulfilled the DSM-III-R criteria (4) of dementia. Laboratory tests included erythrocyte sedimentation rate, complete blood count, serum levels of glucose, bilirubin, urea, creatinine, sodium, potassium, calcium, total protein, lipids, liver enzymes, LDH, thyroid function and TPHA. Ultrasound examination of the neck vessels and CT or MRI of the brain were obtained in each case to rule out
1Requests for reprints should be addressed to W.-D. Heiss.
93
94
MIELKE ~/t ~:~i
hemodynamically significant stenoses of the carotid arteries, ischemic infarcts, hemorrhage and brain tumors. The modified Hachinski score according to Rosen et al. (37) was 3 or less. Patients did not receive any drug like minor or major tranquilizers, antidepressants, nootropics or psycbostimulants. History of intellectual deficits such as memory loss and impairment of abstractive ability lasted from 6 months until 6 years with an average duration of disease of 2 . 5 - 1.3 years. The degree of mental deterioration was quantified based on the Mini-Mental-State Test (MMSE) of Folstein et al. (12) and on the Global-DeteriorationScale (GDS) of Reisberg et al. (34). Average ~cores were: MMSE 16.2_+6.0 (range 3-27). GDS 4.2_+ 1.0 {range 3-6t. Controls
The 20 normal subjects had no clinical evidence of cognitive deficits or neurological disesase and were completely normal in clinical and neuropsychological examination. Average age was 66.3--_8.2 years (range 51-77 years), the group comprised 8 men and 12 women. They were divided into a presenile (4 men and 4 women) and senile (4 men and 8 women) group with average ages 57.4-+3.6 years ( n = 8 ; range 51-63 years) and 72.3 _+3.6 years (n = 12; range 66-77 years), respectively, PET
PET studies were performed under resting conditions with eyes closed and ears unplugged in a room with dim light and low ambient noise. One hundred and eighty-five MBq ~SF-2-fluoro-2-deoxyglucose (FDG) synthesized according to the procedure of Ehrenkaufer et al. (11) or Ido et al. (22) were injected intravenously. Details of the measurement protocol have been described by Heiss et al. (17). Regional cerebral glucose metabolic rates (rCMRGI) were calculated from tomograms recorded 30 to 50 min after tracer injection using an operational equation with adjustment of the local rate constants to measured tissue activity according to the procedure described by Wienhard et al. (42). A total of fourteen slices parallel to the canthomeatal line with a slice thickness of 11 mm and center to center distance of 6.85 mm comprising virtually the whole brain including the cerebellum were analysed. Regions of interest were adapted to major functional-anatomical structures according to a standard scheme using a semiautomatic computer assisted mapping procedure (20). The analysis was focused on regions typically affected by AD (temporo-parietal and frontal association cortex) in contrast to regions typically not affected by AD (cerebellum, brainstem, lentiform nucleus, visual and sensorimotor cortex). In a previous paper (18) we had demonstrated that a ratio of mean rCMRGI in typically affected regions divided by mean rCMRG1 in unaffected regions provides a measure that can discriminate between normals and DAT with high accuracy. It expresses the contrast of the metabolic pattern in DAT in a single quantitative variable. We therefore used this ratio now to analyse the effects of age on the pattern of glucose metabolism in normals and DAT patients. Data were evaluated by analysis of variance (ANOVA) and linear regression on age, treating the various regions as repeated measures within subjects with appropriate Huyn-Feldt corrections of significance probabilities. Average values are given as means _+standard deviation. All statistical procedures were performed using the SAS software package (SAS Institute, Cary, NC). RESULTS
In CT or MRI no signs of ischemic stroke were observed in patients and normals. Small (
