Journal of Cerebral Blood Flow and Metabolism
10:199--206 © 1990 Raven Press, Ltd., New York
Regional Cerebral Glucose Metabolism is Normal in Young Adults with Down Syndrome
Mark B. Schapiro, Cheryl L. Grady, Anand Kumar, tPeter Herscovitch, James V. Haxby, Angela M. Moore, *Beverly White, Robert P. Friedland, and Stanley I. Rapoport Laboratory of Neurosciences, Section on Brain Aging and Dementia, National Institute on Aging, Clinical Center, and *Laboratory of Chemical Biology, National Institute of Diabetes, Digestive and Kidney Diseases, and fDepartment of Nuclear Medicine, National Institutes of Health, Clinical Center, Bethesda, Maryland, U.S.A.
Summary: Regional CMRglc (rCMRg1c) values were mea sured with ['sF]2-fluoro-2-deoxY-D-glucose eSFDG) and positron emission tomography (PET), using a Scan ditronix PC-1024-7B scanner, in 14 healthy, noninstitu tionalized subjects with trisomy 21 (Down syndrome; DS) (mean age 30.0 years, range 25-38 years) and in 13 sex matched, healthy volunteers (mean age 29.5 years, range 22-38 years). In the DS group, mean mental age on the Peabody Picture Vocabulary Test was 7.8 years and de mentia was not present. Resting rCMRglc was determined with eyes covered and ears occluded in a quiet, darkened room. Global gray CMRglc equaled 8.76 ± 0.76 mg/IOO g/min (mean ± SD) in the DS group as compared with
8.74 ± 1.19 mg/lOO g/min in the control group (p > 0.05). Gray matter regional measurements also did not differ between groups. The ratio of rCMRglc to global CMRg1C' calculated to reduce the variance associated with abso lute rCMRglc, and rightlleft ratios did not show any con sistent differences. These results show that healthy young DS adults do not have alterations in regional or global brain glucose metabolism, as measured with lsFDG and PET, prior to an age at which the neuropathological changes in Alzheimer disease are reported to occur. Key Words: Brain-Cerebral glucose metabolism-Down syndrome-[ISFj2-Fluoro-2-deoxY-D-glucose-Mental retardation-Positron emission tomography.
A previous study in our laboratory using positron emission tomography (PET) and eSp]2-fluoro-2-de oxy-D-glucose eSPDG) showed that Down syn drome (DS) adults over the age of 45 years have reductions in parietal and temporal glucose metab olism in comparison with younger DS adults (Scha piro et al., 1987b). However, another study in our laboratory did not find this pattern in four younger DS adults when compared with controls, although these young DS subjects had a mean global increase in glucose metabolism (Schwartz et al., 1983). Be cause this latter study used a calculated correction
for attenuation of radiation in the determination of brain radioactivity with PET (Huang et al., 1981), with an ECAT II tomograph, it was possible that the smaller brain size (Schapiro et al., 1987a) and possibly thinner skulls (Roche and Sunderland, 1960) of the DS subjects accounted for the differ ences in cerebral metabolic rates between DS and controls. To avoid this problem and to reevaluate brain metabolism in DS, we investigated cerebral glucose metabolism in a carefully screened, larger group of young adults with trisomy 2 1, now using a higher resolution PET tomograph (Scanditronix PC1024-7B) with a measured attenuation correction. This was done to see if there are differences in ab solute cerebral metabolic rate between DS adults and age-matched normals or in patterns of regional metabolism prior to the age of 35-40 years, when Alzheimer disease (AD) neuropathology is reported to occur (Haberland, 1969; Burger and Vogel, 1973; Wisniewski et al., 1985).
Received January 31, 1989; revised September 12, 1989; ac cepted September 20, 1989. Address correspondence and reprint requests to Dr. M. B. Schapiro at Laboratory of Neurosciences, National Institute on Aging, Bldg. 10, Rm. 6C414, Bethesda, MD 20892, U. S.A. Abbreviations used: AD, Alzheimer disease; DS, Down syn drome; 18FDG, [18F12-fluoro-2-deoxY-D-glucose; 10M, inferior orbitomeata1; PET, positron emission tomography; rCMRglc' re gional CMRglc; ROI, region of interest.
1 99
M. B. SCHAPIRO ET AL.
200
MATERIALS AND METHODS Selection of subjects
We studied 14 young adult DS subjects aged 25-38 years (10 men and 4 women, mean age 30.0 years). All were recruited from family or group homes and had tri somy 21. Only two subjects had lived in an institution in the past. Mean mental age, as measured on the Peabody Picture Vocabulary Test (Dunn and Dunn, 1981), was 7.8 years (range 4-16 years). All subjects were screened for dementia of the Alzheimer type using modified criteria from the Diagnostic and Statistical Manual, which spec ified an acquired, progressive loss of intellectual function (Schapiro et al., 1987b). Dementia was not present in the DS subjects in this study (Schapiro et al., 1987b). Screen ing included a review of the medical history, a physical and neurological examination, and the following labora tory studies: routine blood counts, electrolytes, liver and renal function tests, cholesterol, triglycerides, thyroid function tests, Venereal Disease Research Laboratories test, antinuclear antibody (ANA), serum folate and B\2, ECG, and chest radiographs. Of the young DS adults, five had functional systolic murmurs and one had a mitral valve prolapse. Two were euthyroid on thyroid replacement and one was biochem ically hypothyroid while on thyroid replacement. Another subject was receiving intramuscular vitamin B 2 for celiac 1 sprue. With these exceptions, the DS subjects had no other disease. None was taking medication (except as noted above). The controls, 13 sex-matched, healthy volunteers aged 22-38 years (9 men and 4 women, mean 29.5 years), were participants in the National Institute on Aging Laboratory of Neurosciences Study on Aging (Duara et al., 1983) and were screened with medical, neurological, and laboratory tests. Subjects had no history of neuro logic, psychiatric, or systemic medical disorder. None was taking medication for at least 2 weeks prior to the study. Controls signed an informed consent describing the purpose of the study, the tests performed, and the risks involved. For the DS subjects, a relative and/or a legal guardian was required to sign the informed consent. The research was conducted under NIH protocols 81-AG-1O (DS) and 80-AG-26 (controls). PET technique
The global gray matter CMRg1c and regional CMRgJc (rCMRgJc) were measured in the resting state using 18FDG and a multislice PET scanner (Scanditronix PC-1024-7B, Uppsala, Sweden). The 18FDG was prepared by the method of Shiue et al. (1982). 18F2 was produced at the 2 NIH cyclotron via the °Ne (a,a) 18F nuclear reaction. The isotonic, buffered product was checked for chemical and radiochemical purity using both high pressure liquid chromatography and thin layer chromatography. Scans of patients were interspersed with scans of control subjects to avoid drifts of metabolic values due to possible changes in machine-operating characteristics or 18FDG synthesis. In the tomograph, there are four rings of de tectors that allow simultaneous collection of data from seven transverse sections separated by 13.8 mm center to center. Spatial resolution (full width at half-maximum) was 6 mm in the image planes, and axial resolution was 11 mm for straight slices and 8 mm for cross-slices in the center of the field of view.
J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
Subjects fasted for at least 2 h and refrained from al cohol, caffeine, and nicotine for at least 24 h prior to the PET scan. After percutaneous cannulation of the radial artery with cutaneous anesthesia and placement of an in travenous catheter, the subject was positioned in the PET scanner. The head was immobilized with a thermoplastic mask that was molded to the shape of the head. A laser light and a line drawn on the mask corresponding to the inferior orbitomeatal (10M) line were used to position the subject in the scanner. With the ears plugged with cotton and eyes patched, the subject lay supine in a quiet and darkened scan room (resting state). A transmission scan, using a ring phantom of 18F, was performed to correct for attenuation of radiation by the skull, with slices parallel to and from 10 to 100 mm above the 10M line. Two transmission scans were performed, offset by 6.9 mm (half-slice separation) in the z-axis, to provide an interleaved set of 14 slices. Mter the transmission scan was completed, the subject received an intravenous bolus injection of 5 mCi of 18FDG. Serial arterial blood samples were drawn during the 45-min uptake period and during the emission scan and were centrifuged to provide plasma for measurement of radioactivity and glucose. After 45 min, lO-min emis sion scans were performed at the same two levels as for the transmission scans. PET slices had at least 2 x 106 coincidence counts. PET images were reconstructed and corrected for attenuation using data from the transmis sion scans. Counts from the brain were converted to brain radioactivities (nCi/cm3), using a calibration factor de rived by scanning a homogeneous, cylindrical phantom with a known 18FDG concentration, within several hours of each scan. Plasma and brain radioactivities were cor rected for physical decay to the time of injection. Data analysis
From the time courses of blood and brain radioactivi ties, of blood glucose concentration, and from the time of scanning, metabolic rates were calculated in units of mil ligrams per 100 g per minute using Brooks' modification (1982) of the operational equation of Sokoloff et al. (1977). A value of 0.418 was used for the lumped constant (Huang et al., 1980). Values for the kinetic rate constants were kl 0.102, k2 0.130, k3 0.062, and k4 0.0068 (in units of min - I ) (Huang et al., 1980). A template was derived from a PET scan of a healthy normal subject. The template consisted of circular re 2 gions of interest (ROls) 8 mm in diameter (50 mm ). From the set of 14 PET slices of this individual, the slices best representative of the neocortical areas of interest were identified through comparison with an atlas of a human brain sectioned in the same plane as the PET scans (Ey cleshymer and Schoemaker, 1911). The ROls were spaced evenly throughout the cortex and also were cen tered in subcortical areas, including the thalamus and len ticular and caudate nuclei. ROls for subcortical struc tures were placed on the single slice in which they were seen best. The white matter ROls were placed in the cen trum semiovale bilaterally. There were 201 ROls on eight slices, which were averaged over larger regions. For the analysis of study subjects, PET slices were compared with anatomical sections from an atlas of a human brain (Eycleshymer and Schoemaker, 1911). Be cause of individual differences in head size and shape, the height above the 10M line of the best-fit slice in the atlas was assigned to each PET slice. By comparison of the =
=
=
=
CEREBRAL GLUCOSE METABOLISM IN DOWN SYNDROME
PET slice with its best-fit slice in the atlas, we identified anatomical ROIs in the PET scans. Since the DS subjects had smaller and more rounded heads, it was not possible to blind the data analysis procedure. The template, which was derived from the normal subject as described above, was placed over the corresponding PET slices for each subjects and adjusted to fit the individual brains. Because ROIs could be individually moved, adjustments could be made for differences in head size or shape, such as the brachycephaly noted in the DS subjects. Global metabolism represented the weighted mean of all gray matter regions, excluding the vermis and mid brain. Two indexes of metabolic patterns were examined: (a) the ratio of rCMRglC to global CMRg!c and (b) an asym metry index of rCMRg!c, calculated as (R - L)/[(R + L)/2], where Rand L represent right and left metabolic rates, respectively. Other measurements
Supine resting heart rate and systolic and diastolic blood pressures were measured on a non-PET day and prior to the injection of 18FDG and after the emission scan on a PET day. Each subject was given a rating of anxiety (Duara et ai., 1984). Arterial blood was obtained during the PET scan for determination of glucose. Noncontrast computed tomography scans of brain were performed with a GE 8800 CT scanner (General Electric Co., Milwaukee, WI, U.S.A.) using previously described methodology for scanning and analysis (Schwartz et al., 1985). Statistics
For physiologic data, group differences were compared with an unpaired t test (SAS Institute, 1985). Within groups, differences in these data between mean values from non-PET and PET days, as well as differences in means from preinjection and postinjection, were com pared with a paired t test (SAS Institute, 1985). For metabolic and computed tomography data, differ ences between mean values in the DS and control groups were analyzed with an unpaired t test (SAS Institute, 1985). Comparisons of high and low intellectually func tioning DS subjects were performed with a Wilcoxon two sample test (SAS Institute, 1985). Within groups, right and left regional metabolic values were compared with a paired t test (SAS Institute, 1985). The criterion of statis tical significance was p < 0.05. RESULTS
Table 1 lists the physiologic data for the DS sub jects and controls. DS subjects were significantly shorter than controls. On a non-PET day, supine resting heart rate and systolic and diastolic blood pressures were significantly lower in the DS group as compared with controls. On the day of the PET scan, the DS and control groups did not differ prior to injection on mean heart rate or blood pressure, but postscan systolic and diastolic blood pressures were significantly lower in the DS group. Supine heart rate and blood pressure were not significantly different on PET and non-PET days for DS sub jects, though diastolic blood pressure was less on
201
TABLE 1. Physiologic values for Down syndrome control and subjects
Age (yrs) Height (m) Weight (kg) Measurements on non-PET day Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Measurements on PET day Preinjection Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Postinjection Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Glucose (mg/IOO ml) Anxiety rating (1-3)
Controls
Down syndrome
29.5 ± 4.5 (13) 1.76 ± 0.06 (8) 71.5 ± 10.9 (II)
30.0 ± 4.3 (14) 1.55 ± 0.08 (l4)a 66.1 ± 11.2 (14)
122 ± 11 79 ± 4 77 ± 6
(10) (10) (10)
107 ± 10 70 ± 7 70 ± 9
(l4)a (l4)a (l4)a
124 ± 11 74 ± 5 70 ± 7
(12) (l2)b (11)
118 ± 13 70 ± 11 71 ± 10
(13) (13) (12)
130 ± 12 80 ± 7 70 ± 4 103 ± 8 0.3 ± 0.6
(10) (lO)C (9) (13) (II)
III 67 68 100 0.6
±
± ± ± ±
8 10 10 9 0.7
(13)a (13)a (13)C (14) (14)
Values are means ± SD, no. of subjects shown in parentheses. PET, positron emission tomography; BP, blood pressure. a Mean in Down syndrome differs from mean in control by unpaired t test (p < 0.05). b Non-PET day significantly different from PET day by paired t test (p < 0.05). Post injection significantly different from preinjection by paired t test (p < 0.05). C
the day of the PET in the controls. The anxiety rating, a measure of stress, did not differ between groups. DS subjects, in comparison with healthy normals, had significantly (p < 0.05) decreased total intracra nial volume [ 1,069 ± 93 vs. 1,236 ± 1 13 (SD) cm3], seven-slice intracranial volume (the 50-mm segment of brain, starting at the lowest slice that contained the body of the third ventricle; 688 ± 4 1 vs. 784 ± 57 cm3), and gray matter plus white matter volume (659 ± 40 vs. 776 ± 64 cm3). There was no signifi cant difference in third ventricle volume or right or left ventricle volume between the DS and control groups, either directly or after normalizing to seven-slice intracranial volume. Table 2 shows absolute values of global CMRg\c and of rCMRglc for the DS and control groups. Global CMRg\c equaled 8.76 ± 0.76 (SD) mg/100 g/min in the DS group as compared with 8.74 ± 1. 19 mgllOO g/min in controls (p > 0.05). Global and gray matter regional measurements did not differ be tween groups (p > 0.05), though white matter rCMRg\c was increased in the DS as compared with the control group (p < 0.05). PET images at two levels (Fig. 1) showed no differences in appearance between a DS and a control subject. Comparison of the six DS subjects above the age of 30 years (3�38 years old) with the eight DS subjects below the age of 30 years (25-29 years old) did not show any dif ferences in global or gray matter regional measure ments. The asymmetry index of rCMRg\c failed to show J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
202
M. B. SCHAPIRO ET AL.
±
1.09 1.10
±
0.96 0.88
±
0.98 1.09
±
0.95 0.93
±
0.76 0.88
±
0.77 1.15
±
1.02 0.93
9.10 9.22
±
0.97 1.05
0.77 0.93
8.01 8.16
±
0.53 0.67
±
1.15 1.32
6.79 7.10
±
0.95 1.01
±
1.09 1.21
8.41 8.44
±
0.73 0.86
significant differences between groups (p > 0.05). Within groups, significant asymmetries were not present, except in the middle temporal and paracen tral regions in the control group (left> right metab olism), although one significant asymmetry would be expected by chance (assuming independence). Table 3 shows mean ratios of rCMRglc to global CMRglc, which were calculated to reduce the vari ance associated with absolute values of rCMRglc (Duara et aI., 1984) and to examine patterns of me tabolism. DS and control groups differed in several scattered regions. The ratio was lower in the DS group in the left orbitofrontal, left inferior parietal, left middle temporal, and right anterior cingulate regions and the cerebellum bilaterally; the ratio was elevated in the DS group in the right sensorimotor and right calcarine regions and the thalamus bilat erally. These differences, however, were very slight (differences between ratios of ±0.04 were statisti cally significant at p < 0.05), and moreover, two would be expected by chance alone at p < 0.05. Further studies would be required to evaluate the biological meaning of each difference. Because of the variation in mental ages, the DS group was divided into high (>8 years) and low «8 years) functioning subjects. There was no differ ence between these subgroups for any measure of rCMRglc•
±
1.23 1.14
9.18 9.25
±
1.13 1.12
DISCUSSION
8.85 8.75 10.10
±
1.47 1.46 1.61
8.12 8.37 9.49
±
1.21 1.01 0.99
9.34 9.32
±
1.31 1.18
9.18 9.13
±
1.15 1.02
8.73 9.25
±
1.10 1.49
8.65 8.82
±
0.99 1.01
8.92 9.41
±
1.67 1.80
9.84 10.03
±
1.10 1.02
±
1.86 1.13
9.87 9.99
±
±
1.37 1.59
±
1.91 1.31
10.29 10.56
±
±
1.40 1.59
5.57 5.65
±
0.90 0.76
5.84 5.44
±
0.65 0.77
7.05 6.87
±
0.99 0.77
7.39 7.04
±
0.63 0.69
±
1.31 1.38 1.30
6.72 6.64 6.75
±
0.77 0.93 0.81
0.61 0.65 1.19
3.07 3.03 8.76
TABLE 2. Absolute CMRglc in control and Down syndrome subjects Regional CMRglc (mg/loo g/min) Brain region
Control (13)
Orbital frontal Right Left Prefrontal Right Left Premotor Right Left Sensorimotor Right Left Superior parietal Right Left Medial parietal Right Left Inferior parietal Right Left Calcarine Right Left Occipital association Right Left Inferior temporal Right Left Middle temporal Right Left Superior temporal Right Left Anterior cingulate Right Left Posterior cingulate Insula Right Left Paracentral Right Left Thalamus Right Left Caudate Right Left Lenticular Right Left Amygdala Right Left Hippocampus Right Left Cerebellum Right Left Vermis White matter (centrum semiovale) Right Left Global gray matter
Down syndrome (14)
±
1.30 1.32
7.58 7.79
±
±
1.40 1.38
9.00 8.91
±
±
1.62 1.68
9.69 9.68
±
±
1.41 1.61
9.66 9.47
9.16 9.28
±
1.49 1.66
9.31 9.16
9.40 9.64
±
1.07 1.64
9.26 9.19
±
0.96 1.44
8.91 8.93
±
0.98 0.90
±
7.93 8.19
±
8.77 8.75
±
9.77 9.89 9.12 9.40
±
±
±
±
9.01 9.30
±
8.50 8.87
±
7.91 8.07 6.67 6.88 8.44 8.83 8.91 8.95
±
±
±
±
±
±
±
±
±
9.24 9.86
±
10.00 10.07
±
7.29 7.26 6.77 2.33 2.23 8.74
±
±
±
± ±
±
±
±
±
±
±
±
±
±
±
±
± ±
±
±
±
±
±
± ±
±
± ±
0.93a 0.99a 0.76
Values are means ± SD. no. of subjects shown in parentheses. a Mean in Down syndrome differs from mean in control by unpaired t test (p < 0.05).
J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
This study examined young adult, otherwise healthy, DS subjects with trisomy 2 1, who for the most part had never been institutionalized and were free of disease other than their primary disorder. All were younger than 40 years and had no evidence of dementia (Schapiro et aI., 1987b), suggesting that extensive AD neuropathology did not exist (Haber land, 1969; Burger and Vogel, 1973). We now have shown that these subjects do not have alterations of brain glucose metabolism, measured at rest and with reduced sensory input, as compared with healthy, age-matched controls. In addition, refer ence ratios, which examine changes in rCMRglc in regions using global gray CMRglc as a within-subject reference, show no consistent difference in the in trahemispheric distribution of rCMRg1c as compared with controls. When significant changes were found, they were biologically small and not imme diately meaningful, except perhaps at the thalamus, which another study, using a correlational matrix analysis (Horwitz et aI., in press), showed to be less functionally integrated in DS subjects, a fact that may underlie a defect in directed attention. In contrast to a previous PET scan study from this laboratory, which showed an increase in mean
CEREBRAL GLUCOSE METABOLISM IN DOWN SYNDROME
203
(top) and 70 (bottom) mm above the inferior orbitomeatal line in a control and a Down syndrome (right) subject. The color scale provides for regional CMRg,c in mg/100 g/min. There were no differences in global or regional glucose metabolism between Down syndrome and control groups. FIG. 1. Positron emission tomography images at 45
(left)
rCMRglc in four young DS subjects using ECAT II-derived data (Schwartz et aI., 1 983), the present results do not demonstrate an increase in cerebral metabolism in DS. However, there are several methodological differences between the two stud ies. First, the studies were done using different PET scanners: the first, a low-resolution ECAT II, and ours, a higher-resolution Scanditronix PC- 1024-7B. In the original study, a correction for attenuation of radiation of the skull was performed using a calcu lated ellipse method (Huang et aI., 1 98 1 ) rather than a measured transmission method. As we have shown that DS subjects have smaller heads com pared with controls ( 1 ,069 vs. 1 ,236 cm3) and others have demonstrated that DS subjects have thinner skulls (Roche and Sunderland, 1 960) than controls, the calculation method may not take into account differences in head size and skull thickness and may introduce an artifactual elevation of rCMRglc in DS subjects as compared with controls. Also, with the brachycephalic head shape in DS, differences in
placement of the ellipse in the DS and control groups might result in erroneous group differences in rCMRglc. To perform the attenuation correction using the ECT A II software the user must interac tively determine the size and position of an ellipse that best fits around the subject's skull by examin ing the 18FDG brain image. Because the DS brain is round rather than ellipitical, it might be more diffi cult to correctly size and place the ellipse. Finally, though the order of scanning of the DS patients and controls in the original study is unclear, the two groups were interleaved in time in the present study. Our results are consistent with previous investi gations that showed no difference in the global ce rebral oxygen consumption using the Kety-Schmidt nitrous oxide saturation technique (Fazekas et aI., 1 958; Lassen et aI., 1 966), and no difference in CBF using I33Xe inhalation (Risberg, 1 980) in young DS adults compared with age-matched controls. In deed, we recently demonstrated that regional CBF J Cereb Blood Flow Metab. Vol.
10. No. 2. 1990
204
M. B. SCHAPIRO ET AL.
TABLE 3. Ratio of regional CMRglc to gray matter CMRglc in control and Down syndrome subjects
Control (13) Orbital frontal Right Left Prefrontal Right Left Premotor Right Left Sensorimotor Right Left Superior parietal Right Left Medial parietal Right Left Inferior parietal Right Left Calcarine Right Left Occipital association Right Left Inferior temporal Right Left Middle temporal Right Left Superior temporal Right Left Anterior cingulate Right Left Insula Right Left Paracentral Right Left Cerebellum Right Left Thalamus Right Left Caudate Right Left Amygdala Right Left Hippocampus Right Left Lenticular Right Left
Down syndrome (14)
0. 91 0. 94
±
0. 07 0. 06
0. 86 0. 89
±
0. 07 0. 06Q
1. 00 1. 00
±
0. 03 0. 04
1. 03 1. 02
±
0. 05 0. 05
±
0. 04 0. 05
1. 11 1. 10
±
0. 04 0. 05
1. 04 1. 07
±
0. 04 0. 06
1. 10 1. 08
±
0. 05Q 0. 05
1. 05 1. 06
±
0. 04 0. 05
1. 06 1. 05
±
0. 06 0. 06
±
0. 10 0. 11
1. 06 1. 05
±
0. 07 0. 10
±
0. 06 0. 05
1. 02 1. 02
±
±
0. 06 0. 05Q
0. 98 1. 02
±
0. 08 0. 06
1. 04 1. 05
±
±
0. 08Q 0. 08
0. 91 0. 93
±
0. 08 0. 07
0. 92 0. 93
±
0. 05 0. 06
0. 76 0. 78
±
0. 06 0. 05
0. 78 0. 81
±
0. 10 0. 10
0. 97 1. 01
±
0. 06 0. 05
0. 96 0. 96
±
0. 08 0. 06Q
±
0. 03 0. 05
1. 05 1. 05
±
0. 07 0. 06
1. 01 1. 00
±
0. 08 0. 07
0. 92 0. 96
±
0. 08Q 0. 08
1. 07 1. 07
±
0. 06 0. 07
1. 05 1. 04
±
0. 09 0. 07
1. 00 1. 06
±
0. 06 0. 07
0. 99 1. 01
±
0. 08 0. 08
0. 83 0. 83
±
0. 05 0. 06
0. 77 0. 76
±
0. 05Q O. 07Q
1. 02 1. 07
±
0. 10 0. 09
1. 12 1. 15
±
0. 05Q O. IOQ
1. 05 1. 13
±
0. 13 0. 08
1. 12 1. 14
±
0. 10 0. 11
0. 64 0. 65
±
0. 07 0. 08
0. 67 0. 62
±
0. 06 0. 08
0. 81 0. 79
±
0. 10 0. 07
0. 85 0. 81
±
0. 05 0. 09
1. 14 1. 15
±
0. 09 0. 05
1. 17 1. 20
±
0. 11 0. 10
1. 11 1. 13
±
±
±
±
±
1. 08 1. 10
±
1. 04 1. 06
±
1. 02 1. 03
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
Values are means ± SD, no. of subjects shown in parentheses. Q Mean in Down syndrome differs from mean in control by unpaired t test (p < 0. 05).
J Cereb Blood Flow Metab. Vol.
10, No.2, 1990
was not different between DS subjects and controls, using many of the same DS subjects as in this report (Schapiro et aI., 1988b). In contrast to our findings with both regional CBF and CMRgIc measurements, Melamed et aI. ( 1987) reported reduced regional CBF in DS adults aged 16-44 years, similar in extent and distribution to changes seen in AD, with 133Xe. However, in their study, the karyotype and health status of the DS subjects were not described, the subjects were more severely mentally retarded than ours, older subjects (>40 years) were included (even though they frequently have AD-type neuropathology), and the presence or absence of dementia was not noted. A previous quantitative computed tomography study from our laboratory (also using the same sub jects as studied in this report) showed that these young DS adults without demetia have smaller brains and intracranial volumes than controls (Schapiro et aI., 1987a), but not out of proportion to their smaller stature. Despite this reduced brain vol ume in DS, our results do not suggest that CMRgIc or rCMRgIc is elevated in these same individuals. In contrast, Hatazawa et aI. ( 1987a,b) concluded that global CMRgIc and cortical rCMRgic were inversely proportional to a brain-size index in control sub jects and ascribed this to increased neuron packing density in smaller brains. However, their studies differed in some respects from our work. Rather than measuring brain volumes, they calculated brain size indexes from linear or area measure ments. Also, ventricular volumes were not mea sured. Since smaller normal brains have dispropor tionately smaller ventricles (Creasey et aI., 1986), this might result in less partial voluming and higher global CMRgIc than in larger brains. An alternative explanation for the discrepancy between studies may be that DS brains have a decreased neuronal density compared with normals, as Ross et aI. ( 1984) and Wisniewski et aI. ( 1986) have shown a decrease in granular cells in the neocortex of young DS children and adults. Further, abnormalities in dendritic arborization and dendritic spine density (Scott et aI., 1983), reduction in synaptic density (Wisniewski et aI., 1986), alterations in the type of synaptic contacts (Scott et aI., 1986), and reduc tions in the synaptic parameters (Scott et aI., 1986; Wisniewski et aI., 1986) also have been described in DS. As neuronal and synaptic elements metabolize the glucose in the brain, abnormalities in both ele ments may result in a relative reduction in glucose utilization in DS. With the smaller brain in DS, these histopathological changes might be reflected in no difference in absolute values in comparison with controls.
CEREBRAL GLUCOSE METABOLISM IN DOWN SYNDROME
Postmortem examination of DS adults generally does not show the neuropathologic changes of AD until after the age of 35-40 years. When the AD changes do occur, they usually are distributed in neocortical association areas (Burger and Vogel, 1973; Ellis et aI., 1 974; Wisniewski et aI., 1985; Schapiro et aI., 1988a). Furthermore, large numbers of neurofibrillary tangles and neuronal cell loss ap pear 10-20 years after large numbers of senile plaques first are demonstratable after the age of 3540 years (Haberland, 1969; Burger and Vogel, 1973). The lack of selective metabolic involvement of particular regions of the neocortex in young DS adults differs from the cerebral metabolic pattern seen in demented older DS subjects (Schapiro et aI., 1987b), where cerebral glucose utilization is re duced selectively in temporal and parietal associa tion neocortices, suggesting that when age-related reductions in brain metabolism in DS do occur, they reflect AD neuropathology. It further indicates that mean rCMRglc differences from control values can not identify young adult DS subjects at risk for AD pathology. The significance of the increased white matter rCMRglc in the DS subjects is not apparent. Less partial voluming of white matter with ventricular cerebrospinal fluid in DS might explain this result, but the DS subjects do not have smaller ventricles despite having smaller brains. Alternatively, if the gray matter/white matter ratio were increased in DS, more partial voluming of gray matter in the white matter in DS might elevate white matter rCMRglc• Also, in smaller brains but with the same resolution, there may be more effective partial vol uming of gray matter and white matter in DS PET scans. The normal bilateral symmetry of rCMRglc in young DS adults in this study is consistent with our recent study using \33Xe (Schapiro et aI., 1988b), with a lack of focal neuropathology in young adult DS brains (Zellweger, 1977) and with the symmetry seen in the controls. Despite mental retardation and cognitive abnormalities, possibly related to re ported abnormalities of dendritic arborization, den dritic spine density, and synaptic density (Scott et aI., 1983; Wisniewski et aI., 1986), young DS sub jects do not appear to show region-by-region differ ences in rCMRglc, which is considered a measure of local cerebral functional activity. Further, these young DS subjects do not show a difference in rCMRglc as a function of mental age. Such findings are consistent with results in healthy normals, where relations of brain metabolism and intelli gence measures are absent in those screened (Duara et aI., 1984).
205
REFERENCES Brooks RA (1982) Alternative formula for glucose utilization us ing labeled deoxyglucose. J Nucl Med 23:538-539 Burger PC, Vogel FS (1973) The development of the pathologic changes of Alzheimer's disease and senile dementia in pa tients with Down's syndrome. Am J Pathol 73:457-476 Creasey H, Schwartz M, Frederickson H, Haxby J, Rapoport S (1986) Quantitative computed tomography in dementia of the Alzheimer type. Neurology 36:1563-1568 Duara R, Margolin RA, Robertson-Tchabo EA, London ED, Schwartz M, Renfrew JW, Koziarz BJ, Sundaram M, Grady C, Moore AM, Ingvar DH, Sokoloff L, Weingartner H, Kessler RM, Manning RG, Channing MA, Cutler NR, Rap oport SI (1983) Cerebral glucose utilization, as measured with positron emission tomography in 21 resting healthy men between the ages of 21 and 83 years. Brain 106:761-775 Duara R, Grady C, Haxby J, Ingvar D, Sokoloff L, Margolin RA, Manning RG, Cutler NR, Rapoport SI (1984) Human brain glucose utilization and cognitive function in relation to age. Ann Neurol 16:702-713 Dunn LM, Dunn LM (1981) Peabody Picture Vocabulary Test Revised. Circle Pines, NM, American Guidance Service Ellis WG, McCulloch JR, Corley CL (1974) Presenile dementia in Down's syndrome. Neurology 24:101-106 Eycleshymer AC, Schoemaker DM (1911) A Cross-Section Anatomy. New York, D. Appleton, pp 2-55 Fazekas JF, Ehrmantraut WR, Shea JG, Kleh J (1958) Cerebral hemodynamics and metabolism in mental deficiency. Neu rology 8:558-560 Haberland C (1969) Alzheimer's disease in Down syndrome: clinical-neuropathological observations. Acta Neurol Belg 69:369-380 Hatazawa J, Brooks RA, Di Chiro G, Bacharach SL (l987a) Glucose utilization rate versus brain size in humans. Neu rology 37:583-588 Hatazawa J, Brooks RA, Di Chiro G, Campbell G (1987b) Global cerebral glucose utilization is independent of brain size: a PET study. J Comput Assist Tomogr 11:571-576 Horwitz B, Schapiro M, Grady C, Rapoport SI (in press) Cere bral metabolic patterns in young adult Down syndrome: al tered intercorrelations between regional rates of glucose uti lization. J Ment Defic Res (in press) Huang SC, Phelps ME, Hoffman EJ, Sideris K, Selin CJ, Kuhl DE (1980) Noninvasive determination of local cerebral met abolic rate of glucose in man. Am J Physiol 238:E69--E82 Huang SC, Carson RE, Phelps ME, Hoffman EJ, Schelbert H, Kuhl DE (1981) A boundary method for attenuation correc tion in positron computed tomography. J Nucl Med 22:627637 Lassen NA, Christensen S, Hoedt-Rasmussen K, Stewart BM (1966) Cerebral oxygen consumption in Down's syndrome. Arch Neurol 15:595--602 Melamed E, Mildworf B, Sharav T, Belenky L, Wertman E (1987) Regional cerebral blood flow in Down's syndrome. Ann Neurol 22:275-278 Risberg J (1980) Regional cerebral blood flow measurements by 133 Xe-inhalation: methodology and applications in neuro psychology and psychiatry. Brain Lang 9:9--34 Roche AF, Sunderland S (1960) Postmortem observations of mongoloid crania. J Neuropathol Exp Neurol 19:554-558 Ross MH, Galaburda AM, Kemper TL (1984) Down's syn drome: is there a decreased population of neurons? Neurol ogy 34:909--916 SAS Institute (1985) SAS User's Guide: Statistics, Version 5. Cary, NC, SAS Institute Schapiro MB, Creasey H, Schwartz M, Haxby JV, White B, Moore A, Rapoport SI (l987a) Quantitative CT analysis of brain morphometry in adult Down's syndrome at different ages. Neurology 37:1424-1427 Schapiro MB, Haxby JV, Grady CL, Duara R, Schlageter NL, White B, Moore A, Sundaram M, Larson SM, Rapoport SI (l987b) Decline in cerebral glucose utilisation and cognitive
J Cereb Blood Flow Metab, Vol.
10, No. 2, 1990
M. B. SCHAPIRO ET AL.
206
function with aging in Down's syndrome. J Neurol Neuro surg Psychiatry 50: 766-774 Schapiro MB, Ball MJ, Grady CL, Haxby JV, Kaye JA, Rapo port SI (1988a) Dementia in Down's syndrome: cerebral glu cose utilization, neuropsychological assessment and neuro pathology. Neurology 38: 938-942 Schapiro MB, Berman KF, Friedland RP, Weinberger DR, Rap oport SI (1988b) Regional cerebral blood flow is not de creased in young adult Down syndrome. Ann Neuro124:310 Schwartz M, Duara R, Haxby J, Grady C, White BJ, Kessler RM, Kay AD, Cutler NR, Rapoport SI (1983) Down's syn drome in adults: brain metabolism. Science 221: 781-783 Schwartz M, Creasey H, Grady CL, DeLeo JM, Frederickson HA, Cutler NR, Rapoport SI (1985) Computed tomographic analysis of brain morphometrics in 30 healthy men, aged 21 to 81 years. Ann Neurol 17: 146--157 Scott BS, Becker LE, Petit TL (1983) Neurobiology of Down's syndrome. In: Progress in Neurobiology, VoI2I. Great Brit ain: Pergamon Press, pp 199-237 Shiue CY, Salvadori PA, Wolf AP, Fowler JS, MacGregor R
J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
(1982) A new improved synthesis of 2-deoxy-2-[18Flfluoro d-glucose from 18F-labeled acetyl hypofluorite. J Nucl Med 23:899-903 Sokoloff L, Reivich M, Kennedy C, DesRosiers MH, Patlak CS, Pettigrew KD, Sakurada 0, Shinohara M (1977) The [14C]-deoxyglucose method for the measurement of local ce rebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neu rochem 28: 897-916 Wisniewski KE, Wisniewski HM, Wen GY (1985) Occurrence of neuropathological changes and dementia of Alzheimer's dis ease in Down's syndrome. Ann NeuroI17:278-282 Wisniewski KE, Laure-Kamionowska M, Connell F, Wen GY (1986) Neuronal density and synaptogenesis in the postnatal stage of brain maturation in Down syndrome. In: The Neu robiology of Down syndrome (Epstein CJ, ed), New York, Raven Press, pp 29-44 Zellweger H (1977) Down syndrome. In: Handbook of Clinical Neurology, Vol 31 (Vinken PJ, Bruyn GW, eds), New York, North Holland Publishing Co., pp 367-469
10:199--206 © 1990 Raven Press, Ltd., New York
Regional Cerebral Glucose Metabolism is Normal in Young Adults with Down Syndrome
Mark B. Schapiro, Cheryl L. Grady, Anand Kumar, tPeter Herscovitch, James V. Haxby, Angela M. Moore, *Beverly White, Robert P. Friedland, and Stanley I. Rapoport Laboratory of Neurosciences, Section on Brain Aging and Dementia, National Institute on Aging, Clinical Center, and *Laboratory of Chemical Biology, National Institute of Diabetes, Digestive and Kidney Diseases, and fDepartment of Nuclear Medicine, National Institutes of Health, Clinical Center, Bethesda, Maryland, U.S.A.
Summary: Regional CMRglc (rCMRg1c) values were mea sured with ['sF]2-fluoro-2-deoxY-D-glucose eSFDG) and positron emission tomography (PET), using a Scan ditronix PC-1024-7B scanner, in 14 healthy, noninstitu tionalized subjects with trisomy 21 (Down syndrome; DS) (mean age 30.0 years, range 25-38 years) and in 13 sex matched, healthy volunteers (mean age 29.5 years, range 22-38 years). In the DS group, mean mental age on the Peabody Picture Vocabulary Test was 7.8 years and de mentia was not present. Resting rCMRglc was determined with eyes covered and ears occluded in a quiet, darkened room. Global gray CMRglc equaled 8.76 ± 0.76 mg/IOO g/min (mean ± SD) in the DS group as compared with
8.74 ± 1.19 mg/lOO g/min in the control group (p > 0.05). Gray matter regional measurements also did not differ between groups. The ratio of rCMRglc to global CMRg1C' calculated to reduce the variance associated with abso lute rCMRglc, and rightlleft ratios did not show any con sistent differences. These results show that healthy young DS adults do not have alterations in regional or global brain glucose metabolism, as measured with lsFDG and PET, prior to an age at which the neuropathological changes in Alzheimer disease are reported to occur. Key Words: Brain-Cerebral glucose metabolism-Down syndrome-[ISFj2-Fluoro-2-deoxY-D-glucose-Mental retardation-Positron emission tomography.
A previous study in our laboratory using positron emission tomography (PET) and eSp]2-fluoro-2-de oxy-D-glucose eSPDG) showed that Down syn drome (DS) adults over the age of 45 years have reductions in parietal and temporal glucose metab olism in comparison with younger DS adults (Scha piro et al., 1987b). However, another study in our laboratory did not find this pattern in four younger DS adults when compared with controls, although these young DS subjects had a mean global increase in glucose metabolism (Schwartz et al., 1983). Be cause this latter study used a calculated correction
for attenuation of radiation in the determination of brain radioactivity with PET (Huang et al., 1981), with an ECAT II tomograph, it was possible that the smaller brain size (Schapiro et al., 1987a) and possibly thinner skulls (Roche and Sunderland, 1960) of the DS subjects accounted for the differ ences in cerebral metabolic rates between DS and controls. To avoid this problem and to reevaluate brain metabolism in DS, we investigated cerebral glucose metabolism in a carefully screened, larger group of young adults with trisomy 2 1, now using a higher resolution PET tomograph (Scanditronix PC1024-7B) with a measured attenuation correction. This was done to see if there are differences in ab solute cerebral metabolic rate between DS adults and age-matched normals or in patterns of regional metabolism prior to the age of 35-40 years, when Alzheimer disease (AD) neuropathology is reported to occur (Haberland, 1969; Burger and Vogel, 1973; Wisniewski et al., 1985).
Received January 31, 1989; revised September 12, 1989; ac cepted September 20, 1989. Address correspondence and reprint requests to Dr. M. B. Schapiro at Laboratory of Neurosciences, National Institute on Aging, Bldg. 10, Rm. 6C414, Bethesda, MD 20892, U. S.A. Abbreviations used: AD, Alzheimer disease; DS, Down syn drome; 18FDG, [18F12-fluoro-2-deoxY-D-glucose; 10M, inferior orbitomeata1; PET, positron emission tomography; rCMRglc' re gional CMRglc; ROI, region of interest.
1 99
M. B. SCHAPIRO ET AL.
200
MATERIALS AND METHODS Selection of subjects
We studied 14 young adult DS subjects aged 25-38 years (10 men and 4 women, mean age 30.0 years). All were recruited from family or group homes and had tri somy 21. Only two subjects had lived in an institution in the past. Mean mental age, as measured on the Peabody Picture Vocabulary Test (Dunn and Dunn, 1981), was 7.8 years (range 4-16 years). All subjects were screened for dementia of the Alzheimer type using modified criteria from the Diagnostic and Statistical Manual, which spec ified an acquired, progressive loss of intellectual function (Schapiro et al., 1987b). Dementia was not present in the DS subjects in this study (Schapiro et al., 1987b). Screen ing included a review of the medical history, a physical and neurological examination, and the following labora tory studies: routine blood counts, electrolytes, liver and renal function tests, cholesterol, triglycerides, thyroid function tests, Venereal Disease Research Laboratories test, antinuclear antibody (ANA), serum folate and B\2, ECG, and chest radiographs. Of the young DS adults, five had functional systolic murmurs and one had a mitral valve prolapse. Two were euthyroid on thyroid replacement and one was biochem ically hypothyroid while on thyroid replacement. Another subject was receiving intramuscular vitamin B 2 for celiac 1 sprue. With these exceptions, the DS subjects had no other disease. None was taking medication (except as noted above). The controls, 13 sex-matched, healthy volunteers aged 22-38 years (9 men and 4 women, mean 29.5 years), were participants in the National Institute on Aging Laboratory of Neurosciences Study on Aging (Duara et al., 1983) and were screened with medical, neurological, and laboratory tests. Subjects had no history of neuro logic, psychiatric, or systemic medical disorder. None was taking medication for at least 2 weeks prior to the study. Controls signed an informed consent describing the purpose of the study, the tests performed, and the risks involved. For the DS subjects, a relative and/or a legal guardian was required to sign the informed consent. The research was conducted under NIH protocols 81-AG-1O (DS) and 80-AG-26 (controls). PET technique
The global gray matter CMRg1c and regional CMRgJc (rCMRgJc) were measured in the resting state using 18FDG and a multislice PET scanner (Scanditronix PC-1024-7B, Uppsala, Sweden). The 18FDG was prepared by the method of Shiue et al. (1982). 18F2 was produced at the 2 NIH cyclotron via the °Ne (a,a) 18F nuclear reaction. The isotonic, buffered product was checked for chemical and radiochemical purity using both high pressure liquid chromatography and thin layer chromatography. Scans of patients were interspersed with scans of control subjects to avoid drifts of metabolic values due to possible changes in machine-operating characteristics or 18FDG synthesis. In the tomograph, there are four rings of de tectors that allow simultaneous collection of data from seven transverse sections separated by 13.8 mm center to center. Spatial resolution (full width at half-maximum) was 6 mm in the image planes, and axial resolution was 11 mm for straight slices and 8 mm for cross-slices in the center of the field of view.
J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
Subjects fasted for at least 2 h and refrained from al cohol, caffeine, and nicotine for at least 24 h prior to the PET scan. After percutaneous cannulation of the radial artery with cutaneous anesthesia and placement of an in travenous catheter, the subject was positioned in the PET scanner. The head was immobilized with a thermoplastic mask that was molded to the shape of the head. A laser light and a line drawn on the mask corresponding to the inferior orbitomeatal (10M) line were used to position the subject in the scanner. With the ears plugged with cotton and eyes patched, the subject lay supine in a quiet and darkened scan room (resting state). A transmission scan, using a ring phantom of 18F, was performed to correct for attenuation of radiation by the skull, with slices parallel to and from 10 to 100 mm above the 10M line. Two transmission scans were performed, offset by 6.9 mm (half-slice separation) in the z-axis, to provide an interleaved set of 14 slices. Mter the transmission scan was completed, the subject received an intravenous bolus injection of 5 mCi of 18FDG. Serial arterial blood samples were drawn during the 45-min uptake period and during the emission scan and were centrifuged to provide plasma for measurement of radioactivity and glucose. After 45 min, lO-min emis sion scans were performed at the same two levels as for the transmission scans. PET slices had at least 2 x 106 coincidence counts. PET images were reconstructed and corrected for attenuation using data from the transmis sion scans. Counts from the brain were converted to brain radioactivities (nCi/cm3), using a calibration factor de rived by scanning a homogeneous, cylindrical phantom with a known 18FDG concentration, within several hours of each scan. Plasma and brain radioactivities were cor rected for physical decay to the time of injection. Data analysis
From the time courses of blood and brain radioactivi ties, of blood glucose concentration, and from the time of scanning, metabolic rates were calculated in units of mil ligrams per 100 g per minute using Brooks' modification (1982) of the operational equation of Sokoloff et al. (1977). A value of 0.418 was used for the lumped constant (Huang et al., 1980). Values for the kinetic rate constants were kl 0.102, k2 0.130, k3 0.062, and k4 0.0068 (in units of min - I ) (Huang et al., 1980). A template was derived from a PET scan of a healthy normal subject. The template consisted of circular re 2 gions of interest (ROls) 8 mm in diameter (50 mm ). From the set of 14 PET slices of this individual, the slices best representative of the neocortical areas of interest were identified through comparison with an atlas of a human brain sectioned in the same plane as the PET scans (Ey cleshymer and Schoemaker, 1911). The ROls were spaced evenly throughout the cortex and also were cen tered in subcortical areas, including the thalamus and len ticular and caudate nuclei. ROls for subcortical struc tures were placed on the single slice in which they were seen best. The white matter ROls were placed in the cen trum semiovale bilaterally. There were 201 ROls on eight slices, which were averaged over larger regions. For the analysis of study subjects, PET slices were compared with anatomical sections from an atlas of a human brain (Eycleshymer and Schoemaker, 1911). Be cause of individual differences in head size and shape, the height above the 10M line of the best-fit slice in the atlas was assigned to each PET slice. By comparison of the =
=
=
=
CEREBRAL GLUCOSE METABOLISM IN DOWN SYNDROME
PET slice with its best-fit slice in the atlas, we identified anatomical ROIs in the PET scans. Since the DS subjects had smaller and more rounded heads, it was not possible to blind the data analysis procedure. The template, which was derived from the normal subject as described above, was placed over the corresponding PET slices for each subjects and adjusted to fit the individual brains. Because ROIs could be individually moved, adjustments could be made for differences in head size or shape, such as the brachycephaly noted in the DS subjects. Global metabolism represented the weighted mean of all gray matter regions, excluding the vermis and mid brain. Two indexes of metabolic patterns were examined: (a) the ratio of rCMRglC to global CMRg!c and (b) an asym metry index of rCMRg!c, calculated as (R - L)/[(R + L)/2], where Rand L represent right and left metabolic rates, respectively. Other measurements
Supine resting heart rate and systolic and diastolic blood pressures were measured on a non-PET day and prior to the injection of 18FDG and after the emission scan on a PET day. Each subject was given a rating of anxiety (Duara et ai., 1984). Arterial blood was obtained during the PET scan for determination of glucose. Noncontrast computed tomography scans of brain were performed with a GE 8800 CT scanner (General Electric Co., Milwaukee, WI, U.S.A.) using previously described methodology for scanning and analysis (Schwartz et al., 1985). Statistics
For physiologic data, group differences were compared with an unpaired t test (SAS Institute, 1985). Within groups, differences in these data between mean values from non-PET and PET days, as well as differences in means from preinjection and postinjection, were com pared with a paired t test (SAS Institute, 1985). For metabolic and computed tomography data, differ ences between mean values in the DS and control groups were analyzed with an unpaired t test (SAS Institute, 1985). Comparisons of high and low intellectually func tioning DS subjects were performed with a Wilcoxon two sample test (SAS Institute, 1985). Within groups, right and left regional metabolic values were compared with a paired t test (SAS Institute, 1985). The criterion of statis tical significance was p < 0.05. RESULTS
Table 1 lists the physiologic data for the DS sub jects and controls. DS subjects were significantly shorter than controls. On a non-PET day, supine resting heart rate and systolic and diastolic blood pressures were significantly lower in the DS group as compared with controls. On the day of the PET scan, the DS and control groups did not differ prior to injection on mean heart rate or blood pressure, but postscan systolic and diastolic blood pressures were significantly lower in the DS group. Supine heart rate and blood pressure were not significantly different on PET and non-PET days for DS sub jects, though diastolic blood pressure was less on
201
TABLE 1. Physiologic values for Down syndrome control and subjects
Age (yrs) Height (m) Weight (kg) Measurements on non-PET day Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Measurements on PET day Preinjection Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Postinjection Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (beats/min) Glucose (mg/IOO ml) Anxiety rating (1-3)
Controls
Down syndrome
29.5 ± 4.5 (13) 1.76 ± 0.06 (8) 71.5 ± 10.9 (II)
30.0 ± 4.3 (14) 1.55 ± 0.08 (l4)a 66.1 ± 11.2 (14)
122 ± 11 79 ± 4 77 ± 6
(10) (10) (10)
107 ± 10 70 ± 7 70 ± 9
(l4)a (l4)a (l4)a
124 ± 11 74 ± 5 70 ± 7
(12) (l2)b (11)
118 ± 13 70 ± 11 71 ± 10
(13) (13) (12)
130 ± 12 80 ± 7 70 ± 4 103 ± 8 0.3 ± 0.6
(10) (lO)C (9) (13) (II)
III 67 68 100 0.6
±
± ± ± ±
8 10 10 9 0.7
(13)a (13)a (13)C (14) (14)
Values are means ± SD, no. of subjects shown in parentheses. PET, positron emission tomography; BP, blood pressure. a Mean in Down syndrome differs from mean in control by unpaired t test (p < 0.05). b Non-PET day significantly different from PET day by paired t test (p < 0.05). Post injection significantly different from preinjection by paired t test (p < 0.05). C
the day of the PET in the controls. The anxiety rating, a measure of stress, did not differ between groups. DS subjects, in comparison with healthy normals, had significantly (p < 0.05) decreased total intracra nial volume [ 1,069 ± 93 vs. 1,236 ± 1 13 (SD) cm3], seven-slice intracranial volume (the 50-mm segment of brain, starting at the lowest slice that contained the body of the third ventricle; 688 ± 4 1 vs. 784 ± 57 cm3), and gray matter plus white matter volume (659 ± 40 vs. 776 ± 64 cm3). There was no signifi cant difference in third ventricle volume or right or left ventricle volume between the DS and control groups, either directly or after normalizing to seven-slice intracranial volume. Table 2 shows absolute values of global CMRg\c and of rCMRglc for the DS and control groups. Global CMRg\c equaled 8.76 ± 0.76 (SD) mg/100 g/min in the DS group as compared with 8.74 ± 1. 19 mgllOO g/min in controls (p > 0.05). Global and gray matter regional measurements did not differ be tween groups (p > 0.05), though white matter rCMRg\c was increased in the DS as compared with the control group (p < 0.05). PET images at two levels (Fig. 1) showed no differences in appearance between a DS and a control subject. Comparison of the six DS subjects above the age of 30 years (3�38 years old) with the eight DS subjects below the age of 30 years (25-29 years old) did not show any dif ferences in global or gray matter regional measure ments. The asymmetry index of rCMRg\c failed to show J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
202
M. B. SCHAPIRO ET AL.
±
1.09 1.10
±
0.96 0.88
±
0.98 1.09
±
0.95 0.93
±
0.76 0.88
±
0.77 1.15
±
1.02 0.93
9.10 9.22
±
0.97 1.05
0.77 0.93
8.01 8.16
±
0.53 0.67
±
1.15 1.32
6.79 7.10
±
0.95 1.01
±
1.09 1.21
8.41 8.44
±
0.73 0.86
significant differences between groups (p > 0.05). Within groups, significant asymmetries were not present, except in the middle temporal and paracen tral regions in the control group (left> right metab olism), although one significant asymmetry would be expected by chance (assuming independence). Table 3 shows mean ratios of rCMRglc to global CMRglc, which were calculated to reduce the vari ance associated with absolute values of rCMRglc (Duara et aI., 1984) and to examine patterns of me tabolism. DS and control groups differed in several scattered regions. The ratio was lower in the DS group in the left orbitofrontal, left inferior parietal, left middle temporal, and right anterior cingulate regions and the cerebellum bilaterally; the ratio was elevated in the DS group in the right sensorimotor and right calcarine regions and the thalamus bilat erally. These differences, however, were very slight (differences between ratios of ±0.04 were statisti cally significant at p < 0.05), and moreover, two would be expected by chance alone at p < 0.05. Further studies would be required to evaluate the biological meaning of each difference. Because of the variation in mental ages, the DS group was divided into high (>8 years) and low «8 years) functioning subjects. There was no differ ence between these subgroups for any measure of rCMRglc•
±
1.23 1.14
9.18 9.25
±
1.13 1.12
DISCUSSION
8.85 8.75 10.10
±
1.47 1.46 1.61
8.12 8.37 9.49
±
1.21 1.01 0.99
9.34 9.32
±
1.31 1.18
9.18 9.13
±
1.15 1.02
8.73 9.25
±
1.10 1.49
8.65 8.82
±
0.99 1.01
8.92 9.41
±
1.67 1.80
9.84 10.03
±
1.10 1.02
±
1.86 1.13
9.87 9.99
±
±
1.37 1.59
±
1.91 1.31
10.29 10.56
±
±
1.40 1.59
5.57 5.65
±
0.90 0.76
5.84 5.44
±
0.65 0.77
7.05 6.87
±
0.99 0.77
7.39 7.04
±
0.63 0.69
±
1.31 1.38 1.30
6.72 6.64 6.75
±
0.77 0.93 0.81
0.61 0.65 1.19
3.07 3.03 8.76
TABLE 2. Absolute CMRglc in control and Down syndrome subjects Regional CMRglc (mg/loo g/min) Brain region
Control (13)
Orbital frontal Right Left Prefrontal Right Left Premotor Right Left Sensorimotor Right Left Superior parietal Right Left Medial parietal Right Left Inferior parietal Right Left Calcarine Right Left Occipital association Right Left Inferior temporal Right Left Middle temporal Right Left Superior temporal Right Left Anterior cingulate Right Left Posterior cingulate Insula Right Left Paracentral Right Left Thalamus Right Left Caudate Right Left Lenticular Right Left Amygdala Right Left Hippocampus Right Left Cerebellum Right Left Vermis White matter (centrum semiovale) Right Left Global gray matter
Down syndrome (14)
±
1.30 1.32
7.58 7.79
±
±
1.40 1.38
9.00 8.91
±
±
1.62 1.68
9.69 9.68
±
±
1.41 1.61
9.66 9.47
9.16 9.28
±
1.49 1.66
9.31 9.16
9.40 9.64
±
1.07 1.64
9.26 9.19
±
0.96 1.44
8.91 8.93
±
0.98 0.90
±
7.93 8.19
±
8.77 8.75
±
9.77 9.89 9.12 9.40
±
±
±
±
9.01 9.30
±
8.50 8.87
±
7.91 8.07 6.67 6.88 8.44 8.83 8.91 8.95
±
±
±
±
±
±
±
±
±
9.24 9.86
±
10.00 10.07
±
7.29 7.26 6.77 2.33 2.23 8.74
±
±
±
± ±
±
±
±
±
±
±
±
±
±
±
±
± ±
±
±
±
±
±
± ±
±
± ±
0.93a 0.99a 0.76
Values are means ± SD. no. of subjects shown in parentheses. a Mean in Down syndrome differs from mean in control by unpaired t test (p < 0.05).
J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
This study examined young adult, otherwise healthy, DS subjects with trisomy 2 1, who for the most part had never been institutionalized and were free of disease other than their primary disorder. All were younger than 40 years and had no evidence of dementia (Schapiro et aI., 1987b), suggesting that extensive AD neuropathology did not exist (Haber land, 1969; Burger and Vogel, 1973). We now have shown that these subjects do not have alterations of brain glucose metabolism, measured at rest and with reduced sensory input, as compared with healthy, age-matched controls. In addition, refer ence ratios, which examine changes in rCMRglc in regions using global gray CMRglc as a within-subject reference, show no consistent difference in the in trahemispheric distribution of rCMRg1c as compared with controls. When significant changes were found, they were biologically small and not imme diately meaningful, except perhaps at the thalamus, which another study, using a correlational matrix analysis (Horwitz et aI., in press), showed to be less functionally integrated in DS subjects, a fact that may underlie a defect in directed attention. In contrast to a previous PET scan study from this laboratory, which showed an increase in mean
CEREBRAL GLUCOSE METABOLISM IN DOWN SYNDROME
203
(top) and 70 (bottom) mm above the inferior orbitomeatal line in a control and a Down syndrome (right) subject. The color scale provides for regional CMRg,c in mg/100 g/min. There were no differences in global or regional glucose metabolism between Down syndrome and control groups. FIG. 1. Positron emission tomography images at 45
(left)
rCMRglc in four young DS subjects using ECAT II-derived data (Schwartz et aI., 1 983), the present results do not demonstrate an increase in cerebral metabolism in DS. However, there are several methodological differences between the two stud ies. First, the studies were done using different PET scanners: the first, a low-resolution ECAT II, and ours, a higher-resolution Scanditronix PC- 1024-7B. In the original study, a correction for attenuation of radiation of the skull was performed using a calcu lated ellipse method (Huang et aI., 1 98 1 ) rather than a measured transmission method. As we have shown that DS subjects have smaller heads com pared with controls ( 1 ,069 vs. 1 ,236 cm3) and others have demonstrated that DS subjects have thinner skulls (Roche and Sunderland, 1 960) than controls, the calculation method may not take into account differences in head size and skull thickness and may introduce an artifactual elevation of rCMRglc in DS subjects as compared with controls. Also, with the brachycephalic head shape in DS, differences in
placement of the ellipse in the DS and control groups might result in erroneous group differences in rCMRglc. To perform the attenuation correction using the ECT A II software the user must interac tively determine the size and position of an ellipse that best fits around the subject's skull by examin ing the 18FDG brain image. Because the DS brain is round rather than ellipitical, it might be more diffi cult to correctly size and place the ellipse. Finally, though the order of scanning of the DS patients and controls in the original study is unclear, the two groups were interleaved in time in the present study. Our results are consistent with previous investi gations that showed no difference in the global ce rebral oxygen consumption using the Kety-Schmidt nitrous oxide saturation technique (Fazekas et aI., 1 958; Lassen et aI., 1 966), and no difference in CBF using I33Xe inhalation (Risberg, 1 980) in young DS adults compared with age-matched controls. In deed, we recently demonstrated that regional CBF J Cereb Blood Flow Metab. Vol.
10. No. 2. 1990
204
M. B. SCHAPIRO ET AL.
TABLE 3. Ratio of regional CMRglc to gray matter CMRglc in control and Down syndrome subjects
Control (13) Orbital frontal Right Left Prefrontal Right Left Premotor Right Left Sensorimotor Right Left Superior parietal Right Left Medial parietal Right Left Inferior parietal Right Left Calcarine Right Left Occipital association Right Left Inferior temporal Right Left Middle temporal Right Left Superior temporal Right Left Anterior cingulate Right Left Insula Right Left Paracentral Right Left Cerebellum Right Left Thalamus Right Left Caudate Right Left Amygdala Right Left Hippocampus Right Left Lenticular Right Left
Down syndrome (14)
0. 91 0. 94
±
0. 07 0. 06
0. 86 0. 89
±
0. 07 0. 06Q
1. 00 1. 00
±
0. 03 0. 04
1. 03 1. 02
±
0. 05 0. 05
±
0. 04 0. 05
1. 11 1. 10
±
0. 04 0. 05
1. 04 1. 07
±
0. 04 0. 06
1. 10 1. 08
±
0. 05Q 0. 05
1. 05 1. 06
±
0. 04 0. 05
1. 06 1. 05
±
0. 06 0. 06
±
0. 10 0. 11
1. 06 1. 05
±
0. 07 0. 10
±
0. 06 0. 05
1. 02 1. 02
±
±
0. 06 0. 05Q
0. 98 1. 02
±
0. 08 0. 06
1. 04 1. 05
±
±
0. 08Q 0. 08
0. 91 0. 93
±
0. 08 0. 07
0. 92 0. 93
±
0. 05 0. 06
0. 76 0. 78
±
0. 06 0. 05
0. 78 0. 81
±
0. 10 0. 10
0. 97 1. 01
±
0. 06 0. 05
0. 96 0. 96
±
0. 08 0. 06Q
±
0. 03 0. 05
1. 05 1. 05
±
0. 07 0. 06
1. 01 1. 00
±
0. 08 0. 07
0. 92 0. 96
±
0. 08Q 0. 08
1. 07 1. 07
±
0. 06 0. 07
1. 05 1. 04
±
0. 09 0. 07
1. 00 1. 06
±
0. 06 0. 07
0. 99 1. 01
±
0. 08 0. 08
0. 83 0. 83
±
0. 05 0. 06
0. 77 0. 76
±
0. 05Q O. 07Q
1. 02 1. 07
±
0. 10 0. 09
1. 12 1. 15
±
0. 05Q O. IOQ
1. 05 1. 13
±
0. 13 0. 08
1. 12 1. 14
±
0. 10 0. 11
0. 64 0. 65
±
0. 07 0. 08
0. 67 0. 62
±
0. 06 0. 08
0. 81 0. 79
±
0. 10 0. 07
0. 85 0. 81
±
0. 05 0. 09
1. 14 1. 15
±
0. 09 0. 05
1. 17 1. 20
±
0. 11 0. 10
1. 11 1. 13
±
±
±
±
±
1. 08 1. 10
±
1. 04 1. 06
±
1. 02 1. 03
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
Values are means ± SD, no. of subjects shown in parentheses. Q Mean in Down syndrome differs from mean in control by unpaired t test (p < 0. 05).
J Cereb Blood Flow Metab. Vol.
10, No.2, 1990
was not different between DS subjects and controls, using many of the same DS subjects as in this report (Schapiro et aI., 1988b). In contrast to our findings with both regional CBF and CMRgIc measurements, Melamed et aI. ( 1987) reported reduced regional CBF in DS adults aged 16-44 years, similar in extent and distribution to changes seen in AD, with 133Xe. However, in their study, the karyotype and health status of the DS subjects were not described, the subjects were more severely mentally retarded than ours, older subjects (>40 years) were included (even though they frequently have AD-type neuropathology), and the presence or absence of dementia was not noted. A previous quantitative computed tomography study from our laboratory (also using the same sub jects as studied in this report) showed that these young DS adults without demetia have smaller brains and intracranial volumes than controls (Schapiro et aI., 1987a), but not out of proportion to their smaller stature. Despite this reduced brain vol ume in DS, our results do not suggest that CMRgIc or rCMRgIc is elevated in these same individuals. In contrast, Hatazawa et aI. ( 1987a,b) concluded that global CMRgIc and cortical rCMRgic were inversely proportional to a brain-size index in control sub jects and ascribed this to increased neuron packing density in smaller brains. However, their studies differed in some respects from our work. Rather than measuring brain volumes, they calculated brain size indexes from linear or area measure ments. Also, ventricular volumes were not mea sured. Since smaller normal brains have dispropor tionately smaller ventricles (Creasey et aI., 1986), this might result in less partial voluming and higher global CMRgIc than in larger brains. An alternative explanation for the discrepancy between studies may be that DS brains have a decreased neuronal density compared with normals, as Ross et aI. ( 1984) and Wisniewski et aI. ( 1986) have shown a decrease in granular cells in the neocortex of young DS children and adults. Further, abnormalities in dendritic arborization and dendritic spine density (Scott et aI., 1983), reduction in synaptic density (Wisniewski et aI., 1986), alterations in the type of synaptic contacts (Scott et aI., 1986), and reduc tions in the synaptic parameters (Scott et aI., 1986; Wisniewski et aI., 1986) also have been described in DS. As neuronal and synaptic elements metabolize the glucose in the brain, abnormalities in both ele ments may result in a relative reduction in glucose utilization in DS. With the smaller brain in DS, these histopathological changes might be reflected in no difference in absolute values in comparison with controls.
CEREBRAL GLUCOSE METABOLISM IN DOWN SYNDROME
Postmortem examination of DS adults generally does not show the neuropathologic changes of AD until after the age of 35-40 years. When the AD changes do occur, they usually are distributed in neocortical association areas (Burger and Vogel, 1973; Ellis et aI., 1 974; Wisniewski et aI., 1985; Schapiro et aI., 1988a). Furthermore, large numbers of neurofibrillary tangles and neuronal cell loss ap pear 10-20 years after large numbers of senile plaques first are demonstratable after the age of 3540 years (Haberland, 1969; Burger and Vogel, 1973). The lack of selective metabolic involvement of particular regions of the neocortex in young DS adults differs from the cerebral metabolic pattern seen in demented older DS subjects (Schapiro et aI., 1987b), where cerebral glucose utilization is re duced selectively in temporal and parietal associa tion neocortices, suggesting that when age-related reductions in brain metabolism in DS do occur, they reflect AD neuropathology. It further indicates that mean rCMRglc differences from control values can not identify young adult DS subjects at risk for AD pathology. The significance of the increased white matter rCMRglc in the DS subjects is not apparent. Less partial voluming of white matter with ventricular cerebrospinal fluid in DS might explain this result, but the DS subjects do not have smaller ventricles despite having smaller brains. Alternatively, if the gray matter/white matter ratio were increased in DS, more partial voluming of gray matter in the white matter in DS might elevate white matter rCMRglc• Also, in smaller brains but with the same resolution, there may be more effective partial vol uming of gray matter and white matter in DS PET scans. The normal bilateral symmetry of rCMRglc in young DS adults in this study is consistent with our recent study using \33Xe (Schapiro et aI., 1988b), with a lack of focal neuropathology in young adult DS brains (Zellweger, 1977) and with the symmetry seen in the controls. Despite mental retardation and cognitive abnormalities, possibly related to re ported abnormalities of dendritic arborization, den dritic spine density, and synaptic density (Scott et aI., 1983; Wisniewski et aI., 1986), young DS sub jects do not appear to show region-by-region differ ences in rCMRglc, which is considered a measure of local cerebral functional activity. Further, these young DS subjects do not show a difference in rCMRglc as a function of mental age. Such findings are consistent with results in healthy normals, where relations of brain metabolism and intelli gence measures are absent in those screened (Duara et aI., 1984).
205
REFERENCES Brooks RA (1982) Alternative formula for glucose utilization us ing labeled deoxyglucose. J Nucl Med 23:538-539 Burger PC, Vogel FS (1973) The development of the pathologic changes of Alzheimer's disease and senile dementia in pa tients with Down's syndrome. Am J Pathol 73:457-476 Creasey H, Schwartz M, Frederickson H, Haxby J, Rapoport S (1986) Quantitative computed tomography in dementia of the Alzheimer type. Neurology 36:1563-1568 Duara R, Margolin RA, Robertson-Tchabo EA, London ED, Schwartz M, Renfrew JW, Koziarz BJ, Sundaram M, Grady C, Moore AM, Ingvar DH, Sokoloff L, Weingartner H, Kessler RM, Manning RG, Channing MA, Cutler NR, Rap oport SI (1983) Cerebral glucose utilization, as measured with positron emission tomography in 21 resting healthy men between the ages of 21 and 83 years. Brain 106:761-775 Duara R, Grady C, Haxby J, Ingvar D, Sokoloff L, Margolin RA, Manning RG, Cutler NR, Rapoport SI (1984) Human brain glucose utilization and cognitive function in relation to age. Ann Neurol 16:702-713 Dunn LM, Dunn LM (1981) Peabody Picture Vocabulary Test Revised. Circle Pines, NM, American Guidance Service Ellis WG, McCulloch JR, Corley CL (1974) Presenile dementia in Down's syndrome. Neurology 24:101-106 Eycleshymer AC, Schoemaker DM (1911) A Cross-Section Anatomy. New York, D. Appleton, pp 2-55 Fazekas JF, Ehrmantraut WR, Shea JG, Kleh J (1958) Cerebral hemodynamics and metabolism in mental deficiency. Neu rology 8:558-560 Haberland C (1969) Alzheimer's disease in Down syndrome: clinical-neuropathological observations. Acta Neurol Belg 69:369-380 Hatazawa J, Brooks RA, Di Chiro G, Bacharach SL (l987a) Glucose utilization rate versus brain size in humans. Neu rology 37:583-588 Hatazawa J, Brooks RA, Di Chiro G, Campbell G (1987b) Global cerebral glucose utilization is independent of brain size: a PET study. J Comput Assist Tomogr 11:571-576 Horwitz B, Schapiro M, Grady C, Rapoport SI (in press) Cere bral metabolic patterns in young adult Down syndrome: al tered intercorrelations between regional rates of glucose uti lization. J Ment Defic Res (in press) Huang SC, Phelps ME, Hoffman EJ, Sideris K, Selin CJ, Kuhl DE (1980) Noninvasive determination of local cerebral met abolic rate of glucose in man. Am J Physiol 238:E69--E82 Huang SC, Carson RE, Phelps ME, Hoffman EJ, Schelbert H, Kuhl DE (1981) A boundary method for attenuation correc tion in positron computed tomography. J Nucl Med 22:627637 Lassen NA, Christensen S, Hoedt-Rasmussen K, Stewart BM (1966) Cerebral oxygen consumption in Down's syndrome. Arch Neurol 15:595--602 Melamed E, Mildworf B, Sharav T, Belenky L, Wertman E (1987) Regional cerebral blood flow in Down's syndrome. Ann Neurol 22:275-278 Risberg J (1980) Regional cerebral blood flow measurements by 133 Xe-inhalation: methodology and applications in neuro psychology and psychiatry. Brain Lang 9:9--34 Roche AF, Sunderland S (1960) Postmortem observations of mongoloid crania. J Neuropathol Exp Neurol 19:554-558 Ross MH, Galaburda AM, Kemper TL (1984) Down's syn drome: is there a decreased population of neurons? Neurol ogy 34:909--916 SAS Institute (1985) SAS User's Guide: Statistics, Version 5. Cary, NC, SAS Institute Schapiro MB, Creasey H, Schwartz M, Haxby JV, White B, Moore A, Rapoport SI (l987a) Quantitative CT analysis of brain morphometry in adult Down's syndrome at different ages. Neurology 37:1424-1427 Schapiro MB, Haxby JV, Grady CL, Duara R, Schlageter NL, White B, Moore A, Sundaram M, Larson SM, Rapoport SI (l987b) Decline in cerebral glucose utilisation and cognitive
J Cereb Blood Flow Metab, Vol.
10, No. 2, 1990
M. B. SCHAPIRO ET AL.
206
function with aging in Down's syndrome. J Neurol Neuro surg Psychiatry 50: 766-774 Schapiro MB, Ball MJ, Grady CL, Haxby JV, Kaye JA, Rapo port SI (1988a) Dementia in Down's syndrome: cerebral glu cose utilization, neuropsychological assessment and neuro pathology. Neurology 38: 938-942 Schapiro MB, Berman KF, Friedland RP, Weinberger DR, Rap oport SI (1988b) Regional cerebral blood flow is not de creased in young adult Down syndrome. Ann Neuro124:310 Schwartz M, Duara R, Haxby J, Grady C, White BJ, Kessler RM, Kay AD, Cutler NR, Rapoport SI (1983) Down's syn drome in adults: brain metabolism. Science 221: 781-783 Schwartz M, Creasey H, Grady CL, DeLeo JM, Frederickson HA, Cutler NR, Rapoport SI (1985) Computed tomographic analysis of brain morphometrics in 30 healthy men, aged 21 to 81 years. Ann Neurol 17: 146--157 Scott BS, Becker LE, Petit TL (1983) Neurobiology of Down's syndrome. In: Progress in Neurobiology, VoI2I. Great Brit ain: Pergamon Press, pp 199-237 Shiue CY, Salvadori PA, Wolf AP, Fowler JS, MacGregor R
J Cereb Blood Flow Metab, Vol.
10, No.2, 1990
(1982) A new improved synthesis of 2-deoxy-2-[18Flfluoro d-glucose from 18F-labeled acetyl hypofluorite. J Nucl Med 23:899-903 Sokoloff L, Reivich M, Kennedy C, DesRosiers MH, Patlak CS, Pettigrew KD, Sakurada 0, Shinohara M (1977) The [14C]-deoxyglucose method for the measurement of local ce rebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neu rochem 28: 897-916 Wisniewski KE, Wisniewski HM, Wen GY (1985) Occurrence of neuropathological changes and dementia of Alzheimer's dis ease in Down's syndrome. Ann NeuroI17:278-282 Wisniewski KE, Laure-Kamionowska M, Connell F, Wen GY (1986) Neuronal density and synaptogenesis in the postnatal stage of brain maturation in Down syndrome. In: The Neu robiology of Down syndrome (Epstein CJ, ed), New York, Raven Press, pp 29-44 Zellweger H (1977) Down syndrome. In: Handbook of Clinical Neurology, Vol 31 (Vinken PJ, Bruyn GW, eds), New York, North Holland Publishing Co., pp 367-469
