- Email: [email protected]
Pattern of Increased Cerebral FDG Uptake in Down Syndrome Patients
Pattern of Increased Cerebral FDG Uptake in Down Syndrome Patients Zsolt Lengyel, MD*, Erzsébet Balogh, MSc, PhD†, Miklós Emri, MSc*, Edit Szikszai, MD†, József Kollár, MD, PhD‡, Judit Sikula, MD‡, Olga Ésik, MD, PhD, DSc§, Lajos Trón, PhD, DSc*, and Éva Oláh, MD, PhD, DSc† Resting cerebral glucose metabolism was assessed by 18 [F]-fluorodeoxyglucose in 11 Down syndrome patients. Standardized uptake values were determined on a pixel-by-pixel basis from the measured tissue-activity data. The results revealed a mean overall 18[F]-fluorodeoxyglucose uptake in the Down syndrome patients close to that observed in the control group, consisting of children and young adults. However, the standard deviation of the standardized uptake values was much higher in the Down syndrome group in almost all voxels relating to the gray matter. The statistical parametric mapping method was applied to compare the cerebral 18[F]-fluorodeoxyglucose accumulation patterns of the Down syndrome and control groups. Six regions (clusters) were found for which the glucose uptake was higher in the Down syndrome patients than in the control group. The anatomic localization of these clusters was based on magnetic resonance investigations and a brain-atlas technique. The localization of the identified clusters with an increased glucose metabolism in the Down syndrome patients suggests that these subjects have an enhanced resting neuronal activity in cortical areas involved in reasoning, cognition, and speech as compared with normal subjects. © 2006 by Elsevier Inc. All rights reserved. Lengyel Z, Balogh E, Emri M, Szikszai E, Kollár J, Sikula J, Ésik O, Trón L, Oláh É. Pattern of increased cerebral FDG uptake in Down syndrome patients. Pediatr Neurol 2006;34:270-275.
Introduction An intellectual disability is defined by an intelligence quotient lower than 70, a deficient or abnormal pattern of
From *PET Centre, †Department of Pediatrics, and ‡Department of Radiology, Medical and Health Science Centre, Debrecen University, Debrecen, Hungary; and §Department of Radiotherapy, Semmelweis University, Budapest, Hungary.
270
PEDIATRIC NEUROLOGY
Vol. 34 No. 4
adaptive functions, and onset before the age of 18. The intellectual disability may result from any genetic abnormality (e.g., chromosome aberrations, monogenic diseases, triplet expansion, abnormal genetic imprinting pattern) or from various diseases involving the central nervous system acquired during prenatal, perinatal, or postnatal life (teratogenic effects, infections, intracranial hemorrhage, injuries, and the like). The clinical picture, i.e. the pattern of the intellectual disability, and the type and severity of the intellectual damage, is characteristic of the specific origin of the intellectual disability. The question arises as to how the different intellectual deficiencies relate to morphologic and functional disturbances in the various cerebral cortical regions. Down syndrome accounts for 1-3% of disorders involving intellectual disability. The clinical picture is characterized by facial dysmorphism, a high frequency of congenital heart diseases and other visceral abnormalities, and early aging. Down syndrome is also associated with developmental abnormalities of the central nervous system that result in general hypotonicity of the muscles, disturbance of movement coordination, mental retardation, and age-dependent Alzheimer-type neurodegeneration [1]. The intellectual disability involves both the cognitive and adaptive functions, relating to attention, abstract thinking, memory, and readiness of speech. Alzheimer disease may be observed much more frequently in Down syndrome patients, and several decades earlier than in the normal population [2]. Some phenotypic features of Down syndrome are caused by the overexpression of genes located within a segment of chromosome 21, termed the Down locus (21q22.3). The appearance of the histopathologic alterations (amyloid plaques in the brain) characteristic of Alzheimer disease can also be found in 40% of Down
Communications should be addressed to: Dr. Oláh; Department of Pediatrics; Medical and Health Science Centre; University of Debrecen; H-4012, Debrecen; Nagyerdei krt. 98; Hungary. E-mail: [email protected] Received December 9, 2003; accepted August 30, 2005.
© 2006 by Elsevier Inc. All rights reserved. doi:10.1016/j.pediatrneurol.2005.08.035 ● 0887-8994/06/$—see front matter
syndrome patients over 40 years of age with dementia, indicating a common failure of gene regulation (affecting segments both within and outside the Down locus of chromosome 21) in the background of the two distinct diseases [3]. At the same time, the histopathologic alterations and the consequent intellectual decline seem to be separate from the mental retardation and phenotypic features of Down syndrome, as this condition is demonstrated by patients with chromosome 21 trisomy and dementia, but without retardation and somatic abnormalities. That is why it seems to be worth studying the morphologic and functional background of the intellectual disability of nondemented Down syndrome patients. Morphologic features of the brain in Down syndrome can be studied by the traditional imaging techniques (computed tomography and magnetic resonance imaging) and by histologic analysis of tissue samples obtained at postmortem examination; these procedures provide information on the macroscopic and microscopic anatomy, respectively, of the brain. These features include the smaller size of the brain, delayed myelinization relative to age, fewer neurons and synapses, disturbances in neuron distribution, a decreased density of neurotransmitter receptors, abnormal dendritic arborization, A-beta amyloid deposition, frequent apoptotic cell death, and progressive cortical atrophy [1,4]. The functional abnormalities of the brain are reflected in the activities of enzymes involved in the glucose metabolism [5], and in the constitutively increased expression of genes (amyloid precursor protein, superoxide dismutase, and S100-beta protein) located in the Downspecific region of chromosome 21 which are responsible for neurodegenerative lesions (such as the accumulation of A-beta amyloid, apoptotic cell death, and abnormal dendritic arborization). In addition to the above-mentioned changes, overexpressions of genes located in non–Downspecific loci (growth-associated protein, nitrogen monoxide synthase 3, and neuronal thread protein) and also that of pro-apoptosis genes (p53, Bax, and interleukin-betaconverting enzyme) are similarly common features of the disease [1]. Published investigations document the usefulness of measurements of the regional cerebral blood flow [6] and certain biochemical parameters (plasma homocysteine and folic acid levels, superoxide dismutase and catalase activities) to characterize the physiologic alterations in the brain tissue and the body, respectively. Positron emission tomography is a new possibility for investigation of the overall and the regional cerebral function in vivo. Because the normal brain function requires energy provided exclusively by glucose and oxygen, the regional energy consumption, i.e. oxygen and glucose utilization, is assumed to reflect the regional neuronal activity. Any disturbance of the regional neuronal metabolism provides information about the location and severity of the altered neuronal activity involved in the mental functions [7,8]. The type and severity of the neuronal dysfunction may aid in the classification of the
disease and hence in the choice of adequate therapy. Parallel application of the morphologic (computed tomography and magnetic resonance imaging) and functional (positron emission tomography) imaging procedures and joint consideration of the information obtained with these methods may provide further data on the pathogenesis of intellectual disabilities of different origins [3,7-10]. The aim of the present study was to investigate the glucose utilization pattern of the brain in Down syndrome patients without dementia as compared with that in normal control individuals. Although the cortical activity can be affected by diverse factors (e.g., the blood flow, the stimulation and activity of enzymes related or nonrelated to the Down chromosomal region), the glucose consumption adequately characterizes the resting functions of the different cortical areas [1,6,10,11]. Static 18[F]-fluorodeoxyglucose (FDG) positron emission tomography examinations readily furnish useful information on the metabolic activity of the resting brain.
Patients and Methods The overall and the regional cerebral glucose metabolism were assessed by positron emission tomography, using 18[F]-fluorodeoxyglucose, in 11 patients (6 females, 5 males) with Down syndrome. Their median age was 13.1 years (range, 5-24 years), with seven subjects younger than 16 years. Informed consent was obtained from the guardian of the patient in all cases. The diagnosis of Down syndrome was set up by karyotyping using the standard method. All the patients proved to have trisomy 21. No translocation in any of them was identified. Of the 11 Down syndrome patients, 3 had congenital heart defect and 1 duodenal atresia. Hyperthyroidism was identified in one case. Their intelligence quotient was studied by using the Budapest test. The average intelligence quotient of the patients was 27.3 (S.D. ⫽ 12). No features of dementia were evident in any of them. The impossibility to obtain data on brain metabolism in healthy children is obvious. Therefore the control group consisted of nine subjects (children and adults of similar age) who had been investigated previously because of diseases not involving the central nervous system. This group included only males, two with malignant lymphoma (4 and 5 years of age), another 5-year-old child with neuroblastoma, a 10-year-old male with a presumed metabolic disturbance (which was excluded later), and a 17-year-old with complaints after suspected concussion of the brain, who was scanned twice. Three healthy young (19-, 22-, and 24-year-old) male volunteers were also included in the normal group. The study of the control group was approved by the Ethical Committee of the University of Debrecen in accordance with the Helsinki Declaration. All subjects in both groups were right-handed. The T1-weighted magnetic resonance imaging scans of the patients were obtained in order to be able to localize the metabolic alterations anatomically by an image fusion technique [10]. 18 [F]-fluorodeoxyglucose was administered in isotonic solution as an intravenous bolus 40-45 minutes before scanning. The dose was 0.15 mCi (5.5 MBq) per body weight in kilograms. During the loading period, the patients were kept resting in a darkened room. A GE 4096 plus positron emission tomography camera with a mean transaxial resolution of 5.5 mm was used; the duration of data collection was 20 minutes. To minimize motion during the scan, a short intravenous dose was administered just after patient positioning. The baseline of the field of view (97.5 mm) was positioned in the orbito-meatal plane. The images were reconstructed by filtered back-projection with a slice thickness of 6.5 mm. The 15 reconstructed slices were corrected for tissue attenuation by a contour-finding method. The dimensionless standardized uptake values
Lengyel et al: Cerebral FDG Uptake Pattern in Down Syndrome 271
Figure 1. “Glass-brain” representation of the clusters with a significantly increased 18[F]-fluorodeoxyglucose accumulation in the Down syndrome group as compared with the control group. The clusters were delineated from the SPM{T} map by a height threshold of T ⬎ 3.61 and by the threshold of cluster size ⬎ 30 voxels with the contrast of the Down syndrome group minus the control group. The number of degrees of freedom was 18, and the voxel-level P threshold, not corrected for the entire volume, was ⱕ0.001. The images are projections from left, posterior, and superior directions, respectively. Darker tones denote higher T values.
(SUVs) were calculated according to Equation 1 on a pixel-by-pixel basis: SUV ⫽ activity of tracer in unit mass of tissue ⁄ injected activity ⁄ body weight
(1)
The resulting three-dimensional standardized uptake value data were spatially standardized into the Talairach space [12] using the standard statistical parametric mapping (SPM) positron emission tomography template with the automated image registration software package [13]. The automated standardization was controlled and in some cases corrected by an interactive image registration and fusion program [14]. Three-dimensional averages of the scans were created separately in the two groups. The anatomically identical slices obtained from the averaged three-dimensional data allowed a comparison of the average overall and regional standardized uptake values in the Down syndrome group with those for the control group. The standard deviation of the standardized uptake value data was generated for each voxel of the standardized volumes, and the results were visualized as parametric images for both study groups to assess the regional variability in 18[F]-fluorodeoxyglucose accumulation. The relatively high variability of the voxel-level standardized uptake values observed in the Down syndrome group relative to the standard deviation of the voxel values in the control group led us to apply statistical parametric mapping for the spatially standardized uptake value images. We used the SPM99 version of the software package available from the Wellcome Department of the Cognitive Neurology Functional Imaging Laboratory and chose the “Compare-populations: 1 scan/subject (two sample t-test)” design. The standardized uptake value data were analyzed after appropriate volume filtering with an isotropic Gaussian smoothing filter (full width at half maximum ⫽ 16 mm) to diminish the variances in individual neuroanatomy of the subjects. To eliminate the disturbing effect of the intersubject variation in global glucose metabolic rate, the standardized uptake value data ascribed to the individual voxels were normalized to the overall glucose metabolic rate values of the subjects by proportional scaling.
272
PEDIATRIC NEUROLOGY
Vol. 34 No. 4
Results Magnetic resonance imaging detected moderate dilatation of the ventricles in seven cases, and in four cases a brain volume smaller than normal was found. Visual evaluation of the individual positron emission tomography images of the Down syndrome patients indicated certain areas (Broca area, cerebellum, hippocampus, and thalami) with a low glucose metabolism (as compared with the surrounding cortical regions), in which the neuronal activity impairment may be related to clinical manifestations of Down syndrome based on the generally known function of these regions [5,15]. However, comparison of the average positron emission tomography images of Down syndrome patients with those of the normal children and young adults did not reveal a pronounced difference in standardized uptake value in these regions or in the overall standardized uptake value. The standard deviation of the standardized uptake value relating to almost all gray matter voxels in the Down syndrome group surpasses that for the corresponding voxels in the control group. In the Down syndrome patients, the statistical parametric mapping analysis identified six regions (clusters) in which there was a significantly higher glucose uptake as compared with that for the control group (Fig 1). These clusters were delineated from the SPM{T} map by a height threshold of T ⬎ 3.61 and by the threshold of a cluster size larger than 30 voxels (the number of degrees of freedom was 18, and P ⱕ 0.001 was the voxel-level P threshold, not corrected for the entire volume). The P values of these highlighted groups of voxels, together with
Table 1.
Summary of the cluster- and voxel-level statistics from the output of the SPM analysis
Set-level P c 0.000
6
Pcorrected
Cluster-level kE Puncorrected
0.043 0.031
540 611
0.027 0.020
0.021
712
0.013
0.567 0.549 0.541
42 47 49
0.517 0.492 0.482
Voxel-level (Z⬅)
Pcorrected
T
0.016 0.111 0.668 0.173 0.582 0.458 0.590 0.596
6.52 5.27 3.86 4.97 4.02 4.24 4.00 3.99
(4.61) (4.05) (3.25) (3.89) (3.35) (3.49) (3.34) (3.33)
x,y,z (mm)
Puncorrected 0.000 0.000 0.001 0.000 0.000 0.000 0.000 0.000
⫺46 ⫺48 ⫺46 ⫺2 10 38 50 54
⫺38 10 20 38 30 ⫺36 ⫺32 6
0 8 ⫺16 8 28 30 ⫺6 4
Column kE contains the extent of the clusters in voxels (2 ⫻ 2 ⫻ 2 mm). The last three columns contain the Talairach coordinates of the local maxima (and submaxima) of the clusters. On the basis of the above coordinates, the involved anatomic structures can be identified. Abbreviations: c ⫽ Number of clusters kE ⫽ cluster size in voxels P ⫽ P value (corrected and uncorrected for the whole volume) SPM ⫽ statistical parametric mapping T ⫽ T value of the voxel x,y,z ⫽ Talairach coordinates of the voxel Z ⫽ Z score of the voxel
the standard coordinates of the maximum T-value voxels of the clusters, are presented in Table 1. The area revealing the most significant difference between the groups is located in the left medial temporal gyrus embodied by Brodmann areas 21 and 22. The references to the original Brodmann areas can be regarded only as tentative points of orientation. The coordinates of the local (and at the same time the overall) maxima are x ⫽ ⫺46, y ⫽ ⫺38, z ⫽ 0. Another region with a significantly increased relative 18 [F]-fluorodeoxyglucose uptake is located close to the left precentral and inferior frontal gyri (Brodmann areas 44 and 45). Areas with lower significance were localized in the contralateral hemisphere, almost symmetrically to those mentioned above. A pronounced difference between the Down syndrome patients and the control group was
observed in the midline in the anterior part of the cingulate gyri (Brodmann areas 24 and 32). The voxel-level statistical inference demonstrated that the left hemispheric differences were more significant. This lateralization is supported by the larger size of the clusters in the left hemisphere as compared with those on the right side. The clusters listed above can be displayed on a representative three-dimensional brain model to facilitate visualization of the spatial relations (Fig 2). The difference between the groups in the mean of the normalized standardized uptake values at the overall maximum of the SPM{T} map was 22% of the overall mean. It is interesting that the inverse statistics performed with the same criteria did not yield any area with a decreased 18 [F]-fluorodeoxyglucose metabolism in the Down syn-
Figure 2. Clusters with a significantly higher 18[F]fluorodeoxyglucose uptake in the Down syndrome patients, visualized in a representative three-dimensional brain model.
Lengyel et al: Cerebral FDG Uptake Pattern in Down Syndrome 273
drome group. This result suggests that there is a considerable individual variance among the Down syndrome patients as concerns the localization of cerebral regions with relative hypometabolism. Discussion Statistical parametric mapping analysis of 18[F]-fluorodeoxyglucose positron emission tomography data obtained on these Down syndrome and control subjects indicates cortical regions with an increased population-level resting glucose metabolism as compared with that in the same location in control subjects. Six areas of this kind were found through the voxel-level comparison, whereas the cluster-level analysis resulted in only three significant (uncorrected and corrected P ⬍ 0.05) differences in 18 [F]-fluorodeoxyglucose uptake (Table 1). The most likely explanation for this might be the small number of investigated subjects and the few measurements per group. This finding contradicts that of Schapiro et al. [11], who did not observe any difference in resting-state glucose metabolism between nondemented young Down syndrome adults and healthy control subjects. The most likely explanation for this contradiction may be the different method used (regional analysis vs statistical parametric mapping). To our knowledge, this is the first study that has applied spatial standardization and statistical parametric mapping on 18[F]-fluorodeoxyglucose positron emission tomography data from Down syndrome patients. We believe that statistical parametric mapping analysis is a more sophisticated method, capable of revealing subtle dissimilarities as compared with the simpler regional survey. Investigations of the background of dementia in adult Down syndrome patients revealed that the glucose uptake in the period preceding the dementia was similar to that in healthy persons; interestingly, no stress-induced increase in this parameter occurred [2,3,6,9,11,16,17]. No data have been published on a decreased intensity of cerebral glucose metabolism in nondemented young Down syndrome patients, but only on signs of a functional disruption of neural circuits associated with directed attention [5]. Further, Cutler reported on marked elevations in brain metabolism in young adult Down syndrome patients [2]. Dani et al. provided evidence [16] of a rapid linear decrease in glucose metabolic rate on the appearance of dementia in the parietal and temporal regions (the most sensitive regions to damage in Alzheimer disease). The population-level data of the present study support these observations, i.e. the resting overall glucose utilization in nondemented Down syndrome patients was not inferior to that in healthy subjects. Moreover, it was demonstrated that certain cortical regions in the Down syndrome patients were distinguished by an increased glucose metabolism. The most interesting finding was that the inverse comparison did not reveal any area with an impaired 18[F]-fluorodeoxyglucose uptake in the patient
274
PEDIATRIC NEUROLOGY
Vol. 34 No. 4
group. This finding is indirect evidence of the nondemented state of the examined patients, based on the observation that reductions in metabolic function of the brain tissue develop before the onset of neuropsychologic deficits [18]. Those areas with an increased metabolic activity can be related to centers of higher mental processes, such as speech (Brodmann areas 44 and 45), reasoning, memory (Brodmann areas 21 and 22), and cognition (Brodmann areas 24 and 32). These centers in Down syndrome patients may require a higher supply of energy in the resting state to compensate for the impaired function of the inadequately developed neural networks, in contrast with that of normally developed brains. This hypothesis is consistent with the fact that all the subjects included in the study were right-handed, with the left hemisphere dominant. The higher than normal relative resting glucose metabolism in the above-mentioned regions suggests the presence of a constantly elevated neuronal activity, which might also account for the lack of an increase in metabolism during brain stimulation in young Down syndrome subjects, reported by Pietrini et al. [17]. Besides the morphologic abnormalities, the brain tissue of Down syndrome patients manifests an impaired glucose metabolism through the downregulation of phosphoglucose isomerase, a key enzyme of glycolysis, in the frontal and temporal lobes and in the cerebellum [4]. This feature is unique to Down syndrome patients, which distinguishes them both from normal subjects and from Alzheimer patients, suggesting a strong link between glucose metabolism and Down syndrome rather than Alzheimer disease. The low phosphoglucose isomerase activity overlaps with cortical areas where clusters with an increased 18[F]fluorodeoxyglucose were revealed by our measurements, which implies possible mechanisms compensating for the diminished effectiveness of glycolysis by maintaining elevated substrate (glucose-6-phosphate) concentrations in the neurons of the most demanded centers. There is another possible explanation. An elevated 18 [F]-fluorodeoxyglucose accumulation is generally regarded as a sign of a higher glucose consumption. However, an intracellular increase in the amount of FDG-6phosphate (FDG6P) indicates a really higher energy demand and glucose utilization only if all the steps in the chain of the glycolytic enzyme reactions are taking place in the same way (with the same rate and efficacy) as under standard conditions. Similarly to glucose, 18[F]-fluorodeoxyglucose is phosphorylated by hexokinase, to FDG6P, which is trapped in the intracellular environment. The extent of FDG6P accumulation is determined by the rate of its production, which is closely related to the rate of glycolysis. The correlation of these entities is guaranteed, among others, by the allosteric regulation of hexokinase. The activity of this enzyme is downregulated by the product G6P. If the anomalous function (a low level of expression) of phosphoglucose isomerase were accompanied by a defect in the allosteric regulation of hexokinase,
the amount of G6P (and hence FDG6P) would not reflect the rate of glycolysis. The impaired regulation of hexokinase would be similar to that described by Galton et al. in gout [19]. The outlined mechanism would account both for the observed apparent hypermetabolism in some frontal and temporal lobe areas and for the impaired mental functions ascribed to these regions. The above hypothesis awaits confirmation by targeted experimentation. The lack of cortical fields with a decreased glucose metabolism in the Down syndrome group can be interpreted as a result of the young age of the investigated Down syndrome patients. Each examined Down syndrome subject was well socialized and was living in a caring family. In another possible interpretation, there are regions with a decreased cerebral glucose metabolic rate, but their locations vary greatly from patient to patient, so that no region with a significantly lower 18[F]-fluorodeoxyglucose uptake is highlighted in the population-level analysis. In an attempt to acquire a deeper insight into differences in functioning between normal and Down syndrome brains, as reflected in metabolic activity, additional experimentation is planned. In one special set of experiments, we intend to carry out studies involving an analysis of stimulation-induced metabolic changes. As another approach, the possible demonstration of a correlation in statistical parametric mapping between certain biochemical parameters or psychological measures of particular mental functions related to Down syndrome, and 18[F]fluorodeoxyglucose accumulation in the reported clusters, would contribute to a more detailed understanding of intellectual disability in Down syndrome. Advances in the clarification of pathologic brain functioning may help towards the provision of a better education and management of patients with intellectual disability in order to improve their quality of life. Conclusions The experimental data in the present study have afforded evidence of a lateralized difference in the regional cerebral 18[F]-fluorodeoxyglucose uptake between Down syndrome and normal subjects. The localization of the identified clusters with an increased glucose metabolism in the Down syndrome patients suggests that these subjects have an enhanced resting neuronal activity in cortical areas involved in reasoning, cognition, and speech as compared with normal subjects. This suggestion is supported by the facts that this difference is confined to the dominant (left) hemisphere, and that the inverse statistics did not yield any region with a decreased cerebral 18[F]fluorodeoxyglucose accumulation in these nondemented and well-socialized Down syndrome patients. This sidedness seems to be a special feature of this intellectual disability.
This study was supported by OTKA grant No. T/026108.
References [1] de la Monte SM. Molecular abnormalities of the brain in Down syndrome: Relevance to Alzheimer’s neurodegeneration. J Neural Transm Suppl 1999;57:1-19. [2] Cutler NR. Cerebral metabolism as measured with positron emission tomography (PET) and 18-F-2 deoxy-d-glucose: Healthy aging, Alzheimer’s disease and Down syndrome. Prog Neuropsychopharmacol Biol Psychiatry 1986;10:309-21. [3] Schapiro MB, Haxby JV, Grady CL, et al. Decline in cerebral glucose utilization and cognitive function with aging in Down’s syndrome. J Neurol Neurosurg Psychiatry 1987;50:766-74. [4] Labudova O, Caims N, Kitzmuller E, Lubec G. Impaired brain glucose metabolism in patients with Down syndrome. J Neural Transm Suppl 1999;57:247-56. [5] Horwitz B, Schapiro MB, Grady CL, Rapoport SI. Cerebral metabolic pattern in young adult Down’s syndrome subjects: Altered intercorrelations between regional rates of glucose utilization. J Ment Defic Res 1990;34(Pt 3):237-52. [6] Schapiro MB, Haxby JV, Grady CL. Nature of mental retardation and dementia in Down syndrome: Study with PET, CT, and neuropsychology. Neurobiol Aging 1992;13:723-34. [7] Casey BJ. Neuroimaging of the child brain: Implications for Studying the biology of intellectual disabilities [proceedings]. 10th IASSID World Conference, Helsinki, 1996, Session 127. [8] Chugani HT. Positron emission tomography in the study of childhood cognitive disorders [proceedings]. 10th IASSID World Conference, Helsinki, 1996; Session 127. [9] Devinsky O, Sato S, Conwit RA, Schapiro MB. Relation of EEG alpha background to cognitive function, brain atrophy, and cerebral metabolism in Down’s syndrome: Age-specific changes. Arch Neurol 1990;47:58-62. [10] Emri M, Ésik O, Repa I, Márián T, Trón L. [Image fusion of different tomographic methods PET/CT/MRI effectively contributes to therapy planning.] [Article in Hungarian] Orv Hetil 1997;138:2919-24. [11] Schapiro MB, Grady CL, Kumar A, et al. Regional cerebral glucose metabolism is normal in young adults with Down syndrome. J Cereb Blood Flow Metab 1990;10:199-206. [12] Talairach J, Tournoux P. Co-planar stereotactic atlas of the human brain. Stuttgart: Georg Thieme Verlag, 1988. [13] Woods RP, Cherry SR, Mazziotta JC. Rapid automated algorithm for aligning and reslicing PET images. J Comput Assist Tomogr 1992;16:620-33. [14] Evans AC, Collins DL, Neelin P, Peters TM. Three-dimensional correlative imaging: Applications in human brain mapping. In: Thatcher RW, Hallet M, Zeffiro T, John ER, Huerta M, eds. Functional neuroimaging. San Diego: Academic Press, 1994:145-61. [15] Azari NP, Horwitz B, Pettigrew KD, et al. Abnormal pattern of cerebral glucose metabolic rates involving language areas in young adults with Down syndrome. Brain and Language 1994;46:1-20. [16] Dani A, Pietrini P, Furey ML, et al. Brain cognition and metabolism in Down syndrome adults in association with development of dementia. Neuroreport 1996;7:2933-6. [17] Pietrini P, Dani A, Furey ML, et al. Low glucose metabolism during brain stimulation in older Down’s syndrome subjects at risk for Alzheimer’s disease prior dementia. Am J Psychiatry 1997;154:1063-9. [18] Cutler NR, Heston LL, Davies P, Haxby LV, Schapiro MB. NIH Conference. Alzheimer’s disease and Down’s syndrome: New insights. Ann Intern Med 1985;103:566-78. [19] Galton DJ, Betteridge DJ, Taylor KG, Holdsworth G, Stocks J. Defects of enzyme regulation in metabolic disease. Clin Sci Mol Med 1977;53:197-203.
Lengyel et al: Cerebral FDG Uptake Pattern in Down Syndrome 275
Introduction An intellectual disability is defined by an intelligence quotient lower than 70, a deficient or abnormal pattern of
From *PET Centre, †Department of Pediatrics, and ‡Department of Radiology, Medical and Health Science Centre, Debrecen University, Debrecen, Hungary; and §Department of Radiotherapy, Semmelweis University, Budapest, Hungary.
270
PEDIATRIC NEUROLOGY
Vol. 34 No. 4
adaptive functions, and onset before the age of 18. The intellectual disability may result from any genetic abnormality (e.g., chromosome aberrations, monogenic diseases, triplet expansion, abnormal genetic imprinting pattern) or from various diseases involving the central nervous system acquired during prenatal, perinatal, or postnatal life (teratogenic effects, infections, intracranial hemorrhage, injuries, and the like). The clinical picture, i.e. the pattern of the intellectual disability, and the type and severity of the intellectual damage, is characteristic of the specific origin of the intellectual disability. The question arises as to how the different intellectual deficiencies relate to morphologic and functional disturbances in the various cerebral cortical regions. Down syndrome accounts for 1-3% of disorders involving intellectual disability. The clinical picture is characterized by facial dysmorphism, a high frequency of congenital heart diseases and other visceral abnormalities, and early aging. Down syndrome is also associated with developmental abnormalities of the central nervous system that result in general hypotonicity of the muscles, disturbance of movement coordination, mental retardation, and age-dependent Alzheimer-type neurodegeneration [1]. The intellectual disability involves both the cognitive and adaptive functions, relating to attention, abstract thinking, memory, and readiness of speech. Alzheimer disease may be observed much more frequently in Down syndrome patients, and several decades earlier than in the normal population [2]. Some phenotypic features of Down syndrome are caused by the overexpression of genes located within a segment of chromosome 21, termed the Down locus (21q22.3). The appearance of the histopathologic alterations (amyloid plaques in the brain) characteristic of Alzheimer disease can also be found in 40% of Down
Communications should be addressed to: Dr. Oláh; Department of Pediatrics; Medical and Health Science Centre; University of Debrecen; H-4012, Debrecen; Nagyerdei krt. 98; Hungary. E-mail: [email protected] Received December 9, 2003; accepted August 30, 2005.
© 2006 by Elsevier Inc. All rights reserved. doi:10.1016/j.pediatrneurol.2005.08.035 ● 0887-8994/06/$—see front matter
syndrome patients over 40 years of age with dementia, indicating a common failure of gene regulation (affecting segments both within and outside the Down locus of chromosome 21) in the background of the two distinct diseases [3]. At the same time, the histopathologic alterations and the consequent intellectual decline seem to be separate from the mental retardation and phenotypic features of Down syndrome, as this condition is demonstrated by patients with chromosome 21 trisomy and dementia, but without retardation and somatic abnormalities. That is why it seems to be worth studying the morphologic and functional background of the intellectual disability of nondemented Down syndrome patients. Morphologic features of the brain in Down syndrome can be studied by the traditional imaging techniques (computed tomography and magnetic resonance imaging) and by histologic analysis of tissue samples obtained at postmortem examination; these procedures provide information on the macroscopic and microscopic anatomy, respectively, of the brain. These features include the smaller size of the brain, delayed myelinization relative to age, fewer neurons and synapses, disturbances in neuron distribution, a decreased density of neurotransmitter receptors, abnormal dendritic arborization, A-beta amyloid deposition, frequent apoptotic cell death, and progressive cortical atrophy [1,4]. The functional abnormalities of the brain are reflected in the activities of enzymes involved in the glucose metabolism [5], and in the constitutively increased expression of genes (amyloid precursor protein, superoxide dismutase, and S100-beta protein) located in the Downspecific region of chromosome 21 which are responsible for neurodegenerative lesions (such as the accumulation of A-beta amyloid, apoptotic cell death, and abnormal dendritic arborization). In addition to the above-mentioned changes, overexpressions of genes located in non–Downspecific loci (growth-associated protein, nitrogen monoxide synthase 3, and neuronal thread protein) and also that of pro-apoptosis genes (p53, Bax, and interleukin-betaconverting enzyme) are similarly common features of the disease [1]. Published investigations document the usefulness of measurements of the regional cerebral blood flow [6] and certain biochemical parameters (plasma homocysteine and folic acid levels, superoxide dismutase and catalase activities) to characterize the physiologic alterations in the brain tissue and the body, respectively. Positron emission tomography is a new possibility for investigation of the overall and the regional cerebral function in vivo. Because the normal brain function requires energy provided exclusively by glucose and oxygen, the regional energy consumption, i.e. oxygen and glucose utilization, is assumed to reflect the regional neuronal activity. Any disturbance of the regional neuronal metabolism provides information about the location and severity of the altered neuronal activity involved in the mental functions [7,8]. The type and severity of the neuronal dysfunction may aid in the classification of the
disease and hence in the choice of adequate therapy. Parallel application of the morphologic (computed tomography and magnetic resonance imaging) and functional (positron emission tomography) imaging procedures and joint consideration of the information obtained with these methods may provide further data on the pathogenesis of intellectual disabilities of different origins [3,7-10]. The aim of the present study was to investigate the glucose utilization pattern of the brain in Down syndrome patients without dementia as compared with that in normal control individuals. Although the cortical activity can be affected by diverse factors (e.g., the blood flow, the stimulation and activity of enzymes related or nonrelated to the Down chromosomal region), the glucose consumption adequately characterizes the resting functions of the different cortical areas [1,6,10,11]. Static 18[F]-fluorodeoxyglucose (FDG) positron emission tomography examinations readily furnish useful information on the metabolic activity of the resting brain.
Patients and Methods The overall and the regional cerebral glucose metabolism were assessed by positron emission tomography, using 18[F]-fluorodeoxyglucose, in 11 patients (6 females, 5 males) with Down syndrome. Their median age was 13.1 years (range, 5-24 years), with seven subjects younger than 16 years. Informed consent was obtained from the guardian of the patient in all cases. The diagnosis of Down syndrome was set up by karyotyping using the standard method. All the patients proved to have trisomy 21. No translocation in any of them was identified. Of the 11 Down syndrome patients, 3 had congenital heart defect and 1 duodenal atresia. Hyperthyroidism was identified in one case. Their intelligence quotient was studied by using the Budapest test. The average intelligence quotient of the patients was 27.3 (S.D. ⫽ 12). No features of dementia were evident in any of them. The impossibility to obtain data on brain metabolism in healthy children is obvious. Therefore the control group consisted of nine subjects (children and adults of similar age) who had been investigated previously because of diseases not involving the central nervous system. This group included only males, two with malignant lymphoma (4 and 5 years of age), another 5-year-old child with neuroblastoma, a 10-year-old male with a presumed metabolic disturbance (which was excluded later), and a 17-year-old with complaints after suspected concussion of the brain, who was scanned twice. Three healthy young (19-, 22-, and 24-year-old) male volunteers were also included in the normal group. The study of the control group was approved by the Ethical Committee of the University of Debrecen in accordance with the Helsinki Declaration. All subjects in both groups were right-handed. The T1-weighted magnetic resonance imaging scans of the patients were obtained in order to be able to localize the metabolic alterations anatomically by an image fusion technique [10]. 18 [F]-fluorodeoxyglucose was administered in isotonic solution as an intravenous bolus 40-45 minutes before scanning. The dose was 0.15 mCi (5.5 MBq) per body weight in kilograms. During the loading period, the patients were kept resting in a darkened room. A GE 4096 plus positron emission tomography camera with a mean transaxial resolution of 5.5 mm was used; the duration of data collection was 20 minutes. To minimize motion during the scan, a short intravenous dose was administered just after patient positioning. The baseline of the field of view (97.5 mm) was positioned in the orbito-meatal plane. The images were reconstructed by filtered back-projection with a slice thickness of 6.5 mm. The 15 reconstructed slices were corrected for tissue attenuation by a contour-finding method. The dimensionless standardized uptake values
Lengyel et al: Cerebral FDG Uptake Pattern in Down Syndrome 271
Figure 1. “Glass-brain” representation of the clusters with a significantly increased 18[F]-fluorodeoxyglucose accumulation in the Down syndrome group as compared with the control group. The clusters were delineated from the SPM{T} map by a height threshold of T ⬎ 3.61 and by the threshold of cluster size ⬎ 30 voxels with the contrast of the Down syndrome group minus the control group. The number of degrees of freedom was 18, and the voxel-level P threshold, not corrected for the entire volume, was ⱕ0.001. The images are projections from left, posterior, and superior directions, respectively. Darker tones denote higher T values.
(SUVs) were calculated according to Equation 1 on a pixel-by-pixel basis: SUV ⫽ activity of tracer in unit mass of tissue ⁄ injected activity ⁄ body weight
(1)
The resulting three-dimensional standardized uptake value data were spatially standardized into the Talairach space [12] using the standard statistical parametric mapping (SPM) positron emission tomography template with the automated image registration software package [13]. The automated standardization was controlled and in some cases corrected by an interactive image registration and fusion program [14]. Three-dimensional averages of the scans were created separately in the two groups. The anatomically identical slices obtained from the averaged three-dimensional data allowed a comparison of the average overall and regional standardized uptake values in the Down syndrome group with those for the control group. The standard deviation of the standardized uptake value data was generated for each voxel of the standardized volumes, and the results were visualized as parametric images for both study groups to assess the regional variability in 18[F]-fluorodeoxyglucose accumulation. The relatively high variability of the voxel-level standardized uptake values observed in the Down syndrome group relative to the standard deviation of the voxel values in the control group led us to apply statistical parametric mapping for the spatially standardized uptake value images. We used the SPM99 version of the software package available from the Wellcome Department of the Cognitive Neurology Functional Imaging Laboratory and chose the “Compare-populations: 1 scan/subject (two sample t-test)” design. The standardized uptake value data were analyzed after appropriate volume filtering with an isotropic Gaussian smoothing filter (full width at half maximum ⫽ 16 mm) to diminish the variances in individual neuroanatomy of the subjects. To eliminate the disturbing effect of the intersubject variation in global glucose metabolic rate, the standardized uptake value data ascribed to the individual voxels were normalized to the overall glucose metabolic rate values of the subjects by proportional scaling.
272
PEDIATRIC NEUROLOGY
Vol. 34 No. 4
Results Magnetic resonance imaging detected moderate dilatation of the ventricles in seven cases, and in four cases a brain volume smaller than normal was found. Visual evaluation of the individual positron emission tomography images of the Down syndrome patients indicated certain areas (Broca area, cerebellum, hippocampus, and thalami) with a low glucose metabolism (as compared with the surrounding cortical regions), in which the neuronal activity impairment may be related to clinical manifestations of Down syndrome based on the generally known function of these regions [5,15]. However, comparison of the average positron emission tomography images of Down syndrome patients with those of the normal children and young adults did not reveal a pronounced difference in standardized uptake value in these regions or in the overall standardized uptake value. The standard deviation of the standardized uptake value relating to almost all gray matter voxels in the Down syndrome group surpasses that for the corresponding voxels in the control group. In the Down syndrome patients, the statistical parametric mapping analysis identified six regions (clusters) in which there was a significantly higher glucose uptake as compared with that for the control group (Fig 1). These clusters were delineated from the SPM{T} map by a height threshold of T ⬎ 3.61 and by the threshold of a cluster size larger than 30 voxels (the number of degrees of freedom was 18, and P ⱕ 0.001 was the voxel-level P threshold, not corrected for the entire volume). The P values of these highlighted groups of voxels, together with
Table 1.
Summary of the cluster- and voxel-level statistics from the output of the SPM analysis
Set-level P c 0.000
6
Pcorrected
Cluster-level kE Puncorrected
0.043 0.031
540 611
0.027 0.020
0.021
712
0.013
0.567 0.549 0.541
42 47 49
0.517 0.492 0.482
Voxel-level (Z⬅)
Pcorrected
T
0.016 0.111 0.668 0.173 0.582 0.458 0.590 0.596
6.52 5.27 3.86 4.97 4.02 4.24 4.00 3.99
(4.61) (4.05) (3.25) (3.89) (3.35) (3.49) (3.34) (3.33)
x,y,z (mm)
Puncorrected 0.000 0.000 0.001 0.000 0.000 0.000 0.000 0.000
⫺46 ⫺48 ⫺46 ⫺2 10 38 50 54
⫺38 10 20 38 30 ⫺36 ⫺32 6
0 8 ⫺16 8 28 30 ⫺6 4
Column kE contains the extent of the clusters in voxels (2 ⫻ 2 ⫻ 2 mm). The last three columns contain the Talairach coordinates of the local maxima (and submaxima) of the clusters. On the basis of the above coordinates, the involved anatomic structures can be identified. Abbreviations: c ⫽ Number of clusters kE ⫽ cluster size in voxels P ⫽ P value (corrected and uncorrected for the whole volume) SPM ⫽ statistical parametric mapping T ⫽ T value of the voxel x,y,z ⫽ Talairach coordinates of the voxel Z ⫽ Z score of the voxel
the standard coordinates of the maximum T-value voxels of the clusters, are presented in Table 1. The area revealing the most significant difference between the groups is located in the left medial temporal gyrus embodied by Brodmann areas 21 and 22. The references to the original Brodmann areas can be regarded only as tentative points of orientation. The coordinates of the local (and at the same time the overall) maxima are x ⫽ ⫺46, y ⫽ ⫺38, z ⫽ 0. Another region with a significantly increased relative 18 [F]-fluorodeoxyglucose uptake is located close to the left precentral and inferior frontal gyri (Brodmann areas 44 and 45). Areas with lower significance were localized in the contralateral hemisphere, almost symmetrically to those mentioned above. A pronounced difference between the Down syndrome patients and the control group was
observed in the midline in the anterior part of the cingulate gyri (Brodmann areas 24 and 32). The voxel-level statistical inference demonstrated that the left hemispheric differences were more significant. This lateralization is supported by the larger size of the clusters in the left hemisphere as compared with those on the right side. The clusters listed above can be displayed on a representative three-dimensional brain model to facilitate visualization of the spatial relations (Fig 2). The difference between the groups in the mean of the normalized standardized uptake values at the overall maximum of the SPM{T} map was 22% of the overall mean. It is interesting that the inverse statistics performed with the same criteria did not yield any area with a decreased 18 [F]-fluorodeoxyglucose metabolism in the Down syn-
Figure 2. Clusters with a significantly higher 18[F]fluorodeoxyglucose uptake in the Down syndrome patients, visualized in a representative three-dimensional brain model.
Lengyel et al: Cerebral FDG Uptake Pattern in Down Syndrome 273
drome group. This result suggests that there is a considerable individual variance among the Down syndrome patients as concerns the localization of cerebral regions with relative hypometabolism. Discussion Statistical parametric mapping analysis of 18[F]-fluorodeoxyglucose positron emission tomography data obtained on these Down syndrome and control subjects indicates cortical regions with an increased population-level resting glucose metabolism as compared with that in the same location in control subjects. Six areas of this kind were found through the voxel-level comparison, whereas the cluster-level analysis resulted in only three significant (uncorrected and corrected P ⬍ 0.05) differences in 18 [F]-fluorodeoxyglucose uptake (Table 1). The most likely explanation for this might be the small number of investigated subjects and the few measurements per group. This finding contradicts that of Schapiro et al. [11], who did not observe any difference in resting-state glucose metabolism between nondemented young Down syndrome adults and healthy control subjects. The most likely explanation for this contradiction may be the different method used (regional analysis vs statistical parametric mapping). To our knowledge, this is the first study that has applied spatial standardization and statistical parametric mapping on 18[F]-fluorodeoxyglucose positron emission tomography data from Down syndrome patients. We believe that statistical parametric mapping analysis is a more sophisticated method, capable of revealing subtle dissimilarities as compared with the simpler regional survey. Investigations of the background of dementia in adult Down syndrome patients revealed that the glucose uptake in the period preceding the dementia was similar to that in healthy persons; interestingly, no stress-induced increase in this parameter occurred [2,3,6,9,11,16,17]. No data have been published on a decreased intensity of cerebral glucose metabolism in nondemented young Down syndrome patients, but only on signs of a functional disruption of neural circuits associated with directed attention [5]. Further, Cutler reported on marked elevations in brain metabolism in young adult Down syndrome patients [2]. Dani et al. provided evidence [16] of a rapid linear decrease in glucose metabolic rate on the appearance of dementia in the parietal and temporal regions (the most sensitive regions to damage in Alzheimer disease). The population-level data of the present study support these observations, i.e. the resting overall glucose utilization in nondemented Down syndrome patients was not inferior to that in healthy subjects. Moreover, it was demonstrated that certain cortical regions in the Down syndrome patients were distinguished by an increased glucose metabolism. The most interesting finding was that the inverse comparison did not reveal any area with an impaired 18[F]-fluorodeoxyglucose uptake in the patient
274
PEDIATRIC NEUROLOGY
Vol. 34 No. 4
group. This finding is indirect evidence of the nondemented state of the examined patients, based on the observation that reductions in metabolic function of the brain tissue develop before the onset of neuropsychologic deficits [18]. Those areas with an increased metabolic activity can be related to centers of higher mental processes, such as speech (Brodmann areas 44 and 45), reasoning, memory (Brodmann areas 21 and 22), and cognition (Brodmann areas 24 and 32). These centers in Down syndrome patients may require a higher supply of energy in the resting state to compensate for the impaired function of the inadequately developed neural networks, in contrast with that of normally developed brains. This hypothesis is consistent with the fact that all the subjects included in the study were right-handed, with the left hemisphere dominant. The higher than normal relative resting glucose metabolism in the above-mentioned regions suggests the presence of a constantly elevated neuronal activity, which might also account for the lack of an increase in metabolism during brain stimulation in young Down syndrome subjects, reported by Pietrini et al. [17]. Besides the morphologic abnormalities, the brain tissue of Down syndrome patients manifests an impaired glucose metabolism through the downregulation of phosphoglucose isomerase, a key enzyme of glycolysis, in the frontal and temporal lobes and in the cerebellum [4]. This feature is unique to Down syndrome patients, which distinguishes them both from normal subjects and from Alzheimer patients, suggesting a strong link between glucose metabolism and Down syndrome rather than Alzheimer disease. The low phosphoglucose isomerase activity overlaps with cortical areas where clusters with an increased 18[F]fluorodeoxyglucose were revealed by our measurements, which implies possible mechanisms compensating for the diminished effectiveness of glycolysis by maintaining elevated substrate (glucose-6-phosphate) concentrations in the neurons of the most demanded centers. There is another possible explanation. An elevated 18 [F]-fluorodeoxyglucose accumulation is generally regarded as a sign of a higher glucose consumption. However, an intracellular increase in the amount of FDG-6phosphate (FDG6P) indicates a really higher energy demand and glucose utilization only if all the steps in the chain of the glycolytic enzyme reactions are taking place in the same way (with the same rate and efficacy) as under standard conditions. Similarly to glucose, 18[F]-fluorodeoxyglucose is phosphorylated by hexokinase, to FDG6P, which is trapped in the intracellular environment. The extent of FDG6P accumulation is determined by the rate of its production, which is closely related to the rate of glycolysis. The correlation of these entities is guaranteed, among others, by the allosteric regulation of hexokinase. The activity of this enzyme is downregulated by the product G6P. If the anomalous function (a low level of expression) of phosphoglucose isomerase were accompanied by a defect in the allosteric regulation of hexokinase,
the amount of G6P (and hence FDG6P) would not reflect the rate of glycolysis. The impaired regulation of hexokinase would be similar to that described by Galton et al. in gout [19]. The outlined mechanism would account both for the observed apparent hypermetabolism in some frontal and temporal lobe areas and for the impaired mental functions ascribed to these regions. The above hypothesis awaits confirmation by targeted experimentation. The lack of cortical fields with a decreased glucose metabolism in the Down syndrome group can be interpreted as a result of the young age of the investigated Down syndrome patients. Each examined Down syndrome subject was well socialized and was living in a caring family. In another possible interpretation, there are regions with a decreased cerebral glucose metabolic rate, but their locations vary greatly from patient to patient, so that no region with a significantly lower 18[F]-fluorodeoxyglucose uptake is highlighted in the population-level analysis. In an attempt to acquire a deeper insight into differences in functioning between normal and Down syndrome brains, as reflected in metabolic activity, additional experimentation is planned. In one special set of experiments, we intend to carry out studies involving an analysis of stimulation-induced metabolic changes. As another approach, the possible demonstration of a correlation in statistical parametric mapping between certain biochemical parameters or psychological measures of particular mental functions related to Down syndrome, and 18[F]fluorodeoxyglucose accumulation in the reported clusters, would contribute to a more detailed understanding of intellectual disability in Down syndrome. Advances in the clarification of pathologic brain functioning may help towards the provision of a better education and management of patients with intellectual disability in order to improve their quality of life. Conclusions The experimental data in the present study have afforded evidence of a lateralized difference in the regional cerebral 18[F]-fluorodeoxyglucose uptake between Down syndrome and normal subjects. The localization of the identified clusters with an increased glucose metabolism in the Down syndrome patients suggests that these subjects have an enhanced resting neuronal activity in cortical areas involved in reasoning, cognition, and speech as compared with normal subjects. This suggestion is supported by the facts that this difference is confined to the dominant (left) hemisphere, and that the inverse statistics did not yield any region with a decreased cerebral 18[F]fluorodeoxyglucose accumulation in these nondemented and well-socialized Down syndrome patients. This sidedness seems to be a special feature of this intellectual disability.
This study was supported by OTKA grant No. T/026108.
References [1] de la Monte SM. Molecular abnormalities of the brain in Down syndrome: Relevance to Alzheimer’s neurodegeneration. J Neural Transm Suppl 1999;57:1-19. [2] Cutler NR. Cerebral metabolism as measured with positron emission tomography (PET) and 18-F-2 deoxy-d-glucose: Healthy aging, Alzheimer’s disease and Down syndrome. Prog Neuropsychopharmacol Biol Psychiatry 1986;10:309-21. [3] Schapiro MB, Haxby JV, Grady CL, et al. Decline in cerebral glucose utilization and cognitive function with aging in Down’s syndrome. J Neurol Neurosurg Psychiatry 1987;50:766-74. [4] Labudova O, Caims N, Kitzmuller E, Lubec G. Impaired brain glucose metabolism in patients with Down syndrome. J Neural Transm Suppl 1999;57:247-56. [5] Horwitz B, Schapiro MB, Grady CL, Rapoport SI. Cerebral metabolic pattern in young adult Down’s syndrome subjects: Altered intercorrelations between regional rates of glucose utilization. J Ment Defic Res 1990;34(Pt 3):237-52. [6] Schapiro MB, Haxby JV, Grady CL. Nature of mental retardation and dementia in Down syndrome: Study with PET, CT, and neuropsychology. Neurobiol Aging 1992;13:723-34. [7] Casey BJ. Neuroimaging of the child brain: Implications for Studying the biology of intellectual disabilities [proceedings]. 10th IASSID World Conference, Helsinki, 1996, Session 127. [8] Chugani HT. Positron emission tomography in the study of childhood cognitive disorders [proceedings]. 10th IASSID World Conference, Helsinki, 1996; Session 127. [9] Devinsky O, Sato S, Conwit RA, Schapiro MB. Relation of EEG alpha background to cognitive function, brain atrophy, and cerebral metabolism in Down’s syndrome: Age-specific changes. Arch Neurol 1990;47:58-62. [10] Emri M, Ésik O, Repa I, Márián T, Trón L. [Image fusion of different tomographic methods PET/CT/MRI effectively contributes to therapy planning.] [Article in Hungarian] Orv Hetil 1997;138:2919-24. [11] Schapiro MB, Grady CL, Kumar A, et al. Regional cerebral glucose metabolism is normal in young adults with Down syndrome. J Cereb Blood Flow Metab 1990;10:199-206. [12] Talairach J, Tournoux P. Co-planar stereotactic atlas of the human brain. Stuttgart: Georg Thieme Verlag, 1988. [13] Woods RP, Cherry SR, Mazziotta JC. Rapid automated algorithm for aligning and reslicing PET images. J Comput Assist Tomogr 1992;16:620-33. [14] Evans AC, Collins DL, Neelin P, Peters TM. Three-dimensional correlative imaging: Applications in human brain mapping. In: Thatcher RW, Hallet M, Zeffiro T, John ER, Huerta M, eds. Functional neuroimaging. San Diego: Academic Press, 1994:145-61. [15] Azari NP, Horwitz B, Pettigrew KD, et al. Abnormal pattern of cerebral glucose metabolic rates involving language areas in young adults with Down syndrome. Brain and Language 1994;46:1-20. [16] Dani A, Pietrini P, Furey ML, et al. Brain cognition and metabolism in Down syndrome adults in association with development of dementia. Neuroreport 1996;7:2933-6. [17] Pietrini P, Dani A, Furey ML, et al. Low glucose metabolism during brain stimulation in older Down’s syndrome subjects at risk for Alzheimer’s disease prior dementia. Am J Psychiatry 1997;154:1063-9. [18] Cutler NR, Heston LL, Davies P, Haxby LV, Schapiro MB. NIH Conference. Alzheimer’s disease and Down’s syndrome: New insights. Ann Intern Med 1985;103:566-78. [19] Galton DJ, Betteridge DJ, Taylor KG, Holdsworth G, Stocks J. Defects of enzyme regulation in metabolic disease. Clin Sci Mol Med 1977;53:197-203.
Lengyel et al: Cerebral FDG Uptake Pattern in Down Syndrome 275