Imaging brain tumor proliferative activity with [124I]iododeoxyuridine.

Iododeoxyuridine (IUdR) uptake and retention was imaged by positron emission tomography (PET) at 0-48 min and 24 h after administration of 28.0-64.4 MBq (0.76-1.74 mCi) of [124I]IUdR in 20 patients with brain tumors, including meningiomas and gliomas. The PET images were directly compared with gadolinium contrast-enhanced or T2-weighted magnetic resonance images. Estimates for IUdR-DNA incorporation in tumor tissue (Ki) required pharmacokinetic modeling and fitting of the 0-48 min dynamically acquired data to correct the 24-h image data for residual, nonincorporated radioactivity that did not clear from the tissue during the 24-h period after IUdR injection. Standard uptake values (SUVs) and tumor:brain activity ratios (Tm:Br) were also calculated from the 24-h image data. The Ki, SUV, and Tm/Br values were related to tumor type and grade, tumor labeling index, and survival after the PET scan. The plasma half-life of [124I]IUdR was short (2-3 min), and the arterial plasma input function was similar between patients (48 +/- 12 SUV*min). Plasma clearance of the major radiolabeled metabolite ([124I]iodide) varied somewhat between patients and was markedly prolonged in one patient with renal insufficiency. It was apparent from our analysis that a sizable fraction (15-93%) of residual nonincorporated radioactivity (largely [124I]iodide) remained in the tumors after the 24-h washout period, and this fraction varied between the different tumor groups. Because the SUV and Tm:Br ratio values reflect both IUdR-DNA incorporated and exchangeable nonincorporated radioactivity, any residual nonincorporated radioactivity will amplify their values and distort their significance and interpretation. This was particularly apparent in the meningioma and glioblastoma multiforme groups of tumors. Mean tumor Ki values ranged between 0.5 +/- 0.9 (meningiomas) and 3.9 +/- 2.3 microl/min/g (peak value for glioblastoma multiforme, GBM). Comparable SUV and Tm:Br values at 24 h ranged from 0.13 +/- 0.03 to 0.29 +/- 0.19 and from 2.0 +/- 0.6 to 6.1 +/- 1.5 for meningiomas and peak GBMs, respectively. Thus, the range of values was much greater for Ki (approximately 8-fold) compared with that for SUV (approximately 2.2-fold) and Tm:Br (approximately 3-fold). The expected relationships between Ki, SUV, and Tm:Br and other measures of tumor proliferation (tumor type and grade, labeling index, and patient survival) were observed. However, greater image specificity and significance of the SUV and Tm:Br values would be obtained by achieving greater washout and clearance of the exchangeable fraction of residual (background) radioactivity in the tumors, i.e., by increased hydration and urinary clearance and possibly by imaging later than 24 h after [124I]IUdR administration.

Grey matter volumetric changes related to recovery from hand paresis after cortical sensorimotor stroke.

Preclinical studies using animal models have shown that grey matter plasticity in both perilesional and distant neural networks contributes to behavioural recovery of sensorimotor functions after ischaemic cortical stroke. Whether such morphological changes can be detected after human cortical stroke is not yet known, but this would be essential to better understand post-stroke brain architecture and its impact on recovery. Using serial behavioural and high-resolution magnetic resonance imaging (MRI) measurements, we tracked recovery of dexterous hand function in 28 patients with ischaemic stroke involving the primary sensorimotor cortices. We were able to classify three recovery subgroups (fast, slow, and poor) using response feature analysis of individual recovery curves. To detect areas with significant longitudinal grey matter volume (GMV) change, we performed tensor-based morphometry of MRI data acquired in the subacute phase, i.e. after the stage compromised by acute oedema and inflammation. We found significant GMV expansion in the perilesional premotor cortex, ipsilesional mediodorsal thalamus, and caudate nucleus, and GMV contraction in the contralesional cerebellum. According to an interaction model, patients with fast recovery had more perilesional than subcortical expansion, whereas the contrary was true for patients with impaired recovery. Also, there were significant voxel-wise correlations between motor performance and ipsilesional GMV contraction in the posterior parietal lobes and expansion in dorsolateral prefrontal cortex. In sum, perilesional GMV expansion is associated with successful recovery after cortical stroke, possibly reflecting the restructuring of local cortical networks. Distant changes within the prefrontal-striato-thalamic network are related to impaired recovery, probably indicating higher demands on cognitive control of motor behaviour.

Complementary positron emission tomographic studies of the striatal dopaminergic system in Parkinson's disease.

OBJECTIVE: To assess the relationship between striatal dopa decarboxylase capacity, D2 dopamine receptor binding, and energy metabolism in Parkinson's disease (PD). DESIGN: Positron emission tomographic (PET) studies of glucose and dopa metabolism and D2 dopamine receptor binding in the caudate nucleus and putamen of patients with PD at different Hoehn and Yahr (HY) stages using PET and the tracers 18F-fluorodeoxyglucose (FDG), 6-18F-fluoro-L-dopa (FDOPA), and 11C-raclopride (RACLO). SETTING: Positron emission tomography research program at the Paul Scherrer Institute. SUBJECTS: Twenty patients with PD at different stages of the disease (HY stages I through IV; five patients for each stage) compared with separate groups of age-matched healthy subjects. MAIN OUTCOME MEASURES: Influx constant (Ki) for specific FDOPA uptake; uptake index ratio for RACLO binding to D2 dopamine receptors; normalized to global FDG metabolic rate for glucose consumption; and semiquantitative score for assessment of tremor, rigidity, and bradykinesia in PD. RESULTS: Patients with PD at HY stages I to II (hereafter HY-I-II PD) revealed reduced FDOPA metabolism, particularly in the putamen. The FDOPA uptake in the putamen and caudate nucleus declined with increasing HY staging and scoring for bradykinesia and rigidity. Putamen RACLO binding to D2 dopamine receptors was up-regulated in patients with HY-I-II PD but declined toward control values, with increasing disease severity. Putamen side-to-side asymmetries of FDOPA metabolism and RACLO binding revealed a significant correlation. Putamen FDG metabolism showed a relative increase in all patients with PD. CONCLUSIONS: Our results show that FDOPA, RACLO, and FDG PET measurements provide complementary information to characterize metabolic and receptor changes in the striatum of PD with different degrees of motor disability. The FDOPA uptake reflects the best motor-related pathologic features, as indicated by the significant correlation between Ki values and clinical scores. The significant association between RACLO and FDOPA in the putamen suggests that D2 dopamine receptor changes are related to the reduction of presynaptic dopaminergic nerve terminals. Putamen FDG increase is probably the result of more complex feedback mechanisms that are primarily induced by striatal dopamine deficiency.

Reduced glucose metabolism in the frontal cortex and basal ganglia of multiple sclerosis patients with fatigue: a 18F-fluorodeoxyglucose positron emission tomography study.

To investigate the pathophysiology of fatigue in MS, we assessed cerebral glucose metabolism (CMR-Glu) in 47 MS patients using PET and 18F-fluorodeoxyglucose. Applying the Fatigue Severity Scale (FSS), we first compared MS patients with severe fatigue (MS-FAT, n = 19, FSS > 4.9) and MS patients without fatigue (MS-NOF, n = 16, FSS < 3.7) on a pixel-by-pixel basis using Statistical Parametric Mapping (SPM95). Second, we compared FSS values of all 47 patients covering the whole range of this scale with CMRGlu using an analysis of covariance (SPM95). In addition, we determined global CMRGlu by region-of-interest analysis. Sixteen healthy subjects served as control subjects (CON). Global CMRGlu was significantly lower in both MS groups compared with CON (CON 43.3 +/- 6.9 mumol/100 mL/min, MS-FAT 34.7 +/- 4.4, MS-NOF 35.4 +/- 4.5) but was not related to fatigue severity. Comparing the two MS groups, SPM95 analysis revealed predominant CMRGlu reductions bilaterally in a prefrontal area involving the lateral and medial prefrontal cortex and adjacent white matter, in the premotor cortex, putamen, and in the right supplementary motor area of MS-FAT. In addition, there were CMRGlu reductions in the white matter extending from the rostral putamen toward the lateral head of the caudate nucleus. FSS values were inversely related to CMRGlu in the right prefrontal cortex. CMRGlu in the cerebellar vermis and anterior cingulate was relatively higher in MS-FAT than in MS-NOF patients. CMRGlu of both regions showed positive correlations with FSS values. Our data suggest that fatigue in MS is associated with frontal cortex and basal ganglia dysfunction that could result from demyelination of the frontal white matter.

Reproducibility of regional metabolic covariance patterns: comparison of four populations.

UNLABELLED: In a previous [18F]fluorodeoxyglucose (FDG) PET study we analyzed regional metabolic data from a combined group of Parkinson's disease (PD) patients and healthy volunteers (N), using network analysis. By this method, we identified a unique pattern of regional metabolic covariation with an expression which accurately discriminated patients from healthy volunteers. To assess the reproducibility of this pattern as a potential marker for PD, we compared the pattern's topography with that of the disease-related covariance patterns identified in three other independent populations of patients with PD and healthy individuals studied in different PET laboratories. METHODS: The following patient populations were studied: group A (original cohort: 22 PD, 20 N; resolution: 7.5 mm full width at half maximum [FWHM]); group B (18 PD, 12 N; resolution: 4.2 mm FWHM); group C (25 PD, 15 N; resolution: 8.0 mm FWHM); and group D (14 PD, 10 N; resolution: 10 mm FWHM). Region weights for the PD-related covariance pattern (PDRP) identified in the group A analysis were correlated with those for the disease-related patterns identified in the analyses of groups B, C and D. In addition, subject scores for the group A PDRP were computed prospectively for every individual in each of the study populations. PDRP scores for PD and N within each cohort were compared. RESULTS: The PDRP topography identified in group A was highly correlated with each of the corresponding topographies identified in the other populations (r2 approximately 0.60, P < 0.0001). Prospectively computed subject scores for the group A PDRP significantly discriminated PD from N in each population (P < 0.004). CONCLUSION: The PDRP topography identified previously in Group A is highly reproducible across patient populations and tomographs. Prospectively computed PDRP scores can accurately discriminate patients from controls in multiple populations studied with different tomographs. Brain network imaging with FDG PET can provide robust metabolic markers for the diagnosis of PD.

Metabolic network abnormalities in early Huntington's disease: an [(18)F]FDG PET study.

UNLABELLED: The identification of discrete patterns of altered functional brain circuitry in preclinical Huntington's disease (HD) gene carriers is important to understanding the pathophysiology of this disorder and could be useful as a biologic disease marker. The purpose of this study was to use PET imaging of regional cerebral glucose metabolism to identify abnormal networks of brain regions that are specifically related to the preclinical phase of HD. METHODS: Eighteen presymptomatic HD gene carriers, 13 early-stage HD patients, and 8 age-matched gene-negative relatives were scanned using PET with [(18)F]FDG to quantify regional glucose utilization. A network modeling strategy was applied to the FDG PET data to identify disease-related regional metabolic covariance patterns in the preclinical HD cohort. The outcome measures were the region weights defining the metabolic topography of the HD gene carriers and the subject scores quantifying the expression of the pattern in individual subjects. RESULTS: Network analysis of the presymptomatic carriers and the gene-negative control subjects revealed a significant metabolic covariance pattern characterized by caudate and putamenal hypometabolism but also included mediotemporal metabolic reductions as well as relative metabolic increases in the occipital cortex. Subject scores for this pattern were abnormally elevated in the preclinical group compared with those of the control group (P < 0.005) and in the early symptomatic group compared with those of the presymptomatic group (P < 0.005). CONCLUSION: These findings show that FDG PET with network analysis can be used to identify specific patterns of abnormal brain function in preclinical HD. The presence of discrete patterns of metabolic abnormality in preclinical HD carriers may provide a useful means of quantifying the rate of disease progression during the earliest phases of this illness.

Activation of the human brain by monetary reward.

With the purpose of studying neural activation associated with reward processing in humans, we measured regional cerebral blood flow in 10 right-handed healthy subjects performing a delayed go-no go task in two different reinforcement conditions. Correct responses were either rewarded by money or a simple "ok' reinforcer. Behaviour rewarded by money, as compared with the "ok' reinforcement, was most significantly associated with activation of dorsolateral and orbital frontal cortex and also involved the midbrain and thalamus. These results may reflect the processing of reward information, although arousal effects cannot be completely excluded. It is suggested that the observed foci are implicated in the assessment of consequences in goal-directed behaviour which agrees with research in non-human primates.

Reduced reward processing in the brains of Parkinsonian patients.

Regional cerebral blood flow (rCBF) in healthy controls and non-demented, non-depressed Parkinsonian patients was measured using H2(15)O PET while subjects performed a prelearned pattern recognition task with delayed response. To investigate differences between the two groups in response to reward, the experimental design consisted of three reinforcement conditions: no reinforcement consisting of nonsense feedback, positive symbolic reinforcement and monetary reward. In the controls, monetary reward activated bilaterally the striatum and anterior cingulate gyrus, as well as unilaterally the left cerebellum, midbrain and medial frontal gyrus. Symbolic reinforcement revealed a similar pattern of activation, except that the striatum and left midbrain showed no activation. The Parkinsonian patients responded to monetary reward with increased activation bilaterally in the cerebellum, medial frontal gyrus, and anterior cingulate gyrus as well as unilaterally in the right fusiform gyrus and midbrain and left caudate nucleus and precentral gyrus. Symbolic reinforcement induced significantly increased rCBF in the right cerebellum only. Compared with symbolic reinforcement, monetary reward produced extended activation of temporoparietal association cortex. The pattern observed in the controls demonstrates the role in reward processing of dopaminergic mesolimbic pathways in the healthy human brain, whereas the pattern in the Parkinsonian patients suggests the involvement of compensatory cortical loops in the diseased brain.

Intermanual transfer of training: blood flow correlates in the human brain.

In a previous study, we found that relearning of a task with one hand might negatively be influenced by previous, opposite hand training of the analogue task, Thut G., et al., Exp. Brain Res., 108 (1996) 321-327. Drawing of a figure with the right hand, following left hand training, was slower than right hand drawing of an unknown figure. These conditions were termed right hand transfer learning (rTL) and right hand original learning (rOL). The present study aimed to identify the cerebral areas associated with these influences by measuring regional cerebral blood flow (rCBF) in 16 right-handed, healthy subjects during rTL and rOL. Positron emission tomography and statistical parametric mapping were used. Compared with rOL, rTL was associated with increased rCBF in the left medial prefrontal cortex and the right prefrontal convexity. Individual rCBF changes in the area homotopic to the right prefrontal convexity furthermore correlated with individual changes in rTL performance. While the smallest rCBF increases were found in subjects with weakest slowing of rTL relative to rOL, highest rCBF increases were present when rTL slowing dominated. Comparisons between rTL and rOL, however, revealed on average no performance differences. Our data suggest that relearning after previous opposite hand training activates neural mechanisms within the prefrontal convexity which might have an inhibitory function but that inhibition does not have to be the net final behavioral result.

Metabolic hyperfrontality and psychopathology in the ketamine model of psychosis using positron emission tomography (PET) and [18F]fluorodeoxyglucose (FDG).

To date, the ketamine/PCP model of psychosis has been proposed to be one of the best pharmacological models to mimic schizophrenic psychosis in healthy volunteers, since ketamine can induce both positive and negative symptoms of schizophrenia. At subanesthetic doses, ketamine has been reported to primarily block N-methyl-D-aspartate (NMDA) receptor complex giving support to a glutamate deficiency hypothesis in schizophrenia. Positron emission tomography was used to study ketamine-induced psychotic symptom formation in relation to cerebral metabolic alterations in healthy volunteers. Our study shows that NMDA receptor blockade results in a hyperfrontal metabolic pattern. Increased metabolic activity in the frontomedial and anterior cingulate cortex correlated positively with psychotic symptom formation, in particular with ego pathology. Analysis of correlations between syndrome scores and metabolic rate of glucose (CMRglu) or metabolic gradients (ratios) revealed that each psychopathological syndrome was associated with a number of metabolic alterations in cortical and subcortical brain regions, suggesting that not a single brain region, but distributed neuronal networks are involved in acute psychotic symptom formation.

Crossed cerebellar diaschisis and brain tumor biochemistry studied with positron emission tomography, [18F]fluorodeoxyglucose and [11C]methionine.

Cerebral gliomas may cause a reduction of glucose metabolism in the cerebellum contralateral to the tumor side (crossed cerebellar diaschisis, CCD). We investigated whether CCD is related to tumor localization, histological grade, size and tumor biochemistry. Cerebellar glucose metabolism was measured in 44 glioma patients and 15 healthy subjects using positron emission tomography and [18F]fluorodeoxyglucose (FDG). CCD was determined by calculating an asymmetry index of cerebellar glucose metabolism. Further, the tumor uptake of FDG and [11C]methionine (MET) was also assessed, and was expressed as ratio of normalized tracer uptake in tumor over contralateral cortex (T/C). Frontal lobe tumors were associated with highest CCD values. For these tumors, CCD was higher in malignant (-11.8+/-9.9%) than in low-grade (-4.3+/-4.1%) gliomas (P=0.010). In addition, frontal lobe tumors showed increasing CCD values with increasing size. In tumors of the parietal or temporal lobe, CCD was less marked or absent. T/C ratios of tumor tracer uptake were higher in malignant than in low-grade gliomas, but were not correlated with CCD. Our data indicate that the magnitude of CCD is mainly determined by tumor localization and size, the latter being associated with tumor grade. These findings raise the question whether CCD provides a measure of expansion or progression particularly in low-grade tumors of the frontal lobe.

Artificial neural network Radon inversion for image reconstruction.

Image reconstruction techniques are essential to computer tomography. Algorithms such as filtered backprojection (FBP) or algebraic techniques are most frequently used. This paper presents an attempt to apply a feed-forward back-propagation supervised artificial neural network (BPN) to tomographic image reconstruction, specifically to positron emission tomography (PET). The main result is that the network trained with Gaussian test images proved to be successful at reconstructing images from projection sets derived from arbitrary objects. Additional results relate to the design of the network and the full width at half maximum (FWHM) of the Gaussians in the training sets. First, the optimal number of nodes in the middle layer is about an order of magnitude less than the number of input or output nodes. Second, the number of iterations required to achieve a required training set tolerance appeared to decrease exponentially with the number of nodes in the middle layer. Finally, for training sets containing Gaussians of a single width, the optimal accuracy of reconstructing the control set is obtained with a FWHM of three pixels. Intended to explore feasibility, the BPN presented in the following does not provide reconstruction accuracy adequate for immediate application to PET. However, the trained network does reconstruct general images independent of the data with which it was trained. Proposed in the concluding section are several possible refinements that should permit the development of a network capable of fast reconstruction of three-dimensional images from the discrete, noisy projection data characteristic of PET.

Thalamic glucose metabolism in temporal lobe epilepsy measured with 18F-FDG positron emission tomography (PET).

Thalamic glucose metabolism has been studied in 24 patients suffering from temporal lobe epilepsy (TLE) using interictal 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET). A total of 17 patients had a unilateral TL seizure onset, 11 of these patients had a mesial temporal lobe epilepsy syndrome (MTLE), with mesial gliosis and a mesial TL seizure origin. Three patients had a lateral TL seizure origin, and 3 patients had mesial TL tumors. Bilateral TLE was assumed in 7 patients. Only in the patient group with MTLE (n = 11), the ipsilateral thalamic glucose uptake showed a statistically significant lower value when compared to the thalamus of the contralateral side (Wilcoxon paired sign test, P = 0.012). There was a more pronounced hypometabolism in right TLE compared to left TLE. A 'hypersynchronous seizure onset pattern' in ictal EEG was only seen in 6 (26%) patients (1 patient with bilateral, 5 with unilateral TLE). No correlation existed between the thalamic, temporal glucose metabolism and the 'hypersynchronous seizure onset pattern'.

Neural activity related to the processing of increasing monetary reward in smokers and nonsmokers.

This study investigated the processing of increasing monetary reward in nonsmoking and smoking subjects. The choice of the subject populations has been motivated by the observation of differences between nonsmokers and smokers in response to rewarding stimuli in a previous study. Subjects performed a pattern recognition task with delayed response, while rCBF was measured with [H215O] PET. Correct responses to the task were reinforced with three different amounts of monetary reward. The subjects received the sum of the rewards at the end of the experiment. The results show that a cortico-subcortical loop, including the dorsolateral prefrontal cortex, the orbitofrontal cortex, the cingulate gyrus and the thalamus is involved in processing increasing monetary reward. Furthermore, the striatal response differentiates nonsmokers from smokers. Thus, we found significant correlations between rCBF increases in striatum and increasing monetary reward and between striatal rCBF increases and mood in nonsmokers, but not in smokers. Moreover, no significant mood changes among the different monetary rewards could be observed in smokers. We infer that the response of the striatum to reward is related to changes in subjective feelings. The differences between smokers and nonsmokers confirm our previous conclusions that the association between blood flow, performance, mood and amount of reward is more direct in nonsmokers.

How does the human brain deal with a spinal cord injury?

The primary sensorimotor cortex of the adult brain is capable of significant reorganization of topographic maps after deafferentation and de-efferentation. Here we show that patients with spinal cord injury exhibit extensive changes in the activation of cortical and subcortical brain areas during hand movements, irrespective of normal (paraplegic) or impaired (tetraplegic patients) hand function. Positron emission tomography ([15O]-H2O-PET) revealed not only an expansion of the cortical 'hand area' towards the cortical 'leg area', but also an enhanced bilateral activation of the thalamus and cerebellum. The areas of the brain which were activated were qualitatively the same in both paraplegic and tetraplegic patients, but differed quantitatively as a function of the level of their spinal cord injury. We postulate that the changes in brain activation following spinal cord injury may reflect an adaptation of hand movement to a new body reference scheme secondary to a reduced and altered spino-thalamic and spino-cerebellar input.

Changes in brain activation associated with reward processing in smokers and nonsmokers. A positron emission tomography study.

Tobacco smoking is the most frequent form of substance abuse. Several studies have shown that the addictive action of nicotine is mediated by the mesolimbic dopamine system. This system is implicated in reward processing. In order to better understand the relationship between nicotine addiction and reward in humans, we investigated differences between smokers and nonsmokers in the activation of brain regions involved in processing reward information. Using [H2(15O)] positron emission tomography (PET), we measured regional cerebral blood flow (rCBF) in healthy smokers and nonsmokers while they performed a prelearned, pattern-recognition task. We compared two conditions involving nonmonetary reinforcement or monetary reward with a baseline condition in which nonsense feedback was presented. With monetary reward, we found activation in the frontal and orbitofrontal cortex, occipital cortex, cingulate gyrus, cerebellum, and midbrain in both groups. Additionally, monetary reward activated typical dopaminergic regions such as the striatum in nonsmokers but not in smokers. We found a similar pattern of activation associated with nonmonetary reinforcement in nonsmokers, whereas activation was found in smokers only in the cerebellum. The different patterns of activation suggest that the brains of smokers react in a different way to reward than those of nonsmokers. This difference involves in particular the regions of the dopaminergic system including the striatum. In principle these observations could be interpreted either as a consequence of tobacco use or as a primitive condition of the brain that led people to smoke. Supported by related nonimaging studies, we interpret these differences as a consequence of tobacco smoking, even if a short-term effect of smoking prior to the experiment cannot be excluded.

Complementary PET studies of striatal neuronal function in the differential diagnosis between multiple system atrophy and Parkinson's disease.

We used PET with the tracers [18F]fluorodeoxyglucose (FDG), [18F]fluorodopa (FDOPA) and [11C]raclopride (RACLO) to study striatal glucose and dopa metabolism, and dopamine D2 receptor binding, respectively, in nine patients with multiple system atrophy. Ten patients with classical Parkinson's disease were investigated with the same three PET tracers' and three separate groups, each of 10 healthy subjects, served as control populations. We found that striatal FDOPA values separated all healthy subjects from patients with parkinsonism but they were not useful in distinguishing multiple system atrophy from Parkinson's disease. Conversely, striatal RACLO as well as FDG values discriminated all multiple system atrophy from Parkinson's disease patients as well as from healthy control subjects. Metabolic and receptor binding decrements in the putamen of multiple system atrophy patients were significantly correlated. Stepwise regression analysis revealed that a linear combination of putamen RACLO and FDOPA values accurately predicted clinical measures of disease severity in the multiple system atrophy group. Our findings suggest that striatal FDG and particularly RACLO are sensitive and effective measures of striatal function and may help characterizing patients with multiple system atrophy. In contrast, FDOPA measurements are accurate in detecting abnormalities of the nigrostriatal dopaminergic system but may not distinguish among different forms of parkinsonism.

Influence of spinal cord injury on cerebral sensorimotor systems: a PET study.

OBJECTIVES: To assess the effect of a transverse spinal cord lesion on cerebral energy metabolism in view of sensorimotor reorganisation. METHODS: PET and 18F-fluorodeoxyglucose were used to study resting cerebral glucose metabolism in 11 patients with complete paraplegia or tetraplegia after spinal cord injury and 12 healthy subjects. Regions of interest analysis was performed to determine global glucose metabolism (CMRGlu). Statistical parametric mapping was applied to compare both groups on a pixel by pixel basis (significance level P = 0.001). RESULTS: Global absolute CMRGlu was lower in spinal cord injury (33.6 (6.6) mumol/100 ml/min (mean (SD)) than in controls (45.6 (6.2), Mann-Whitney P = 0.0026). Statistical parametric mapping analysis disclosed relatively increased glucose metabolism particularly in the supplementary motor area, anterior cingulate, and putamen. Relatively reduced glucose metabolism in patients with spinal cord injury was found in the midbrain, cerebellar hemispheres, and temporal cortex. CONCLUSIONS: It is assumed that cerebral deafferentiation due to reduction or loss of sensorimotor function results in the low level of absolute global CMRGlu found in patients with spinal cord injury. Relatively increased glucose metabolism in brain regions involved in attention and initiation of movement may be related to secondary disinhibition of these regions.