Activation And Regional Cbf Patterns In Normal Individuals And In Pd

The classical view of the basal ganglia function under normal circumstances is that of a structure in which five separate parallel circuits, under dual control by the direct and indirect pathways, subserve specific areas of the neocortex. These segregated pathways thus process cortical information from the supplementary motor area (SMA) and premotor cortex (PMC); dorsolateral prefrontal cortex (DLPFC); lateral orbitofrontal cortex; anterior cingulate area; and oculomotor cortex. Their role seems to be the activation of cortical areas mediating motor, oculomotor, or cognitive behaviors, as well as alternating the focus of attention towards novel and rewarding extrapersonal stimuli (10).

Since the introduction of regional CBF (rCBF) and activation studies with PET, it became possible to analyze cortical and subcortical patterns during motor performance. Therefore, either rCBF PET with H215O to detect changes in local blood delivery during performance of motor or cognitive tasks or glucose metabolism with 18F-fluorodeoxyglucose (18F-FDG) can be used.

Several interrelated aspects of limb movement have been investigated by means of specific paradigms. Jenkins et al. (11) studied changes in rCBF related to movement frequency in six right-handed normal subjects moving a joystick with freely selected directions and found focal activation related to increasing speed in posterior SMA, bilateral lateral PMC, contralateral sensorimotor cortex, and ipsilateral cerebellar hemisphere and vermis (11). Also, from the Hammersmith group, a study of H215O rCBF evaluated patterns of cortical-subcortical activation during performance of a prelearned sequence of unimanual finger movements at a constant pace. The authors observed that progressive complexity of the task was positively correlated with regional increases in rostral SMA, pallidothalamic loop (bilateral pallidal and contralateral thalamus), right precuneus (Brodmann area 7), and ipsilateral PMC (ref. 12) . Dettmers et al. (13) evaluated the neural correlate of strength of finger movement by means of H215O PET scan. In their study, increasing force of right index finger flexion in six right-handed subjects was related to increases of rCBF in primary sensorimotor cortex, the posterior part of SMA and cin-gulate sulcus, and the cerebellar vermis.

It is hypothesized that the BG function by activating parts of the cortex into similar frequencies, thereby facilitating the performance of a motor or cognitive action. This pattern of activation, referred to as "focused attention," would be the physiological substrate for the harmonic performance of movement. In PD, the imbalance between direct and indirect pathways as a consequence of dopamine deficiency leads to overactivation of BG output nuclei and subsequent failure to achieve the state of focused attention (10). Several reports have addressed the changes in CBF and 18F-FDG uptake during different tasks in PD patients. In an inhaled C15O2 CBF PET study, Playford et al. (14) evaluated six PD patients and six controls at rest during the execution of freely chosen movements and with programmed repetitive forward movement of a joystick. For the free-selection task, both groups had similar increases in blood flow in left sensorimotor and bilateral premotor cortices. However, PD patients failed to activate the SMA and anterior cingulate areas, putamen, thalamus, and cerebellum. In the repetitive task, the control patients had lesser degrees of activation of right DLPFC, whereas for PD patients this was observed for right inferior parietal association cortex and right premotor area (14). Imagination of movement, when examined in normal individuals, resulted in activations in the sensorimo-tor, inferior parietal, bilateral dorsal premotor (PMC), caudal supplementary motor area, bilateral ventral premotor, right Ml and left superior parietal cortices, left putamen, and right cerebellum. However, a relationship between imagined movement complexity and enhanced activation was observed in contralateral PMC and ipsilateral superior parietal cortex and cerebellar vermis (15). In PD patients, imagination of movements did not translate into greater activation of DLPFC or mesial frontal areas (homolog to SMA in nonhuman primates), and the execution of movement did not activate DLPFC. Other components of motor control evaluated by functional imaging include the differential aspects of activation during unilateral or bilateral movement and the observation of cerebellar overactivation in PD. To categorize the rCBF distribution according to complex ity of movement, Goerres et al. (16) created a paradigm for H215O PET, in which six individuals were scanned while performing unimanual, bimanual-symmetric, and bimanual-asymmetric ballistic finger movement. When subjects moved one finger, there was contralateral activation of primary sen-sorimotor cortex, inferior parietal cortex, and precuneus, whereas bilateral symmetric movement led to bilateral increases in rCBF. Bilateral asymmetric finger movement, in comparison with the other tasks, was associated with augmented CBF in rostral SMA. In a Xenon single-photon emission computed tomography (SPECT) CBF study, Rascol et al. (17) compared 12 normal subjects, 12 PD patients not currently on medication, and 16 PD patients on medication (the two latter groups exhibiting mainly the akinetic-dominant form) during performance of unilateral finger-to-thumb movements. Regions of interest were placed over the cerebral and cerebellar hemispheres. Both groups of PD patients on medication and controls had similar values of cerebellar increases in CBF, whereas PD individuals who were studied off medication had significantly enhanced activation of ipsilateral cerebellar hemisphere (17).

Findings of overactivation of extrastriatal areas in PD have been further reproduced and converge with the "preimaging era" description of the paradoxical gait of PD, in which patients with advanced akinetic status and a typical short-stepped gait may experience benefit when helped by visual cues. The possible physiological mechanism underlying this discrepancy was investigated by Hanakawa et al. (18) in a study where 10 PD patients on medication and 10 controls underwent Tc-99m exametazime (HMPAO) SPECT imaging after walking on a treadmill. This treadmill was equipped with either parallel or transverse lines. After walking with each of the line orientations, the individuals were scanned. The purpose was to observe changes in rCBF correspondent to changes in cadence (number of steps/minute), more so in PD patients walking on transverse visual cues. While walking on the treadmill equipped with transverse lines, PD patients had marked improvement of cadence as previously reported and showed significantly greater activation of lateral premotor cortex, a region known to have abundant connections to the cerebellum and posterior parietal cortex (18). In the setting of BG dysfunction, leading to a state of impaired activation of SMA and PMC loops responsible for automatic gait, the use of such an alternate pathway may represent an adaptation process.

These reports of distant sites activated during execution of different motor tasks and their underuse in the parkinsonian state led to the proposal of a rather attractive putative imaging network of corticosubcortical structures known to be altered in PD (19). In one of the first articles addressing this issue, 18F-FDG PET was used in three groups: PD patients, presumed multiple system atrophy (MSA), and normal controls. The authors used the scaled subprofile model, a form of principal components analysis, to determine a topographical covariance pattern that would allow differentiation of the groups. Individuals with PD, as a group, had enhanced 18F-FDG uptake in the pallidum, thalamus, and pons, combined with hypometabo-lism in the SMA and lateral premotor, DLPFC, and parietooccipital association cortices. Most interestingly, this covariance profile correlated with Hoehn and Yahr (H&Y) stage, rigidity, and akinesia scores, but not with tremor scores (20). The same model, when applied to early typical and atypical PD patients, resulted in covariance patterns that were able to distinguish between the groups of patients (21).

This pattern of subcortical hypermetabolism with hypome-tabolism in related cortical areas was referred to as the Parkinson's disease-related pattern (PRDP) and further showed quantitative potential. The authors compared 23 early PD patients (mean H&Y 2.4 ± 1.3) with 14 patients with more advanced disease (mean H&Y 3.2 ± 1.2) using clinical assessment and 18F-FDG PET. The magnitude of the PDRP was calculated on an individual basis using topographic profile rating. This tool disclosed statistically significant results in the distinction of PDRPs for early and advanced disease and for rigidity and bradykinesia, but once again not for tremor (22). Why were tremor scores not related to PDRP in these two studies? To answer this question, the same group conducted 18F-FDG PET of 16 PD patients, 8 with and 8 without tremor, the former with Unified Parkinson's Disease Rating Scale (UPDRS) tremor score of at least 4, and the latter scoring 0. The groups were otherwise matched for age, duration of disease, H&Y, and composite motor UPDRS (exclusive of tremor). Compared with akinetic patients, the tremulous group had hypermetabolism of the thalamus, pons, and motor association cortex. A cerebellar-midbrain-thalamic-cortical network is thought to be independently active in PD patients featuring resting tremor (23). These results are in keeping with neuropathologic findings in tremor-dominant PD patients, who display more severe neuronal cell loss in medial SNc and retrorubral field A8, which are related to dorsolateral striatum and ventromedial thalamus (24). Patients with essential and writing tremor also were shown to display bilateral cerebellar, red nucleus, and thalamic overactivation (25). Interestingly, although olivary metabolism was not increased in these circumstances, it showed a negative correlation with improvement of tremor and cerebellar activation after alcohol intake in essential tremor individuals (26).

The PDRP was further reproduced in other centers (27), and its imaging substrate was then compared with electrophysi-ological testing when 42 PD patients underwent 18F-FDG PET and pallidotomy with intraoperative spontaneous single unit activity recording of the internal globus pallidus (28). There was a positive correlation between spontaneous GPi firing rates and ipsilateral anterior thalamic glucose metabolism, but no other significant relationship could beestablished between putamen and caudate. Although GPi activity is expected to inhibit thalamic activity through GABAergic neurotransmission, it should be noted that 18F-FDG measures synaptic rather than cell body activity and that metabolism may be influenced to a greater extent by excitatory activity than by inhibitory input (29). Moreover, other excitatory influences over GPi, such as those from subthalamic nucleus, may also play a role.

In PD patients under treatment with levodopa (L-dopa), PET assessment disclosed attenuation of the covariance profile of PDRP, with decreased metabolism in left putamen, right thalamus, bilateral cerebellum, and left PMC (30). Similarly, these pleiades of CSPTC loop alterations have shown improvement with surgical treatments, such as pallidotomy

(31), pallidal deep brain stimulation (32), and subthalamo-tomy (33), with varying degrees of increased uptake in cortical areas and diminished activity in the BG after the procedures, and statistically significant correlation with UPDRS motor scores. Nevertheless, medical and surgical treatments may apparently benefit patients by different mechanisms: pallidotomy leads to enhanced 18F-FDG uptake in primary motor cortex, lateral premotor cortex and DLPFC (31), whereas L-dopa treatment is associated with reduced metabolism in prefrontal cortex in nondemented patients, especially in the orbitofrontal cortex (34). It has been suggested that these different patterns of metabolic improvement may explain differences in cognitive outcome following the different interventions (35). Of course, comparisons of this kind should always be viewed with caution because they refer to studies conducted with different populations and using diverse PET protocols.

0 0

Post a comment