Background: Multiple System Atrophy (MSA) is an atypical parkinsonian syndrome including cerebellar impairment and poor response to dopatherapy. The objective of the study is to assess right-hand motor activation in MSA patients before and after an acute levodopa challenge and to compare these data with those obtained in patients with Parkinson Disease (PD) and healthy volunteers (HV).

Methods: Eighteen MSA patients, eight PD patients and 10 age-matched HV will be included. rCBF measurements with H215O PET will be performed at rest and during a right hand movement. Statistical parametric mapping will be used to analyze motor versus rest in OFF and ON condition and effect of levodopa on motor activation.

Hypothesis: MSA and PD patient should recruited different motor networks.

Subjects. In this prospective study, MSA patients will be included if they met Gilman criteria for probable MSA. All those subjects will be distinguished between parkinsonian form (MSA-P) and cerebellar form (MSA-C). All will underwent Unified Parkinson's Disease Rating Scale UPDRS and International Cooperative Ataxia Rating Scale ICARSS. All patients will have a poor response to levodopa (<30% of the UPDRS score). Patients with PD will be included if they suffered from idiopathic PD according to the criteria of UKPDSBB and had a positive response to levodopa (≥ 30% improvement on UPDRS part III). All healthy subjects will have normal neurological examination and none will have a history of neurological, cardiovascular or psychiatric disturbance. For all subjects, handedness will be determined by the Edinburg test. For all patients (MSA and PD) a MRI brain scan will be realized PET activation study PET investigations will be performed during two pharmacological conditions: OFF (e.g after 12 hours of usual dopaminergic treatment discontinuation) and ON (after an acute oral levodopa challenge) in all subjects. During each PET there will be two motor conditions: rest (no movement, hand and wrist lying on the joystick) and a right-hand movement, consisting of moving joystick in 4 four different directions avoiding sequence repetition. Movement will be paced by an auditory stimulus at a frequency of 0.33 Hz. Each patient will be trained to perform the joystick movement before the PET. During PET investigation, angular wrist speed and angular wrist acceleration will be recorded using a computer recording connected to a joystick. The movement will start 30 seconds before image acquisition. Rest and Movement scan conditions will be replicated, making a total number of 6 six scans per patient in OFF condition and 6 six scans per patient in ON condition. The order of OFF and ON sessions and motor conditions will be fully counterbalanced across subjects to eliminate time and order effects.

H215O will be intravenously injected in the arm contralateral to the hand movement. PET measurements will be performed with an EXACT HR+ tomograph (CTI/Siemens, Knoxville, TN, USA) allowing the simultaneous 3D acquisition of 63 transaxial slices. Spatial resolution after reconstruction reached 4.5 and 4.1 mm in the transaxial and axial direction, respectively {Bendriem, 1996 #39}(19). To measure rCBF, 300 MBq of H215O will be administered for each 80-second emission scan. To allow complete decay of injected tracer activity, image acquisitions will be performed 10 minutes apart.

Image analysis will be performed on a personal computer station (DELL inc, Round Rock, Texas, USA) using the "statistical parametric mapping" package (SPM2, Wellcome Department of Cognitive Neurology, London, United Kingdom). Images of each subject will be realigned to the first volume and normalized to the MNI standard proportional stereotaxic space, which is based on that of Talairach and Tournoux (1988). The images will be coregistered on a template and spatially smoothed with a Gaussian kernel of 12 mm full width at half maximum (FWHM) to take into account variations in gyral anatomy and individual variability in structure-function relationships, and to improve the signal-to-noise ratio.

Statistical analysis :

All baseline characteristics in MSA, PD and healthy volunteers will be compared by the Man-Whitney U test. For PET imaging, the 36 subjects will be included in the same statistical analysis on a voxel-by-voxel basis. Statistical parametric maps will be generated using an ANCOVA model implemented through the General Linear Model formulation of SPM after normalization for global effect by proportional scaling. We will analyze three main effects 1) The "movement effect" consists of comparing the images obtained during hand movement with those acquired at rest for each group (MSA, PD and Healthy subjects) using the Family Wise Error (FWE) statistical threshold at P < 0.05 for peak height in OFF and ON conditions. 2) Difference between motor activation of the three groups in OFF condition. 3) Difference between motor activation during OFF and ON condition in each group reflecting levodopa effect on motor activation. For inter groups comparisons statistical threshold will set at p<0.01 with clusters>10 voxels.

Inclusion Criteria:

• MSA patients will be included if they met Gilman criteria for probable MSA. All those subjects will be distinguished between parkinsonian form (MSA-P) and cerebellar form (MSA-C). All will underwent Unified Parkinson's Disease Rating Scale UPDRS and International Cooperative Ataxia Rating Scale ICARSS. All patients will have a poor response to levodopa (<30% of the UPDRS score).
• Patients with PD will be included if they suffered from idiopathic PD according to the criteria of UKPDSBB and had a positive response to levodopa (≥ 30% improvement on UPDRS part III).
• All healthy subjects will have normal neurological examination and none will have a history of neurological, cardiovascular or psychiatric disturbance.
• For all subjects, handedness will be determined by the Edinburg test. For all patients (MSA and PD) a MRI brain scan will be realized