Neurology, Vol 45, Issue 11 1995-2004, Copyright © 1995 by American Academy of Neurology

ARTICLES

Early differential diagnosis of Parkinson's disease with 18F- fluorodeoxyglucose and positron emission tomography

D Eidelberg, JR Moeller, T Ishikawa, V Dhawan, P Spetsieris, T Chaly, A Belakhlef, F Mandel, S Przedborski and S Fahn
Department of Neurology, North Shore University Hospital/Cornell University Medical College, Manhasset, NY 11030, USA.

Early-stage Parkinson's disease (EPD) is often clinically asymmetric. We used 18F-fluorodeoxyglucose (FDG) and PET to assess whether EPD can be detected by a characteristic pattern of regional metabolic asymmetry. To identify this pattern, we studied 10 EPD (Hoehn and Yahr stage I) patients (mean age 61.1 +/- 11.1 years) using 18F-FDG and PET to calculate regional metabolic rates for glucose. The scaled subprofile model (SSM) was applied to metabolic asymmetry measurements for the combined group of EPD patients and normal subjects to identify a specific covariation pattern that discriminated EPD patients from normal subjects. To determine whether this pattern could be used diagnostically, we studied a subsequent group of five presumptive EPD patients (mean age 50.9 +/- 18.3), five normal subjects (mean age 44.6 +/- 15.3), and nine patients with atypical drug-resistant early-stage parkinsonism (APD) (mean age 44.6 +/- 14.0). In each member of this prospective cohort, we calculated the expression of the EPD-related covariation pattern (subject scores) on a case-by-case basis. We also studied 11 of the EPD patients, five patients with APD, and 10 normal subjects with 18F-fluorodopa (FDOPA) and PET to measure presynaptic nigrostriatal dopaminergic function, and we assessed the accuracy of differential diagnosis with both PET methods using discrimination analysis. SSM analysis disclosed a significant topographic contrast profile characterized by covariate basal ganglia and thalamic asymmetries. Subject scores for this profile accurately discriminated EPD patients from normal subjects and APD patients (p < 0.0001). Group assignments into the normal or parkinsonian categories with FDG/PET were comparable to those achieved with FDOPA/PET, although APD and EPD patients were not differentiable by the latter method. Metabolic brain imaging with FDG/PET may be useful in the differential diagnosis of EPD.