Intrastriatal gradient and VOI analyses of 18F-FDOPA PET scans for differentiation of Parkinsonian disorders

Objectives L-3,4-dihydroxy-6-18F-fluorophenylalanine (FDOPA) PET allows quantification of presynaptic dopaminergic functioning in vivo. The striatal-to-occipital ratio has been successfully utilized to discriminate subjects with parkinsonism from healthy controls. The aim of this study was to determine if intrastriatal analysis of FDOPA could facilitate further differentiation, discriminating idiopathic parkinson’s disease (IPD) from atypical parkinsonisms (AP) and subjects without parkinsonism.

Methods Scans of 56 patients (27 PD, 12 AP and 17 controls) who underwent FDOPA PET as part of their clinical evaluation and reached a final diagnosis by a movement disorder specialist, were retrospectively analyzed. PET images were spatially normalized to a FDOPA template and uptake was determined using predefined volumes-of-interests (VOI’s) (putamen and caudate nucleus). Diagnostic accuracy was retrospectively evaluated (ANOVA , ROC curves) using three methods: striatal-to-occipital ratio’s (SOR), caudate nucleus-to-putamen ratio’s (CNPR) and the slope of multiple in-line VOI’s from anterior-to posterior through the putamen (gradient analysis).

Results The SOR, CNPR and gradient were statistically significantly different between controls and subjects with parkinsonism in all regions investigated. Only the SOR of the left (p=0.02) and right caudate (p=0.03), and the CNPR bilaterally (both sides p <0.01) were statistically significantly different between subjects with IPD and AP. The area’s under the curve for the ROC curve to differentiate IPD from AP were 0.663 and 0.657 for the left and right caudate, and 0.757 and 0.713 for the left and right CNPR respectively. Using the optimal combination of predictors (left caudate and left CNPR), 76% of subjects with IPD and AP could be correctly classified, whereas 70% could be correctly classified using the single optimal predictor (left CNPR).

Conclusions The caudate nucleus-to-putamen ratio of FDOPA scans may be useful as an adjunct to differentiate subjects with IPD from AP, although it has limited diagnostic accuracy.