Integrated PET/MRI With C-CFT and F-FDG for levodopa response difference in Parkinson's disease.

Aim: Parkinson's disease is one of the most common neurodegenerative diseases. Excellent levodopa responsiveness has been proposed as a characteristic supporting feature in substantiating the PD diagnosis. However, a small portion of clinically established PD patients shows poor levodopa response. This study aims to investigate brain function alterations of PD patients with poor levodopa responsiveness by PET/MRI.

Method: A total of 46 PD patients were recruited. They all completed C-CFT PET/MRI scans and the acute levodopa challenge test. Among these 46 PD patients, 42 participants further underwent F-FDG PET/MRI scans. Clinical variables regarding demographic data, disease features and cognition scales were also collected. Based on the improvement rate of UPDRS-III, PD patients were divided into non-responders (improvement rate < 33 %) and responders (improvement rate ≥ 33 %). Statistical parametric zapping was performed to analyze molecular imaging. Dopaminergic uptake and metabolism of 70 brain regions were converted to quantitative values and expressed as standard uptake value (SUV). SUV was further normalized by the cerebellum. The resulting SUV ratios and clinical variables were then compared by SPSS.

Results: The difference between levodopa non-responders (n = 17) and responders (n = 29) in the UPDRS III baseline was statistically significant and the former had a lower UPDRS III baseline (19 (10, 32), p＜0.05). In contrast, no statistical difference between these two groups was found in age, gender, disease duration, cognition, motor subtype and Hoehn-Yahr stage. Dopaminergic uptake differences between levodopa non-responders (n = 17) and responders (n = 29) were shown in the left inferior frontal cortex (1.00 ± 0.09 vs 1.07 ± 0.08, p < 0.05 and FDR < 0.2), the right posterior cingulum (1.10 ± 0.10 vs 1.20 ± 0.13, p < 0.05 and FDR < 0.2) and the right insula (1.21 ± 0.12 vs 1.30 ± 0.10, p < 0.05 and FDR < 0.2). The metabolic alterations between levodopa non-responders (n = 16) and responders (n = 26) were shown in the right supplementary motor area (1.30 (1.18, 1.39) vs 1.41 (1.31, 1.53), p < 0.05 and FDR < 0.2), right precuneus (1.37 ± 0.10 vs 1.47 ± 0.18, p < 0.05 and FDR < 0.2), right parietal cortex (1.14 ± 0.15 vs 1.27 ± 0.21, p < 0.05 and FDR < 0.2), right supramarginal gyrus (1.16 (1.12, 1.26) vs 1.25 (1.14, 1.46), p < 0.05 and FDR < 0.2), right postcentral gyrus (1.15 (1.08, 1.32) vs 1.24 (1.17, 1.39), p < 0.05 and FDR < 0.2), medulla (0.75 ± 0.07 vs 0.80 ± 0.07, p < 0.05 and FDR < 0.2), right rolandic operculum (1.25 (1.18, 1.32) vs 1.33 (1.25, 1.50), p < 0.05 and FDR < 0.2), right olfactory (0.95 (0.91, 1.01) vs 1.01 (0.95, 1.15), p < 0.05 and FDR < 0.2), the right insula (1.15 (1.06, 1.22) vs 1.21 (1.12, 1.35), p < 0.05 and FDR < 0.2) and the left cerebellum crus (0.96 (0.91, 1.01) vs 0.92 (0.86, 0.96), p < 0.05 and FDR < 0.2).

Conclusions: PD patients with poor response to levodopa showed less severe impairment of baseline motor symptoms, more severe dopaminergic deficits in the left inferior frontal, right posterior cingulate cortex and the right insula, and lower metabolism in the right supplementary motor area, right precuneus, right parietal cortex, right supramarginal gyrus, right postcentral gyrus, medulla, right rolandic operculum, right olfactory, the right insula and higher metabolism in the left cerebellum crus.