According to the following inclusion and exclusion criteria, from April 2019 to June 2021, 75 PD patients were retrospectively included in this study, who underwent ¹¹C-CFT PET/CT at the Department of Nuclear Medicine, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. Inclusion criteria: (1) The patients were diagnosed with PD by 2 neurologists according to the “Diagnostic Criteria for Brain Bank of the Parkinson’s Society of London, UK” [15]; (2) The patients’ clinical data and scales were complete, reliable and clear. Exclusion criteria: (1) Non-primary PD; (2) Presence of other brain diseases, such as stroke in the neostriatum, which might affect the uptake of ¹¹C-CFT.

Anti-parkinsonian medications were withdrawn for more than 12 hours before clinical evaluation and imaging acquisition. The clinical assessment for individual patient was conducted using the UPDRS II-III [16], modified H-Y scale and MMSE. All patients were classified into different stages according to the H-Y scale. The study was approved by the Ethics Committee of Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. All research was performed in accordance with relevant guidelines/regulations. The study was carried out in accordance with the guidelines of the Declaration of Helsinki. All patients gave written informed consent before ¹¹C-CFT PET/CT scanning.

PET/CT examinations were performed on a Biograph 64 system (Siemens Healthineers, BIOGRAPHM CT/S, Erlangen, Germany) with a 21.6 cm axial field of view. ¹¹C-CFT was injected intravenously with a dose of 8–15 mCi, which was produced by chemists from the Nuclear Medicine Department of Shanghai Jiao Tong University School of Medicine using the Sumitomo HM-10 cyclotron and Sumitomo Carbon-11 multifunctional synthesis module (radiochemical purity $>$90%). About 60–80 minutes after injection, a 20-minute brain PET/CT scan was performed in 3D mode. PET/CT images in the transverse, sagittal, and coronal planes were obtained using an iterative reconstruction algorithm with a thickness of 5 mm, and attenuation correction of CT images.

All images were analyzed by two experienced nuclear medicine physicians who were blinded to the clinical diagnosis. Plane image-based DAT parameters were obtained through semi-automatic software processing as described previously [17]. In brief, SPM software (version 5, Statistical Parametric Mapping; Wellcome Department of Imaging Neuroscience, Institute of Neurology, London, UK) implemented in Matlab7.4 (Mathworks Inc., Sherborn, MA, USA) and ScanVP software Version 5.9.1 (Centre for Neuroscience, the Feinstein Institute for Medical Research, Manhasset, NY, USA) were used for data processing [18, 19]. All images were spatially normalized into Montreal Neurological Institute (MNI) brain space with 3D transformations. A Gaussian filter of 10 mm full width at half maximum was adopted for smoothing the normalized PET images.

Then, on the mean image summed over central slices [20], the standard ROIs were drawn in the caudate nucleus, anterior and posterior putamen, and occipital cortex (as reference) to calculate the regional DAT bindings, plane parameter, which was named as DAT binding of caudate nucleus (DBC), DAT binding of anterior putamen (DBAP) and DAT binding of posterior putamen (DBPP), respectively. ¹¹C-CFT PET is a crucial tool for DAT imaging, as it specifically binds to the striatum, while the occipital lobe exhibits hardly uptake of this specific imaging agent. Consequently, we used the occipital lobe as the reference region [21]. The following formula was used: (SUVavg of ROI- SUVavg of occipital lobe)/SUVavg of occipital lobe. Furthermore, the DAT binding asymmetry index of the caudate, anterior and posterior putamen (Cai, APai and PPai) were calculated: the difference between the left and right parameters divided by the average of the left and right parameters.

The 3D parameters were directly obtained using TrueD software (VE 13A, Siemens, Erlangen, Germany), without the need for further imaging processing and were easy to execute. In fact, our image sketching process adopted a semi-automated approach. Firstly, manually draw a circular 3D ROI with an appropriate diameter in the corresponding area of the cross-sectional slice of the PET image. The software would provide various semi quantitative indicator values for this area, such as maximum standardized uptake value (SUVmax). Before conducting this study, we performed extensive testing and comparisons on numerous images and determined that when the threshold was set at 40% of SUVmax, the software could automatically and accurately delineate areas with ¹¹C-CFT functionality within the manually defined 3D ROI. We used this threshold to manually delineate the left and right neostriatum (the whole caudate nucleus and putamen), putamen and posterior putamen, respectively. After that, the software would automatically give the volume of DAT in the corresponding region that effectively binded ¹¹C-CFT. These indicators reflected the functional volume of the neostriatum. At the same time, the software would automatically calculate the total binding amount of DAT to ¹¹C-CFT (Total binding amount of DAT in neostriatum (TNDB), Total binding amount of DAT in putamen (TPDB) and Total binding amount of DAT in posterior putamen (TPPDB)) in the above regions. These indicators could better reflect the residual function of the neostriatum, and also relatively balanced the anatomical volume difference of the human brain (Fig. 1). Finally, the asymmetry index (ai) of the regions were calculated according to the formula as mentioned above for plane parameters. For each 3D parameter, the mean value and the lower value of the left and right regions were selected for analysis, respectively, and any value was with statistical significance, the parameter would be considered to be statistically significant.

Fig. 1.

An illustration example of delineating 3D parameters. First, draw a 3D-ROI in the neostriatum. Taking 40% of SUVmax in the neostriatum region as the threshold, delineate the ROIs in the left and right neostriatum (a), putamen (b) and posterior putamen (c), respectively, and then the effective binding volume and total binding amount of DAT are obtained. (d) 3D-ROI in the neostriatum with each layer display. ROI, regions of interest; SUVmax, maximum standardized uptake value; DAT, dopamine transporter; DBNV, DAT binding volume of neostriatum; TNDB, Total binding amount of DAT in neostriatum; DBPV, DAT binding volume of putamen; TPDB, Total binding amount of DAT in putamen; DBPPV, DAT binding volume of posterior putamen; TPPDB, Total binding amount of DAT in posterior putamen.

Data were expressed as the mean $\pm$ SD. Bivariate correlation linear analysis was used to analyze the consistency of 3D parameters and plane parameters. Similarly, bivariate correlation linear analysis was also used to correlate DAT binding parameters with clinical data and scores. Pearson’s correlation coefficient (r²) was used to measure the degree of consistency or correlation. Bonferroni was adopted to the correction of multiple comparisons. An independent sample t-test was used to compare the difference of imaging parameters between H-Y early stage and lately stage groups. p 0.05 was considered statistically significant. SPSS 21.0 Windows software (IBM Corp., Armonk, NY, USA) was used for statistical analysis.

The clinical and demographic characteristics of all 75 (40 males, 35 females; 64 $\pm$ 10 years) PD patients were listed in Table 1. The average duration of disease of 75 patients was 3.3 $\pm$ 2.7 years (0.5–17 years). The average UPDRS II and III scores of 75 patients were 10 $\pm$ 7, 19 $\pm$ 12. The average modified H-Y scale scores of 75 patients were 2 $\pm$ 0.8, including 27 patients in stage 1, 7 in stage 1.5, 27 in stage 2, 12 in stage 3 and 2 in stage 4. Patients with H-Y scores of 1, 1.5 and 2 were defined as early course of disease (n = 61), and patients with H-Y scores of 3 and 4 were defined as late course of disease (n = 14). The average MMSE scores of 75 patients were 27 $\pm$ 3.

The 3D parameters of DAT PET/CT imaging were mostly consistent with the plane parameters (Table 2). The binding volume and total binding amount of DAT in neostriatum and putamen were statistically consistent with DBAP (Bonferroni Corrected p 0.05, r²_DBNVlower = 0.286, r²_DBNVmean = 0.251, r²_TNDBlower = 0.363, r²_TNDBmean = 0.362, r²_DBPVlower = 0.292, r²_DBPVmean = 0.265, r²_TPDBlower = 0.446, r²_TPDBmean = 0.454). The binding volume and total binding amount of DAT in putamen and post putamen were statistically consistent with DBPP (Bonferroni Corrected p 0.05, r²_TPDBlower = 0.421, r²_TPDBmean = 0.421, r²_TPPDBlower = 0.544, r²_TPPDBmean = 0.510). Similarly, the asymmetry index of binding volumes and total binding amount of DAT in neostriatum and putamen were statistically consistent with APai (Bonferroni Corrected p 0.05, r²_DBNVai = 0.254, r²_TNDBai = 0.254, r²_DBPVai = 0.321, r²_TPDBai = 0.439). The asymmetry index of the total DAT binding amount in putamen was consistent with PPai in statistics (Bonferroni Corrected p 0.05, r²_TPDBai = 0.369).

DBC was negatively correlated with UPDRS II and positively correlated with MMSE in statistics (Bonferroni Corrected p 0.05, r²_{UPDRS II} = 0.095, r²_MMSE = 0.099, Table 3). DBAP was negatively correlated with UPDRS II-III, modified H-Y score in statistics (Bonferroni Corrected p 0.05, r²_{UPDRS II} = 0.102, r²_{UPDRS III} = 0.117, r²_{modified H-Y score} = 0.123, Table 3). But DBPP was only negatively correlated with modified H-Y score in statistics (Bonferroni Corrected p 0.05, r² = 0.118, Table 3). The difference of DBAP between early and late H-Y course was statistically significant (p 0.05, Table 3). None of the asymmetry index of plane parameters was correlated with the clinical characteristics.

The binding volume and total binding amount of DAT in all regions were negatively correlated with UPDRS II in statistics (Bonferroni Corrected p 0.05, r²_DBNVlower = 0.160, r²_DBNVmean = 0.160, r²_TNDBlower = 0.115, r²_TNDBmean = 0.118, r²_DBPVlower = 0.131, r²_DBPVmean = 0.112, r²_TPDBlower = 0.124, r²_TPDBmean = 0.128, r²_DBPPVlower = 0.121, r²_DBPPVmean = 0.088, r²_TPPDBlower = 0.111, r²_TPPDBmean = 0.081, Table 3). The binding volume and total binding amount of DAT in all regions were negatively correlated with UPDRS III (Bonferroni Corrected p 0.05, r²_DBNVlower = 0.117, r²_DBNVmean = 0.109, r²_TNDBlower = 0.115, r²_TNDBmean = 0.113, r²_DBPVlower = 0.111, r²_DBPVmean = 0.093, r²_TPDBlower = 0.127, r²_TPDBmean = 0.123, r²_TPPDBlower = 0.094, r²_TPPDBmean = 0.097, Table 3). The binding volume of DAT in neostriatum and putamen were negatively correlated with modified H-Y score in statistics (r²_DBNVlower = 0.099, r²_DBNVmean = 0.111, r²_DBPVlower = 0.096, r²_DBPVmean = 0.112, Table 3) and the total binding amount of DAT in putamen and post putamen were negatively correlated with modified H-Y score in statistics (r²_TPDBlower = 0.089, r²_TPDBmean = 0.121, r²_TPPDBmean = 0.083) (Bonferroni Corrected p 0.05, Table 3). All these 3D parameters in patients with H-Y early course were significantly bigger than those with late course (p 0.05). In addition, the DAT binding volume in neostriatum were negatively correlated with the disease duration in statistics (Bonferroni Corrected p 0.05, r²_DBNVmean = 0.098, Table 3).

Among various asymmetry indexes, DAT binding volume of posterior putamen asymmetry index (DBPPVai) and TPPDBai were positively correlated with UPDRS II (r²_DBPPVai = 0.066, r²_TPPDBai = 0.068, Table 3), TPDBai was positively correlated with UPDRS III in statistics (r² = 0.237) (Bonferroni Corrected p 0.05, Table 3). Interestingly, the statistics showed MMSE was negative correlation with two asymmetry indexes, including DAT binding volume of neostriatum asymmetry index (DBNVai), TNDBai (Bonferroni Corrected p 0.05, r²_DBNVai = 0.134, r²_TNDBai = 0.105, Table 3).