Discussion
Since discovered 200 years ago, the practice of exploring PD has never stopped. According to the various and systemic symptoms of PD populations, herein we attempted to break out the limitation of single studies confined to the brain, with the application of [11C]CFT tracer targeting DAT and advantages of the integrated dynamic total-body PET/CT imaging, aimed to systematically evaluate the potential patterns related to the dopaminergic changes and interactions among the digestive and cerebral regions. Dopamine serves as a protective agent for human GI system, preventing gastric and intestinal mucosa from injury, resisting against peptic ulcer, and promoting ulcer healing29. PD patients also have a relatively high incidence of duodenal ulcers, while the GI dysfunction or dysmotility are often likely due to the DA deficiency or peripheral DAT depletion30. Most of the previous reports on DA and DAT focused on their neurotransmitter function only in the central nerve system (CNS). However, certain peripheral organs also have abundant DAT expression. Recent studies have indicated that at the time when pathological α-synuclein aggregation was found in the submucosal and myenteric neurons, there also existed abnormal changes of dopaminergic neurons in these enteric plexuses both in PD patients and animal models31-33. However, it is unknown whether peripheral DAT uptake capacity and/or its level are changed. A potential relationship between peripheral and central DAT may assess the disease progression and the effectiveness of pharmacotherapies in PD34. Therefore, it is necessary to employ a macroscopic, whole-body approach to make an integrated evaluation of the DA signaling and homeostasis both in the CNS and ENS.
From the SUVmean curve analysis of the cerebral core areas (caudate, putamen, pallidum and substantial nigra), the tracer activities of the four nuclei in healthy volunteers continuously increased over time, suggesting the integrity of the dopaminergic function of the nigrostriatal system in healthy brains. Moreover, the strong positive correlations between the quantitative parameters among these brain nuclei also yielded similar network connections. Instead, the [11C]CFT uptakes of the targeted nuclei in PD patients tended to approach the “plateau” in advance, especially in the putamen with the earlier occurrence of dopamine loss. This is in accordance with the caudal-rostral gradient rule of striatal dopamine depletion in the disease progress. In the pancreas and duodenum, the peak [11C]CFT activity of healthy participants was higher than that of the patients, indicating a potential peripheral dopaminergic neuron impairment in these two regions that may be related to PD. In the stomach, due to the natural peristole influence of this motional organ, resulted in a consistent fluctuation of the SUVmean curve over time. Despite this, the final tracer uptake remained higher in the healthy volunteers, suggesting a potential binding reduction of the stomach in PD patients.
The quantitative method incorporated with kinetic modeling provides direct information on the specific biologic process, representing the tissue density of DAT and relative transporter rate constant of the tracer , whereas the semi-quantitative method of SUVR requires shorter scan acquisition and thus is often favored in clinical setting to estimate such parameters21. Our prior study found certain correlations between the pancreas, duodenum, stomach and the nigrostriatal nuclei in the healthy participants rather than PD patients in the correlation analysis based on [11C]CFT total-body dynamic PET/CT imaging with SUVmean measurements for the “brain-GI” targeted areas35. In this study, we further introduced voxel-based quantitative analysis based on SRTM2 modeling together with the SUVR semi-quantitative analysis to explore the interplay between the digestive and cerebral regions. For the healthy population, there was a negative correlation of the relative blood perfusion estimate R1 between stomach and the certain brain nuclei, whereas in terms of the DAT density estimate DVR, the significant positive correlation between pancreas and the specific brain nuclei were also demonstrated. In addition, both the semi-quantitative estimates, SUVREP and SUVRLP, indicated similar correlation results from the quantification approach in most of the region pairs. This indicated that there may exist an original intrinsic correlation upon the “brain-GI” neuronal chain in healthy individuals rather than the PD patients. Previous research on animal models of DAT specific SPECT tracers (99mTc-TRODAT-1) suggested specific high uptakes in the gastric mucosa 36. Herein the correlation relationship between the stomach and core brain nuclei probably may attribute to the early perfusion of the tracer from the gastric mucosal vascular epithelium. Huang et al.37 found a high pancreas uptake by using standard [11C]CFT PET/CT in healthy individuals. The correlation discrepancies between the HC and PD groups in our findings indicated the native “healthy” interplay pattern related to the digestive and cerebral dopaminergic interaction might be diminished with the disease occurrence.
The simplified reference region model demonstrated a higher computational efficiency and voxel-level stability compared to that of nonlinear least squares method in describing the kinetic of [11C]CFT in rats, suggested by Gunn et al.27. Other advantages of the SRTM2 method also include the avoidance of arterial cannulation and metabolite measurements38. As the compartmental modeling approach with plasma input are usually considered the “gold standard” in kinetic analysis, we additionally performed the VOI-based 1T2K kinetic modeling to verify the estimates from the SRTM221, 26, 27, 38. The results of the highly correlated DVR estimated from SRTM2 using the occipital lobe as the reference and the VT generated from the 1T2K models suggested the accurate tracer quantification without arterial sampling. The significant correlation in VT of pancreas and putamen in the HC group rather than PD group was consistent to the results represented by DVR and SUVRLP, providing more evidence towards the interplay alteration due to the PD occurrence.
Using the correlation analysis between the brain nuclei and digestive organs is a primary attempt to investigate the complex interplay between the digestive and cerebral regions in PD based on noninvasive molecular imaging. Prospective studies with more participants and more records of GI symptoms will be needed to confirm the findings. Since long scanning time in dynamic PET/CT acquisition may be challenging for the PD patients, our results on the high consistency between the quantitative and semi-quantitative estimates will support the semi-quantitative measurements to be the potential surrogates for estimating tracer kinetics with shorter scanning time, facilitating the participant recruitment in the future prospective studies. Moreover, several image-based systematic approaches in brain analysis have been applied to the study of organ-level total-body metabolic network of rodent39, which may have potential prospect in the interpretation of PD etiology and disease evaluation in the clinical practice.