Maccioni Lucia, Brusaferri Ludovica, Barzon Leonardo, Schubert Julia J, Nettis Maria A, Cousins Oliver, Rosenzweig Ivana, Mizuno Yuya, Vicente-Rodríguez Marta, Singh Nisha, Marques Tiago R, Harrison Neil A, Fryer Tim, Bullmore Edward T, Cash Diana, Mondelli Valeria, Pariante Carmine, Howes Oliver, Turkheimer Federico E, Loggia Marco L, Veronese Mattia
Department of Information Engineering, University of Padova, Padova, Italy.
Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, USA.
J Cereb Blood Flow Metab. 2025 Jul 31:271678X251361261. doi: 10.1177/0271678X251361261.
Positron Emission Tomography (PET) of 18 kDa translocator protein (TSPO) has been investigated as putative marker of neuroinflammation but faces substantial methodological challenges. These include issues with arterial blood sampling for kinetic modeling, the absence of suitable reference regions, genetic polymorphisms affecting tracer affinity, altered blood-to-brain tracer delivery in inflammatory conditions, and high signal variability. This study presents a novel blood-free reference-free method for TSPO PET quantification, leveraging a logistic regression model to estimate the probability of TSPO overexpression across brain regions. Validation was performed on 323 human brain scans from five datasets and three radiotracers. The quantified TSPO topology in healthy controls showed strong concordance with constitutive TSPO gene expression for all tracers. When using [C]PBR28 PET data, the method replicated previous findings in schizophrenia, Alzheimer's disease, chronic pain, and XBD173 blocking. However, model extension to [F]DPA-714 and [C]-(R)-PK11195 revealed small effect sizes and high variability, suggesting the need for tracer-specific model optimization. Finally, validation in a rat model of lipopolysaccharide-induced neuroinflammation confirmed previous evidence of increased brain TSPO uptake after systemic challenge. This novel non-invasive method provides individualized TSPO PET quantification, demonstrating broad applicability across TSPO PET tracers and imaging sites, assuming sufficient training data for model development.
18 kDa转运蛋白（TSPO）的正电子发射断层扫描（PET）已被研究作为神经炎症的假定标志物，但面临重大的方法学挑战。这些挑战包括用于动力学建模的动脉血采样问题、缺乏合适的参考区域、影响示踪剂亲和力的基因多态性、炎症条件下血脑示踪剂递送的改变以及高信号变异性。本研究提出了一种用于TSPO PET定量的新型无血无参考方法，利用逻辑回归模型估计脑区TSPO过表达的概率。在来自五个数据集和三种放射性示踪剂的323例人脑扫描上进行了验证。健康对照中量化的TSPO拓扑结构显示，所有示踪剂与组成性TSPO基因表达具有很强的一致性。使用[C]PBR28 PET数据时，该方法重复了先前在精神分裂症、阿尔茨海默病、慢性疼痛和XBD173阻断方面的研究结果。然而，将模型扩展到[F]DPA - 714和[C]-(R)-PK11195显示效应量小且变异性高，表明需要针对示踪剂进行特定的模型优化。最后，在脂多糖诱导的神经炎症大鼠模型中的验证证实了先前关于全身攻击后脑TSPO摄取增加的证据。这种新型非侵入性方法提供了个体化的TSPO PET定量，表明在有足够的模型开发训练数据的情况下，该方法在TSPO PET示踪剂和成像部位具有广泛的适用性。
