基于脑代谢网络分析的阿尔茨海默病和路易小体痴呆的早期鉴别诊断研究

Early differential diagnosis is difficult between Alzheimer’s disease (AD) and dementia with Lewy bodies (DLB) in clinical practice, for which there is a lack of ideal biomarkers. Our preliminary work demonstrated that different metabolic brain networks including metabolic pattern and metabolic connectivity based on 18F-FDG PET imaging exit in both AD and DLB. However, the values of metabolic brain networks in discriminating AD from DLB and their underlying pathological basis are still unknown. In this study, both metabolic pattern and metabolic connectivity analyses will be performed firstly in a cohort of AD and DLB patients to assess the accuracy of AD- and DLB-related metabolic networks for differential diagnosis. Then the validation will be done in another cohort of patients with AD and DLB and the strategy for single-subject classification will be identified. Following that, the prognosis prediction in individual patient with mild cognitive impairment (MCI) based on the single-subject analyses of baseline 18F-FDG imaging will be compared with the final clinical diagnosis to demonstrate the early diagnostic performance for discrimination between AD and DLB. Lastly, the longitudinal changes with time in metabolic brain networks and Aβ deposit and the correlation between them in patients with MCI due to AD or those due to DLB will be utilized to evaluate their values of course monitoring in the prodromal stage. This study will provide valuable biomarkers for early differential diagnosis and deciding treatment option as well as for identifying participants for clinical trials of AD and DLB.

阿尔茨海默病（AD）和路易小体痴呆（DLB）的早期鉴别诊断是临床一大难点，目前尚无理想的客观标志物。我们前期的18F-FDG PET显像研究发现， AD和DLB在脑葡萄糖代谢网络方面存在不同模式和连接，但脑代谢网络在两者鉴别诊断中的价值及相关病理基础尚不清楚。本研究将首先从脑代谢模式和代谢连接两方面分别对AD和DLB进行网络特征分析，评价脑代谢网络对两者鉴别诊断的准确性；然后在另一组AD和DLB患者中对脑代谢网络进行验证，建立个体化诊断策略；继而根据基线18F-FDG显像对轻度认知损害（MCI）患者的转归进行个体化预测，并与临床随访结果对比，评价脑代谢网络对AD和DLB早期鉴别诊断的准确性；最后分别纵向观察转化为AD或DLB的MCI的脑代谢网络和Aβ沉积随时间的变化及相互关系，评价其监测前驱期病程的价值。本研究将为AD和DLB的早期鉴别诊断和治疗决策以及临床试验对象的选择提供有用的标志物。

阿尔茨海默病（AD）和路易小体痴呆（DLB）的早期鉴别诊断目前尚无理想的客观标志物。我们在前期研究的基础上拟评价AD和DLB相关脑葡萄糖代谢网络在两者鉴别诊断中的价值及相关病理基础。本研究首先从脑代谢模式和代谢连接两方面对AD和DLB进行网络特征分析和验证；继而对轻度认知损害（MCI）患者进行随访，纵向观察转化为AD或DLB的MCI的脑代谢网络和Aβ沉积随时间的变化及相互关系，评价脑代谢网络的早期诊断和病程监测价值。结果如下：1.DLB相关脑代谢模式（DLBRP）表现为双侧颞中回、枕中回、舌回、楔前叶、楔叶、角回、顶上回、顶下回、额中回、额下回、扣带回和尾状核相对低代谢，而小脑、壳核、丘脑、中央前后回、旁中央小叶相对高代谢，与帕金森病痴呆相关脑代谢模式（PDDRP）具有高度相似的图像特征和模式表达值以及疾病严重度评估价值。这种高度一致性为两者属于同一疾病谱系理论提供了新的证据。2.AD相关脑代谢模式（ADRP）表现为双侧额中回、楔前叶、扣带回、颞中回相对代谢减低，伴小脑、丘脑、舌回、中央后回相对代谢增高，而不典型AD即后部皮层萎缩（PCA）相关脑代谢模式（PCARP）表现为双侧顶枕叶、颞中回相对代谢减低，伴小脑、丘脑相对代谢增高；ADRP表达值在AD和PCA组均与记忆评分呈显著负相关，但仅PCARP表达值与PCA组的视空间功能评分呈显著负相关。这提示PCARP是PCA病情严重度评估的最佳指标。3.后扣带回/（楔前叶+楔叶）或后扣带回/枕叶的葡萄糖代谢半定量值比值可以准确区分AD和DLB。4.进展为AD/DLB的MCI患者基线期ADRP/DLBRP表达值即显著增高，且随病程进展进一步增高，这提示ADRP/DLBRP可用于AD/DLB的早期诊断和病情进展评价；进展为AD的MCI患者基线期Aβ沉积水平即显著增高，随病程进展其顶叶Aβ沉积水平较基线期显著增加，且其变化与ADRP表达值变化呈显著正相关。5.稀疏逆协方差评估（SICE）分析显示AD和DLB的脑代谢连接特征存在显著差异，有望用于两者的鉴别诊断。本研究将为AD和DLB的早期诊断与鉴别诊断、治疗决策以及临床试验对象的选择提供了有用的标志物。