Different z-score cut-offs for striatal binding ratio (SBR) of DaT SPECT are needed to support the diagnosis of Parkinson's Disease (PD) and dementia with Lewy bodies (DLB)
Different z-score cut-offs for striatal binding ratio (SBR) of DaT SPECT are needed to support the diagnosis of Parkinson's Disease (PD) and dementia with Lewy bodies (DLB)
复制标题
DOI：
10.1007/s00259-022-06069-0
复制
发表时间：
2022-12-06
影响因子：
9.1
通讯作者：
Morbelli, Silvia
中科院分区：
文献类型：
--
作者：
Lanfranchi, Francesco;Arnaldi, Dario;Morbelli, Silvia
目的 迄今为止，纹状体或壳核 SBR 的 -2 z 分数截止值已被任意用于定义疑似神经退行性帕金森病患者的异常 DaT SPECT。我们的目的是通过实验确定纹状体和纹状体下区域 SBR 最准确的 z 分数截止值，以分别独立区分特发性震颤 (ET) 和阿尔茨海默病 (AD) 的 PD 和 DLB。方法 纳入 225 名接受 DaT SPECT 的患者（75 名从头 PD、80 名 ET、50 名 DLB 和 20 名 AD）。半定量通过 Datquant (R) 软件进行计算，该软件返回属于帕金森病进展标记计划 (PPMI) 的 118 名健康志愿者的纹状体 SBR 和 z 分数测量值。 ROC 分析用于确定纹状体和纹状体下区域 z 分数最准确的截止值（随访时的临床诊断作为金标准）。结果 z 分数截止值为 -1.27 的受影响最严重的半球 (MAH) 的后壳核表现出区分 PD 和 ET 的最高准确度（敏感性 0.97，特异性 0.94）。整个壳核（z 分数截止值 -0.96）是支持 DLB 诊断的最准确参数（敏感性 0.74，特异性 0.95）。壳核与尾状核的比率可以准确地检测 PD（尤其是在早期阶段），但不能准确地检测 DLB 患者。结论 我们通过实验证明，应考虑不同的纹状体区域和 SBR z 分数的截止值来支持 PD 或 DLB 的诊断。所确定的不太保守的截止值显示出更高的灵敏度，而对于任意 -2 z 分数，特异性没有明显降低。
Purpose A cut-off of -2 z-score for striatal or putaminal SBR has been to date arbitrarily used to define an abnormal DaT SPECT in patients with suspected neurodegenerative parkinsonism. We aimed to experimentally identify the most accurate z-score cut-offs for SBR of striatal and substriatal regions to independently discriminate PD and DLB, with respect to essential tremor (ET) and Alzheimer's disease (AD) respectively. Methods Two-hundred twenty-five patients undergoing DaT SPECT were enrolled (seventy-five de novo PD, eighty ET, fifty DLB, and twenty AD). Semiquantification was computed by means of Datquant (R) software which returns measures of striatal SBR and z-scores with respect to 118 healthy volunteers belonging to the Parkinson's Progression Markers Initiative (PPMI). ROC analysis was used to identify most accurate cut-offs for z-score for striatum and substriatal regions (clinical diagnosis at follow-up as gold standard). Results Posterior putamen of the most affected hemisphere (MAH) with a z-score cut-off of -1.27 demonstrated the highest accuracy to differentiate between PD and ET (sensitivity 0.97, specificity 0.94). The whole putamen (z-score cut-off -0.96) was the most accurate parameter to support the diagnosis of DLB (sensitivity 0.74, specificity 0.95). Putamen to caudate ratio was accurate to detect PD (especially in early stages) while not DLB patients. Conclusion We experimentally demonstrated that different substriatal regions and cut-offs for z-score of SBR should be considered to support the diagnosis of either PD or DLB. The identified less conservative cut-offs showed higher sensitivity without a measurable reduction in specificity with respect to the arbitrary -2 z-score.