Lai Yijie, He Naying, Wei Hongjiang, Deng Lifu, Zhou Haiyan, Li Jun, Kaiser Marcus, Zhang Chencheng, Li Dianyou, Sun Bomin
1Department of Neurosurgery, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China.
2Department of Radiology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China.
J Neurosurg. 2022 May 6;138(1):27-37. doi: 10.3171/2022.3.JNS212732. Print 2023 Jan 1.
Functional connectivity shows the ability to predict the outcome of subthalamic nucleus deep brain stimulation (DBS) in Parkinson disease (PD). However, evidence supporting its value in predicting the outcome of globus pallidus internus (GPi) DBS remains scarce. In this study the authors investigated patient-specific functional connectivity related to GPi DBS outcome in PD and established connectivity models for outcome prediction.
The authors reviewed the outcomes of 21 patients with PD who received bilateral GPi DBS and presurgical functional MRI at the Ruijin Hospital. The connectivity profiles within cortical areas identified as relevant to DBS outcome in the literature were calculated using the intersection of the volume of tissue activated (VTA) and the local structures as the seeds. Combined with the leave-one-out cross-validation strategy, models of the optimal connectivity profile were constructed to predict outcome.
Connectivity between the pallidal areas and primary motor area, supplementary motor area (SMA), and premotor cortex was identified through the literature as related to GPi DBS outcome. The similarity between the connectivity profile within the primary motor area, SMA, pre-SMA, and premotor cortex seeding from the VTA-GPi intersection from an out-of-sample patient and the constructed in-sample optimal connectivity profile predicts GPi DBS outcome (R = 0.58, p = 0.006). The predictions on average deviated by 13.1% ± 11.3% from actual improvements. On the contrary, connectivity profiles seeding from the GPi (R = -0.12, p = 0.603), the VTA (R = 0.23, p = 0.308), the VTA outside the GPi (R = 0.12, p = 0.617), or other local structures were found not to be predictive.
The results showed that patient-specific functional connectivity seeding from the VTA-GPi intersection could help in GPi DBS outcome prediction. Reproducibility remains to be determined across centers in larger cohorts stratified by PD motor subtype.
功能连接显示出预测帕金森病（PD）丘脑底核深部脑刺激（DBS）结果的能力。然而，支持其在预测苍白球内侧部（GPi）DBS结果方面价值的证据仍然稀少。在本研究中，作者调查了与PD中GPi DBS结果相关的患者特异性功能连接，并建立了用于结果预测的连接模型。
作者回顾了21例在瑞金医院接受双侧GPi DBS和术前功能磁共振成像的PD患者的结果。使用激活组织体积（VTA）与局部结构的交集作为种子，计算文献中确定与DBS结果相关的皮质区域内的连接图谱。结合留一法交叉验证策略，构建最佳连接图谱模型以预测结果。
通过文献确定苍白球区域与初级运动区、辅助运动区（SMA）和运动前皮质之间的连接与GPi DBS结果相关。来自样本外患者的VTA-GPi交集中的初级运动区、SMA、前SMA和运动前皮质内的连接图谱与构建的样本内最佳连接图谱之间的相似性可预测GPi DBS结果（R = 0.58，p = 0.006）。预测结果与实际改善情况的平均偏差为13.1%±11.3%。相反，发现从GPi（R = -0.12，p = 0.603）、VTA（R = 0.23，p = 0.308）、GPi外的VTA（R = 0.12，p = 0.617）或其他局部结构作为种子的连接图谱无预测性。
结果表明，从VTA-GPi交集中获取的患者特异性功能连接有助于预测GPi DBS结果。在按PD运动亚型分层的更大队列中，各中心之间的可重复性仍有待确定。
