An Integrated Biomarker Approach to Personalized, Adaptive Deep Brain Stimulation in Parkinson Disease

DBS治疗帕金森病[PD]是FDA批准的主要手术方法，已在临床试验中被证明是有效的。然而，这种持续刺激疗法仅限于治疗运动症状的一部分(即震颤、僵直、运动迟缓和运动障碍)，并需要相当大的术后临床调整来治疗症状。对DBS治疗PD的改进正在测试中，包括刺激模式、额外靶点和多个电极的变化。然而，一种关键的新方法涉及使用与临床症状相关的替代生理生物标记物进行自主参数调整[适应性DBS]。这些生物标志物和自主控制可能对参数的动态调整、后续编程和长期优化有用。自适应DBS包括记录代理信号并制定控制策略，通过参数调整将这些信号与活动联系起来。这种方法可以在多个时间尺度上改善DBS治疗，包括短期动态(即几分钟以上)、初始规划(几周到几个月)和长期，具体取决于对DBS的反应时间进程。然而，哪些生物标志物在这些不同的时间尺度上是有用的，以及适当的、多层次的控制政策，将需要与持续的DBS相比进行相对有效性和效率的测试。我们假设，在多个时间尺度上集成多个生物标志物(除了β带振荡之外)将提供更有效的自适应DBS控制。为了验证这一假设，我们将使用植入的先进可植入脉冲发生器[IPGs]对来自人类的多个相关生物标志物进行长期记录，将内部控制模式与高度复杂的外部控制模式进行比较。这些临床实验将集中在一项小型的试验性临床研究(n=6名PD患者)，他们接受了双侧丘脑底核和苍白球DBS电极的植入，所有4个电极都连接到一个单一的美敦力顶峰RC+S记录和刺激IPG。我们已经正式分析了这一队列在1年内的疗效，显示联合STN+GP刺激既是首选的，也比单独使用任何一个部位更好。我们将分析内部(嵌入的，在IPG内可用的)简单自适应DBS与外部(分布式)自适应DBS的比较效果，后者允许集成多个生物标记物和使用复杂的多时间尺度控制策略。我们将进一步开发比例控制反馈方案，具体整合PD症状的时间多个时间常数，以优化控制PD症状。在一个独特的研究患者队列中进行的这些临床试验将产生多种新的结果，继续对STN、GP和联合DBS疗效进行直接的、面对面的比较，分析用于提高DBS疗效的替代生物标志物的最佳组合，并定义用于不同时间尺度的治疗的最佳标量反馈比例控制系统。

DBS therapy for Parkinson Disease [PD], the primary, FDA-approved surgical approach, has proven efficacious in clinical trials. However, this continuous stimulation therapy is limited to treatment of a subset of motor symptoms (i.e., tremor, rigidity, bradykinesia and dyskinesias) and requires considerable postoperative clinical adjustment to treat symptoms. Improvements to DBS for PD are being tested, including changes in patterns of stimulation, additional targets, and multiple electrodes. However, a critical new approach involves autonomous parameter adjustment [adaptive DBS] using surrogate physiological biomarkers relevant to clinical symptoms. These biomarkers and autonomous control may be useful for dynamic adjustment, subsequent programming, and long-term optimization of parameters. Adaptive DBS involves recording surrogate signals and developing a control policy that relates these signals to activity through parameter adjustments. This approach could improve DBS therapy across multiple time scales, including short-term dynamics (i.e., over minutes), initial programming (over weeks to months), and long-term, depending on the time course of response to DBS. However, which biomarkers are useful at these various time scales and appropriate, multi- layered control policy, will require testing in comparison to continuous DBS for relative efficacy and efficiency. We hypothesize that integrating multiple biomarkers (in addition to beta band oscillations) across multiple time scales will provide more efficacious adaptive DBS control. To test this hypothesis we will perform long-term recordings of multiple, relevant biomarkers from humans with implanted, advanced implantable pulse generators [IPGs], comparing internal control modes to highly complex external control modes. These clinical experiments will focus on a small, pilot clinical study (n = 6 PD patients) who have undergone implantation of bilateral subthalamic nucleus [STN] and globus pallidus [GP] DBS electrodes, with all 4 electrodes connected to a single Medtronic Summit RC+S recording and stimulating IPG. We have formally analyzed this cohort for efficacy at 1 year, showing that combined STN + GP stimulation is both preferred and better compared to either site alone. We will analyze the comparative efficacy of internal (embedded, available within the IPG) simple adaptive DBS to external (distributed) adaptive DBS, which allows both integrating multiple biomarkers and using a complex, multiple time scale control policy. We will further develop a proportional control feedback program, which specifically integrates the time multiple time constants of PD symptoms, to optimally control PD symptom. These clinical experiments in a unique cohort of research patients will lead to multiple novel outcomes, continuing a direct, within-person comparison of STN, GP, and combined DBS efficacy, analyzing an optimal mix of surrogate biomarkers for enhancing DBS efficacy, and defining an optimal, scalar feedback, proportional control system for treatment on various time scales.