Cognitive impairment is a common and disabling nonmotor symptom of Parkinson's disease (PD) and is associated with decreased quality of life and increased mortality. Deep brain stimulation (DBS) is used to treat the motor symptoms of PD. However, subthalamic nucleus DBS can also cause cognitive impairment in some patients. The subthalamic nucleus is divided into regions important for motor, cognitive, and emotional processes. Cognitive side effects of DBS may be related to alteration of activity in the cognitive division of this nucleus by stimulation intended to target the motor division. Increases in beta power in motor circuitry are associated with motor symptoms in PD patients. I have previously found that beta power changes in cognitive circuitry including dorsolateral prefrontal cortex similarly are associated with cognitive symptoms. Research has suggested that a closed loop DBS technique using beta power, where DBS stimulation turns on when beta power is elevated, may improve motor symptoms in PD. The goal of the present proposal is to identify neural activity biomarkers of cognitive changes induced by DBS. This will allow us to develop closed loop DBS paradigms to minimize cognitive side effects and treat cognitive symptoms in PD patients in the future.

We will use neural activity recordings in PD patients to identify these biomarkers. DBS surgery is commonly performed with patients awake in order to perform neural recordings and stimulation to optimize the position of the DBS electrode. We will put an additional electrode on the exposed surface of the brain during surgery to record from the dorsolateral prefrontal cortex. This technique has previously been reported with excellent safety profile. We will record from this electrode while subjects are resting quietly and while they participate in a working memory task, both with no stimulation and with stimulation at the DBS target turned on. We will identify whether patients develop postoperative cognitive impairment from routine postoperative cognitive testing and compare changes in dorsolateral prefrontal cortex activity associated with stimulation between patients who develop cognitive impairment and those who do not. We will also recruit patients who had previously been implanted with DBS and have an existing DBS battery capable of recording neural activity to participate in a study where we record neural activity from their implanted electrodes with DBS turned on and off, both at rest and during a working memory task. We will compare neural activity between patients who have stimulation associated cognitive impairment and those who do not.

These studies will allow us to identify neural activity signatures in dorsolateral prefrontal cortex and subthalamic nucleus associated with DBS related cognitive impairment. We anticipate that these findings will immediately be able to help identify DBS patients at risk of stimulation related cognitive impairment, which could help guide management of these patients and ensure that they receive appropriate interventions. Such signatures could also be used in the intraoperative setting to guide electrode placement: if the signature is seen with stimulation the electrode could be repositioned to find a position with similar motor benefit but without the signature of cognitive impairment, similar to how neural activity and stimulation findings are currently used to guide electrode placement to optimize motor benefit and minimize motor side effects. Finally, the signatures could be used to guide DBS programming. If the signature is identified during DBS programming the contact being stimulated or stimulation parameters could be adjusted until the patient has improvement in motor symptoms without cognitive side effects.

In the longer term, we anticipate that these findings will be used to help develop closed loop DBS paradigms. Identifying neural signatures of stimulation associated cognitive impairment would allow us to develop DBS algorithms where stimulation is temporarily turned off or adjusted in response to the neural signature. The technology to do this exists in clinical practice and closed loop approaches are currently being studied to optimize PD motor symptoms. We therefore anticipate that a paradigm to minimize cognitive symptoms could be in clinical use within a few years of completion of this proposal and identification of the relevant signatures of cognitive impairment. In the more distant future, we hope that identifying signatures of neural activity associated with cognitive impairment may allow us to develop DBS algorithms to change disordered neural activity to a state associated with normal cognitive functioning, and therefore to develop DBS as a new treatment for PD cognitive impairment in the larger PD population.

In summary, we expect that the proposed research will allow us to identify neural signatures in cognitive brain structures associated with DBS related cognitive impairment. This will help us to understand the underlying mechanism of DBS related cognitive impairment, help to identify patients at risk for this side effect, guide electrode placement during surgery, and improve programming to decrease its occurrence. We also anticipate that these findings will allow us to develop improved DBS stimulation algorithms to minimize this side effect and ultimately to treat PD patients with cognitive impairment.