Hi, I'm Nicole Bentley and I'm going to be speaking about some work that we've been doing in our Parkinson's patients looking at an inhibitory control network by investigating the function of the dorsolateral prefrontal cortex as well as subcortical structures. The motivation behind this is that in Parkinson's disease, motor impairment is relatively well treated now with medications and deep brain stimulation, but many of these patients will complain about their cognitive issues even once their motor impairment is well controlled. Cognitive impairment in Parkinson's is very prevalent and disabling. Mild cognitive impairment is diagnosed in 25 to 30 percent of patients at the time of their Parkinson's diagnosis and overt dementia occurs in about 80 percent of patients over the course of their disease. Both of these statistics far outpace the rates seen in the general population. In Parkinson's disease, the types of cognitive deficits seen are various, but specifically involve executive function and processing speed, which is sometimes referred to as bradyphrenia or slow processing. It also affects memory and attention networks. And these localize to several areas including prefrontal cortex and parietal cortex as well as temporal lobes, but many of these functions do consistently activate various areas in the prefrontal cortex and are measured with various objective tests frequently employed by neuropsychologists. And these tests have their drawbacks but are intended to test specific domains of cognitive function. The other main motivation besides the fact that cognitive impairment is very prevalent in Parkinson's is that it lacks effective treatment. So while motor impairment, as stated before, has good established treatments, the treatments for cognitive impairment really rely heavily on medications that are used in Alzheimer's disease. These have a lot of side effects and are not very effective. And traditional DPS, we know, isn't really thought to improve cognition at all, and in some studies has shown to worsen cognition. The overall goal with this project, then, is to first of all study the network electrophysiology between basal ganglia structures and areas to which these functions localize, namely the prefrontal cortex, and to try and discern what types of electrophysiologic signatures underlie both relatively intact and cognitively impaired networks. And then to also, as a second aim, compare various patterns of stimulation that may be used to improve upon preexisting cognitive impairment. So first, the range of cognitive deficits, as I've said, is quite large. In this particular project, we're most interested in response selection or inhibition because patients with Parkinson's have a particular difficulty with this type of cognitive control. And response inhibition can be thought of as the ability to suppress an action or inappropriate response in the face of changing or conflicting environmental demands. So the example that I like to use is that if a person with intact response inhibition were about to cross a road on a crosswalk and they saw a car coming, they could easily step back and suppress that action of moving forward, whereas a patient with Parkinson's or anyone with impaired response inhibition would have difficulty with changing that action once it's already sort of in place. And one formal definition of this is the containment or suppression of prepotent behavioral responses. And there are many tasks that claim to measure this. Some of the most common are the Stroop or Simon tasks, stop signal, go, no-go, and there are others. But patients with Parkinson's have well-described deficits in these areas specifically. So towards the first aim, we wanted to study an area that we felt consistently and reliably subserved these functions. And after looking through fMRI literature as well as electrophysiologic data, there are several areas involved. There is a lot of action in the inferior prefrontal cortex as well as studies looking at the medial prefrontal cortex. But it did seem that the dorsolateral prefrontal cortex was relatively understudied and seemed to have some top-down control over some of the other areas. And it's also conveniently very accessible during awake DBS surgery. So because of this, mainly because of its consistent activation, we decided to look at this area in particular. And one study that came out of Vanderbilt with a group led by Dan Klassens was that in Parkinson's patients, they have deficits in a Simon task, measuring response inhibition, and both performance and the bold signal increased after administration of dopamine. So this does seem to be an area that is involved in the hypodopaminergic state of Parkinson's. It also has direct connections to both STN and basal ganglia. The STN connections are more well described, but the GPI also has connections to the DLPFC. And both of these, of course, are common targets already in deep brain stimulation. And so the thought is that by delivering a DBS electrode and then testing alternative patterns, we could potentially drive a network that is deficient in Parkinson's patients. So to begin looking at the electrophysiology, we're looking at all frequency bands, but in particular, we're interested in the theta band and also the delta band because these are prominent in Parkinson's patients and in older patients in general. And we know that theta underlies many of these functions of cognitive control based on multiple previous EEG and intracranial EEG studies. So the first thing we did was to measure the resting state in a subset of Parkinson's patients in whom we implanted temporarily a six-contact subdural strip a six-contact subdural strip over DLPFC, which is depicted here. And basically showed that in some patients, some subjects, there was a prominent theta peak in some electrodes, as you can see in these examples. In others, there were more prominent beta peaks, which was pretty consistent across these subjects. Here are another four examples. And so theta is seen at rest in some but not all subjects and in some but not all contacts along the strip that's placed sort of down the long axis of what is typically defined as the DLPFC, which corresponds roughly to middle frontal gyrus, depending on which atlas you use. So in the interest of time, I'll go on to the next step here and then our second aim. So obviously, we need to incorporate not just resting data, but data from during performance of a cognitive task. And so we're testing the Simon task and GoNoGo task and the OR. And then correlating the results with pre-existing cognitive status, whether they have MCI, dementia, or cognitively normal, and against different measures performed preoperatively, such as preop, task performance, dementia rating scales, etc. So as our second aim, we want to compare alternative patterns of deep brain stimulation to see if we can essentially derive what would be a normal electrophysiologic pattern in patients with cognitive deficits. This is just an example of what our setup looks like in the OR with the strip in place, DBS in place, and then we're recording from both locations. Again, at rest and during the task. We are looking at two main alternative types of deep brain stimulation. This is just a schematic showing what standard high frequency DBS pulses look like over time as compared to just a continuous 4 hertz or a low theta stimulation. And then we're interested in intermittent theta burst stimulation for various reasons that I won't go into here, but ITBS is borrowed from the TMS literature, where it's FDA approved for depression over when applied over the DLPFC. It also seems to have some effects on cognition as well. So these are just spectrograms from those that cohort of patients I just showed. This is an individual's data showing without stimulation as compared to with ITBS and high frequency. And the main conclusion here is that ITBS seems to interact with the DLPFC to a greater extent than high frequency. This is showing the individual burst of stimulation. And the scalogram showing that it does seem to be driving activity in the theta range in the DLPFC when delivered from the subcortical targets. The preliminary data was recently published in Frontiers of Neuroscience, and basically the take home message here was that when contrasting the GPI and STN, we consistently had more activity in the DLPFC when stimulating from the GPI. So this is just an example showing pre-stimulation and then with ITBS, so not 4 hertz here, and there was a rise in theta power, whereas in the STN over a group of patients, this did not seem to be as robust a result. We have a small sample, so we are replicating these data. But it serves as the motivation for looking at both targets, but we're very interested in what the GPI to DLPFC network may be doing. A lot of work has been done in STN looking at the hyperdirect pathway, but this would suggest that there's also reasons to be looking at the GPI as well. So next steps would be, again, incorporating the cognitive task, looking at changes both in performance and the electrophysiologic activity during various types of stimulation, including 4 hertz and ITBS. We also are interested in measures of connectivity between the subcortical and cortical structures, and ultimately determining if driving theta activity or, you know, whatever the normal native pattern may be, if driving that does lead to improved cognition and the subset of patients who have cognitive impairment. So that's where we are currently with this project, and I appreciate your time and your attention.

Parkinson's patients

cognitive impairment

inhibitory control network

dorsolateral prefrontal cortex

subcortical structures