Long-term Stability of LFP Recorded From the STN and the Effects of DBS

Overview

Movement disorders such as Parkinson Disease, dystonia, and tremor are related to abnormalities of part of the brain known as the basal ganglia. Recently, it has been suggested that the basal ganglia works by oscillations (group of neurons cycle between activation/deactivation when stimulated) of electrical signals. A treatment that involves insertion of electrodes in the subthalamic nucleus (STN) for electrical stimulation, known as deep brain stimulation (DBS), is an established treatment for advanced Parkinson's disease. However its mechanism of action is still not completely understood. Currently, DBS utilizes an "open loop" system whereby the stimulation settings are manually adjusted depending on the patients' clinical response. 1. Determine whether the local field potential (LFP) recorded from the STN is stable over a 1.5 year period. 2. Address whether STN LFP is a suitable control signal, and how it should be used to change DBS parameters

Study Design and Methods Ten Parkinson's disease (PD) patients who are scheduled for bilateral STN DBS will be recruited. DBS surgery will be performed at the Toronto Western Hospital as per standard clinical care. The studies will be performed at two sites: Toronto Western Hospital and London Mobility Laboratory at Western University. Recording of STN LFP using externalized DBS leads, the effects DBS at different frequencies on PD motor symptoms will be studied at the Toronto Western Hospital. Gait studies with full body kinematic measurements, recording of LFP and effects of different stimulation frequencies will be studied at the London Mobility Laboratory (LML) at Western University. An outline of the study timeline is provided in Table (2, 3 and 4). Standardized tasks performed are provided in Table 5.Ten PD patients with bilateral DBS in STN will be studied for six visits to each site (TWH and LML). There will also be a presurgical baseline visit to LML. Patients will first be studied 1-3 days after DBS electrode implantation and before the implantation of the pulse generator. Local field potential (LFP) will be recorded from bilateral STN DBS electrodes using externalized leads. Following internalization of the DBS leads and insertion of the pulse generator (Activa PC+S, Medtronic), patients will be studied within 7 days for a second visit. DBS programming will be performed as per usual clinical practice. Patients will then be studied at 1, 2, 6, 12 and 18 months postoperatively. Visits at the Toronto Western Hospital (TWH) Visit 1 (TWH V1) This visit will obtain "acute" LFP recording for comparison to chronic recordings. Patients will be studied 1-3 days after electrode insertion when the leads are externalized. This is a two day visit, with one day in the off medication state (overnight withdrawal) and one day in the on medication state in random order. For the on medication state, 125% of the levodopa equivalent of the first morning preoperative dose will be administered to increase the chance of recording levodopa-induced dyskinesia. Stage 1: LFP will be recorded from bilateral STN DBS electrodes together with scalp EEG. Fig. 3 shows a schematic representation of the recording setup. Surface EMG and accelerometery will be used to measure the tremor frequencies and levodopa-induced dyskinesias. mUPDRS and the Rush dyskinesia rating scale (for on medication recordings) assessments will be performed. Recordings will be made at rest and while the patients perform self-paced wrist extension movements. The most prominent LFP frequencies in the θ, β and γ band oscillations at rest and during voluntary movements in both on and off medication states (individualized frequencies) as well as the electrode contact that shows the most prominent LFP oscillations will be determined for each patient. This recording will take about one hour. The data analysis will take one hour during which the patient will have a break. Stage 2: This will involve stimulation. The DBS leads will be connected to an external stimulator and a "filter DBS" device to record LFP during DBS. Four DBS conditions will be studied in random order: 1\) high frequency (185 Hz), 2) the individualized γ frequency that show maximum peak at rest (γ rest, determined from stage 1), 3) the individualized γ frequency that show maximum increase with voluntary movements (γ movement, determined from stage 1), and 4) no stimulation. The contact will be selected based on location within the STN from postoperative MRI using methods published by our group and maximum changes in the θ and β bands with voluntary movements (Stage 1). The selected DBS contact will be the cathode (-ve) and anode (+ve) will be a surface electrode placed on the chest wall (similar to contact -ve, case +ve commonly used in therapeutic DBS). Pulse width will be set at100 μs and the current will be the highest (maximum 5 mA) without inducing adverse effect. LFP will be recorded from the electrode contacts adjacent to the one used for DBS using a bipolar montage to allow for maximum cancellation of stimulus artifacts. For example, if contact 2 is used for DBS, LFP will be recorded from contacts 1 and 3. For each condition, 15 min of DBS while resting will be followed by mUPDRS assessment while DBS is maintained. Patients will be video recorded to allow subsequent video scoring of mUPDRS and the Rush dyskinesia rating scale by blinded raters. DBS for the right and left STN will be studied in random order, with one side in the morning and one side in the afternoon. Investigators have performed more complex stimulation protocols in PD patients with STN DBS. The second day of study will occur the next day. The procedure will be the same as Day 1 except that the medication status (on or off medication) will alternate with Day 1 in random order. Visit 2 (TWH-V2) The purpose of this visit is to obtain early LFP recording from the Activa PC+S device before DBS programming begins. This visit will be conducted within 4 weeks of internalization of the DBS leads and insertion of the pulse generator (Activa PC+S), but before the pulse generator is turned on for programming. The visit will be conducted over two days, one day in the off medication state (overnight withdrawal) and one day in the on medication state in random order as in Visit 1. The procedure is identical for Visit 1 with recording of LFP at rest and during a self-paced wrist extension task, followed by testing of 4 different stimulation conditions. However, LFP will be recorded from the Activa PC+S device using the real-time uplink mode to provide continuous recording rather than from externalized leads. Since the device can only record LFP from one pair of contacts from each electrode (right and left) at a time, for recordings in the resting state to examine the difference between on and off medications, investigators will use the "montage sweep" mode that will record alternately from all six possible combinations of bipolar contacts (0-1, 1-2, 2-3, 0-2, 0-3, 1-3) for each electrode. For the wrist extension movements, we will record from the bipolar contact pair that produced the greatest change in γ oscillations with dopaminergic medications and voluntary movements based on the results of Visit 1. In addition, we will record from the adjacent contact pair. For example, if contact 1-2 was found to produce the maximum γ peak, contacts 1-2 as well as contacts 0-1 and 2-3 will be studied. Each pair of electrode contacts will be studied in separate runs. Following the LFP recording (stage 1), the individual γ frequency peaks will be established. Stage 2 will study the effects of DBS at high frequency, γ rest, γ movement and no stimulation assessed in random order with recording of STN LFP during stimulation as in Visit 1. The same contact used for DBS in Visit 1 will be used with monopolar montage (contact -ve, case +ve). LFP with be recorded from the adjacent contacts with bipolar montage as described in Visit 1 but the implanted Activa PC+S device will be used for LFP recording Visits 3- 6 (TWH-V3, V4, V5 and V6) After Visit 2, DBS programming will be performed as per usual clinical practice and takes about 2-4 weeks to complete. Patients will then be studied at 2, 6, 12, and 18 months after surgery. The assessment procedure will be identical to that for Visit 2 except that the "high" frequency stimulation setting will be the setting normally used by the patient for clinical benefit (130-185 Hz) instead of 185 Hz. Visits at London Mobility Laboratory (Western University) Visit 1 will occur before surgery to obtain baseline gait and kinematic measurements data in the on and off medication states. For visits 2, 3 and 7, the TWH visits will occur 1-5 days before the London Mobility Laboratory visits. The γ frequency stimulation to be used in the London Mobility Laboratory visits will be the "γ movement" frequency determined from the corresponding TWH visit.