NCT03180515

Neural and Kinematic Features of Freezing of Gait for Adaptive Neurostimulation

Completed NA Results posted Last updated 21 November 2019

What this trial tests

NA trial testing Activa PC+S Neurostimulator in Parkinson Disease in 12 participants. Completed in 31 October 2018.

Timeline

15 May 2017

Primary endpoint
31 October 2018

31 October 2018

Quick facts

Lead sponsor	Stanford University
Phase	NA
Status	Completed
Study type	INTERVENTIONAL
Allocation	na
Design	single group
Masking	none
Primary purpose	basic science
Enrollment	12
Start date	15 May 2017
Primary completion	31 October 2018
Estimated completion	31 October 2018
Sites	1 location across United States

Drugs / interventions tested

• Activa PC+S Neurostimulator

Conditions studied

• Parkinson Disease — all drugs for Parkinson Disease →

Sponsor

Stanford University

Who can join

Adults 18 to 80, any sex, with Parkinson Disease. Patients with the condition only — healthy volunteers not accepted.

Results — posted to ClinicalTrials.gov

Per-arm endpoint measurements with 95% confidence intervals where reported. Source: trial results section.

Incidence of Treatment-Emergent Adverse Events [Safety and Tolerability] Related to aDBS Primary · 30 min - 2 hours

Safety, tolerability and feasibility of aDBS

Group	Value	95% CI
(OFF) Activa PC+S Neurostimulator	0
(cDBS) Activa PC+S Neurostimulator	0
(aDBS) Activa PC+S Neurostimulator	0

Aim 1: Alpha Power Secondary · 30 minutes

Subthalamic nucleus (STN) local field potentials (LFP) recordings demonstrate oscillatory neuronal activity in both the alpha (8-12 Hz) and beta (13-30 Hz) bands in the resting state in PD. Spectrograms were generated using a short-time Fourier transform, with a 1 second Hanning window and a 0.5 second overlap, creating a frequency resolution of 1 Hz. Power spectral densities were calculated using the Welch method with the same window and overlap parameters. Power was summed in the beta and alpha bands. This power can be representative of the magnitude of oscillatory activity in this frequency

Stepping in Place Task

Group	Value	95% CI
Freezers	0.845	± 0.543
Non-Freezers	1.051	± 0.511

Forward Walking

Group	Value	95% CI
Freezers	1.710	± 1.280
Non-Freezers	2.753	± 3.261

Turning and Barrier Course

Group	Value	95% CI
Freezers	1.393	± 0.901
Non-Freezers	1.890	± 1.351

Aim 1: Beta Power Secondary · 30 minutes

Subthalamic nucleus (STN) local field potentials (LFP) recordings demonstrate oscillatory neuronal activity in both the alpha (8-12 Hz) and beta (13-30 Hz) bands in the resting state in PD. Spectrograms were generated using a short-time Fourier transform, with a 1 second Hanning window and a 0.5 second overlap, creating a frequency resolution of 1 Hz. Power spectral densities were calculated using the Welch method with the same window and overlap parameters. Power was summed in the beta and alpha bands. This power can be representative of the magnitude of oscillatory activity in this frequency

Stepping in Place Task

Group	Value	95% CI
Freezers	4.085	± 2.077
Non-Freezers	11.258	± 8.493

Forward Walking

Group	Value	95% CI
Freezers	9.143	± 6.835
Non-Freezers	8.400	± 5.529

Turning and Barrier Course

Group	Value	95% CI
Freezers	7.321	± 4.429
Non-Freezers	6.993	± 0.207

Aim 1: Alpha Sample Entropy Secondary · 30 minutes

The predictability of the local field potentials (band-pass filtered between 8-12 Hz for alpha) was analyzed using Sample Entropy (SampEn), a nonlinear measure suitable for physiological time series. SampEn may be a more consistent measure and more suitable to shorter time series data than approximate entropy, partially due to the elimination of counting self matches. SampEn is calculated as the negative logarithm of the estimated conditional probability that if consecutive subseries of length m are similar according to some preset tolerance r, the consecutive subseries of length m+1 will be s

Stepping in Place Task

Group	Value	95% CI
Freezers	0.251	± 0.002
Non-Freezers	0.253	± 0.004

Forward Walking

Group	Value	95% CI
Freezers	0.248	± 0.007
Non-Freezers	0.247	± 0.007

Turning and Barrier Course

Group	Value	95% CI
Freezers	0.253	± 0.003
Non-Freezers	0.250	± 0.004

Aim 1: Beta Sample Entropy Secondary · 30 minutes

The predictability of the local field potentials (band-pass filtered between 15-30 Hz for beta) was analyzed using Sample Entropy (SampEn), a nonlinear measure suitable for physiological time series. SampEn may be a more consistent measure and more suitable to shorter time series data than approximate entropy, partially due to the elimination of counting self matches. SampEn is calculated as the negative logarithm of the estimated conditional probability that if consecutive subseries of length m are similar according to some preset tolerance r, the consecutive subseries of length m+1 will be s

Stepping in Place Task

Group	Value	95% CI
Freezers	0.407	± 0.031
Non-Freezers	0.337	± 0.062

Forward Walking

Group	Value	95% CI
Freezers	0.380	± 0.050
Non-Freezers	0.376	± 0.053

Turning and Barrier Course

Group	Value	95% CI
Freezers	0.377	± 0.050
Non-Freezers	0.376	± 0.052

Aim 2: Asymmetry Secondary · 30 minutes

Asymmetry during both forward walking and stepping in place was calculated using periods of walking or stepping when the subject was not freezing. According to previous studies, asymmetry is defined as: 100\*(absolute value of the natural log of the shorter average swing time over the longer average swing time) or mathematically: 100\*\| ln (SSWT/LSWT) \| where SSWT = shorter mean swing time LSWT = longer mean swing time

Stepping in Place Task

Group	Value	95% CI
Freezers (OFF Stim)	26.74	± 23.44
Freezers (60 Hz Stim)	27.1	± 21.42
Freezers (140 Hz Stim)	15.99	± 12.12
Non-Freezers (OFF Stim)	9.54	± 6.87
Non-Freezers (60 Hz Stim)	7.86	± 7.10
Non-Freezers (140 Hz Stim)	7.25	± 9.03

Forward Walking

Group	Value	95% CI
Freezers (OFF Stim)	6.11	± 4.11
Freezers (60 Hz Stim)	5.01	± 2.97
Freezers (140 Hz Stim)	4.52	± 3.39
Non-Freezers (OFF Stim)	3.32	± 1.88
Non-Freezers (60 Hz Stim)	2.41	± 2.50
Non-Freezers (140 Hz Stim)	4.50	± 2.23

Aim 2: Arrhythmicity Secondary · 30 minutes

Arrhythmicity during both forward walking and stepping in place was calculated using periods of walking or stepping when the subject was not freezing. According to previous studies, arrhythmicity is defined as the mean stride time coefficient of variation of both legs, and a greater stride time CV implies less rhythmic gait or stepping. Higher arrhythmicity corresponds to more arrhythmic, or more impaired, gait.

Stepping in Place Task

Group	Value	95% CI
Freezers (OFF Stim)	54.04	± 50.46
Freezers (60 Hz Stim)	27.49	± 33.23
Freezers (140 Hz Stim)	29.34	± 56.18
Non-Freezers (OFF Stim)	4.01	± 0.81
Non-Freezers (60 Hz Stim)	4.10	± 1.40
Non-Freezers (140 Hz Stim)	4.00	± 0.73

Forward Walking

Group	Value	95% CI
Freezers (OFF Stim)	6.76	± 3.29
Freezers (60 Hz Stim)	5.18	± 2.25
Freezers (140 Hz Stim)	6.41	± 2.68
Non-Freezers (OFF Stim)	5.98	± 4.37
Non-Freezers (60 Hz Stim)	4.96	± 1.35
Non-Freezers (140 Hz Stim)	4.94	± 1.38

Aim 2: Stride Time Secondary · 30 minutes

Kinematic Features associated with Freezing of Gait

Stepping in Place Task

Group	Value	95% CI
Freezers (OFF Stim)	1.71	± 0.68
Freezers (60 Hz Stim)	1.44	± 0.53
Freezers (140 Hz Stim)	1.21	± 0.44
Non-Freezers (OFF Stim)	1.07	± 0.15
Non-Freezers (60 Hz Stim)	1.05	± 0.14
Non-Freezers (140 Hz Stim)	1.06	± 0.17

Forward Walking

Group	Value	95% CI
Freezers (OFF Stim)	1.14	± 0.19
Freezers (60 Hz Stim)	1.11	± 0.15
Freezers (140 Hz Stim)	1.10	± 0.14
Non-Freezers (OFF Stim)	1.16	± 0.06
Non-Freezers (60 Hz Stim)	1.15	± 0.07
Non-Freezers (140 Hz Stim)	1.18	± 0.10

Aim 2: Percent Time Freezing Secondary · 30 minutes

Freezing of gait episodes during stepping in place were identified using a validated computerized algorithm, and during forward walking by a blinded rater. The percent time freezing was calculated by dividing the time spent freezing by the total time to complete the task then multiplying by 100 to get a percent. If no freezing was observed, then the percent time freezing reported was 0.0%.

Stepping in Place Task

Group	Value	95% CI
Freezers (OFF Stim)	29.37	± 38.27
Freezers (60 Hz Stim)	11.01	± 29.57
Freezers (140 Hz Stim)	18.13	± 34.58
Non-Freezers (OFF Stim)	0.0	± 0.0
Non-Freezers (60 Hz Stim)	0.0	± 0.0
Non-Freezers (140 Hz Stim)	0.0	± 0.0

Forward Walking

Group	Value	95% CI
Freezers (OFF Stim)	0.17	± 0.48
Freezers (60 Hz Stim)	0.0	± 0.0
Freezers (140 Hz Stim)	0.0	± 0.0
Non-Freezers (OFF Stim)	0.0	± 0.0
Non-Freezers (60 Hz Stim)	0.0	± 0.0
Non-Freezers (140 Hz Stim)	0.0	± 0.0

Percent Time Freezing Secondary · 30 minutes

Freezing of gait episodes during stepping in place were identified using a validated computerized algorithm. The percent time freezing was calculated by dividing the time spent freezing by the total time to complete the task then multiplying by 100 to get a percent. If no freezing was observed, then the percent time freezing reported was 0.0%. The percent time spent freezing was compared while the participant was doing the stepping in place task on continuous deep brain stimulation (cDBS) and while the participant was doing the stepping in place task on adaptive deep brain stimulation (aDBS).

Group	Value	95% CI
(OFF) Activa PC+S Neurostimulator	44	± 0
(cDBS) Activa PC+S Neurostimulator	2	± 0
(aDBS) Activa PC+S Neurostimulator	0	± 0

Sponsor's own description

Continuous deep brain stimulation (cDBS) is an established therapy for the major motor signs in Parkinson's disease, however some patients find that it does not adequately treat their freezing of gait (FOG). Currently, cDBS is limited to "open-loop" stimulation,without real-time adjustment to the patient's state of activity, fluctuations and types of motor symptoms, medication dosages, or neural markers of the disease. The purpose of this study is to determine if an adaptive DBS system,responding to patient specific, clinically relevant neural or kinematic feedback related to FOG, is more effective than continuous DBS on the motor Unified Parkinson's Disease Rating Scale (UPDRS III) and gait measures of PD.

Publications & conference data

No peer-reviewed publications indexed yet for this trial. Completed trials usually publish results within 12-18 months.