Introduction: Spinal cord stimulation (SCS) is proposed as a novel therapy to treat Parkinson’s Disease (PD) gait disorders — particularly freezing of gait (FOG). However, the exact neurophysiological effects of SCS on supraspinal brain circuitry are not fully understood. Beta frequency oscillation and its modulation by SCS in animals is known to play a role in symptom relief and, thus, is proposed as a biomarker in closed-loop applications. Alternative analytical methods that measure cross-frequency coupling between the phase and amplitude of specific frequency bands – phase-amplitude coupling (PAC) – shed light on information exchange within and across brain regions. Studies have shown that, in PD patients, beta-gamma, alpha-beta, and theta-beta PAC are considerably elevated, and their reduction or modulation ameliorates motor impairment. Thus, it needs to be investigated whether the reduction or modulation of supraspinal PAC facilitates SCS-induced gait improvement in PD patients undergoing unrestrained locomotion.
Materials and
Methods: We recorded local field potentials (LFPs) from bilateral STN electrodes in a PD patient with freezing of gait (FOG) symptoms before and 3 days after implantation of SCS percutaneous leads. LFP recordings were sampled at 250Hz and acquired by the Medtronic Percept PC system while the patient performed six straight-line walks (SLWs) with and without SCS. Phase-amplitude coupling (PAC) analysis was performed in Python using the open-source toolbox tensorpac. Before implementing PAC, local field potential (LFP) recordings from the left and right hemispheres of the STN were preprocessed using a forward-backward Butterworth filter with a passband of 0.5-100Hz. To access the presence of synchronization between the phase of low-frequency neural oscillations and the amplitude of high-frequency oscillations, Gaussian-Copula PAC, a mutual information-based technique, was computed on LFP signals from each STN hemisphere across treatment conditions during individual straight-line walks (SLWs). The frequency for phase ranged from 4-13Hz (0.25Hz steps, 0.5Hz bandwidth) and amplitude from 13-40Hz (0.5Hz steps, 4Hz bandwidth), which were extracted using the Hilbert transform. For comparisons across conditions, the resulting PAC were normalized to the maximum and minimum values within each condition. Mean PAC for phase and amplitude pairs of interest were then extracted by averaging across the corresponding frequency bands and finally across SLWs. Statistical analysis was performed in GraphPad Prism using one-way ANOVA with multiple comparisons.
Results, Conclusions, and Discussions: Preliminary results indicated that when PAC was studied between STN theta (ϑ) and alpha (α) band phase to low beta (Lβ) and high beta (Hβ) band amplitude we observed differential coupling between OFF-SCS and SCS conditions. In the left STN, significant reduction of mean PAC from both pre- and post- surgery OFF-SCS conditions for ϑ-Lβ, α-Lβ, and α-Hβ was observed during the application of SCS. In the right STN, ϑ-Lβ, α-Lβ, and α-Hβ PAC was significantly reduced during SCS when compared to pre-surgery OFF condition. Additionally, when compared with pre-surgery, α-Lβ, and α-Hβ PAC in the right STN was reduced in the post-surgery OFF condition, suggesting a chronic effect of 3-day continuous SCS application. These results demonstrate that SCS-induced gait improvements could be facilitated by a reduction in STN low-frequency PAC. Thus, STN PAC could serve as a useful biomarker to evaluate the therapeutic effect of SCS on PD symptoms. Additional work on the relationship between PAC and specific phases of the gait cycle needs to be studied to understand whether SCS has a phasic effect on PAC modulation. Overall, SCS is emerging as a promising new technology to improve motor symptoms by modulating supraspinal neural activity.
