Merrill J. Birdno1, Alexis M. Kuncel1, Alan D. Dorval1, Dennis A. Turner2, Robert E. Gross3, Warren M. Grill1
1Biomedical Engineering, Duke University, Durham, North Carolina, USA
2Neurosurgery, Duke University Medical Center, Durham, North Carolina, USA
3Department of Neurosurgery, Center for Neurodegenerative Diseases, Emory University School of Medicine Atlanta, Georgia, USA.
Deep brain stimulation (DBS) is an established therapy for the treatment of movement disorders, including essential tremor (ET) and Parkinson’s disease (PD). While the clinical benefits of DBS are well documented, fundamental questions remain about its mechanisms of action. This lack of understanding may limit full development and optimization of this promising treatment. Recent studies demonstrate that the ability of high-frequency DBS to suppress tremor decreased as the temporal irregularity of stimulation increased. However, it was unclear which characteristics of temporally irregular DBS trains reduced their effectiveness as compared to regular DBS at the same average frequency. The purpose of this study was to test which of the following characteristics of the temporally irregular stimulus trains reduced their effectiveness: long pauses, bursts of stimuli, or the irregularity, per se of the stimulus trains.
We conducted intraoperative measurements of the effect of DBS using stimulation trains with variable IPIs in seven subjects with tremor (6 essential tremor and 1 multiple sclerosis). Tremor was measured in 20 s trials using a 3-axis linear accelerometer taped to the back of the hand. DBS was on for 60 s, with tremor measurements recorded after ~30 s of stimulation. Stimulation trials were separated by one-minute epochs without stimulation.
The set of experimental stimulus patterns included three irregular patterns with high entropies (~5.2-5.6 bits/pulse), and two low-entropy patterns (<1 bit/pulse). The first stimulus train (Uniform) was highly irregular, but contained no interpulse intervals (IPIs) outside the established therapeutic frequency range (90 – 380 Hz), therefore no pauses or bursts. The other two high-entropy trains (Unipeak and Bimodal) included IPIs outside the established therapeutic frequency range, and therefore included both pauses and bursts. The two low-entropy patterns consisted of constant rate pulses interrupted by either long pauses (Pause train) or a short burst of pulses at twice the base frequency (Burst train). We also conducted experiments during regular 185 Hz DBS and with DBS off.
A linear mixed effects model of all tremor responses to the experimental stimulus conditions, with subject as the random effect, revealed a significant effect of stimulus condition on tremor (p < 0.0005). As compared to tremor with stimulation off, tremor was suppressed significantly during regular 185 Hz DBS, DBS with the Uniform train, and DBS with the Burst train (p < 0.05, post hoc Tukey’s honest significant difference (HSD) test). However, DBS with the Unipeak, Bimodal, and Pause trains did not suppress tremor significantly as compared to stimulation off. These results indicate that the decreased effectiveness of irregular DBS trains is due to the long IPIs in these trains.
This work was supported by NIH R01-NS040894 , NIH R21-NS055320, and NIH K25-NS053544.