Common neural structures activated by epidural and transcutaneous lumbar spinal cord stimulation: Elicitation of posterior root-muscle reflexes

Abstract

Epidural electrical stimulation of the lumbar spinal cord is currently regaining momentum as a neuromodulation intervention in spinal cord injury (SCI) to modify dysregulated sensorimotor functions and augment residual motor capacity. There is ample evidence that it engages spinal circuits through the electrical stimulation of large-to-medium diameter afferent fibers within lumbar and upper sacral posterior roots. Recent pilot studies suggested that the surface electrode-based method of transcutaneous spinal cord stimulation (SCS) may produce similar neuromodulatory effects as caused by epidural SCS. Neurophysiological and computer modeling studies proposed that this noninvasive technique stimulates posterior-root fibers as well, likely activating similar input structures to the spinal cord as epidural stimulation. Here, we add a yet missing piece of evidence substantiating this assumption. We conducted in-depth analyses and direct comparisons of the electromyographic (EMG) characteristics of short-latency responses in multiple leg muscles to both stimulation techniques derived from ten individuals with SCI each. Post-activation depression of responses evoked by paired pulses applied either epidurally or transcutaneously confirmed the reflex nature of the responses. The muscle responses to both techniques had the same latencies, EMG peak-to-peak amplitudes, and waveforms, except for smaller responses with shorter onset latencies in the triceps surae muscle group and shorter offsets of the responses in the biceps femoris muscle during epidural stimulation. Responses obtained in three subjects tested with both methods at different time points had near-identical waveforms per muscle group as well as same onset latencies. The present results strongly corroborate the activation of common neural input structures to the lumbar spinal cord-predominantly primary afferent fibers within multiple posterior roots-by both techniques and add to unraveling the basic mechanisms underlying electrical SCS.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Post-activation depression of evoked EMG responses to transcutaneous and epidural SCS.

(A) Exemplary EMG responses to paired stimuli (stim., black triangles) applied transcutaneously with a 50-ms interstimulus interval evoked in rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA), and triceps surae muscle group (TS); responses of one lower limb of subjects 5 and 8. Shown are three repetitions superimposed. The initial brief biphasic peaks are stimulation artifacts consistently generated with the surface-electrode based stimulation. Two examples are shown to illustrate the intraindividual variability of post-activation depression. (B) Group results across subjects tested with transcutaneous SCS (tSCS). (C) Exemplary EMG responses to the first two pulses of a 20-Hz train applied epidurally, subjects 14 and 17. (D) Group results across subjects with epidural SCS (eSCS) considering data obtained with both bipolar electrode combinations tested. Responses to the second pulses (light purple bars in (B) and light blue bars in (D)) were significantly depressed compared to those evoked by the first pulses (purple bars in (B) and blue bars in (D)) for both transcutaneous and epidural SCS. Bars are group means, error bars indicate SE. Asterisks indicate significant effects (***, P

Fig 2. EMG characteristics of responses to transcutaneous and epidural SCS.

(A) Characteristic EMG waveforms of responses to single-pulse transcutaneous (tSCS) and 2-Hz epidural SCS (eSCS) with respective common threshold intensities, obtained by stimulus-triggered averaging of the available set of evoked potentials derived from rectus femoris (RF), subjects 8 (transcutaneous stimulation) and 15 (epidural stimulation); biceps femoris (BF), subjects 4 and 13; tibialis anterior (TA), subjects 6 and 16; and the triceps surae muscle group (TS), subjects 9 and 13. The individual examples were selected to illustrate the most representative muscle-specific shapes of the EMG potentials. Black arrowheads indicate the onsets of the effective phases of the stimulation pulses. (B) Group results of onset latencies, offset latencies, and EMG potential durations of the responses to transcutaneous (purple bars) and epidural SCS (blue bars). Bars are group means, error bars indicate SE. Asterisks indicate significant differences (*, P

Fig 3. Normalized thresholds of responses to transcutaneous and epidural SCS and peak-to-peak amplitudes at common-threshold intensity.

(A) Normalized thresholds for rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA), and the triceps surae muscle group (TS) for transcutaneous SCS (tSCS, purple crosses), epidural SCS (eSCS) with a rostral cathode site (mid-blue crosses), and epidural SCS with a caudal cathode site (light blue crosses). (B) Group results of response sizes of RF, BF, TA, and TS to transcutaneous SCS (purple bars) as well as epidural SCS with a bipolar electrode combination using a rostral cathode site (mid-blue bars) and a caudal cathode site (light blue bars) applied with the respective common threshold intensities. Crosses in (A) and bars in (B) are group means, error bars indicate SE. Significant results of the pairwise post-hoc tests are indicated with asterisks (*, P

Fig 4. EMG characteristics of responses to transcutaneous and epidural SCS obtained within same individuals.

(A) The evoked responses to transcutaneous (tSCS; solid, purple lines) and epidural SCS (eSCS; dashed, blue lines) in rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA), and triceps surae muscle group (TS) appeared in the EMG largely as bi- or triphasic waveforms, with near-identical shapes for both stimulation methods for most muscles. Shown are mean waveforms of the respective evoked potentials after stimulus-triggered averaging; responses in each row are derived from one lower limb each of subjects (S) 8–10. Black arrowheads indicate the respective effective phases of the stimulation pulses. (B) Individual mean onset latencies, offset latencies, and EMG potential durations of the responses to transcutaneous (purple bars) and epidural SCS (blue bars). Error bars indicate SD, asterisks significant differences (*, P