Characterizing spinal reflexes evoked by sensory spinal cord stimulation in people with lower-limb amputation.

Background: People with lower-limb amputation lack sensory inputs from their missing limb, which increases their risk of falling. We recently demonstrated that spinal cord stimulation (SCS) can restore sensation in the missing lower limb. Previous studies have shown that SCS can affect motor control by exciting spinal reflex pathways after stroke or spinal cord injury. The effects of SCS on spinal reflex activation have not been studied in people with lower-limb amputation. Furthermore, it is unknown if SCS-evoked spinal reflex activation would perturb walking. Therefore, the goal of this study was to characterize SCS-evoked spinal reflexes in people with lower-limb amputation and quantify effects on gait parameters, including step cycle duration and limb alternation symmetry.

Methods: We implanted percutaneous SCS electrodes over the lumbosacral enlargement in 3 people with transtibial amputation (2 diabetic neuropathy; 1 traumatic) for 28 or 84 days. SCS was delivered to restore sensation in the missing limb during walking based on signals from a pressure-sensing insole in the shoe under the prosthesis. We used electromyography (EMG) to record posterior root-muscle (PRM) reflexes in the residual limb while participants were seated, standing, or walking. We characterized rate-dependent depression and recruitment properties of the PRM reflexes. We used pressure data from instrumented insoles to measure the step cycle duration and limb alternation symmetry with and without SCS.

Results: SCS evoked PRM reflexes in the residual limb muscles in all participants, which was confirmed by the presence of rate-dependent depression at stimulation frequencies ≥ 2 Hz. Overall, there was broad activation of residual limb muscles with SCS that varied with the position of the stimulating electrode relative to the lumbar spinal cord. PRM reflexes were also activated during walking, as confirmed by the presence of rate-dependent depression. However, SCS-evoked PRM reflexes did not disrupt gait, with similar step cycle duration or limb alternation symmetry with and without SCS.

Conclusions: Restoring sensation in the missing limb using SCS excites spinal reflexes according to the expected rostral-caudal myotomes but does not disrupt the step cycle duration or limb alternation symmetry in people with transtibial amputation. Therefore, SCS can restore sensory feedback in the missing limb without disrupting the existing motor activation of the residual limb muscles.