Electrophysiological And Histological Evaluation Of Composite Regenerative Peripheral Nerve Interfaces For Closed-loop Neuroprosthetic Control
Jarred V. Bratley, Dan C. Ursu, Shelby R. Svientek, Carrie A. Kubiak, Jana D. Moon, Parag G. Patil, Theordore A. Kung, Paul S. Cederna, Stephen WP Kemp.
University of Michigan, Ann Arbor, MI, USA.
PURPOSE: Current bionic limbs are capable of multi-degree-of-freedom, anthropomorphic motor function. However, insensate hardware without intuitive somatosensory feedback is visual/auditory cue dependent and burdensome. Both neuromuscular-like control and neurocutaneous-like feedback are important prosthetic qualities. The composite regenerative peripheral nerve interface (C-RPNI), constructed by implanting a transected mixed sensorimotor peripheral nerve between autologous free muscle and de-epithelialized skin graft components, is an innovation for bidirectional signal transduction. The aim of the current study was twofold: first, to determine electrophysiological signal transduction capabilities and second, to histologically characterize C-RPNI tissue viability, regeneration, and selective axon-to-target organ reinnervation. The overall goal is to develop a multifunctional C-RPNI that amplifies volitional efferent signals and simultaneously transduces sensory input.
METHODS: Thirty rats had C-RPNIs surgically constructed by implanting the distal end of a transected common peroneal nerve between a contralaterally transferred extensor digitorum longus graft and a de-epithelialized glabrous skin graft harvested from an isogenic donor rat hindpaw. Animals were randomly assigned to one of three experimental endpoint groups (3, 6, or 9 months postoperatively) for ex-vivo electrophysiological testing. Electrodes were acutely placed. Three experimental models were evaluated: electrically stimulate 1) proximal nerve, 2) muscle, 3) skin while simultaneously recording a) muscle and skin, b) nerve and skin, c) nerve and muscle signals respectively. C-RPNI constructs were harvested and weighed at all endpoints. H&E stained cross-sections were evaluated for surgical construct health. Additional samples were immunolabeled and imaged using the three-dimensional iDISCO solvent cleared organ method to visually characterize reinnervation.
RESULTS: Three month interval evaluation of C-RPNI electrophysiological parameters recorded CMAP amplitudes and conduction velocities of 8.7▒1.6 mV and 10.0▒1.2 m/s, and evoked peak-to-peak CSNAP amplitudes and conduction velocities of 140▒35 ÁV and 9.1▒1.4 m/s (Fig. 1). Longer-term average recorded CMAP amplitudes and conduction velocities were 6.1▒1.6 mV and 12.0▒2.0 m/s at 6 months, and 10.2▒2.1 mV and 9.5▒0.6 m/s at 9 months. Evoked peak-to-peak CSNAP amplitude and conduction velocity averages were 278▒163 ÁV and 11.1▒1.3 m/s at 6 months, and 202▒6.3 ÁV and 8.8▒1.1 m/s at 9 months. All endpoint C-RPNI histology demonstrated healthy vascularized grafts maintaining 73▒9% of original construct mass, and self-selective motor and sensory axon reinnervation of muscle and dermal components respectively (Fig. 2).
CONCLUSION: The CRPNIs physiologically sort mixed sensorimotor nerve axons. Motor axons selectively reinnervate muscle and sensory axons selectively reinnervate skin target organs. Immunolabeled, three-dimensional imaging spatially mapped specific muscular and dermal components. Bimodal topography allows independent EMG recording and sensory stimulation. CRPNI components electrophysiologically demonstrated appropriate efferent CMAP and afferent CSNAP signaling. Constructs were stable over the 9 month period without neuroma formation or disabling scar tissue. The results support C-RPNI potential for closed-loop neuroprosthetic control.
Back to 2020 Abstracts