In Situ Biocompatibility of Thin-Film Polyimide Neural Electrodes In A Peripheral Nerve Interface Setting

Purpose: Intuitive control of advanced bioengineered neuroprostheses requires bidirectional information exchange across a biotic-abiotic interface. It relies in part on optimization of the biomaterials comprising the neural electrode. Currently, most implantable bioelectronics are stiff and hermetically packaged. In this study, we investigate the in situ biocompatibility of a flexible, custom, 32-channel, low-profile electrode with thermoset polyimide as the substrate. Our hypothesis is polyimide neural electrodes (1) do not alter underlying muscle viability and (2) cause minimal scarring at the muscle-electrode interface.

Methods: In nine rats, the epimysium of the left extensor digitorum longus (EDL) muscle was implanted with either electrode alone (n = 3), silk suture alone (n = 3), or electrode with silk suture (n = 3). Each group including the right EDL muscle (control) was encircled with a single layer of decellular small intestinal submucosa. The neurovascular pedicle was left intact and electrode position was marked with a microsuture. Postoperatively, the hindlimbs were stressed daily with passive flexion/extension exercises. After a convalescent period of 30 days, needle electromyography (EMG) and nerve conduction studies were performed, and muscles were harvested for histology.

Results: All electrodes remained intact and none migrated from the implantation site despite the imposed passive leg movement. Needle EMG demonstrated no sustained fibrillations in any of the EDL muscles, and no differences in rheobase, amplitude, latency, and conduction velocity were found. A significant increase in wet muscle mass was noted in all experimental groups compared to control (p < 0.0001) (Fig. 1). Scanning electron microscopy (Fig. 2) and histology indicated minimal scarring at the muscle-electrode interface.

Conclusions: Polyimide neural electrodes demonstrate superb in situ biocompatibility with no untoward effects on underlying muscle viability. Minimal scarring occurs at the muscle-electrode interface making them an attractive option for inclusion in the regenerative peripheral nerve interface.

This work was supported by the DoD MURI program under grant W911NF-06-1-0218, the DARPA RPI program under grant N6601-11-C-4190, and the PSF under grant 197630.

Figure 1. Box plot of wet muscle mass. *, significantly different compared to control (p < 0.0001).

Figure 2. Scanning electron microscopy of the electrode group demonstrating loose, trabeculated, filmy connective tissue at the muscle-electrode interface. E, electrode; M, muscle; *, muscle-electrode interface.