Evaluation of a conducting elastomeric composite material for intramuscular electrode application

Abstract

Electrical stimulation of the muscle has been proven efficacious in preventing atrophy and/or reanimating paralyzed muscles. Intramuscular electrodes made from metals have significantly higher Young's Moduli than the muscle tissues, which has the potential to cause chronic inflammation and decrease device performance. Here, we present an intramuscular electrode made from an elastomeric conducting polymer composite consisting of PEDOT-PEG copolymer, silicone and carbon nanotubes (CNT) with fluorosilicone insulation. The electrode wire has a Young's modulus of 804 (±99) kPa, which better mimics the muscle tissue modulus than conventional stainless steel (SS) electrodes. Additionally, the non-metallic composition enables metal-artifact free CT and MR imaging. These soft wire (SW) electrodes present comparable electrical impedance to SS electrodes of similar geometric surface area, activate muscle at a lower threshold, and maintain stable electrical properties in vivo up to 4 weeks. Histologically, the SW electrodes elicited significantly less fibrotic encapsulation and less IBA-1 positive macrophage accumulation than the SS electrodes at one and three months. Further phenotyping the macrophages with the iNOS (pro-inflammatory) and ARG-1 (pro-healing) markers revealed significantly less presence of pro-inflammatory macrophage around SW implants at one month. By three months, there was a significant increase in pro-healing macrophages (ARG-1) around the SW implants but not around the SS implants. Furthermore, a larger number of AchR clusters closer to SW implants were found at both time points compared to SS implants. These results suggest that a softer implant encourages a more intimate and healthier electrode-tissue interface. STATEMENT OF SIGNIFICANCE: Intramuscular electrodes made from metals have significantly higher Young's Moduli than the muscle tissues, which has the potential to cause chronic inflammation and decrease device performance. Here, we present an intramuscular electrode made from an elastomeric conducting polymer composite consisting of PEDOT-PEG copolymer, silicone and carbon nanotubes with fluorosilicone insulation. This elastomeric composite results in an electrode wire with a Young's modulus mimicking that of the muscle tissue, which elicits significantly less foreign body response compared to stainless steel wires. The lack of metal in this composite also enables metal-artifact free MRI and CT imaging.

Keywords: Conducting polymer; Foreign body response; Intramuscular electrodes.

Fig. 1.

Depiction of the fabrication process for implantable electrodes with an electroactive area of defined length.

Fig. 2.

SEM of the intramuscular soft wires at 190× and at 5000× magnifications. The wire has uniform and smooth fluorosilicone insulation whereas the conducting band of the wire exhibit rough surface topology typical of a PEDOT-PEG/CNT composite.

Fig. 3.

Electrochemical characterization. (A) Impedance modulus of SW and SS wires with similar 1 cm exposed geometric surface area (n = 5 wires, error bars = SEM). SWs showed similar values to SS wires for impedance modulus (1 Hz to 1000 Hz) with a frequency independent region between 1 kHz and 100 kHz. (B) Phase plot of SS wires and SWs. SS wires exhibited a near 80-degree phase shift across the spectrum whereas SWs exhibited a lower magnitude of phase shift at high frequencies. (C) and (D) Voltage excursion of SS wires and SWs when pulsed with 1:1 charge balanced cathodic leading square pulse at 70 μA. Red traces are current pulses with the vertical scale bar representing 70 μA.Black traces are the resulting voltage transients. Access voltage (Va) and polarization voltage (Emc) are labelled in C.

Fig. 4.

Functional evaluation of SW in vivo. (A) and (B) Intraoperative images of an adult rat with bilateral gastrocnemius muscles implanted with a SS wire and SW. The wires were sutured to the epimysium and electrical connection was made with the gold pin connector anchored subcutaneously. (C) Muscle contraction threshold for SW was significantly lower than that required of SS wires (n = 5, error bar = SEM, *p <0.05 paired Student t-test). (D) Impedance at 1 kHz for SS and SW over 4 weeks. *p <0.05 comparing impedance between SS wire and SWs on the day of surgery. Two-way ANOVA (mixed-effect model).

Fig. 5.

Maximum intensity projection of Micro-CT of gastrocnemius muscles implanted SW (A) and SS wire (B). SWs are radiopaque and do not impose metal artifacts whereas SS wires imposed metal artifacts (widening of the implant) in the micro-CT. The small arrow in (B) points to SS wire debris during extraction of the muscle. Scale bars represent 1 cm.

Fig. 6.

Maximum intensity projections of MRI scans of SS wire and SW implanted gastrocnemius in PBS. SS wires exhibited widening metal artifacts (demarked by the dashed white arrows). SWs did not elicit widening artifacts. The grey dashed line enable identification of the location of the soft wire electrode. Scale bars represent 1 cm.

Fig. 7.

Hematoxylin and eosin (H&E) stains of the gastrocnemius muscle implanted with SS or SW. (A) and (C) Images of SS wire implanted tissues at 1 month and 3 months, respectively. (B) and (D) Images of SW implanted tissues at 1 and 3 months, respectively. The tissue implanted with the SW electrode has dramatically less inflammation and a reduced presence of non-muscle cells around the implant compared to the SS electrode. All scale bars represent 100 μm. (E) Quantification of distance from electrode to healthy muscle fibers. The SS electrodes were significantly farther away from healthy muscle fibers than SW electrodes at both time points. Error bars are SEM and n = 27–50 (from N = 5 animals for each time point). Two-way ANOVA with Tukey’s post-hoc analysis was used as the statistical test with ****p<0.0001.

Fig. 8.

Immunohistochemistry of muscle implanted with SS wires or SW at 1 month and 3 months. Tissue samples were stained with DAPI (cell nuclei), IBA-1 (macrophage), iNOS (pro-inflammatory macrophage), and ARG-1 (pro-healing macrophage). (A), (B), (C), and (D) Images of a muscle sample implanted with a SS wire implanted for 1 month. (E), (F), (G), and (H) Images of a muscle sample implanted with a SS wire for 3 months. (I), (J), (K), and (L) Images of a muscle sample implanted with SW for 1 month. (M), (N), (O), and (P) Images of a muscle sample implanted with a SW for 3 months. All scale bars represent 100 μm. Insets are a zoomed-in views of cells co-labeled with DAPI in the indicated region.

Fig. 9.

Quantification of immunohistochemistry. (A) Total IBA-1 cell counts for both wires at both time points. (B) and (C) Representative co-labeled images of iNOS + or ARG-1+ macrophages, respectively. Images shown are from a SW implant at 1 month, with arrows pointing to examples of co-labeled cells. Scale bars represent 100 μm. (D) Quantification of iNOS+ IBA-1 cells. (E) Quantification of ARG-1 positive IBA-1 cells. (F, G) Percentage of iNOS and ARG-1+ IBA-1 cells, respectively. All error bars represent SEM and n = 13–18. Two-way ANOVA with Tukey’s post-hoc analysis. *p < 0.05, **p<0.01, ***p<0.001., ****p<0.0001.

Fig. 10.

Immunohistochemistry images of acetylcholine receptor clusters. (A) SS-1 month, (B) SS-3 month, (C) SW-1 month, and (D) SW-3 month implanted muscle samples. Scale bars in (A-D) represent 100 μm. (E) and (H) are the 100× images of individual Ach receptor clusters. Scale bars represent 5 μm. (I) and (J) The distribution of the clusters as a function of distance from the electrode center is shown in for SS samples, and for SW samples, respectively.