Retinitis pigmentosa leads to progressive photoreceptor degeneration and loss of vision. Therapeutic approaches using electrical stimulation were successful in improving photoreceptor longevity in transgenic rats and human trials. Transcorneal electrical stimulation (TES) gained attention thanks to being minimally invasive and having successful outcomes in clinical trials. This study aims to address the current challenges in TES and optimize the electrical stimulation parameters for maximum efficacy. This is accomplished using computational methods and in vivo validation on rats and enucleated porcine eyes.
We developed a realistic 3D computational model of a rat’s head to investigate the electric field distribution in the body during TES. Multiple electrode configurations were tested to maximize the current density generated in the retina. The current density, which consists of the electric field and the dielectric tissue properties, was calculated as the indicator of neuroprotection. The simulation results were then tested in vivo using the same electrode setup. We measured the voltage difference across the rats’ retina and estimated the current density using resistivity values obtained previously.
The simulations suggested that placing the return electrode behind the globe generated up to 65% greater current density in the retina compared to temporal, subcutaneous placement. Increasing the electrode contact area with the globe by using a disk design further increased the current density up to 43%. The simulations successfully predicted the stimulation amplitude vs. retina current density relationship, which the in vivo measurements corroborated. An input amplitude of 200-300 μA was shown to generate the target current density of 20 A/m^2 in rats. The same setup was also tested on enucleated porcine eyes, revealing that an amplitude of 1 mA was needed to reach the same current density. This outcome of this work encourages the use of computational models for TES optimization studies where testing on live subjects is not feasible.