Responses evoked by monopolar and bipolar photovoltaic arrays in the S334ter-3 rat model of retinitis pigmentosa

Abstract

Purpose: Evaluate and compare monopolar and bipolar photovoltaic array (PVA) designs and three pixel sizes as prosthetic replacements for degenerate photoreceptors. Specifically, we compared thresholds and spatial aspects of evoked responses between monopolar (mPVA) and bipolar (bPVA) PVAs implanted in the transgenic S334ter-3 rat model of retinitis pigmentosa (RP).

Methods: PVAs were implanted monocularly and subretinal placement confirmed using OCT. Rats were implanted at two different ages (35 & 75 days old) and assessed at two time ranges post-surgery (50-90 & 120-150 days). A scanning IR laser (905 nm) stimulated the PVA while multicellular responses were recorded in the superior colliculus (SC). PVA-evoked responses were characterized by varying the stimulus duration (0.25-20ms), intensity (1.3-21 mW/mm2) and spot diameter (1mm or 0.5mm). BPVAs had three different pixel sizes: 70, 140 and 280 µm. Localization of activation was assessed by stimulating different regions of the PVA while recording at one SC location with robust PVA-evoked responses. We determined if PVAs stimulated the inner retina or ganglion cells (RGCs) directly using an intravitreal injection of synaptic blockers.

Results: Evoked responses were present in 19/20 (95%) rats with bPVAs and in 24/32 (75%) with mPVAs. Thresholds for PVA-evoked responses are at least two orders of magnitude below ocular safety limits. Monopolar PVA-evoked thresholds were similar regardless of implantation age or duration. MPVAs and bPVAs with medium or large pixels had similar thresholds, while bPVAs with small pixels had significantly higher thresholds, except with 0.25 and 0.5ms pulses. Responses recorded at a single SC site while varying the stimulation region of a bPVA showed 95% (22/23) of sites with at least one quadrant having a significantly different response. In contrast, only 17% (3/17) of mPVA sites showed any difference. In the presence of synaptic blockers, spontaneous activity and PVA-evoked responses were eliminated.

Conclusions: The current required to induce PVA-evoked responses is below established safety limits and only pixel size influences threshold. Spatial localization of activation is significantly better in bPVAs than in mPVAs. The current generated by the PVAs stimulates the existing circuitry in the INL and not RGCs directly. The bPVA design represents a significant improvement over the older mPVA designs.

Keywords: 696 retinal degenerations: hereditary • 508 electrophysiology: non-clinical • 727 superior colliculus/optic tectum