First-in-human trial of a novel suprachoroidal retinal prosthesis

Abstract

Retinal visual prostheses ("bionic eyes") have the potential to restore vision to blind or profoundly vision-impaired patients. The medical bionic technology used to design, manufacture and implant such prostheses is still in its relative infancy, with various technologies and surgical approaches being evaluated. We hypothesised that a suprachoroidal implant location (between the sclera and choroid of the eye) would provide significant surgical and safety benefits for patients, allowing them to maintain preoperative residual vision as well as gaining prosthetic vision input from the device. This report details the first-in-human Phase 1 trial to investigate the use of retinal implants in the suprachoroidal space in three human subjects with end-stage retinitis pigmentosa. The success of the suprachoroidal surgical approach and its associated safety benefits, coupled with twelve-month post-operative efficacy data, holds promise for the field of vision restoration.

Trial registration: Clinicaltrials.gov NCT01603576.

Figure 1. Potential anatomical locations for retinal prosthesis implantation.

To date, clinical trials have been performed with devices in the A: epiretinal position , B: subretinal space and D: intrascleral space . Image modified with permission from Bionic Vision Australia.

Figure 2. The intraocular electrode array of the suprachoroidal device (A) and the entire device (B), showing the array connected to the percutaneous connector via a helical lead wire.

The electrodes on the intraocular array (C) were numbered for analysis, with the black electrodes (21a to 21m) being ganged to provide an external ring for common ground and hexagonal stimulation parameter testing. Note electrodes 9, 17 and 19 were smaller (400 µm vs. 600 µm). The percutaneous connector protruded through the skin behind the ear (D), allowing direct connection to the neurostimulator via a connecting lead (E). The scleral incision was made 9 mm to 10 mm posteriorly from the sclero-corneal limbus.

Figure 3. OCT scan of the electrode array in situ, taken 2 months postoperatively in Patient 1.

The horizontal arrow on the infrared image indicates the direction of the OCT scan (A). The cross-sectional OCT image (B) shows the silicone and platinum electrode components of the array, the retina structure and choroidal vasculature, and the electrode to retina distance used for analysis (double-headed arrow). Scale bars  =  200 µm.

Figure 4. Time course of subretinal hemorrhage in P1, as documented with retinal fundus photography.

Complete resolution of the hemorrhage occurred in this subject by 55 days post-operatively. Note the electrode array with individual electrodes can be seen more clearly over time as the blood clears in the temporal retina (arrow).

Figure 5. Retinal thickness measurements over time, showing no observable change in the maximum retinal thickness above the electrode array in the initial twelve months in all three patients.

Each boxplot includes the maximum (upper whisker, excluding outliers), upper quartile (top of box), median (horizontal line in box), lower quartile (bottom of box) and minimum values (lower whisker). Open circles are outliers. The numbers represent the electrode location (see Fig. 2), and the horizontal lines on the graph of P3 represent single data points.

Figure 6. Time series of IR images of the electrode array in all subjects, showing no significant lateral movement of the array over a twelve-month period.

The leading edge of the array is marked with a white arrow in the 3-day images. The position of the fovea is marked with a white cross in the 6-month images.

Figure 7. Distance between the electrodes and retina over time.

This distance was relatively constant in P1, but increased up to two-fold in P2 and P3 over time. Note significant nystagmus in P3 made the measurements difficult, leading to greater variation in the values recorded. Outliers are identified by open circles, and the numbers represent the electrode location (see Fig. 2).

Figure 8. Stability of the intraocular array (arrow) and helical lead wire (star) over the initial 12 months of implantation, as documented by X-ray images.

Additional scans were taken to monitor the percutaneous connector (not shown), which also stayed stable over this time period.

Figure 9. Impedances for the 600 µm platinum electrodes over time in the three subjects.

Impedances were measured with charge-balanced biphasic current pulses (pulse phase width: 25 µs; amplitude: 75 µA). The dotted lines represent the date of first stimulation. In P1 & P2, the impedances were stable over the implantation and stimulation period. Impedances measured in P3 decreased over the course of the implantation period. Outliers are identified by open circles, and the numbers represent the electrode location (see Fig. 2).

Figure 10. Different monopolar electrode stimulations used in the psychophysical testing of the implanted subjects (A & B), where the active electrodes are shown in red and the return electrodes in black.

Table (C) shows the number of electrodes that were capable of eliciting a visual percept using the given stimulation parameters in each subject. PW =  phase width, IPG =  interphase gap, pps =  pulses per second. Stimulus duration was 2 seconds in all cases.

Figure 11. Optotype recognition results for P1 using the Landolt-C test, which gives an indication of acuity (but should not be directly interpreted as standard visual acuity).

Whiskers extend to minimum and maximum recorded thresholds and the box extents show inter-quartile range. The Lanczos2 vision processing performed significantly better than System Off (P = .01). Note that the visual acuity exceeded the 3.24 logMAR software limit in all trials with device off (floor effect).