Following a feasibility study of 5 participants, results from the PRIMAvera clinical trial show that a subretinal photovoltaic implant restored central vision in patients with advanced geographic atrophy (GA) due to age-related macular degeneration (AMD). The researchers, led by Frank G. Holz, MD, of the University of Bonn, evaluated the safety and efficacy of the photovoltaic retina implant microarray (PRIMA) system, which replaces lost photoreceptors. The system is compared with pegcetacoplan and avacincaptad pegol, which can only slow GA. The trial results were recently published in the New England Journal of Medicine.
The PRIMA system includes 3 components: a subretinal photovoltaic implant, a pocket processor, and glasses that project near-infrared (880 nm) light onto the implant. The implant converts this light into electrical pulses that stimulate inner retinal neurons, thereby restoring visual signal transmission to the brain. “Unlike a wired prosthesis,” the researchers described, “the photovoltaic nature of the implant enables wireless operation combined with a straightforward implantation technique. The lenses in the PRIMA glasses are transparent, so participants have natural vision and prosthetic central vision simultaneously.” Wearers can adjust zoom settings from 1x to 12x magnification, as well as brightness.
The open-label, multicenter, prospective study enrolled 38 participants (mean age = 78.9 years; 18 men, 20 women) across 17 sites in 5 European countries. All participants had bilateral GA involving the fovea and a baseline visual acuity of at least 1.2 logMAR (20/320 Snellen equivalent) in the study eye. The PRIMA implant—a 2×2 mm, 30 µm thick silicon microarray with 378 photovoltaic pixels (each 100 µm wide)—was placed subretinally within the atrophic lesion.
The trial’s primary efficacy measure was a clinically meaningful improvement in visual acuity (at least 0.2 logMAR, equivalent to 10 letters) at 12 months after implantation. The primary safety measure was the number and severity of serious adverse events related to the procedure or device within 12 months.
At 12 months, 26 of 32 evaluable participants achieved a clinically meaningful improvement in visual acuity. After multiple imputation for missing data, 80% of all participants were estimated to meet this threshold. The mean improvement at 12 months was 0.49 logMAR (24.5 letters) with PRIMA glasses and 0.51 logMAR (25.5 letters) using the participant’s preferred testing condition (with the PRIMA glasses or without).
A post hoc analysis found that 78% of participants (25 of 32) had an improvement of at least 0.3 logMAR (at least 15 letters). The maximum improvement recorded was 1.18 logMAR (59 letters). At 6 months, the mean improvement was 0.38 logMAR with participant’s choice and 0.32 logMAR with PRIMA glasses. Without the device, visual acuity remained unchanged from baseline.
Prosthetic central vision was detected in 30 of 32 participants at 12 months. Implant-resolution testing demonstrated a mean visual acuity of 1.32 ± 0.16 logMAR (20/417 Snellen equivalent) that was consistent with the implant’s 100-µm pixel size. Eighty-four percent of participants reported reading letters, numbers, or words at home with the device, while 69% reported medium to high satisfaction. Further, with the help of the zoom function on the PRIMA glasses, “participants were able to read fonts smaller (Snellen equivalent, up to 20/42) than the theoretical resolution achievable with the 100-μm pixels (Snellen equivalent, approximately 20/400),” the researchers noted.
Compared with extraocular magnifiers or implantable telescopes, which enlarge images to enable use of the retina beyond the edges of atrophy, they continued, the PRIMA system enables vision in the area of scotoma and only magnifies prosthetic central vision. “The mean native resolution (ie, the resolution independent of external factors) among the participants as assessed with the implant-resolution test was similar to the theoretical resolution achievable with the PRIMA implant (approximately 1.3 logMAR), and the ability to zoom, enhance contrast, and make other modifications to the image allowed the participants to improve visual acuity beyond this resolution.”
PRIMA also compares to previous attempts at vision restoration, including the epiretinal Argus II implant, the subretinal Alpha IMS implant, and a 44-channel suprachoroidal implant, which were limited because of their use of remote retinal cells and maximum resolution that could be achieved. PRIMA stimulates bipolar cells, rather than ganglion cells, and uses eye movement to be reoriented, which improves resolution beyond the theoretical 100-µm pixels in the implant.
Nineteen participants experienced 26 serious adverse events related to the procedure or device, most of which were within 2 months postoperatively. Twenty-one events (81% of events) occurred in the first 8 weeks, and 20 (95% of events) resolved within 2 months of onset. The most common serious adverse event was ocular hypertension (6 cases, all resolved). Other events included peripheral retinal breaks (5 cases), full-thickness macular holes (3 cases), subretinal hemorrhage (4 events in 3 participants), and choroidal neovascularization (CNV; 2 participants who were treated with anti-VEGF therapy). CNV incidence has been observed in 16% among dry eyes with AMD at 2.6 years after vitrectomy treatment, the authors wrote, compared with the 5% incidence in the present study. They added that advanced intraoperative imaging techniques may benefit the implantation of subretinal prostheses. No serious events were attributed to the device alone.
They also noted that “the mean area of atrophy among the study eyes increased by 8.5 mm2 from baseline to 12 months, as compared with an increase of 2.5 mm2 in eyes without the implant; in the study eyes, the increase was attributed to atrophy associated with subretinal surgery, including retinotomy and bleb formation.”
Adverse events were also mild or moderate in 85% of cases. Structural imaging showed stable inner retinal layers over the implant site, and peripheral visual acuity remained similar to baseline. The researchers did not assess changes from baseline in visual impairment as assessed with the IVI questionnaire, and noted that the questionnaire may not be sensitive enough to detect such changes in a cohort like this one with extremely low vision. The data and safety monitoring board concluded that the benefits outweighed the procedural risks.
The PRIMAvera study was funded by Science Corporation (formerly Pixium Vision) and the Moorfields National Institute for Health and Care Research Biomedical Research Centre (NIHR203322). Ongoing follow-up will assess outcomes through 36 months. RP