In October 2025, Holz et al published an impressive paper in the New England Journal of Medicine evaluating the impact of a subretinal photovoltaic implant to restore vision in geographic atrophy (GA).1 To better understand both the photovoltaic retina implant microarray (PRIMA) system and the PRIMAvera study, I spoke with Frank Brodie, MD, MBA, the medical director for vision and PRIMA of Science Corporation, which is developing the PRIMA system and funded the European study.2 I also asked Nitish Mehta, MD, an assistant clinical professor of ophthalmology at NYU Langone in New York City, to provide some insights; Dr. Mehta did not participate in this study and thus would be ideal to provide his independent assessment of the study and system. Our conversation has been edited for length and clarity.
Figure 1. The PRIMA photovoltaic implant, with a pencil for scale. Image courtesy Science Corporation.
Dr. Brodie, can you tell us briefly about the PRIMA system? What makes it different than the Argus II implant beyond just the location (subretinal) and the wireless setup?
The PRIMA implant is a 2x2 mm silicone microarray made up of 378 individual 100 µm photovoltaic cells that are sensitive to near-infrared light. It is placed subretinally in patients with advanced GA who have intact inner retinal layers. The patient wears augmented reality glasses, which have a front-facing camera and an infrared projector facing the pupil. The glasses capture, process, and project the image onto the implant, which in turn stimulates the overlying bipolar cells (similar to our native photoreceptors) to restore the visual cascade.
There are several key differences in both the design and performance of PRIMA compared to Argus II. The Argus II implant was placed epiretinally in patients with Retinitis Pigmentosa. It measured approximately 4x5 mm made up of 60 individual 200 µm electrodes. It required an external power loop and processing source be sutured to the sclera and then connected with a cable.. The implant was paired with a set of glasses with a front-facing camera used to wirelessly communicate with the episceral portion of the device. The implant stimulated the ganglion cells directly. Because that approach bypasses the important image-processing functions of the inner retina, including some loss of retinal topography due to simultaneous stimulation of multiple axon segments, visual outcomes were representative phosphenes without the ability to distinguish letters or numbers of any size. By contrast, because PRIMA stimulates the bipolar cells directly in normal topographic representation, patients were able to see formed vision and 80% of patients were able to read letters and numbers.
The Argus II and other prosthetic vision devices required a wired (transcleral) connection to power the device, where the PRIMA implant uses the infrared light itself to power the implant, thus obviating the need for any wires.
The size of the PRIMA implant allows for minimally invasive implantation, and the higher electrode count allows for higher display resolutions. Moreover, each PRIMA implant is independent, which allows for the potential to implant multiple chips to cover a larger area of the atrophy or upgrade implanted chips more easily.
Dr. Mehta, I know you have read this paper in detail. Can you summarize the topline results? What was most remarkable to you? Are there things that you are concerned about?
The field of retina has been blessed by continued innovation. However, visual restoration in those with severe vision loss has not yet been properly realized, especially in patients with atrophic retinal conditions such as GA secondary to age-related macular degeneration. I was very intrigued by the results of the recent NEJM article reporting 12 months of outcomes of a cohort of patients implanted with the PRIMA device.
The team has designed an innovative product. The subretinal location of the implant, the higher density of photovoltaic pixels, thin silicone array, wireless powering and transmission of data, and use of transparent, intelligent glasses all seem to provide meaningful iteration from prior solutions.
The PRIMA device demonstrated seemingly remarkable restoration of central vision in patients with GA due to AMD, with 81% of participants achieving clinically meaningful improvement (≥0.2 logMAR or ≥10 letters) at 12 months in the reported open-label, multicenter, prospective, single-group clinical study. Thirty-eight patients received the device, and 32 patients completed the study at 12 months. The mean improvement was 0.51 logMAR (25.5 letters), with the maximum improvement reaching 1.18 logMAR (59 letters). However, it is important to note that this mean final visual acuity was around 20/400 (which matched the theoretical pixel resolution), and reports of reading fonts as small as 20/42 was achieved with the use of zoom and contrast enhancement features. Impressively, 84% of the subjects reported using the device at home for reading.
As with any new device—and especially one that is surgical—careful review of adverse events is paramount before implementation. Surgical complications occurred in 19 of 38 participants, with 26 total serious adverse events. Four severe adverse events occurred, which included macular hole, ocular hypertension, retinal detachment, and proliferative vitreoretinopathy. Most of these events were specifically related to the implantation procedure, given their time of occurrence being within 2 months after surgery. It will be curious to see if the surgical implant technique will be modified over time. As this device relies on an intact inner retina, it was reassuring to see no overlying inner retinal structure change on optical coherence tomography (OCT) over time. Atrophy did increase in area, however, over the course of the study in the case eyes compared to the control fellow eyes.
In summary, my interpretation of the data is that in the correct patient, the benefits of visual restation seems to outweigh the risks at this time, but more data to corroborate these initial results would be beneficial.
Dr. Brodie, can you share with us some tricks on the surgery (eg, how to center the implant on the fovea with GA present)?
The surgery took investigators about two and a half hours on average and used many steps familiar to vitreoretinal surgeons. After the vitrectomy, a subretinal bleb is created adjacent to the macula—the location is preplanned to avoid the arcades and the preferred retinal locus of fixation. Using a combination of diathermy and retinal scissors, a 3 mm retinotomy is created and then the subretinal space in the area of atrophy is bluntly dissected is as necessary. Investigators were able to use a spatula or pick for this.
The implant was introduced with a custom injector through a 3 mm sclerotomy at the pars plana. Because the implant is so light and thin (30 µm), fluid turbulence from the open sclerotomy can cause it to float around the vitreous cavity, so good control of the fluidics and pre-placed scleral sutures can be a big help. After the implant was placed under the retina, perfluorocarbon was immediately placed to stabilize the implant. Then, the implant could be nudged transretinally into its final position . Investigators used either gas or silicone oil, and we found the patient needed to lay supine for at least an hour after surgery to prevent any slippage. Overall, investigators found the surgery straightforward and all 38 implantations in the trial (100%) were successful.
Dr. Brodie, do you have a sense of what the patients actually see? How much work is required for visual rehabilitation for the patient?
We did a variety of assessments to get a sense of what the actual visual experience is with PRIMA. As opposed to earlier prosthetic devices, PRIMA patients are actually seeing objects and letters as they appear to the native retina. When asked to draw letters and shapes seen, patients will draw the appropriate letter as shown. The external processor allows patients to control things like brightness and zoom, which is very helpful and allows the reading gains we see in the study. Many patients report a slight yellowish color to the letters. We’ve found patients can comfortably perceive both their natural peripheral vision and central prosthetic vision simultaneously.
Like other devices in low-vision patients, there is a significant amount of rehabilitation required, both in terms of device use (adjusting gaze so the projected infrared light targets the implant) and maximizing function with it both at home and when patients are out and about. The company is developing next-generation glasses for patients which will include pupil tracking to allow more natural gaze patterns and likely less rehabilitation needed as a result.
Dr. Mehta, who would be the ideal patient for this given the visual rehabilitation required?
Based on my interpretation of the published research, a candidate for this device must possess profound central vision loss secondary to GA with foveal involvement with preserved inner retinal layers.
I would be looking for someone without significant concurrent ocular pathology that would confound the potential visual improvement from the device. The candidate would have to be motivated to undergo a novel procedure with some known and some likely unknown adverse events. They would have to have the capacity to undergo training and commit to continued monitoring of the device, potentially for years.
Dr. Brodie, what are the next steps for Science Corporation after this study and what additional data does the company plan to acquire?
The company plans to make PRIMA available to patients initially in Europe, where the PRIMAvera study was conducted. CE Mark application is under review and expected in 2026. We plan to bring it to the United States as soon as possible—discussions with the US Food and Drug Administration are ongoing. The team is also exploring additional indications for PRIMA, including for use in Stargardt disease as well as other inherited retinal diseases like retinitis pigmentosa.
Science Corporation is investing heavily in vision restoration and next-generation implants are being developed with the goals of increasing resolution and restoring field of view. Next-generation glasses are being developed with internal processing and eye tracking functions to reduce complexity and improve real-world usability.
Upcoming key studies will focus on the functional performance of these enhancements as well as improvement in quality of life with the PRIMA system. RP
References
1. Holz FG, Le Mer Y, Muqit MMK, et al. Subretinal photovoltaic implant to restore vision in geographic atrophy due to AMD. N Engl J Med. 2026;394(3):232-242. doi:10.1056/NEJMoa2501396
2. Restoration of central vision with the PRIMA system in patients with atrophic AMD (PRIMAvera). ClinicalTrials.gov identifier: NCT04676854. Updated November 29, 2024. Accessed March 27, 2026. https://clinicaltrials.gov/study/NCT04676854