Science Corp has reported that the first patient with Stargardt disease to receive its PRIMA retinal implant has recovered a fragment of functional reading: the man, implanted in May at Sydney Eye Hospital in Australia, picked out a single letter during his first rehabilitation session and read a full word during his second. The company disclosed the milestone on October 13, presenting it as the first reading result for its subretinal chip in an inherited form of macular degeneration rather than in the age-related condition the device was originally designed around.
- One patient, not a trial result: the update covers a single participant in the Australian PRIMAlia study, and no standardised visual acuity figures for him have been released. Reading one word under supervision is not the same as recovering everyday reading.
- A platform validated elsewhere first: PRIMA has already generated peer-reviewed data in geographic atrophy, the advanced dry form of age-related macular degeneration, with mean letter gains that kept growing between the 12-month and 36-month marks.
- Regulatory distance remains: the device is not approved in the United States, where the FDA has granted humanitarian use device designation for geographic atrophy and Stargardt, while its European CE mark covers geographic atrophy only.
How PRIMA turns light into vision
PRIMA is a subretinal implant: a chip measuring 2 × 2 millimetres and 30 microns thick that is placed beneath the retina, occupying the space where photoreceptors have died. Its 378 photovoltaic pixels, each 100 microns across, convert incoming light into electrical signals that the surviving retinal cells pick up and pass along the optic pathway.
The rest of the system lives outside the body. The patient wears camera-equipped glasses that capture the scene in front of him and project it onto the chip as 880-nanometre near-infrared light, with magnification adjustable between 1x and 12x. Because the implant runs on light rather than wires or a battery, the hardware left inside the eye stays minimal — the zoom, contrast and brightness adjustments that shape what the wearer perceives sit in the glasses and their software. Science Corp acquired the technology from the French company Pixium Vision.
Inside the PRIMAlia trial — and the regulatory gap
The Stargardt case belongs to PRIMAlia (registered as NCT07266584), an Australian study running the device under investigational status and supervised by Dr Matthew Simunovic. The company has not said how many patients will be enrolled.
Stargardt disease is an inherited macular degeneration that destroys central vision, the part of sight used for reading and recognising faces. Science Corp estimates its prevalence at 1 in 10,000 people. In the United States, PRIMA is not approved: the FDA has granted it humanitarian use device designation for geographic atrophy and for Stargardt — a preliminary step, not an authorisation. Within the European Union the chip holds a CE mark, but for geographic atrophy alone.
Why the Stargardt result draws its weight from older data
The real substance behind the announcement comes from the PRIMAvera trial in geographic atrophy, published in the New England Journal of Medicine in October 2025. 38 patients were implanted, and 26 of the 32 evaluable participants improved in a clinically significant way. The study spanned 17 centres across five European countries and recorded a mean gain of 24.5 letters on an acuity chart at 12 months.
At 36 months, figures presented on 5 October at the EURETINA congress in Vienna — not yet published in a journal — put the mean gain at 29 letters, with 14 of 18 evaluated patients gaining at least 15 letters and no new safety signals reported.
“In clinical trials, PRIMA works better at month 36 than at month 6.” — Max Hodak, chief executive of Science Corp.
It is the same territory other neuroprostheses are exploring, as brain-computer interfaces edge toward restoring movement, touch and sight. What sets PRIMA apart is not the ambition but the evidence trail: unlike gene therapies, where each mutation tends to demand its own programme, a device that works with shared retinal circuitry could serve several conditions at once.
What still has to be proven
Two things should be kept separate. The technology carries a trial published in one of medicine’s leading journals plus three-year follow-up — unusually deep ground for this field. The Stargardt case, by contrast, is a promising anecdote released by the company itself. Rehabilitation also plays a decisive role: research on low-vision devices in macular disease, including the LOVIT trials, has shown that structured training alongside a device improves reading outcomes far more than hardware alone, which is why the glasses, the software and the training programme matter almost as much as the chip. What clinicians, payers and investors will now watch is how quickly patients learn to read with PRIMA, how durable those gains turn out to be, and whether rehabilitation can be delivered at scale — the practical questions that will decide if a single word becomes everyday reading.