A Vision Chip Just Crossed Into New Territory: Science Corp's Retinal Implant Gives a Stargardt Patient Back a Word

A letter in the first session. A word in the second. For most people that is a trivial thing, but for the first person with Stargardt disease to receive Science Corporation's PRIMA retinal implant, it is the first evidence that a device built for age-related vision loss may work for an inherited one too. The company announced the milestone on October 13, 2026, making it the first clinical result reported for PRIMA outside geographic atrophy, the form of macular degeneration the device is already sold for in Europe.

Stargardt disease is a rare inherited condition that usually starts stealing central vision in childhood, and there is no treatment that restores what is lost. Science Corporation estimates that about 150,000 people worldwide have lost vision to it. The disease destroys the eye's light-sensing photoreceptor cells while, in many patients, sparing the inner retina that carries signals on to the brain. That pattern is the key to everything that follows: an implant can only work if there is still a pathway to plug into.

How the chip works

PRIMA is a chip 2 millimeters by 2 millimeters placed under the retina where the photoreceptor cells have died. The patient wears glasses fitted with a camera that captures the scene and projects it onto the implant in near-infrared light. The chip's photovoltaic pixels convert that light into electrical signals, which the surviving retinal cells pass on to the brain. The implantation procedure takes about an hour, according to TechCrunch's coverage of the program.

Speaking at TechCrunch Disrupt on October 13, chief executive Max Hodak described the resulting image as high contrast and sharp, enough for patients to "see a couple letters at a time." Hodak co-founded Neuralink and started Science Corporation in 2021 after leaving it, which makes PRIMA something of a sibling project to the brain-interface work happening across the neurotechnology sector. "The retina is part of the central nervous system," Hodak told TechCrunch, framing the eye as the most accessible window into the brain.

A letter in the first session, a word in the second: for a disease with no treatment that restores vision, even preliminary signals are worth watching closely.

The Sydney trial

The patient received the implant in May 2026 at Sydney Eye Hospital, as part of a trial called PRIMAlia led by Matthew Simunovic, head of the Save Sight Institute's Retinal Disease and Rescue Group. The trial began on April 22, 2026 and plans to enroll five adults with inherited retinal degeneration affecting the macula in both eyes, including Stargardt disease and retinitis pigmentosa.

Its main measures at 12 months are the share of participants whose visual acuity improves by at least 0.2 logMAR, about two lines on an eye chart, and the number and severity of serious adverse events tied to the device or the procedure. The registry lists an estimated primary completion date of March 2029, so this first signal is arriving very early in the story.

The caveats, stated plainly

The evidence is thin, and the company has not pretended otherwise. Everything comes from one patient's first two rehabilitation sessions, reported by the company itself in a video update, and nothing has been published in a peer-reviewed journal. Science has not disclosed the size of the letter or word, the patient's visual acuity before or after implantation, whether he could read with the implant's glasses switched off, how many PRIMAlia participants have been implanted so far, or any data from patients with retinitis pigmentosa.

There is also a wrinkle in the trial design worth noting. Participants can enter the trial with vision as good as 20/320 in the study eye, so some natural sight may remain, and the first sessions of rehabilitation are exactly when enthusiasm and learning effects are strongest. The honest read is that this is a feasibility signal, not a proven therapy. What would change that is published acuity data across multiple patients, and whether the gains persist at the 12-month mark.

What it means for a device already on sale

Unlike most neural implants in the news, PRIMA is already a commercial product. It has been on sale in Europe since July 2026 for patients blinded by geographic atrophy, holds a CE mark, and is registered for supply in the UK. In the United States it carries a Breakthrough Device designation plus Humanitarian Use Device designations for advanced geographic atrophy and for Stargardt disease, a status reserved for conditions affecting no more than 8,000 Americans a year. That status opens a route to approval, the Humanitarian Device Exemption, which waives the usual proof of effectiveness but restricts profit and use. Science applied to the FDA for approval in March 2026 and described PRIMA as pending FDA approval on October 5. It has not said which condition its pending application covers or when it expects a decision.

There is commercial muscle behind it too: Science closed a $230 million Series C financing in March 2026. And Hodak said at Disrupt that the next step is a bigger implant with a wider field of view, with a next-generation device reaching human testing within the next year. Earlier this month the company also reported that PRIMA patients in its geographic atrophy program averaged a 29-letter vision gain at three years, the long-term data that anchors the commercial case.

The broader trend is bigger than one device or one disease. Retinal implants are becoming platforms, not single-disease gadgets: the same underlying technology is being tested across macular degeneration, Stargardt disease, and retinitis pigmentosa, with improvements delivered through the external glasses and software rather than new surgery. If that model holds, the first word a Stargardt patient reads this month matters less as a clinical result than as proof that the platform can travel. The real verdict comes when the data does.