Four patients locate objects after eye gene therapy

People blinded by an inherited eye disease learned to detect and locate objects in front of them using a gene-therapy injection, light-stimulating goggles and supervised practice with both. In a trial of ten patients with advanced retinitis pigmentosa, four of the eight who completed object tests improved (IFLScience). The results, published in the New England Journal of Medicine on 7 October, extend to a small group a technique first reported in a single man (NEJM paper).

Some participants also found a doorway or followed a line on the floor while wearing the goggles, though the researchers give no count for those tasks (research team release). These are modest abilities. For someone who perceived little more than light, they are new ones.

Teaching the messengers to see

Start with how a healthy eye works. At the back of the eye, cells called rods and cones catch light, much like the sensor in a digital camera. Retinitis pigmentosa destroys them slowly, over years.

Further along the relay sit ganglion cells, the messengers that carry signals from the eye to the brain. These can outlive the light-catching cells (2021 Nature Medicine paper). They still work, but they have nothing left to pass on.

The injection gives these messengers a new job. A harmless engineered virus acts as a courier, delivering genetic instructions that make the ganglion cells produce a light-sensitive protein. With that protein in place, light can switch the cells on directly, skipping the photoreceptors that are gone.

The faulty gene that caused the blindness stays untouched. That is why researchers hope the approach could help people with many different mutations, a rationale shared across this family of eye therapies (review of gene-agnostic retinal strategies). Ten patients cannot prove it yet.

The goggles supply the light signal the treated cells respond to, and participants practised using them under supervision (NEJM paper). The trial, sponsored by the French company GenSight, enrolled people whose layer of ganglion cells was relatively preserved (trial registry).

Two different scorecards

The trial measured improvement in two separate ways, and they answer different questions.

The first was light sensitivity: how faint a light a person can notice. Seven of ten participants improved, and six crossed a threshold the researchers had set in advance for a clinically meaningful gain (IFLScience). That threshold describes noticing light, not functioning in daily life.

The second scorecard came from object tasks: telling whether something sits on a table in front of you, and pointing to where it is. Here four of the eight participants tested improved. This is the result closer to practical sight. Nobody regained normal vision, read text or recognised faces.

The finding builds on earlier work. In 2021 the same research programme reported that one man could perceive, locate, count and touch objects with his treated eye and the goggles, which he could not do without them (2021 Nature Medicine paper). The new paper adds evidence from more participants, showing the effect was not unique to one person.

Safety came first

This was an early trial, and its main purpose was to check that the treatment is safe. Most eye-related side effects were mild or moderate (research team release). One severe event occurred right after an injection and resolved within minutes.

The design limits how much the vision results can show. Every participant received the treatment, with no randomisation and no comparison group of untreated or masked patients (trial registry). Without that comparison, it is hard to separate the therapy's effect from practice, motivation or chance.

The registry lists the study's completion for October 2027 (trial registry). For now this is a research result rather than an available treatment. The trial found limited, goggle-assisted object detection in some participants with advanced retinitis pigmentosa, in a small study without a control group. Turning that into useful sight in daily life is a question GenSight and its partners can only answer with a larger, controlled trial.