The Eye Was Supposed to Be Gene Therapy's Easy Win

A Pattern, Not a One-Off

This isn't J&J's first stumble in eye gene therapy. It's their second.

Earlier, the company's botaretigene sparoparvovec (bota-vec), a gene therapy for X-linked retinitis pigmentosa (a rare inherited blindness), failed its primary endpoint in the Phase 3 LUMEOS trial. The test? Whether patients could navigate a virtual maze better after treatment. They couldn't, at least not in a statistically meaningful way.

There were some positive secondary signals: 22 of 55 treated patients showed improvement on two or more backup measures, versus zero in the control group. That's not nothing. But "not nothing" doesn't get you an FDA approval. J&J handed the program back to its original developer, MeiraGTx, for just $25 million upfront, a fraction of the deal's original potential value of up to $415 million. MeiraGTx is trying to push forward anyway, raising about $100 million to take a shot at regulatory filings. Their stock dropped 16% on the news.

So in the span of about a year, J&J has exited both of its late-stage eye gene therapy programs. Two swings, two misses.

The Graveyard Gets Crowded

J&J is in good company, which is the problem. The list of failed ocular gene therapy programs is getting uncomfortably long.

Biogen's Phase 3 STAR trial in choroideremia (another inherited retinal disease) missed its primary endpoint of a three-line improvement in visual acuity. Novartis pulled the plug on GT005, a gene therapy for geographic atrophy, in 2023 after a data monitoring committee determined the program was futile.

In wet AMD, several programs testing gene therapies that produce anti-VEGF proteins inside the eye have been abandoned. One trial saw only 4 of 19 patients maintain meaningful improvement at a year. Immune responses, unreliable protein production, and outcomes worse than standard injections kept sinking these efforts.

Meanwhile, Luxturna remains the first broadly approved retinal gene therapy, almost a decade after its landmark results. It's starting to look less like a trailblazer and more like lightning in a bottle.

Why the "Easy" Organ Keeps Humbling Everyone

The eye's supposed advantages are real but overrated. Yes, it's small and accessible. Yes, the subretinal space (the gap between photoreceptors and the supportive cell layer beneath them) offers some immune protection. But that doesn't mean the biology cooperates.

The core challenges boil down to a few stubborn problems.

Getting the therapy to the right cells is harder than it sounds. Intravitreal injection is simple, basically an office visit. But the inner limiting membrane, a thin barrier inside the eye, blocks most viral vectors from reaching the photoreceptors and pigment cells in the outer retina where damage actually occurs. It's like trying to water a garden by spraying the roof; some gets through, but not nearly enough.

Subretinal injection bypasses that barrier by placing the therapy directly where it's needed. But it requires full-blown surgery (vitrectomy, retinal detachment, the works), and the treatment only reaches cells right next to the injection site. For diseases that affect the entire retina, you're bringing a garden hose to a forest fire.

The endpoints are brutal. Luxturna worked because RPE65 deficiency is a clean, well-understood genetic problem with a dramatic visual deficit. Treating it produced obvious, measurable improvement. Geographic atrophy and many other retinal diseases are slower, more complex, and harder to measure. Trials keep showing hints of biological activity without clearing the bar of clinical significance. Several Phase 3 programs have failed primary endpoints while showing "directionally supportive" secondary signals, a phrase that roughly translates to "something happened, but we can't prove it matters."

The competition is already here. In geographic atrophy specifically, complement inhibitors (traditional injectable drugs) are already approved and offer a known, if modest, benefit. For a one-time gene therapy to justify its cost and surgical risk, it needs to be clearly, unambiguously better. PARASOL's data apparently weren't even in the conversation.

What J&J's Retreat Really Means

J&J isn't abandoning ophthalmology entirely. The company says it's "applying learnings from this program" to its early-stage pipeline. Translation: we're going back to the drawing board, but we're not ready to give up the dream.

The bigger signal is strategic. J&J also cut two CAR-T programs from its pipeline in 2026. The company is clearly in pruning mode, trimming high-risk, high-cost modalities that aren't delivering clean data. Resources are flowing toward areas with better odds: oncology, immunology, programs where the path to approval isn't a coin flip.

For the broader gene therapy field, J&J's exit from ocular programs is a reality check. The eye was the proving ground. If gene therapy can't consistently win here, with all the delivery advantages the organ offers, the road in other tissues gets even steeper.

Luxturna showed us what's possible when the biology lines up perfectly: the right gene, the right disease, the right delivery. The lesson of every trial since? That alignment is extraordinarily rare. And no amount of corporate firepower can force it.

Sarepta's Worst Week: Two Deaths, a Black Box, and 500 Pink Slips

Sarepta's Duchenne gene therapy Elevidys just got the FDA's most severe safety warning after fatal liver failures, while the company slashes 36% of its workforce and pivots away from gene therapy entirely. It's a brutal week that raises big questions about the future of the entire field.