Gene therapy for deafness now targets its most common genetic cause

This summer, at a hospital in the northeastern United States, a surgeon placed a gene therapy at a child’s round window, the small membrane that leads into the cochlea, and let it diffuse inside. The treatment carried a working copy of one gene, GJB2. The child was the first person ever to receive it.

The trial is called SONIX, and it is run by Skylark Bio, a 19-person company in Cambridge, Massachusetts. It is small, early and designed mainly to find out whether the treatment is safe. It also matters more than its size suggests. GJB2 is the gene most often behind inherited deafness, and nobody has tried to replace it in a person before.

The first approval, and its limits

In April the US Food and Drug Administration gave accelerated approval to Otarmeni, Regeneron’s gene therapy for deafness caused by faulty copies of a gene called OTOF. It was the first approved gene therapy for any genetic form of hearing loss. In the trial behind it, 16 of 20 participants reached hearing of 70 dB or better, the level the trial set as its main goal. Regeneron offers it free to US patients, and must keep collecting data as a condition of the approval. We wrote in October about two siblings in Cincinnati who were in that trial.

The limit is how few people it can help. Cincinnati Children’s, which treated patients in the trial, puts OTOF variants at about 1% of congenital deafness. Genetic causes as a whole account for about half of congenital hearing loss, according to the American Society of Gene & Cell Therapy. Otarmeni is a breakthrough for a small slice of that half.

OTOF was also a tidy first target. Otoferlin is a protein the inner ear’s hair cells use to pass signals to the hearing nerve. Otarmeni delivers the missing gene to one cell type, the inner hair cells, and lets them do the rest.

Why GJB2 is the big one

GJB2 makes connexin 26, a protein that forms tiny channels between neighboring cells. In the cochlea those channels link the supporting cells around the hair cells and the cells of the cochlea’s outer wall into one network. Part of that network’s job, as Skylark’s team describes it, is recycling potassium through the fluid that bathes the hair cells. Without connexin 26 the network breaks and the cochlea cannot work properly.

The resulting hearing loss is usually present at birth, severe to profound, and stable. Today it is managed with hearing aids and cochlear implants.

How common is it? The companies call GJB2 “the most common cause of genetic deafness”. The most careful number we found is from the National Library of Medicine’s MedlinePlus: GJB2 variants account for about half of all cases of autosomal recessive nonsyndromic hearing loss, the inherited kind that comes without other symptoms, with the share varying by population. Next to OTOF’s 1%, that is a different scale of problem. Sensorion, a French company that had its own OTOF therapy, called the GJB2 need “an order of magnitude larger” when it dropped the OTOF program in June to concentrate on GJB2.

It is also a different kind of problem. A GJB2 therapy has to reach the supporting cells that naturally make connexin 26, and rebuild a network rather than repair a single link. Skylark’s entry rules hint at the stakes: the treated ear must still have intact cochlear cells, and it must never have had an implant.

What the trial is doing

The design, from Skylark’s release and an interview its team gave Hearing Health & Technology Matters on October 28:

• Who: 10 children aged 9 months to 7 years with two faulty copies of GJB2. The older group, 2 to 7, goes first, then 9 months to 2 years.
• Why 9 months: the youngest age is tied to when cochlear implantation is usually considered.
• How: one injection into one ear only. The other ear can have an implant or anything else the family chooses.
• What is measured: safety and tolerability first. Hearing is checked with behavioral tests.
• Day to day: children keep wearing their hearing aids. If hearing improves, the team said, parents may notice a child pulling the aid off, and a hearing test would confirm it before anything changes.
• When: dosing continues through 2026. Preliminary data are promised by the end of the year, with more in 2027.

After that, Skylark says, the company and the FDA will decide whether to expand the trial or treat both ears. “Although current interventions provide meaningful benefits, they do not address the underlying genetic cause,” said Daniel Lee, a surgeon at Massachusetts Eye and Ear who advises the company.

Skylark is not alone. Sensorion has filed to start its own GJB2 trial, called Hearconnex, in France and Canada, and aims to file in the US and Australia by year end.

The honest read

Three cautions belong next to the excitement.

First, there are no results. Not one hearing test from SONIX has been published, and the company has not said how many children have been treated so far. Everything above is design and intention.

Second, a first-in-human trial is mainly a safety test. Ten children, one ear each, no comparison group: even a clear signal would be a reason for a bigger trial, not a verdict. Skylark’s own team describes the path from here to a pivotal trial as “a multi-year process”.

Third, OTOF’s success does not transfer automatically. Otarmeni puts back one protein in one cell type. GJB2 asks a gene therapy to restore a network spread across several. It may work. Nothing yet shows that it does.

None of this changes care today. Hearing aids and cochlear implants remain the treatments for children with GJB2 hearing loss. Gene therapy is one of several attempts to go further, alongside cell therapy for the hearing nerve and a light-driven cochlear implant still in animal testing.

What to watch

Skylark’s preliminary SONIX data, promised by the end of December. The questions to ask of it are simple: was the surgery safe in every child, and did any treated ear hear measurably better than before? Then Sensorion’s first dosing in France. If both say the same thing, GJB2 becomes the second deafness gene with a working treatment, and the first that reaches a large share of families.