Gene therapy has restored hearing in children with inherited OTOF-related deafness, but hair cells lost to noise or age cannot yet be regrown in humans, and no trial has tested either approach on tinnitus.
It is easy to be misled about this, because the honest answer contains a spectacular success that is not the success most people are looking for.
Why the question exists at all
The cochlea converts sound into nerve signals using sensory hair cells. For people born with normal hearing, losing those cells is the most common cause of hearing loss — and in humans they do not come back.
That is not a universal rule of biology. A 2025 review in Hearing Research sets out the comparison that drives the entire field: birds, fish and amphibians naturally regenerate hair cells after damage, and the regeneration restores function. Some limited regeneration also happens in the immature cochlea of mice.
The mature mammalian cochlea does not. No replacement cells, no recovery — which is why noise damage and age-related loss are permanent, and why the tinnitus that follows them is too.
So the question is not whether hair cell regeneration is biologically possible. It obviously is; other animals do it routinely. The question is whether the mechanism can be restarted in a mammal, and that is where the field currently sits — the same review's title names current limitations alongside recent advances.
What has worked, and why it deserves to be taken seriously
In 2024 and 2026 two trials restored hearing in deaf children, published in the Lancet and the New England Journal of Medicine. These were not preliminary or equivocal.
The Lancet trial (2024) used AAV1-hOTOF in six children. Five of six recovered hearing, with average auditory brainstem response thresholds improving by 40 to 57 dB across 0.5–4 kHz. One child went from above 95 dB at baseline to 68 dB at four weeks, 53 dB at thirteen, and 45 dB at twenty-six. Speech perception improved in those who recovered. No dose-limiting toxicity and no serious adverse events.
The NEJM trial (2026) used DB-OTO in twelve children. Nine of twelve — 75% — met the primary endpoint at 24 weeks (p = 1.1×10⁻¹³ for both primary and key secondary endpoints). Six could hear soft speech without any assistive device. Three reached average normal hearing sensitivity. Sixty-seven adverse events occurred, none leading anyone to leave the study.
Children who were deaf can now hear speech. That is a real medical achievement and nothing below is intended to diminish it.
Why it is not the thing you are hoping for
Both trials treated OTOF-related deafness — also called DFNB9 — and the mechanism matters enormously.
In this condition the hair cells are present and structurally intact. What is broken is a gene for a protein called otoferlin, which the cell needs in order to release its signal to the auditory nerve. The ear has working microphones with a disconnected wire.
Gene therapy delivers a functioning copy of the gene. The wire is reconnected. Nothing is regenerated, because nothing was missing.
Now compare that with noise-induced or age-related hearing loss, which is what almost everyone reading this has. There the hair cells are gone. There is no cell sitting there waiting for a working gene, and a therapy that supplies one has nothing to supply it to.
These are different problems that both present as deafness, and conflating them is the single commonest error in reporting on this subject.
How narrow the eligible group currently is
One number captures it. The Lancet trial screened 425 people to enrol six.
That is what a treatment for a specific recessive genetic condition looks like in practice. It is not a comment on the quality of the work — it is a description of how many people the work currently applies to.
What any of this would mean for tinnitus
Suppose the harder problem were solved and hair cells could be restored. Would tinnitus go with them?
Nobody knows, and no trial has asked.
The reason to think it might is the leading model: the auditory brain raises its own gain to compensate for input it is no longer receiving, and that compensatory overactivity is what gets perceived as sound. Restore the input and the compensation should no longer be needed.
The reason to be careful is that the same model is known to be incomplete. Gain changes happen in people who never develop tinnitus, so gain alone does not explain it — and whatever else is involved might not reverse when hearing does. There is also the possibility that long-established tinnitus becomes self-sustaining in a way that removing the original trigger does not undo, which is a pattern seen in other chronic conditions.
The honest position: restoring hearing is a plausible route to reducing tinnitus, it is not a demonstrated one, and it would need its own trials.
Where the research stands
The 2026 Nature Reviews Disease Primers volume on tinnitus lists inner ear regeneration among the field's active research directions, alongside biomarkers, tinnitus genetics, closed-loop neuromodulation and digital therapeutics.
That is the correct category, and it is worth reading precisely. A research direction is a place effort is being spent. It is not a treatment awaiting approval, and the distance between the two is usually measured in decades rather than years.
What this means if you are living with tinnitus now: it is reasonable to be encouraged and unreasonable to wait. The gene therapy results show the inner ear is more tractable than it looked twenty years ago. They do not shorten the time before anything applies to noise-related or age-related loss, and the treatments with evidence behind them today — covered in the treatment comparison — are still the ones worth pursuing in the meantime.
For what else is in development and how to read announcements about it, see where tinnitus research is headed and tinnitus drugs in development, which covers the drug candidates that have reported and what happened to them.