Biotech & Longevity · Thursday, 13 August 2026

01 · Briefing · what happened

A deafness gene therapy reaches a child, in a week ruled by the delivery problem

Biotech & Longevity 3 min 15 sources

Getting a working drug to its target, past the body's own barriers, is often harder than inventing it - and this week's biotech news kept proving the point.

200,000+

people with GJB2 deafness

in the US and Europe, no approved treatment

$90m

raised by Epicrispr

for a muscle-disease gene therapy given by vein

10-3

advisory vote for Replimune

approved after two earlier FDA rejections

9 months

youngest child dosed

single injection into the inner ear

At a glance

A young child got the first dose of Skylark Bio's SKY-GJB2, a gene therapy for the most common inherited deafness.
It works only by injecting straight into the cochlea, the sealed inner ear - no pill or vein can reach it.
The same week: a melanoma virus (Replimune) injected into tumors won FDA approval after two rejections.
Epicrispr raised $90m for a muscle-disease drug given by vein to reach muscle throughout the body.
The FDA rejected ITM's radioactive cancer drug - not on science, but on the fragile chain that makes and ships it.
The lesson under all of it: getting a working drug to its target is often harder than inventing the drug.

Forces in play

Body's barriers High→

the gut, liver, and walls around each organ block most drugs from reaching their target

Delivery science Building↑

money and papers pouring into carriers that survive the trip - nanoparticles, engineered viruses

Manufacturing chains Building↑

radioactive drugs decay by the hour; ITM's rejection shows how fragile the supply line is

Gene therapy reach Building↑

first therapies now reaching hard-to-reach tissue - the inner ear, muscle, tumors

In play Skylark Bio — dosed first child with an inner-ear deafness gene therapy Replimune — won FDA approval for a melanoma virus injected into tumors Epicrispr — raised $90m for a muscle-disease drug given by vein ITM — rejected by the FDA over how its radioactive drug is made

How it unfolded

Apr 2026 first-ever deafness gene therapy (Regeneron's Otarmeni) approved
Aug 6 FDA approves Replimune's tumor-injected melanoma virus
Aug 10 FDA rejects ITM's radioactive cancer drug on manufacturing
Aug 11 first child dosed with Skylark's inner-ear deafness gene therapy; Epicrispr raises $90m

Where this points

Watch Skylark's first readout, expected before year-end, for whether an inner-ear injection safely restores hearing - the test of whether delivery, not the gene, was the wall.

Full briefing

The first child

On August 11, doctors gave a young child the first dose of an experimental gene therapy aimed at the inner ear [1][2]. The therapy, SKY-GJB2 from a company called Skylark Bio, tries to fix GJB2-related deafness [1]. It is the most common inherited cause of deafness, affecting more than 200,000 people in the US and Europe, with no approved treatment [1]. A gene therapy delivers a working copy of a broken gene, packaged inside a harmless virus that ferries it into cells [1].

But the cells that need fixing sit inside the cochlea - the fluid-filled spiral of the inner ear - sealed behind a barrier that keeps most drugs out [2]. So there is only one way in. A surgeon injects the therapy directly into the cochlea, through a single needle, in children as young as nine months [1]. Swallowing a pill or injecting into a vein would do nothing; the medicine would never arrive. It follows the first-ever deafness gene therapy, Regeneron’s Otarmeni, approved in April, which targeted a different, rarer gene - and Skylark’s target is the harder engineering job [2].

The hard part is arrival, not invention

The idea that a drug’s real challenge is getting there ran through the whole week. Designing a molecule that works in a dish is one problem. Getting it into a living body, past the gut, the liver, and the barriers around each organ, at the right strength, is often the bigger one.

The FDA’s approval of Replimune’s melanoma drug, Tudriqev, on August 6 is the vivid case [3][4]. It is an oncolytic virus - a virus engineered to infect and burst cancer cells - and it cannot be given as a pill or a drip [3]. The virus would be destroyed in the bloodstream before it reached a tumor. So a doctor injects it straight into the tumor with a needle [4]. The approval came after two earlier rejections and a 10-to-3 advisory vote, one of the year’s more tumultuous regulatory paths [3][4].

Different tissues force different answers. Epicrispr, which raised $90 million on August 11, treats a muscle-wasting disease called facioscapulohumeral muscular dystrophy [5][6]. It uses “epigenetic editing” - CRISPR tools tuned to switch a gene off rather than cut DNA [6]. Muscle is spread through the whole body, so a single injection into one spot won’t do. Their therapy goes in by vein and rides the blood to muscle everywhere; early data showed three patients gaining lean muscle six months after one infusion [5][6].

When the delivery chain breaks

Delivery can fail even when the drug works. On August 10, the FDA rejected ITM’s radiopharmaceutical - a cancer drug that carries a radioactive atom to a tumor on a targeting molecule - for neuroendocrine tumors [7]. The rejection wasn’t about the science: the drug had beaten a Novartis rival in a large trial [7]. It was about manufacturing [7]. Radioactive drugs decay by the hour, so making them and getting them to patients in time is a fragile supply chain [8]. That chain is now one of the field’s main brakes [8].

Small molecules hit the wall too. Sionna’s experimental cystic fibrosis pill fell flat in a mid-stage trial the same week [9]. And the research world is pouring money into the arrival problem itself. Scientists reported a way to track where nanoparticles - tiny fatty bubbles that carry drugs - actually end up inside a living animal [10]. Others built an oral carrier that shields a cancer-fighting compound through the gut [11], and virus-inspired shells designed to survive the trip [12][13]. Even the gene therapies already reaching children, like an AAV-delivered treatment for the muscle disease Pompe, depend on a virus engineered to home in on the right cells [14].

The through-line is old but easy to forget. One gene-editing founder, taking his company public this week, said the goal is to bring CRISPR “to the masses” [15]. The thing standing between the lab and the masses is almost always delivery.

02 · Lesson · why it matters

The medicine that works everywhere except inside you

A drug can be perfect in a test tube and useless in a body, because getting it there is harder than making it.

How it works

A drug works perfectly in a dish or a test tube
But in a living body it must survive the gut and the liver
And cross the barriers walling off each organ
And not be cleared before it reaches the target
So it arrives too weak, or never - and fails a body it should have helped

The twist

A medicine that works in the lab can still fail a patient for a reason that has nothing to do with the medicine: it simply never gets there.

Where you've seen this

Insulin

must be injected, not swallowed - the gut would digest it before it worked

Brain drugs

most are blocked by the barrier that walls the brain off from the blood

Chemotherapy

much of the dose is lost or hits healthy tissue before reaching the tumor

Eye drops

most of the drop drains away in minutes; only a trace ever soaks in

The catch

Solving delivery for one tissue rarely helps the next - the trick that reaches muscle won't reach the inner ear, so each target is its own fresh problem.

Full lesson

The needle in the ear

To fix a child’s deafness this week, a surgeon did not hand over a pill. They drilled toward the inner ear and injected a gene therapy straight into the cochlea, the coiled, fluid-filled chamber where sound becomes signal.

That sounds extreme until you ask the obvious question: why not just swallow it? Because the medicine would never arrive. The gene it carries works. It has been proven in cells in a dish. But a body is not a dish. Between the mouth and the inner ear stand a dozen defenses, and the cochlea sits behind a wall built to keep almost everything out.

So the whole design collapses into one problem. Not “does the drug work?” but “can we get it to the right place, at the right strength, before the body destroys it?”

Four ways a good drug never arrives

A living body treats a swallowed drug the way a customs system treats a package. It inspects, taxes, and often confiscates.

First, the gut. Stomach acid and gut enzymes are built to break down anything protein-shaped into scraps. That is why insulin has to be injected - swallow it and you digest it, like a meal.

Second, the liver. Everything absorbed from the gut passes through the liver first, and the liver’s job is to break down foreign molecules. A drug can lose most of its dose on that single pass before it ever reaches the blood.

Third, the barriers. The brain, the eye, the testes, and the inner ear each sit behind a tight wall of cells that blocks most drugs. The brain’s wall is the famous one: it is why so many drugs that fix brain cells in a dish do nothing for the person.

Fourth, clearance. Even a drug that gets in doesn’t stay. The kidneys filter the blood constantly, and the immune system hunts anything unfamiliar. A molecule can be flushed out before it does its work.

Design a drug and you are fighting all four at once.

The same problem, four different answers

Watch how this week’s science bends around that single wall.

The melanoma virus the FDA approved cannot travel through blood - the immune system would shred it - so doctors inject it directly into the tumor. The muscle-disease drug faces the opposite shape: muscle is everywhere, so a single injection won’t reach it, and the therapy has to ride the bloodstream to muscle throughout the body. The deafness therapy needs one sealed room, so it goes in by needle through the ear.

None of these routes is a detail. Each is the answer to a different delivery problem, and the answer that works for one tissue almost never works for the next. The trick that reaches muscle will not reach the cochlea. Every target is its own fresh puzzle.

Why the vehicle matters as much as the cargo

This is why so much of modern biotech is really about the carrier, not the drug.

A raw gene, injected into blood, is destroyed in minutes. So scientists hide it inside something - a harmless virus, or a tiny fatty bubble - that shields the cargo and knocks on the right cells’ doors. This week researchers reported new ways to track where those bubbles actually end up in a living animal, and new shells built to survive the gut. That work is unglamorous. It wins fewer headlines than a cure. But it is the difference between a molecule that works in theory and one that works in a person.

Even failure follows the pattern. A radioactive cancer drug was rejected this week - not because it failed patients. It was rejected because the fragile chain that makes and ships it, before the radioactivity decays, could not be trusted. Getting the drug there, on time, at strength, was the wall it hit.

What the whole shows

It is tempting to think medicine is a search for the right molecule, and that once you find it, the hard part is over. The delivery problem says the opposite. Finding the molecule is often the beginning. Getting it, intact and concentrated, to a specific room inside a living body that is built to keep it out - that is where most drugs die.

There is a quiet humility in that. The body is not a passive container waiting to be dosed. It is an active system with its own defenses, filters, and walls, and it does not distinguish a cure from a threat. The people injecting a gene into a child’s ear are not just chemists. They are navigators, working out how to cross a territory that was never designed to let them in. And each crossing they manage teaches only that one route, not the map.

03 · Lab · your turn

Get the drug there

Pick a target tissue and a delivery route, and see how much of a working drug actually arrives - the delivery problem, lived.

04 · Hope · carry this

The wall around the inner ear kept every drug out for as long as medicine has existed. This week a child got one through it - the barriers are hard, but not permanent.