A one-time gene edit cut a rare disease's attacks by 87% — the first phase 3 proof that CRISPR works inside the body

Biotech & Longevity 5 min 80 sources

Intellia's in-vivo CRISPR treatment for hereditary angioedema posted the first late-stage results for editing genes inside a living person. Plus the FDA's softer line on rare-disease gene therapy, a new antifungal as resistance climbs, and an $11bn deal.

Key takeaways

A single CRISPR dose cut a rare disease's attacks by 87% in the first late-stage trial of gene editing done inside the body — a one-time fix where patients used to take drugs for life.
The FDA keeps lowering the evidence bar for rare-disease gene therapies, trading certainty for speed for patients who have no other options.
A new antifungal is the first with a fresh mechanism in over 20 years, arriving as resistance — bred partly by farm chemicals — outruns our existing drugs.

This week the big result came from a rare disease most people have never heard of — and a technique that, until now, had only ever shown its hand in early trials. A company called Intellia reported the first phase 3 results for editing genes inside the body, not in cells removed and re-infused. A phase 3 trial is the large, final test before a company asks regulators to approve a drug; it compares the treatment against a placebo or the current standard in dozens or hundreds of patients. [14]

The lead: a single dose, an 87% drop in attacks

The disease is hereditary angioedema, or HAE — a rare genetic condition that causes sudden, unpredictable swelling of the face, limbs, gut and airway. An airway attack can be fatal. Patients usually manage it with drugs taken for life, on a schedule, to keep attacks at bay. [14]

Intellia’s treatment, lonvo-z, is given once. It uses CRISPR — a tool that finds a specific gene and cuts it — to switch off a gene called KLKB1 inside liver cells. That gene makes kallikrein, an enzyme that triggers the chemical cascade behind the swelling. Turn the gene off, and the body makes far less of the trigger. [14]

In the Haelo study, which randomly assigned 80 patients to the treatment or a placebo, the edited group had 87% fewer monthly attacks than the placebo group. Sixty-two percent of treated patients were attack-free and off all other HAE drugs through the six-month measurement window, against 11% on placebo. Intellia’s chief executive called it the first phase 3 evidence that in-the-body CRISPR works. [14]

The caveat: 80 patients is small, six months is short, and switching a gene off permanently is exactly that — permanent. The long-term safety of a one-time, irreversible edit is the open question, and only years of follow-up will answer it. A strong six-month result is a beginning, not a verdict.

A regulator that keeps softening on rare disease

Gene therapy got a second piece of good news. UniQure, a Dutch company, said the FDA — the US drug regulator — reversed an earlier stance and agreed that three-year data from a single early-to-mid-stage trial could be the main basis for filing its gene therapy for Huntington’s disease. Huntington’s is an inherited, fatal brain disorder with no treatment that slows it. [12][6]

The FDA had pushed back in March. After the agency’s leadership changed this spring, it changed its tune; UniQure’s shares jumped 80%. Analysts read the reversal as the FDA growing more flexible on how much evidence a rare disease must show before approval. [6]

That flexibility cuts two ways. For patients with a fatal disease and no options, a faster path is a lifeline. But “accelerated approval” means a drug reaches people on thinner evidence, with a promise to prove it fully later. The trade between speed and certainty is real, and it is being recalibrated in public, deal by deal.

As fungi outrun our drugs, a new weapon

A quieter result matters more than it sounds. F2G and Shionogi reported phase 3 data for olorofim, the first antifungal with a genuinely new mechanism for invasive aspergillosis in more than 20 years. Aspergillosis is a life-threatening lung infection that mostly strikes cancer and transplant patients whose immune systems are down. [30]

Why now: the standard antifungals, called azoles, are losing their grip. The fungus has grown resistant — partly because the same chemicals are sprayed on crops, breeding resistant strains in the soil before they ever reach a hospital. Olorofim attacks a different target inside the fungus, so existing resistance doesn’t apply. In the trial, deaths at six weeks were about even with the standard IV drug (23.8% versus 24.3%), but olorofim is a pill and caused far fewer drug-related side effects (35.8% versus 63.9%). [30]

The number that should land is that 20-year gap. New antibiotics and antifungals are slow and unprofitable to develop, while resistance keeps rising. A drug that merely matches the old one on survival counts as a win, because the old one is failing.

The money is moving toward inflammation

Two big deals show where large drugmakers are placing bets. AbbVie is closing in on an roughly $11bn acquisition of Apogee Therapeutics, whose lead drug treats eczema — at a reported 60% premium to Apogee’s share price. Biogen agreed to buy RayThera for up to $1bn to pick up early-stage immunology programs. [25][49][16]

Both deals chase the same thing: the next wave of drugs for inflammatory and immune diseases, the field that GLP-1 weight-loss drugs and immune-system cancer treatments have made the hottest in pharma. These are bets on assets years from patients — most of the $1bn Biogen deal is tied to future milestones, not paid upfront. A deal is a vote of confidence, not a result.

The under-covered story: a fight over the lab mouse

Worth watching beneath the headlines: a debate over how ageing research should be done at all. The NIH, the largest funder of US biomedical research, has said it will stop creating new grants focused only on animal models, citing how poorly mouse results translate to people. [36]

The frustration is fair — more than 300 treatments have eased Alzheimer’s in mice, yet almost none have worked in humans. But researchers writing in Nature this week argue that abandoning animal studies wholesale is its own mistake: some questions about a whole living body simply can’t be asked in a dish. The point isn’t that mice are perfect or useless — it’s that the tool has to fit the question, and a blanket rule fits neither. [36]

02 · Lesson · why it matters

The difference between paying a toll forever and paying once to take the bridge down

Some problems we manage every day for a lifetime; a few we can solve once and be done. Which one a problem is shapes everything — the cost, the risk we'll accept, and how hard the choice becomes.

Two ways to deal with the same swelling

A person with hereditary angioedema lives by a schedule. The disease makes their body swell without warning — a hand, a throat — so they take a drug, again and again, for the rest of their life, to keep the attacks down. The treatment is a toll. You pay it every day, and the day you stop paying, the problem comes back.

This week a company reported a different deal. One dose of a gene-editing treatment, given once, cut attacks by 87% over six months. It works by switching off the single gene that starts the swelling. Not a toll paid daily — a switch flipped once.

That gap is the whole lesson. The same disease, two completely different shapes of solution. One you rent. One you own.

Why most of medicine is rent

Almost everything we call treatment is a toll. Blood-pressure pills, insulin, the inhaler, the daily tablet for a long condition — they don’t remove the problem, they hold it down while you keep paying. Stop, and it returns. We are so used to this that we barely notice it’s a choice. It’s just what “treatment” means.

There’s a reason renting dominates. A toll is forgiving. If the drug turns out to have a problem, you stop taking it and the effect fades. The system stays adjustable. You can change your mind tomorrow. Renting keeps the door open, and an open door is worth a lot when you’re not sure.

Why owning is rare, and harder

Flipping a switch once is a different kind of act. The edit that turned off that gene doesn’t wear off — that’s the point of it. But it also means there’s no taking it back. You don’t stop paying the toll; there is no toll. You also can’t undo the choice if it turns out wrong.

So the question stops being “does it work this month” and becomes “am I willing to live with this forever.” That’s a heavier question, and it explains why the trial’s six-month win is only a beginning. You can judge a rental in six months. You cannot judge a thing meant to last a lifetime in six months. The very feature that makes the one-time fix valuable — it’s permanent — is what makes it slow and hard to prove.

The same shape shows up far from gene editing. A regulator deciding how much proof to demand before approval is choosing between rent and ownership too. Demand thin evidence and act fast, and you’ve given a dying patient a chance — but you’ve made a near-permanent decision on a six-month look. Demand years of proof, and some people die waiting for certainty you finally have. There is no setting that is purely safe. There is only which mistake you’d rather risk.

Who else is standing at this fork

It isn’t only the patient at the fork. The drugmaker pricing a one-time cure has to make a lifetime’s worth of revenue from a single dose, which is part of why these treatments cost so much. The health system that pays once instead of forever has to find the whole sum up front, even though the saving comes over decades. The next patient with a different disease is watching this trial, because if a one-time edit holds here, the same approach gets tried on their gene next.

And the choice reaches people who will never take the drug. When fungi outrun our antifungals — bred resistant partly by chemicals sprayed on distant crops — a hospital patient with a lung infection inherits a problem that started in a field. We are all paying small tolls we never agreed to, set in motion by decisions made somewhere we can’t see.

On the whole

It’s tempting to read a one-time cure as simply better than a daily drug — the upgrade, the finish line. But “solve it once” and “manage it forever” aren’t better and worse. They’re different bargains, each with a cost the other doesn’t carry: the rental keeps you free to change your mind; the cure asks you to give that freedom up in exchange for being done.

The honest position is that nobody at this fork — not the patient, not the regulator, not the company — can see far enough down either road to be sure. A six-month result is a real reason for hope and a thin basis for a forever decision, both at once. Seeing that should make us slower to call any of these choices obvious, and a little gentler with the people who have to make them without knowing how it ends.

03 · Lab · your turn

Rent or Own the Cure

Live twelve years of a condition, choosing each year between a daily toll and a one-time irreversible edit — and feel why permanence is both the value and the risk.

04 · Hope · carry this

For the first time, a condition that meant a lifetime of daily management met a treatment given once — and held for six months. We are slowly learning not just to quiet our hardest problems, but to reach the place they start.