Day Lila

Biotech & Longevity · Wednesday, 28 October 2026

01 Briefing what happened

US regulator drafts one licence for gene-editing treatments built for each patient's own faulty gene

14 stories 20 sources

The FDA proposed that a single licence could cover many versions of a gene-editing treatment, each aimed at a different patient's mutation, and that new versions could be added on lab tests instead of new trials.

0

new patient trials needed to add a version inside the agreed limits

lab tests that predict safety and activity would be enough, the draft says [1]

3

FDA drafts this year on gene treatments made for small groups of patients

February's covered developing them, June's reusing earlier data, and October's the licence [1]

75

industry insiders' score for the US regulatory climate, where 100 means neutral

up from 50 in the spring and 23 a year ago, in a survey of 569 people [4]

The lead story — what happened

  • On Tuesday 27 October the FDA, the US drug regulator, published draft rules for gene-editing treatments built for each patient's own faulty gene. [1][2]
  • Gene editing changes DNA at a chosen spot inside a person's cells. A short molecule called a guide RNA steers the editing tool to that spot. [1]
  • Many inherited diseases come from different faults in the same gene, and some faults are found in only one patient or a handful. [1]
  • The draft says one licence can cover many versions of a treatment, each aimed at a different fault but sharing the same core design. [1]
  • A company would set limits in advance on what may change, such as the guide's sequence. The dose, the formula and the package that carries the editor into the body stay fixed. [1]
  • A new version inside those limits would count as a manufacturing change, not a new medicine, if lab tests predict it is safe and active. [1]
  • That lets a company add versions on lab data without new clinical trials, the FDA wrote. [1]
  • The limits themselves must still rest on data from patients. The draft does not say how large the first trials must be, and the FDA plans separate guidance on that. [1]
  • Rare disease groups welcomed it. A PKD Foundation official called it putting a flag in the ground. [2]
  • It came out on the second day of a rare disease summit run by NORD, a US patient group, where the heads of the FDA and NIH were due to speak. [2][3]
  • The FDA is run by an acting head, Kyle Diamantas, since its commissioner Marty Makary resigned in May. [4]
  • The draft is open for public comment and is not yet in force. [1]

Who is involved

  • The FDA

    the US drug regulator; its biologics centre wrote the draft

  • Kyle Diamantas

    the FDA's former food chief, running the agency on an acting basis since May

  • PKD Foundation

    a US charity for polycystic kidney disease, an inherited illness; it welcomed the draft

  • NORD

    the National Organization for Rare Disorders, a US patient group; it held the summit that week

How it unfolded

  1. Jan 2024 FDA guidance on gene treatments that use gene editing [1]
  2. Feb 2026 draft on developing treatments for specific genetic conditions [1]
  3. May 2026 FDA commissioner Marty Makary resigns [4]
  4. Jun 2026 draft on reusing earlier data across gene-editing programmes [1]
  5. 27 Oct draft on one licence covering many versions [1][2]

Where this points

The draft now takes public comment, and the FDA still owes separate guidance on how much patient data the first version needs; that second document decides how much cheaper this route really is. [1]

Signals on this beat

The bar and the word are our reading of how hard each one is pushing, and the arrow is where it is heading.

Deaths in cell treatment trials High↑

Artiva stopped its donor-cell treatment for immune diseases after two patients died. [5] Novartis paused its own cell treatment for immune diseases this summer after three deaths. [9] Artiva says the first death was likely caused by fludarabine, a chemotherapy drug given before many cell treatments. [6]

The cost of proving a drug Building↑

Novartis took a $1.9 billion charge to cover one more final trial of a drug for a rare muscle disease. [9] A doctor in Pennsylvania is building a company to revive treatments that drugmakers dropped because too few patients would make a profit. [3] The FDA draft would let new versions of a gene-editing treatment skip new trials. [1]

Empty seats at US health agencies Steady→

The deputy head of the CDC, the US disease-control agency, leaves on 12 November after four months in the job. [11] The FDA has been run by an acting head since May. [4] The nominee to lead it is still waiting for a vote in a US Senate committee, and the US Senate is away until 9 November. [4]

The rest of the day

13 more stories on this beat.

Each with its own sources. None of these is a link to the story above.

  1. 02

    Artiva stops its cell treatment after two deaths

    Artiva Biotherapeutics, a San Diego biotech, stopped work on Tuesday on AlloNK, a treatment made from donors' natural killer cells, a kind of immune cell. [5][6] It was being tested in rheumatoid arthritis and other immune diseases. [6] One patient died of brain damage that a safety review linked to fludarabine, a chemotherapy drug given first to make room for the new cells. [6][8] A second died suddenly about two months after starting, cause unknown. [5] The effects also faded faster than hoped, and Artiva is cutting 94% of its staff. [6][7]

    Why it matters — Artiva is the third company to report deaths in a cell treatment for immune diseases, after Novartis and Bristol Myers Squibb. [5] Fludarabine is given before many cell treatments, so Artiva says it will publish the case for others to study. [5]

  2. 03

    Novartis writes down a muscle-disease drug by $1.9 billion

    Novartis, the Swiss drugmaker, took a $1.9 billion charge on del-desiran, a drug for myotonic dystrophy type 1, a rare disease that wastes muscles. [9] The drug failed its final trial last month. [9] It came with Novartis's $12 billion purchase of Avidity Biosciences a year ago. [9] On a results call on Tuesday, chief executive Vas Narasimhan called the failure a bump. [9] He said the charge reflects the extra time and money needed for another final trial. [9]

    Why it matters — Two more drugs from the same purchase are still moving: one in a final trial for another muscle disease, and one under a fast FDA review for Duchenne muscular dystrophy. [9]

  3. 04

    US approves a pill pair for some prostate cancers

    The FDA approved Pfizer's talazoparib, sold as Talzenna, with the hormone-blocking drug enzalutamide on 26 October. [10] It is for prostate cancer that has spread and carries faults in genes that repair DNA, found by an approved test. [10] In a trial of 599 men, adding talazoparib cut the risk of the cancer growing or death by 52%. [10] Data on how long the men live are not yet mature. [10]

    Why it matters — It is approved for men whose cancer has not yet had, or still responds to, drugs that block male hormones. [10] Australia's and Canada's regulators are reviewing it alongside the FDA. [10]

  4. 05

    CDC's deputy head leaves after four months

    Jennifer Shuford, deputy director and chief medical officer of the CDC, the US disease-control agency, is leaving. [11][12] President Trump appointed her in April, after she ran Texas's health department, and she started in June. [11] Her last day is 12 November. [11][12] She said she wants more time with her family. [12] Sources told the Washington Post she had clashed over changes to vaccine advice. [11] Satish Pillai will act in her place. [12]

    Why it matters — The CDC has lost a third of its workforce and is facing 48 measles outbreaks, the Guardian reports. [12] Only half of people in the US trust its health advice, a recent survey found. [11]

  5. 06

    A UK pain drug for MS runs out until December

    Nabilone, a drug used to ease pain in multiple sclerosis, is out of stock at wholesalers across the UK, Northern Ireland's health department said. [13] In MS the immune system attacks the coating around nerves. [13] The 1mg capsules ran short in July and the 0.25mg ones more recently, and supply is not expected until December. [13] Jill Kenning, who has taken it for 17 years, has spent most days in bed since running out. [13]

    Medicines in short supply in the UK, by the MIMS tracker.

    Why it matters — The department would not say why supply stopped, calling it commercially sensitive. [13] Medicines in short supply in the UK reached 157 this week, up from 107 in January, by the MIMS tracker. [13]

  6. 07

    Judge lets Lilly keep its weight-loss drug ads

    A US federal judge, Zahid Quraishi, refused on Monday to block Eli Lilly's adverts comparing its Zepbound with Novo Nordisk's Wegovy. [14] Novo, the Danish maker of Wegovy, said the ads leaned on an older head-to-head trial and ignored a newer, higher dose of Wegovy. [14] The judge found Wegovy prescriptions did not fall while the ads ran. [14] He noted Lilly's newer ads mention the higher dose, and he also faulted Novo's own Ozempic ads. [14]

    Why it matters — The ruling only refuses an early ban, so Novo's claim that the ads mislead is still undecided. [14] Analysts expect sales of weight-loss drugs to pass $100 billion a year by 2030. [14]

  7. 08

    AI chatbots name the same drug brand almost every time

    Researchers asked five AI chatbots, including GPT-5.4 and Claude Sonnet 4.6, to recommend medicines for 177,120 made-up patients, in a study in Nature Health on 26 October. [15] Each case named a disease and a medicine that several approved products contain. [15] On average one product took 97.5% of the chatbots' first choices, whatever the patient's background. [15] The picks differed from what US doctors actually prescribe and often cost more. [15]

    • The top product97.5%
    • Every other product2.5%
    Average share of the chatbots' first choice, when several approved products hold the same medicine.

    Why it matters — About a fifth of US adults already use such chatbots to look up health information, the authors note. [15] When told to prefer the cheapest or generic product, the chatbots often switched to a generic one. [15]

  8. 09

    Pennsylvania families build a gene-treatment company

    Kevin Strauss, a doctor who treats Amish and Mennonite families in Pennsylvania, is building a biotech company by and for them, STAT reports. [3] These communities are hit by an unusual set of rare diseases. [3] The company aims to revive treatments that drugmakers dropped or left in university freezers, often because too few patients would make a profit. [3] Strauss's team gave a gene treatment to a calf with a cattle form of maple syrup urine disease, and she is now a healthy five-year-old heifer. [3][20] None of his patients has had it yet, even in a trial. [20]

    Why it matters — The FDA draft in today's lead lets new versions of a treatment skip new trials. [1] This company tries a second route: owners who will pay for a medicine that would never make money. [3]

  9. 10

    Ferring buys an egg-freezing startup for $1.4 billion

    Ferring Pharmaceuticals, a privately held Swiss maker of drugs for IVF, the treatment that joins egg and sperm outside the body, signed a $1.4 billion deal on Tuesday to buy Gameto. [16] Gameto, a biotech, is best known for a product meant to improve egg freezing and fertility for women. [16] Ferring plans to sell it through its reach into fertility clinics around the world. [16]

    Why it matters — The product still needs approval from the FDA, so the price is a bet on an approval that has not happened. [16]

  10. 11

    A sea-bacterium compound kills a brain-eating amoeba in the lab

    Researchers isolated a compound from a marine bacterium that killed several strains of Naegleria fowleri faster than the current treatment, in lab tests. [3] N. fowleri is a tiny parasite that lives in warm fresh water and can infect the brain. [3] Infections are hard to treat and nearly always fatal. [3]

    Why it matters — The team's next step is to test whether the compound can cross from the blood into the brain. [3] Much work remains before it could become a medicine. [3]

  11. 12

    A vitamin B5 relative eases psoriasis in a small trial

    Researchers in Beijing found that pantethine, a form of vitamin B5 sold as a supplement, calmed the immune cells that drive psoriasis, a skin disease. [17] It stopped psoriasis developing in mice. [17] In a 12-week trial of 24 patients, all used a standard skin ointment and half also took pantethine by mouth. [17] Those on pantethine improved more, especially if they started with high cholesterol. [17] The study is in the Journal of Experimental Medicine. [17]

    12 of 24

    patients who also took pantethine

    All 24 patients used the same skin ointment for 12 weeks; the 12 filled dots also took pantethine.

    Why it matters — The lead researcher calls the trial a sign the idea can be tested, not proof of how it works. [17] Pantethine is already sold widely as a supplement to lower cholesterol. [17]

  12. 13

    An old leukaemia drug is linked to fewer relapses

    A study from one hospital looked back at 496 adults with acute myeloid leukaemia, a fast-growing blood cancer. [18] Some had been given gemtuzumab ozogamicin, an antibody that carries a poison into cancer cells, in three small doses with chemotherapy. [18] Where the cancer had a change in a gene called NPM1, 57.9% of those on the drug were alive without relapse at two years, against 36.0% without it. [18] Liver damage was no more common. [18]

    Why it matters — The patients were not split at random, and those who got the drug were younger with less risky cancers. [18] So the gap may partly come from who received it, which only a randomised trial can settle.

  13. 14

    Weight-loss drugs cut demand for liposuction

    Liposuction fell 11.4% worldwide in 2025, after a 12.6% fall the year before, in a survey by the International Society of Aesthetic Plastic Surgery. [19] It asked more than 2,900 surgeons in 51 countries. [19] Cosmetic procedures fell 14% for men and 7.7% for women. [19] Surgeons in Dubai told The National that weight-loss drugs are a big reason. [19]

    Why it matters — One Dubai surgeon said the drugs are also creating a new group of patients, who want their face and body reshaped after losing weight. [19]

02 Lesson why it matters

Why proving a drug works costs each patient more when the disease is rare

Much of the cost of testing and approving a medicine stays the same however many people need it, so each patient's share grows when the disease is rare.

The twist

Many rare diseases could be treated, but too few patients share the cost of proving the treatment works.

The picture

Made-up numbers. The same $100 million bill for proving a drug, split per patient for a rare disease and for a common one.

How it works

  1. Proving a drug works takes trials, factories and checks
  2. Much of that bill stays the same for 500 patients or 5 million
  3. Each patient's share is the bill divided by the patients
  4. For a rare disease the share is huge, so firms walk away
  5. Cheaper proof, or owners who need no profit, can bring it back

The same force, elsewhere today

Where this chain is also running, in today's other stories.

  • Novartis writes down a muscle-disease drug

    One more final trial in a rare muscle disease cost Novartis a $1.9 billion charge, and only a small group of patients can buy the drug that pays it back.

  • Pennsylvania families build a gene-treatment company

    Drugmakers dropped treatments for these families' rare diseases because too few patients would make a profit, so the families are building an owner that needs none.

Where you've seen this

Books in rare languages

translating a book costs the same whether 500 or 5 million people will read it

Buses to small villages

the bus and the driver cost the same with three riders as with thirty

Spare parts for old machines

making the mould costs the same for ten parts as for ten thousand

The catch

Cheaper proof also means fewer checks, and a lab test can miss a harm that only shows up in a living patient.

And the whole of it

A family in Lancaster County, a reviewer at the FDA and an investor choosing which drug to fund each see one part of this bill. Someone with a common illness rarely sees it at all, because millions of other patients share their part.

03 Truth what's really going on

What is really going on

The US FDA proposed that one licence can cover many gene-editing versions, each built for a different patient's faulty gene, and that new versions can be added on lab tests instead of new trials. [1] It is still a draft, and the FDA has not yet said how much patient data the first version must show. [1]

Why it works on us — An advert saying one drug beat another in a head-to-head trial sounds final, even when the doses compared were not the newest ones. [14]

Who gains

  • Families with very rare inherited diseases — A single licence covering many versions could let a treatment reach a new patient's fault on lab tests, without a fresh trial each time. [1]
  • Companies building custom gene editors — They could add new versions to an approved licence as a manufacturing change instead of filing a new medicine. [1]
  • Eli Lilly — Its adverts comparing Zepbound with Wegovy can keep running while Novo's case goes on. [14]
  • Ferring — It gets Gameto's egg-freezing product to sell through its fertility-clinic network, if the FDA approves it. [16]
  • Makers of the brands chatbots name — Five chatbots put one product first 97.5% of the time on average, often a dearer one. [15]

Who pays

  • Artiva's staff — The company is cutting 94% of its workforce after stopping its main treatment. [6][7]
  • People with MS in the UK who rely on Nabilone — Both doses are out of stock until December, and one patient has been in hospital after alternatives failed. [13]
  • Patients waiting for a drug for myotonic dystrophy — Del-desiran needs another final trial, which Novartis says takes extra time and money. [9]
  • Novartis — It took a $1.9 billion charge on a drug from its $12 billion purchase of Avidity. [9]
  • Patients and insurers paying for medicines chosen by chatbots — The chatbots' picks often carried higher projected costs than what doctors actually prescribe. [15]

What nobody knows yet

Open questions from across today’s stories — ours included.

  • 01

    How much patient data the first version of a custom gene-editing treatment must have.

    The draft leaves out trial design, and the FDA says it will cover it in further guidance. [1]

  • 02

    When the draft becomes final, and in what form.

    It is open for public comment, and the FDA is run by an acting head while its nominee waits for a US Senate committee vote. [1][4]

  • 03

    What killed the second patient in Artiva's trial.

    The cause of the sudden death has not been determined. [5][8]

  • 04

    Whether fludarabine adds risk across other cell treatments.

    It is given before many of them, and Artiva has only promised to publish the details of its case. [5]

  • 05

    Whether del-desiran can pass a second final trial.

    Novartis says it saw signs of activity on other measures, but the trial's main goal was missed. [9]

  • 06

    Why Nabilone ran out across the UK.

    Northern Ireland's health department called the reason commercially sensitive and would not give it. [13]

  • 07

    Why Jennifer Shuford is leaving the CDC, and how long she served.

    She cites family, while the Washington Post's sources point to a clash over vaccine advice. Scientific American says she leaves after four months in the role, the Guardian after six. [11][12]

  • 08

    Whether Lilly's adverts are misleading.

    The judge only refused an early ban, and Novo says the main question is still open. [14]

04 Hope carry this

Men with prostate cancer that carries faults in DNA-repair genes have a newly approved US treatment: in a trial of 599 men, adding talazoparib to enzalutamide cut the risk of the cancer growing or death by 52%.

Also true today

  • A calf in Pennsylvania with a cattle form of maple syrup urine disease was given a new gene treatment, and she is now a healthy five-year-old heifer.
  • In lab tests, a compound from a sea bacterium killed several strains of the brain-eating amoeba Naegleria fowleri faster than the current treatment.
  • In a 12-week trial in Beijing, 24 psoriasis patients used the same skin ointment, and the half who also took pantethine improved more.

Across the beats