A man with ALS was given a drug written for his mutation alone. A year later he is still working as a doctor.

16 stories 18 sources

Signals on this beat

Diseases splitting into smaller ones Building ↑

A common epilepsy drug is linked to longer survival in children with diffuse midline glioma, a brainstem tumour, but not in the other childhood brain cancers. [5]
The US drug regulator approved a breast cancer pill only for tumours carrying one mutation, found by a blood test. [6]
Researchers proposed splitting Alzheimer's into variants by which brain network goes first. [7]

Old drugs given new jobs Building ↑

Levetiracetam, prescribed for seizures, turned out to slow tumour growth in mice with a children's brainstem cancer. [5]
Sunitinib, a pill that starves tumours of blood vessels, was added to a radioactive drug in a small group of patients with pancreatic tumours. [15]
Amikacin, an old injected antibiotic, is now breathed in as a mist for a slow lung infection. [9]

Small firms carrying the risk High →

Beacon Therapeutics, with about 80 staff, won the trial that Biogen and Johnson and Johnson lost, and says it may need partners to sell the treatment. [4]
North Immunology reached the stock market by merging into Aethlon Medical, a blood-filter company. [14]
Cue Biopharma bought the rights to its hives drug from the company that was running the trial in China. [10]

Treatments made for one patient Building ↑

Twenty professional bodies and 18 experts published the first agreed rules for treating one patient with viruses picked to match their own infection. [8]
A gene therapy carries a full-length copy of one gene, so it fits only the people whose blindness comes from that gene. [4]

The stories

16 stories on this beat today.

Every story told the same way: what happened, who is involved, why it matters, how it works, the truth and the good news.

01
A man with ALS was given a drug written for his mutation alone. A year later he is still working as a doctor.

3 years

to design the drug, test it in animals and give it to the patient

antisense drugs for commoner ALS mutations took at least ten [1]

5-10%

of people with ALS have a gene fault anyone can name

for everyone else there is no target to aim a drug at [1]

251

patients in the Novartis trial that ended its ALS programme

the drug missed its main and secondary targets [2]

31%

of men and boys on the higher dose of Beacon's eye gene therapy read 15 more letters

nobody in the untreated group managed it [4]
What happened
- A man with a slowly progressing form of ALS was given a drug designed for the exact gene fault causing his disease, and a year later he was still working as a doctor. [1]
- ALS kills the nerve cells that carry instructions from the brain to the muscles.
- Most people die within two to five years of diagnosis, and there is no cure. [1]
- His illness comes from a fault in a gene called CHCHD10, carried by fewer than one in a hundred people who inherit ALS. [1]
- The drug is an antisense oligonucleotide, a short strand of genetic material that sticks to the message a faulty gene sends out so the cell makes less of the harmful protein. [1]
- He was given three 50-milligram doses into his spine and then three of 75 milligrams, between April 2024 and April 2025, with no serious side effects reported. [1]
- A year after the first dose, neurofilament light chain in his blood was back in the normal range.
- That protein leaks out of damaged nerve cells, so it is used to track how fast ALS is moving. [1]
- Antisense drugs for commoner ALS mutations took at least a decade to build.
- This one took three years, a co-author at the Mayo Clinic in Florida said. [1]
- This is one patient with nobody to compare him against, and a few people with ALS improve for a while on their own.
- The team must follow him for another two or three years and treat other people. [1]
- In the same week Novartis ended its ALS programme after its drug missed every target in 251 patients, and Axoltis said its drug missed the blood marker it was built around in 82 patients. [2][3]
251 peopleA drug aimed at ALS
1 manA drug for his own gene fault
Novartis tested its drug in 251 people who all have ALS, and it missed every target. A Mayo Clinic team made a drug for one man's mutation and gave it to him.
Who is involved
- The Mayo Clinic team doctors and scientists at the US hospital group; they designed the drug for one man's mutation and published the result in the journal Med
- Novartis one of the world's biggest drugmakers; it ended its ALS programme after the drug missed in 251 patients
- Beacon Therapeutics a biotech company of about 80 people; its gene therapy for an inherited blindness passed its final test
- Eli Lilly the US drugmaker whose breast cancer pill was approved only for women whose tumour carries an ESR1 mutation
How it unfolded
1. April 2024 the man is given the first of six doses into his spine
2. April 2025 the last dose; no serious side effects reported
3. A year on his nerve-damage protein is back in the normal range and he is still working as a doctor
4. This week two ALS drugs aimed at everyone with the disease fail, and Novartis drops its programme
5. Next two or three more years of follow-up, and other patients with their own mutations
Why it matters

Watch whether a second person with a different rare ALS mutation gets a drug of their own, because that is what would show this is a method rather than one lucky case.
The truth
- What is really going on A drug written for one man's gene fault brought his nerve-damage marker back to normal, while Novartis's ALS drug did nothing in 251 people and the programme was dropped.The one man was chosen because his ALS comes from a fault in a gene called CHCHD10.
- Why it works on people One person getting better is easy to picture and 251 people not getting better is a table, so a first case can feel like more evidence than a failed trial.
- Who gains The Mayo Clinic team and their patient — They showed a drug can be designed, tested in animals and given to one named person in three years, where earlier antisense drugs took ten. [1]
- Who pays People with ALS whose cause nobody can name — Only 5 to 10% of people with ALS have a known gene fault, and a drug aimed at a fault needs one to aim at. [1]
- Who pays The 251 people in the Novartis trial — They took an experimental drug for months, and the result was the end of the programme. [2]
- Nobody knows yet
  Whether the drug is why the man improved.
  
  He is one patient with nobody to compare him against, and a few people with ALS improve for a while on their own. His own doctors say it needs two or three more years and more patients. [1]
- Nobody knows yet
  Whether anyone else can be given this drug.
  
  Fewer than one in a hundred people who inherit ALS carry the CHCHD10 fault, and no second patient has been treated. [1]
- Nobody knows yet
  Why the Novartis drug failed.
  
  The company said only that the trial missed its main and secondary targets. No figures have been published. [2]
Good news
- A man whose ALS comes from a fault in a single gene was given a drug written for that fault. A year after the first dose the protein that leaks from dying nerve cells was back in the normal range in his blood, and he was still working as a doctor.

02
Gene therapy passes its test in an inherited blindness
What happened
- Beacon Therapeutics said its gene therapy improved sight in a final-stage trial of 85 men and boys with X-linked retinitis pigmentosa, an inherited disease that kills the light-sensing cells at the back of the eye. [4]
- A year after a single treatment, 31% of the higher-dose group read at least 15 more letters on an eye chart in dim light, and 24.1% of the lower-dose group did. [4]
- Nobody in the untreated group managed it. [4]
- Biogen's gene therapy for the same disease missed its target in 2021 and Johnson and Johnson's failed last year. [4]
Untreated0%
Low dose24.1%
High dose31%
Share of patients who could read 15 more letters on a dim eye chart a year after treatment.
Why it matters

It is the first treatment to work in a disease that has taken sight from hundreds of thousands of people and beaten two much larger companies.
The company has about 80 staff and says it may need partners to reach patients outside the US.
The truth
- Who gains Beacon Therapeutics — A company of about 80 staff now holds the only gene therapy that has passed its final test in this inherited blindness, after Biogen and Johnson and Johnson missed. [4]
- Nobody knows yet
  Whether Beacon's gene therapy helps people see in daily life.
  
  The measure that passed was letters read on a chart in dim light. The company has not published what changed in the things the patients actually do. [4]
Good news
- A year after one treatment, 31% of the men and boys on the higher dose of Beacon's gene therapy could read at least 15 more letters on an eye chart in dim light. Nobody in the untreated group could read one more.

03
A cheap epilepsy drug and a children's brain tumour
What happened
- Levetiracetam is an ordinary antiseizure medicine, given to many people with brain tumours because the tumours cause fits.
- Researchers reported in Nature Medicine that children with diffuse midline glioma, a tumour in the brainstem that almost nobody survives, lived longer if they were taking it. [5]
- The same pattern was absent in children with high-grade gliomas elsewhere in the brain. [5]
- In mice, the drug slowed the brainstem tumours by quieting the nerve-to-tumour connections those tumours grow on, an effect separate from how it stops seizures. [5]
Brainstem tumourlonger survival
Elsewhere in the brainno link
The same medicine, two childhood brain cancers. For tumours in the brainstem the drug was linked to longer survival; for tumours elsewhere in the brain it was not.
Why it matters

The human half is a look back at records rather than a trial, so it shows a link and not yet a cause.
It matters because the drug is cheap, already licensed and already in these children's hands.
How it works
A disease is named after what it does, not after what causes it

ALS is the name for motor nerve cells dying, and there are several different reasons they die, so one drug can help one person and do nothing at all for the next.

The twist

The word ALS describes nerve cells dying. It does not say why they are dying, and a drug has to be aimed at the why.

The picture

1 of 100
of every hundred people who inherit ALS carry the gene fault this drug was built for
The other ninety-nine have their own reasons for the same diagnosis, and most of those reasons have no name yet.
Step by step
1. A disease gets its name from what the patient shows
2. Several different faults can produce those same signs
3. A drug fixes one fault, not the name
4. Given to everyone with the name, its effect is diluted away
5. Given to the people with that one fault, it shows up
The same force, elsewhere today

Where this chain is also running, in today's other stories.
- A cheap epilepsy drug and a children's brain tumour
  
  The same drug is linked to longer survival in brainstem tumours and not in tumours elsewhere, because only the brainstem kind grows on nerve connections the drug quiets
- Breast cancer pill approved for one mutation only
  
  The prescription follows a blood test for an ESR1 mutation rather than the diagnosis, so the group the drug is aimed at is defined by the fault
- Alzheimer's proposed as several illnesses
  
  Researchers want patients sorted by which brain network fails first, because one name is holding several different illnesses
- Childhood brain cancer differs by age and sex
  
  Tumours with one name split into groups with different molecules and different survival, which is the same split one step earlier
Where you've seen this

A car that will not start

it can be a flat battery, a dead starter or no fuel, and jump leads only fix one of the three

Back pain

one phrase covers a worn disc, a pulled muscle and a squeezed nerve, and the treatments do not swap

Slow internet at home

the router, the cable to the street and the company selling it can each be the cause, and each fix does nothing for the other two

The catch

Splitting a disease into smaller ones makes every group smaller. A group of one person is very hard to prove anything from, which is exactly the problem this man's doctors now have.

And the whole of it

Everyone here is working with the names they were handed - the doctors, the drug companies, the regulator, and the patient who joins a trial because he matches a word on a form. From any one of those seats you cannot see how many different illnesses that word is holding.
The truth
- Who pays Children with high-grade glioma outside the brainstem — The survival link with levetiracetam was absent in their tumours, in both the records and the mice. [5]
- Nobody knows yet
  Whether levetiracetam really lengthens life in a children's brainstem tumour.
  
  The human half is a look back at medical records, not a trial, and children on the drug may differ from those not on it in ways the records cannot show. [5]
Good news
- Children with a brainstem tumour that almost nobody survives lived longer if they were taking levetiracetam, a cheap antiseizure drug many of them were already prescribed.

04
Breast cancer pill approved for one mutation only
What happened
- The US drug regulator approved Eli Lilly's pill imlunestrant, sold as Inluriyo, with another Lilly drug, abemaciclib, for advanced breast cancer that is hormone-driven, HER2-negative and carries a mutation in the ESR1 gene. [6]
- That mutation turns up in tumours that have already been treated with hormone-blocking drugs. [6]
- The regulator approved a blood test, Guardant360 CDx, alongside it, so the prescription depends on the test finding the mutation first. [6]
- The decision rests on EMBER-3, a trial of 874 patients who had all been treated before. [6]
Why it matters

It ties the drug to a genetic label rather than to the organ the cancer started in.
Women whose tumours do not carry the mutation are outside the approval, even with the same diagnosis.

How it works

The prescription follows a blood test for an ESR1 mutation rather than the diagnosis, so the group the drug is aimed at is defined by the fault
The idea, in story 03: A disease is named after what it does, not after what causes it →
The truth
- Who gains Eli Lilly — Its pill imlunestrant is approved with its own older drug abemaciclib, so one approval sells two Lilly medicines. [6]
- Who gains Guardant Health — The regulator approved its blood test as the gatekeeper for the new breast cancer prescription, so every prescription needs a test first. [6]
- Who pays Women with hormone-driven breast cancer and no ESR1 mutation — The new approval only covers tumours where the test finds that mutation, so the same diagnosis does not reach the drug. [6]
05
Alzheimer's proposed as several illnesses
What happened
- Alzheimer's is defined by two deposits in the brain, amyloid plaques and tau tangles, but they do not hit every region equally. [7]
- Writing in Nature Reviews Neurology, researchers set out a framework for the atypical forms, where the first symptom is trouble seeing, speaking or planning rather than forgetting. [7]
- They propose sorting patients on four axes: the symptoms, the biology present, which brain network is affected, and how fast it moves. [7]
- Those forms tend to start younger than the memory-first kind. [7]
Why it matters

Trials recruit people by diagnosis, so mixing four patterns into one group can bury a drug that helps one of them.
The authors argue a shared framework would make separate studies comparable.

How it works

Researchers want patients sorted by which brain network fails first, because one name is holding several different illnesses
The idea, in story 03: A disease is named after what it does, not after what causes it →
06
First agreed rules for treating infections with viruses
What happened
- Phages are viruses that infect bacteria and nothing else, and doctors have been using them one patient at a time against infections that antibiotics no longer touch. [8]
- Until now there was no agreed way to choose a phage, prepare it, check its quality or record what happened. [8]
- Twenty professional societies, patient groups and regulators plus 18 international experts have published more than 60 recommendations in Nature Medicine, under a German medical framework. [8]
- They cover the whole path from the laboratory bench to the hospital bed. [8]
Why it matters

Every treatment is matched to one patient's own bacteria, which is precisely why it never fitted the usual approval process.
Written rules are what let a regulator inspect something that is different every time.
The truth
- Who pays Patients needing phage treatment outside a research hospital — The new rules describe what a properly equipped pharmacy and clinic must do, which most places cannot yet do at all. [8]
- Nobody knows yet
  Whether the new phage rules will be used outside Germany.
  
  The guideline was written under a German medical framework, and each country's regulator decides separately what it will accept. [8]
07
Inhaled antibiotic gets a faster review
What happened
- Insmed said the US drug regulator has agreed to a priority review of Arikayce, a form of the old antibiotic amikacin packed into fatty bubbles so it can be breathed in as a mist. [9]
- It treats MAC lung disease, a slow infection by bacteria related to tuberculosis that mainly strikes damaged lungs. [9]
- The company's ENCORE trial put 425 patients on either Arikayce plus standard tablets or the tablets alone, and reported better breathing symptoms and more patients clearing the bug. [9]
- A decision is due by 28 January 2027. [9]
Why it matters

The drug was cleared years ago on a conditional basis, and this trial is the proof it was required to produce.
The figures come from the company's own announcement, not yet from a journal or the regulator.
The truth
- Nobody knows yet
  What the inhaled antibiotic did in the trial, in figures.
  
  The numbers come from Insmed's own announcement rather than from a journal or the regulator, and the decision is not due until 28 January 2027. [9]
08
Hives drug beats the standard injection on numbers
What happened
- Cue Biopharma reported that CUE-221 cleared chronic spontaneous urticaria, a condition of unexplained hives that can last years, in 43% to 54% of patients at twelve weeks against 11% on placebo, in a 145-person trial run in China. [10]
- One group in the trial received Xolair, the injection doctors now use, which cleared 41%. [10]
- By week 28 the gap had widened, with 60% clear on the top dose of CUE-221 and 24% on Xolair. [10]
- The Xolair comparison was not part of the statistical plan. [10]
Why it matters

A numerical edge over the standard treatment is not a proven one, and the trial was never designed to test it.
It matters because Xolair is the only widely used option for patients whose hives ignore antihistamines.
The truth
- Nobody knows yet
  How Cue's hives drug really compares with Xolair.
  
  The trial dosed a Xolair group but never planned a statistical comparison, so the gap at week 28 is a number and not a result. [10]
Good news
- Hives that had refused to go away cleared completely in 43% to 54% of patients twelve weeks into Cue Biopharma's trial, against 11% of those given a dummy injection.
09
A sleep-pathway drug tried in ADHD
What happened
- Alkermes gave ALKS 7290, a drug that switches on the orexin system the brain uses to stay awake, to adults with ADHD for two weeks. [11]
- Scores on a 54-point symptom scale fell by an average of 14 points on the lower dose and 19 on the higher one. [11]
- The study was a first test in humans and was not built to prove the drug works, so there is no statistical comparison. [11]
- Side effects included dizziness, constipation and needing to urinate more often. [11]
Why it matters

Orexin drugs were developed to keep people with narcolepsy awake, and this is the first attempt to move the class into another condition.
A four-week phase 2 trial is already enrolling.
10
Childhood brain cancer differs by age and sex
What happened
- Researchers at the Icahn School of Medicine at Mount Sinai and colleagues examined the proteins and genes of tumours from more than 100 patients aged 0 to 40 with high-grade glioma, an aggressive brain cancer. [12]
- They found two distinct groups among adolescents and young adults, with different molecular make-up and different survival. [12]
- Age and sex shape normal brain development, so the team separated those effects from the cancer's own. [12]
- The work came out of a US National Cancer Institute consortium with children's brain tumour networks. [12]
Why it matters

One tumour name is carrying several biologies, and a trial that mixes them is testing a drug on more than one disease at once.
The groups they found are candidates for separate trials.

How it works

Tumours with one name split into groups with different molecules and different survival, which is the same split one step earlier
The idea, in story 03: A disease is named after what it does, not after what causes it →
11
Blocking a cell's recycling makes a tumour feel radiation
What happened
- Some brain tumours carry a mutation in a gene called IDH1, which rewires how the cell makes energy. [13]
- Researchers found those tumour cells lean on autophagy, the process by which a cell digests its own worn-out parts for fuel. [13]
- In mice, silencing a gene the process needs, using tiny protein packages carrying an interfering RNA, made the tumours shrink under radiation and left the animals with lasting immunity to the cancer. [13]
- The work was published in Nature Communications. [13]
Why it matters

Radiation is standard for these tumours and works poorly, and this points at why.
It is mice and cells, which is a long way from a treatment people receive.
12
A blood-filter company becomes an eczema company
What happened
- North Immunology, a private biotech backed by ADAR1 Capital Management, is merging into Aethlon Medical, a Nasdaq-listed company that had been developing a device to filter patients' blood. [14]
- The combined company will be built around North's lead drug, NOR-101, an antibody that blocks two inflammation signals at once for atopic dermatitis, the commonest form of eczema. [14]
- The drug is due to enter the clinic early next year, with first data expected by mid-year. [14]
- Existing eczema drugs block only one of the two pathways. [14]
Why it matters

A reverse merger is how a small biotech reaches the stock market without a listing of its own.
The shell it used had been trying to sell a machine that filters a patient's blood.
The truth
- Who gains Aethlon Medical's shareholders — A blood-filter company becomes the stock market listing for a funded eczema drug. [14]
13
Radioactive drug paired with a pill in pancreatic tumours
What happened
- Peptide receptor radionuclide therapy sends a radioactive atom into a tumour by attaching it to a molecule the tumour's cells grab. [15]
- Doctors added sunitinib, a pill that starves tumours of the blood vessels they grow, to the standard radioactive treatment in a small group of heavily pretreated patients with advancing pancreatic neuroendocrine tumours. [15]
- Three patients had near-complete responses on scans and two had partial ones, and the combination was tolerated. [15]
- Laboratory work showed each drug made the other bite harder. [15]
Why it matters

It is a pilot with a handful of patients and no comparison group, so it sets up a trial rather than settling anything.
These tumours usually come back after the radioactive treatment alone.
14
A cancer pill tried in a rare womb cancer
What happened
- Uterine leiomyosarcoma grows in the muscle wall of the womb, is aggressive and has no established treatment. [16]
- Researchers tested amebaciclib, a drug that blocks the switches a cell uses to start dividing, in mice bred to develop the tumour on their own. [16]
- The drug shut down that signalling chain and slowed the tumours. [16]
- The same class of drug is already licensed in breast cancer, where it is given with hormone treatment. [16]
Why it matters

A licensed class reaching a cancer with nothing available is the cheapest kind of progress, because the safety work is done.
It is a mouse result, and most mouse results do not carry over.
15
A cheaper way to read cancer's traces in blood
What happened
- Fragments of DNA from dying cells float in the blood, and their chemical tags reveal which tissue they came from and whether a tumour is present. [17]
- The cheap way of reading those tags, MeDIP sequencing, tells you a region is tagged but not how heavily, which has limited what it can be used for. [17]
- A team published a statistical model in Communications Biology that combines those cheap reads with reference maps to recover the missing quantities. [17]
- They tested it on simulations, matched datasets and patient samples. [17]
Why it matters

Blood tests for cancer live or die on cost, because they only pay off when used widely.
The code is free to download.
16
Children's gene therapy starts commercial production
What happened
- Fayuvi, approved last week as the first treatment for Sanfilippo syndrome type A, is now being made for sale at Andelyn Biosciences' plant in Columbus, Ohio, alongside its developer Ultragenyx's own site in Bedford, Massachusetts. [18]
- Sanfilippo type A is an inherited disease that strips away speech and thinking in young children; between 3,000 and 5,000 people worldwide have it and half do not reach 15. [18]
- It is the first approved gene therapy made on Andelyn's production platform. [18]
Why it matters

An approval only reaches a child if somebody can make the doses.
Gene therapies are grown in living cells, which is why supply is a separate problem from approval.
Good news
- Fayuvi, the first treatment for Sanfilippo syndrome type A, is now being made for sale at two plants. Until last week there was nothing for those children but help with their symptoms.

Across the day

What is true of the whole day.

The truth, across the day

The 251 were chosen because they have ALS.

A man with ALS was given a drug written for his mutation alone. A year later he is still working as a doctor.

16 stories on this beat today.

A disease is named after what it does, not after what causes it

What is true of the whole day.