A drug for the world's third-biggest killer cut attacks by 30%, and it was tested in the patients the other new drugs leave out

The stories

21 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.

1. 01

   A drug for the world's third-biggest killer cut attacks by 30%, and it was tested in the patients the other new drugs leave out

   ~30%

   fewer sudden worsenings of COPD over a year on tozorakimab

   measured across all patients in the two final-stage trials [1]

   3rd

   COPD's rank among causes of death worldwide

   and the disease still has few treatments that work well [1]

   2 of 2

   final-stage trials reporting together

   shown in Barcelona and published in the New England Journal of Medicine on the same day [1]

   What happened

   • AstraZeneca said on Tuesday that two final-stage trials of tozorakimab cut sudden worsenings of COPD by about 30% over a year. [1]
   • COPD is a lung disease that makes breathing steadily harder.
   • A worsening, or flare-up, is any sudden decline that needs medical care - a severe cough, or breathlessness. [1]
   • It is the third-biggest cause of death in the world and, in STAT's words, greatly underserved: there is still very little that works well. [1]
   • The results were presented at the European Respiratory Society's annual meeting in Barcelona and published in the New England Journal of Medicine. [1]
   • A 30% cut is about as good as the other new COPD drugs of recent years.
   • What is different is who was in the trials. [1]
   • Tozorakimab was tested in patients who are not eligible for those other new drugs, and the benefit seen in them could hand AstraZeneca a large part of the market with no competitor. [1]
   • The drug is an injection.
   • AstraZeneca is also testing it in asthma, and analysts expect billions of dollars a year in sales if it is approved. [1]
   • The company needed this.
   • It has had several notable trial failures in recent months, the biggest of them a heart-disease drug. [1]

   Who is involved

   • AstraZeneca a British drugmaker and one of the world's largest; it makes tozorakimab and has had several trial failures in recent months [1]
   • People with COPD patients whose lungs get worse over years; a flare-up is a sudden decline that needs a doctor or a hospital [1]
   • The European Respiratory Society the body for Europe's lung doctors; the results were presented at its annual meeting in Barcelona [1]
   • The New England Journal of Medicine one of the most-read journals in medicine; it published both trials the same day [1]

   How it unfolded

   1. Recent months AstraZeneca's heart-disease drug fails in a late-stage trial [1]
   2. Tuesday the company releases results from two final-stage trials of tozorakimab in COPD [1]
   3. The same day the data are presented in Barcelona and published in the New England Journal of Medicine [1]
   4. Next AstraZeneca has to file for approval, and is also testing the drug in asthma [1]

   Why it matters

   Watch which patients a regulator's label covers rather than the 30% figure - tozorakimab's value rests on whether it is cleared for the people who cannot take the other new COPD drugs. [1]

   The truth

   • What is really going on AstraZeneca's lung drug matters less for its 30% figure than for who was in the trials: it enrolled COPD patients who cannot take the other new drugs, and that group is where the money is. [1]
   • Why it works on people A percentage sounds like a fact about a drug when it is really a fact about the people who were in the trial - AstraZeneca's 30% and Inhibrx's 48.3% each describe a completely different set of patients. [1][10]
   • Who gains AstraZeneca — Two final-stage wins in the world's third-biggest cause of death, in a patient group its rivals' drugs are not cleared for, which analysts value in the billions of dollars a year. [1]
   • Nobody knows yet

     How well tozorakimab works in the patients who cannot take the other new COPD drugs.

     AstraZeneca reported about 30% fewer flare-ups across all patients and has published no separate figure for that group, which is the group the whole commercial case rests on. [1]

   Good news

   • A new AstraZeneca injection cut sudden worsenings of COPD, a lung disease that makes breathing steadily harder and is the third-biggest cause of death in the world, by about 30% over a year in two final-stage trials.

1. 02

   Novartis loses a second big trial in three days

   What happened

   • Novartis, one of the world's biggest drugmakers, said on Tuesday that its drug del-desiran had failed the final-stage HARBOR trial in myotonic dystrophy type 1, a disease that slowly stiffens and weakens muscles. [2]
   • The trial's main measure was video hand opening time - how fast a hand relaxes after squeezing - and the drug did no better than a dummy treatment. [3]
   • It was the company's second late-stage failure in days, after its cholesterol drug missed on Friday. [2][3]
   • The shares fell as much as 10%, and Novartis left its sales forecast to 2030 unchanged. [3]

   Why it matters

   Novartis is heading into a patent cliff, when its biggest sellers lose their protection, and the drugs it has bought are what is meant to replace them. [2]

   The truth

   • Who pays People with myotonic dystrophy type 1 — Del-desiran had reached a final-stage trial for a disease that progressively stiffens and weakens muscles, and it failed its main test. [2]
   • Who pays Novartis shareholders — The shares fell as much as 10% on Tuesday after the muscle drug missed, a day after a 3% fall on the cholesterol drug. [2][3]
   • Nobody knows yet

     Why Novartis's muscle drug missed.

     The trial's main measure was video hand opening time, how fast a hand relaxes after squeezing, and Novartis has released nothing beyond the miss while reaffirming its sales forecast to 2030. [2][3]

2. 03

   Roivant's lung drug cuts a pressure measure by 56%

   What happened

   • Roivant said on Tuesday that mosliciguat met every goal of a mid-stage trial in pulmonary hypertension linked to interstitial lung disease - high blood pressure in the vessels between heart and lungs, in people whose lung tissue is scarring. [5]
   • Over 16 weeks it cut pulmonary vascular resistance, a measure of how hard blood must push through the lungs, by 56% against a dummy treatment. [5]
   • Patients also walked 35 metres further in six minutes. [5]
   • Roivant has already started a final-stage trial, and one analyst put possible sales above $10 billion a year. [4]

   Why it matters

   The 56% describes pressure inside the lungs and the 35 metres is a patient walking further in six minutes; the final-stage trial now has to repeat both. [5]

1. 04

   Bristol Myers reports a win and no numbers

   What happened

   • Bristol Myers Squibb said its cell therapy arlo-cel hit the main goal of the pivotal Quintessential trial in multiple myeloma, a cancer of the plasma cells in bone marrow. [6][7]
   • The treatment takes a patient's own immune cells, edits them in a laboratory to recognise a protein called GPRC5D on the cancer, and returns them in one infusion. [7]
   • Everyone in the trial had already failed four classes of standard treatment. [6]
   • The company released no figures at all, saying only that the result was statistically significant and clinically meaningful. [6][7]

   Why it matters

   Johnson & Johnson's rival drug against the same protein reported a 72% response rate, so until Bristol Myers publishes a number nobody outside the company can tell whether arlo-cel beats it. [6]

   How it works

   Why a drug looks better the fewer people it is tested in

   Leave the patients least likely to respond out of a trial and the average result improves, without the drug doing anything more.

   The twist

   The quickest way to improve a drug's result is to shrink the group it is tested in, and that same choice shrinks the number of people who will ever be allowed to take it.

   Step by step

   1. 1 A drug helps some patients much more than others
   2. 2 A trial chooses who is allowed in
   3. 3 Leave out the ones least likely to respond
   4. 4 The average result rises, and the drug looks stronger
   5. 5 The licence follows the trial, so fewer people can ever get it

   The same force, elsewhere today

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

   • Inhibrx's head and neck cancer result

     Everyone in the trial already had high PD-L1, and the biggest effect came from a smaller group inside that - patients with HPV-positive tumours, at 80% against 33.3% - so the company is recruiting 50 more of exactly those.

   • Bristol Myers's cell therapy for multiple myeloma

     The trial took only people who had already failed four classes of treatment, which is the same step run backwards: against a group with nothing left, almost any response counts as a win.

   • The Alzheimer's prevention trial in Surrey

     Blood tests pick the 1,600 healthy volunteers most likely to benefit before anyone is enrolled, so whatever the trial finds will be about them and not about everyone at risk.

   • The pancreatic cancer drug patients cannot get

     Here the line is not drawn by biology at all - approval shut the free programme, and who gets the drug now depends on whether an insurer has finished its paperwork.

   Where you've seen this

   Job-training programmes

   a course that only takes people already close to work reports a high success rate

   Bank lending

   a lender that only lends to safe borrowers has almost no defaults, and turns most applicants away

   Driving instructors

   a teacher who only enters ready pupils for the test has a pass rate near 100%

   The catch

   Narrowing is often the right call. If a drug only works on one gene fault, giving it to everyone would bury a real benefit under the people it was never going to help.

   And the whole of it

   The company wants a clear signal, the regulator wants proof, and the doctor wants this patient to get better. Every one of them is being sensible, and the patients who did not qualify are counted in nobody's numbers.

   The truth

   • Nobody knows yet

     What Bristol Myers's cell therapy actually achieved.

     The company said the result was statistically significant and clinically meaningful and released no numbers at all, saying full data will come at a later medical meeting. [6][7]

1. 05

   Pharvaris cuts swelling attacks by 83%

   What happened

   • Pharvaris, a Swiss biotech, said its once-daily pill deucrictibant cut the rate of swelling attacks by 83% over 24 weeks in 85 people with hereditary angioedema. [8]
   • The condition is inherited and causes sudden swelling of the face, limbs and airway with no warning, and it can kill. [8]
   • Among the 80 patients with the two commonest forms, the reduction was 87%. [8]
   • It is the drug's second final-stage win; the US drug regulator is due to rule on a different version of it in April 2027, and the shares rose 14%. [8]

   Why it matters

   Pharvaris says the pill could be the first single drug used both to stop an attack and to prevent them; the existing oral rival, KalVista's Ekterly, only treats attacks. [8]

   The truth

   • Who gains Pharvaris — An 83% cut in swelling attacks in a second final-stage trial, with the shares up 14% and a regulator's decision on another version already booked for April 2027. [8]

   Good news

   • Pharvaris said a once-daily pill cut swelling attacks by 83% over 24 weeks in 85 people with hereditary angioedema, an inherited condition in which the face, limbs and airway swell without warning.
2. 06

   Inhibrx's result came from a small group inside the trial

   What happened

   • Inhibrx said its drug INBRX-106, given with Merck's Keytruda, shrank tumours in 48.3% of head and neck cancer patients against 26.5% on Keytruda alone, in a 68-person mid-stage trial. [10]
   • Everyone enrolled already had high levels of PD-L1, a protein that marks out who tends to respond to this kind of immunotherapy. [10]
   • The strongest effect was in patients whose tumours carried HPV, the human papillomavirus: 80% responded, against 33.3%. [10]
   • Inhibrx is now adding 50 more HPV-positive patients to chase a faster approval. [10]

   Why it matters

   STAT reports the benefit was largely driven by that subgroup, which could sharply limit who the drug is ever licensed for. [9]

   How it works

   Everyone in the trial already had high PD-L1, and the biggest effect came from a smaller group inside that - patients with HPV-positive tumours, at 80% against 33.3% - so the company is recruiting 50 more of exactly those.

   The idea, in story 04: Why a drug looks better the fewer people it is tested in →

   The truth

   • What is really going on On the same day, Inhibrx said its cancer drug's best result came from a small set of HPV-positive patients, and announced it would go and recruit 50 more of them. [9][10]
   • Who gains Inhibrx — Fifteen years of failure at the OX40 target ended with a result strong enough to justify a new 50-patient arm aimed at a faster approval. [9][10]
   • Who pays Head and neck cancer patients whose tumours do not carry HPV — Inhibrx's benefit was largely driven by the HPV-positive group, and the expansion arm takes only those patients. [9][10]
   • Nobody knows yet

     Whether Inhibrx's drug helps anyone whose tumour does not carry HPV.

     Responses were 48.3% against 26.5% overall, but STAT reports the benefit was largely driven by the HPV-positive subgroup, and the trial is being expanded only in that arm. [9][10]

   Good news

   • Pfizer, Genentech, GSK and AstraZeneca all tried and failed for fifteen years to turn a protein called OX40 into a cancer treatment. Inhibrx reported a randomised trial in which it worked, nearly doubling the share of head and neck cancer patients whose tumours shrank.
3. 07

   Spyre's third gut-disease win in five months

   What happened

   • Spyre Therapeutics said its antibody SPY003 met its goals in a mid-stage trial in 44 people with moderate to severe ulcerative colitis, an inflammatory disease of the large bowel. [11]
   • Disease activity fell by ten points on the Robarts index, a score doctors read off a sample of bowel lining under a microscope. [11]
   • It is the third of three Spyre antibodies to succeed since April, each blocking a different part of the immune signal that drives the disease. [11]
   • The company's plan is to combine them rather than pick one. [11]

   Why it matters

   Spyre's chief executive called it three potential best-in-class readouts in five months, and the plan is to use the three antibodies together rather than pick a winner. [11]

4. 08

   $116m for a children's brainstem tumour

   What happened

   • BrainChild Bio, spun out of Seattle Children's Hospital, raised $116 million to finish a pivotal trial of BCB-276 in diffuse intrinsic pontine glioma, a tumour that grows into the brainstem of children. [12][13]
   • Surgeons cannot reach it, no drug is approved for it, and radiation adds only a few months of life. [12]
   • There are 250 to 300 new cases a year, and the money came from the hospital itself and one local fund. [12]
   • Unlike every approved cell therapy, BrainChild's cells are given in repeated doses. [12]

   Why it matters

   Brain tumours are now the leading cause of cancer death in children, and a disease this small struggles to attract investors at all. [12]

   The truth

   • Who gains Seattle Children's Hospital and WRF Capital — They funded BrainChild Bio's $116m round, paying for a final trial in a children's brainstem tumour that has no approved drug at all. [12][13]

   Good news

   • Seattle Children's Hospital and one local fund put $116 million into finishing a trial for diffuse intrinsic pontine glioma, a tumour in the brainstem of children that surgeons cannot reach and for which no drug is approved.
5. 09

   $70m to switch off a gene in fat cells

   What happened

   • Moonwalk Biosciences raised $70 million and said its first drug will use RNA interference - switching a gene off before it can make its protein - aimed at fat tissue rather than at appetite. [14]
   • The company launched in 2024 to mine the epigenome, and changed course because RNA drugs reach patients faster, its chief executive said. [14]
   • The technology is licensed from Suzhou Siran Biotechnology in China. [14]
   • Human testing is planned by the end of 2027. [14]

   Why it matters

   GLP-1 drugs dominate the obesity market and work by suppressing appetite; Moonwalk's chief executive says acting on fat cells instead could avoid the gut side effects and help keep muscle. [14]

   The truth

   • Who gains Suzhou Siran Biotechnology — Moonwalk licensed its RNA-silencing technology and has now raised $70m to take it into obesity trials. [14]
6. 10

   Approval closed the door for pancreatic cancer patients

   What happened

   • When the US drug regulator approved Rasonque for pancreatic cancer last month, its maker Revolution Medicines immediately shut the programme that had been supplying the drug free before approval. [15]
   • Insurers take weeks or months to start paying for a newly approved medicine, so people already on the waiting list now have neither route. [15]
   • Kelsey Solano, a nurse in Washington state, was partway through that process for her husband Juan, whose pancreatic cancer has spread. [15]
   • One patient on Medicare, the US health scheme for older people, was refused cover in writing. [15]

   Why it matters

   The drug has been shown to nearly double how long patients live on average, and pancreatic cancer gives them very little time to wait for paperwork. [15]

   How it works

   Here the line is not drawn by biology at all - approval shut the free programme, and who gets the drug now depends on whether an insurer has finished its paperwork.

   The idea, in story 04: Why a drug looks better the fewer people it is tested in →

   The truth

   • Who gains Revolution Medicines — Its pancreatic cancer drug is now approved and sold rather than given away, and the free-access programme closed the day the approval landed. [15]
   • Who pays Pancreatic cancer patients on the waiting list — Kelsey Solano's husband Juan was partway through Revolution Medicines' free programme when approval shut it, and one patient on Medicare was refused cover in writing. [15]
   • Nobody knows yet

     How long the pancreatic cancer patients caught between the two systems will wait.

     Approval closed Revolution Medicines' free programme at once while insurers take weeks or months to start paying, and nobody has published how many people are stuck in between. [15]

7. 11

   Caffeine acts on a different switch than expected

   What happened

   • Researchers at Queen Mary University of London reported that caffeine slows cell ageing by switching on AMPK, a sensor that tells a cell it is running low on fuel. [16]
   • The same group had previously thought caffeine acted directly on TOR, the switch that decides when a cell grows. [16]
   • The work was done in fission yeast, a single-celled organism used as a stand-in for human cells, and published in the journal Microbial Cell. [16]
   • AMPK is the same system the diabetes drug metformin works through. [16]

   Why it matters

   Nothing here shows that coffee slows human ageing; what it gives is a specific target that exists in people too. [16]

   The truth

   • Nobody knows yet

     Whether caffeine does anything to human ageing.

     The Queen Mary work was done in fission yeast, a single-celled organism used as a stand-in for human cells, and there is no human result. [16]

8. 12

   Semaglutide linked to fewer psychiatric admissions

   What happened

   • Researchers led by Griffith University in Australia followed nearly 15,000 people in Sweden with bipolar disorder who had been prescribed GLP-1 drugs, the weight-loss and diabetes family, over 15 years. [17]
   • Periods when someone was taking semaglutide - sold as Ozempic and Wegovy - carried a 21% lower risk of being admitted to hospital for their mental illness than periods off it. [17]
   • Other GLP-1 drugs showed no such link. [17]
   • About 37 million people worldwide live with bipolar disorder. [17]

   Why it matters

   The study compares the same people at different times rather than assigning the drug at random, so it cannot show that semaglutide kept anyone out of hospital. [17]

   The truth

   • Nobody knows yet

     Whether semaglutide lowers psychiatric admissions or the people taking it were already doing better.

     The Swedish study compared periods on and off the drug in the same 15,000 people rather than assigning it at random, and other GLP-1 drugs showed no such link. [17]

9. 13

   Sudan's health system is close to collapse

   What happened

   • Medecins Sans Frontieres said Sudan's health system is close to collapse because the money that ran it has gone. [18]
   • The US government closed USAID, its aid agency and Sudan's largest donor, last year, and grants to many organisations it funded ran out in June. [18]
   • In central Darfur 45 health centres lost their funding in May alone, and admissions to facilities MSF supports rose 22.5% early this year. [18]
   • In one camp holding 18,400 displaced people, the maternity and surgical wards closed, so women needing a caesarean travel three hours. [18]

   Why it matters

   More than 33 million people in Sudan need medical help after three years of war, and international aid fell 21.3% worldwide in 2025. [18]

   The truth

   • Who pays People in Sudan — Forty-five health centres in central Darfur lost their funding in May alone, and in one camp of 18,400 people the maternity and surgical wards closed. [18]
10. 14

    A tropical mosquito bred in east London

    What happened

    • Aedes aegypti, the mosquito that carries dengue, chikungunya and Zika in hot countries, has been found breeding in homes in east London. [19]
    • It is the first time the species has bred in the UK and the fourth time it has been found there. [19]
    • The UK Health Security Agency said the risk is low, no human infections are linked to it, and the British climate is still too cold for the species to settle. [19]
    • It most likely arrived in vehicles, luggage or imported plants. [19]

    Why it matters

    The agency expects more arrivals as the climate warms, and a study this year found chikungunya can now be passed on in most of Europe. [19]

11. 15

    Congo's Ebola deaths pass 3,000, and schools reopen anyway

    What happened

    • More than 3,000 people have now died in the Democratic Republic of Congo's worst Ebola outbreak, spread across six provinces, with 48.6% of those infected dying. [21]
    • Schools reopened this week even in Ituri province's red zones, where transmission is active - more than 1,000 schools - after Congo's government dropped a plan to replace lessons with worksheets and radio broadcasts. [21]
    • Officials said pupils would otherwise fall too far behind. [21]
    • Separately, AFC/M23 rebels reported a new case in North Kivu, two months after declaring their territory free of the virus. [20]

    Why it matters

    The new patient is a 21-year-old man who had arrived from government-held Butembo, and nobody yet knows where he was infected. [20]

    The truth

    • Nobody knows yet

      Where the 21-year-old man in North Kivu caught Ebola.

      He arrived from Butembo in government-held territory, and the rebel administration that reported the case said it was not clear where he was infected. [20]

12. 16

    An Alzheimer's trial in people with no symptoms

    What happened

    • An NHS trust in Surrey has become the first site in Europe to enrol patients in a trial of trontinemab, a drug meant to delay or prevent Alzheimer's before any symptoms appear. [22]
    • The trial is recruiting 1,600 healthy adults aged 55 to 80 worldwide, and blood tests are used to pick the people most likely to benefit. [22]
    • The drug clears beta amyloid, a sticky protein that builds up between brain cells in the disease. [22]
    • Almost all Alzheimer's research so far has treated people who already have symptoms. [22]

    Why it matters

    About 982,000 people in the UK are living with dementia, of which Alzheimer's is 60% to 80% of cases. [22]

    How it works

    Blood tests pick the 1,600 healthy volunteers most likely to benefit before anyone is enrolled, so whatever the trial finds will be about them and not about everyone at risk.

    The idea, in story 04: Why a drug looks better the fewer people it is tested in →
13. 17

    Three genes in a blood sample to find TB

    What happened

    • Researchers testing samples from 175 adults narrowed a 12-gene blood signature down to three genes that separate active tuberculosis from other lung diseases. [23]
    • In a first clinical check on 56 people, the signatures were right for more than 77% of TB patients and more than 94% of those without it. [23]
    • Almost every TB test in use needs a sputum sample coughed up from the lungs, which many patients cannot produce. [23]
    • The authors say a blood method could be cheap enough for primary care clinics. [23]

    Why it matters

    Tuberculosis is among the ten biggest infectious causes of death worldwide, and this has only been checked in small groups so far. [23]

14. 18

    Gut infections leave immune memory at the brain's border

    What happened

    • Giving mice gut bacteria and parasites changed the immune cells sitting in the dura mater, the tough outer wrapping of the brain. [24]
    • Immune cells trained in the gut travelled to that border and settled there for the long term, and when the same germ was later injected into the blood they reacted straight away. [24]
    • The dura's blood vessels are leaky, which is how germs carried in blood can reach the brain. [24]
    • The work was published in Nature Neuroscience. [24]

    Why it matters

    It is a mouse result, and what it means for human brain inflammation is not known. [24]

15. 19

    GBP500,000 each for two UK biotechs

    What happened

    • Medicines Discovery Catapult, a UK research organisation, put GBP500,000 each into Flip Therapeutics and Nascent Therapeutics, using money from Innovate UK, the UK government's innovation agency. [25]
    • Flip is trying to reprogramme immune T cells inside the body so they stop driving inflammation and start calming it, in diseases such as psoriasis and alopecia areata. [25]
    • Nascent, a spinout from the University of East Anglia, works on the cell's power packs in ageing-related disease. [25]
    • Both also get access to the Catapult's laboratories and advisers. [25]

    Why it matters

    About one in ten people in the UK has a chronic autoimmune or inflammatory disease of the kind Flip is aiming at. [25]

16. 20

    A duck cell line instead of chickens' eggs

    What happened

    • ProBioGen widened its licence to Transgene for AGE1.CR.pIX, a cell line grown from a duck embryo and used to make vaccines from viruses. [26]
    • Much viral vaccine is still grown in hens' eggs, which is slow and depends on a steady egg supply; a cell line grows in a defined liquid with no animal ingredients. [26]
    • Every batch then starts from the same fully characterised stock. [26]
    • Transgene makes personalised cancer vaccines, including one for head and neck cancer. [26]

    Why it matters

    This is manufacturing rather than science, and it is the part that decides whether a personalised vaccine can be made for more than a handful of people. [26]

17. 21

    More exercise, worse results, in one genetic group

    What happened

    • Researchers followed 682 adults aged 65 and over through annual health checks. [27]
    • Women who carry two copies of one version of the ACTN3 gene did worse at standing up from a chair five times if their activity went above a certain level. [27]
    • That version leaves them without a muscle protein called alpha-actinin-3. [27]
    • Below that level, more activity went with better results, making a U-shaped curve. [27]
    • The gene by itself showed no link to muscle loss. [27]

    Why it matters

    The authors call the exact threshold exploratory and say it needs checking in other groups before anyone reads advice into it. [27]