Biotech, in brief

The US Food and Drug Administration, the country's drug regulator, held a public meeting this week on how psychedelic medicines should be given to patients if it approves one. [1][2]

Almost every psychedelic drug is still a Schedule I controlled substance in the United States, the same class as heroin, meaning it is officially treated as having no medical use. [1]

Compass Pathways, a British company, is filing for approval of a lab-made version of psilocybin, the active chemical in magic mushrooms, for depression that has not responded to other treatments. [1][2]

In its late-stage trial of 258 people, patients who kept taking doses through a year fell an average of 13 points on a standard depression scale; about 30% were in remission after an open-label dose. [4]

Definium Therapeutics said on Monday that its LSD-based drug beat a dummy pill in a second late-stage anxiety trial. Fierce Biotech called it the company's third late-stage win in twelve weeks. [3][2]

Patients on the 100-microgram dose fell 9.8 points on the Hamilton anxiety scale against 4.7 on the dummy, and some changed by the second day after dosing. [3]

The argument at the meeting was about who may supervise a dosing session. Compass and Definium have moved away from the therapist model to a monitor who need not be a therapist. [2]

A guidance the regulator issued in July says that if the lead monitor is not a doctor, an on-call doctor must be able to reach the patient within 15 minutes. [2]

Kevin Sabet, a former White House drug-policy adviser, told the meeting that psychedelics should earn approval "through science and trials, not politics or podcasters". [2]

Susan Taymor Sagy told the meeting that her 21-year-old daughter Shira died after taking a psychedelic, and asked for health screening before anyone is treated. [2]

US President Donald Trump signed an order on 18 April telling the regulator to speed up psychedelic reviews, after the podcaster Joe Rogan visited the Oval Office. Three developers were then given vouchers that shorten the agency's review. [2]

Australia already approves MDMA for post-traumatic stress and psilocybin for hard-to-treat depression. Switzerland allows limited psilocybin therapy, and Mexico allows ibogaine given by medical staff. [1]

In 2024 the US regulator refused an MDMA treatment for post-traumatic stress from Lykos Therapeutics, citing problems with the study methods, and asked for another late-stage trial. [1]

Why it matters. Watch whether the US regulator writes down who may supervise a dosing session, because that decision, more than any trial result, sets how many people could reach the treatment if it is approved.

Peter Sands, who runs the Global Fund to Fight AIDS, Tuberculosis and Malaria, said global aid cuts last year hit prevention work harder than treatment. [5] The fund pays for programmes in more than 100 countries and published its annual report on Wednesday. [5] Treatment held up: 1.3 million more people started HIV medicines, reaching 26.9 million, and 7.4 million were treated for tuberculosis. [5] But 300,000 fewer people got HIV prevention pills, down to 1.1 million. [5]

Why it matters. Countries protected the patients already on their books and cut the outreach that keeps new people off them. Sands said the effect on infections and deaths will not be measurable for several months. [5]

United Nations officials said on Tuesday that the response to Ebola in eastern Congo is showing results, while the epidemic remains massive. [6] Congo's health minister said daily new cases have fallen from about 120 at the August peak to around 80, and ten of the 62 affected health zones have gone more than 22 days with no new case. [6] Cases passed 7,000 on 11 September as the virus reached another province. [7] The total now stands at 7,258 confirmed cases and 3,510 deaths across seven provinces. [6]

Why it matters. Jean-Jacques Muyembe, who runs Congo's biomedical research institute, said the indicators in Ituri province have turned green. The World Health Organization's Olivier le Polain said an outbreak can improve and then pick up again, so nobody will call a peak. [6]

Gilead Sciences said on Tuesday it had agreed with the Pan American Health Organization to widen access to lenacapavir, an HIV prevention injection given twice a year. [8] Fourteen countries including Brazil, Mexico and Argentina can now buy it through that organisation's regional purchasing system. [8] None of them were covered by the royalty-free licences Gilead granted six generic makers in 2024 for poorer countries. [8] New HIV infections in the region rose 13% between 2010 and 2024. [8]

Why it matters. The drug already existed and was already cheap in the poorest countries. What these fourteen lacked was a way to buy it, and Gilead says it is also discussing making the drug in Brazil. [8]

The US drug regulator opened applications on Tuesday for a pilot meant to shorten the time before a new medicine can first be tested in people. [9][10] It will pick eight to ten pairings of a drug company with a research institution and review their paperwork in rolling pieces rather than all at once. [10] Getting a drug into trials in the United States typically takes up to two years; in Australia a trial can start less than 70 days after the protocol is filed. [10] Applications close on 30 October. [10]

Why it matters. Trials run in China rose from under 1,000 a year in 2010 to more than 5,000 in 2024, while the US number stayed near 3,500. [10] Mike Davis, who runs the regulator's drug evaluation centre, said the country risks losing early access to new treatments. [10]

Taiho and Cullinan presented full results on Monday in Seoul for zipalertinib, a pill for lung cancer driven by a particular fault in a gene called EGFR. [11] In 279 previously untreated patients, the pill with chemotherapy held the cancer back for a median of 14.5 months against 8.5 months on chemotherapy alone. [11] Tumours shrank in 65% of patients against 40.3%. [11] Survival figures are not yet mature but show a 28% lower risk of dying. [11] Serious side effects hit 87.1% of the combination group against 54.4%. [11]

Why it matters. About a third of these lung cancers carry an EGFR fault, and about 4% of advanced cases carry the insertion this pill is built for. [11] Cullinan receives $100m if it is approved as a first treatment. [11]

BioNTech said on Monday that gotistobart produced a median survival of 18.5 months in the first stage of a late-stage trial. [12] The drug is an antibody aimed at CTLA-4, a brake on the immune system, and the patients had squamous lung cancer that had come back. [12] Patients on docetaxel, the chemotherapy that has been standard for these patients for years, lived a median of 10 months. [12] Gilead's Trodelvy reached 11.1 months in this setting and a rival from AstraZeneca and Daiichi Sankyo 12.9. [12]

Why it matters. Serious side effects hit 44.4% of the gotistobart group against 29.3% on chemotherapy, and 15.6% stopped treatment because of them. [12] The second stage of the trial has to repeat the survival result, and BioNTech expects its first look around the end of the year. [12]

Sionna Therapeutics said on Monday it will lay off nearly half its workforce, five weeks after its cystic fibrosis drug missed its goal. [13][14] The drug was meant to steady a protein that folds wrongly in the commonest form of the disease, and it failed to cut sweat chloride, the standard measure, by enough. [13] Investors drove the shares down 90%. [13] The company now says an after-the-fact analysis found factors that muddied the trial, and it will move a different pair of its drugs into a mid-stage study. [13][14]

Why it matters. The job cuts stretch Sionna's money into the second half of 2029. [14] A Stifel analyst said the company's explanation seems reasonable but is hard to have confidence in. [14]

The US regulator approved Bexlutry on Tuesday, a copy of Novartis's Lutathera made by Curium. [14] Both carry radioactive lutetium to a protein called SSTR that sits on the surface of neuroendocrine tumours, so the radiation lands on the tumour rather than the whole body. [14] It is the first copy of any radioligand therapy to be cleared. [14] The approval covers tumours of the foregut, midgut and hindgut, and came after Curium won a patent case against Novartis in June. [14]

Why it matters. Until this week every patient given this kind of treatment received Novartis's version. Curium had to win the patent case before a copy could be cleared at all. [14]

The US regulator approved Pixclara, made by Telix, for imaging gliomas, the tumours behind about 80% of malignant brain cancers and around 24,000 US cases a year. [15] Used with a PET scanner, it helps tell a returning or growing tumour apart from changes left behind by its treatment. [15] The scan is already recommended in international guidelines, and no drug had been approved in the United States to carry it out. [15] The regulator rejected the same application last April, asking for more confirmatory evidence. [15]

Why it matters. Doctors there were following a guideline with no approved product to do it with. Telix's precision medicine head said the drug gives doctors more certainty in diagnosis and in planning treatment. [15]

Vera Therapeutics said on Tuesday that Trutakna steadied kidney function and cut the risk of kidney-related death and other complications, in final two-year results from a late-stage trial. [16] The drug treats IgA nephropathy, a long-running disease in which faulty immune antibodies build up in the kidney's tiny filters and slowly destroy them, sometimes ending in dialysis. [16] The US regulator gave the drug an early, conditional approval in July. [16] Vera says it will ask for full approval in the last quarter of this year. [16]

Why it matters. An early approval is granted before the long-term evidence exists, on the understanding that it will arrive. This is that evidence arriving, two years on. [16]

Contineum Therapeutics said on Monday that JNJ-5120, a drug it is developing with Johnson & Johnson, failed a mid-stage trial in major depression. [17][14] 107 patients were randomly given one of two doses or a dummy pill, and at day five those on the drug scored no better on a depression scale. [17] Johnson & Johnson paid $50m up front for the drug in 2023 and promised up to $1bn more if it worked. [17] Ten months ago the same drug failed a trial in multiple sclerosis. [17]

Why it matters. Contineum's shares fell only 3.5%, because analysts say most of the company's value sits in a different drug for lung scarring. [17] Johnson & Johnson says it is still reading the results before deciding what to do next. [17]

Axoltis said on Tuesday that NX210c, a synthetic piece of a protein, did not significantly change neurofilament light chain in 82 people with ALS in France. [18] That protein leaks into the blood when nerve cells are injured, and it was the trial's main measure, so the trial is a miss. [18] ALS, also called motor neurone disease, destroys the nerve cells that tell muscles to move. [18] Axoltis points instead to improvements on a movement rating scale and a fall in a marker of a leaky barrier around the brain. [18]

Why it matters. The company says it will take the drug into the next stage anyway, and named Alzheimer's, Parkinson's and multiple sclerosis as further targets. [18] Those readings were not what the trial was set up to test. [18]

Two antibody-drug conjugates aimed at a protein called B7-H3 were shown at the lung cancer meeting in Seoul, both discovered in China and licensed to Western drugmakers. [19] Roche's cut the risk of death by 54% against the chemotherapy topotecan in a China-only trial, and GSK's cut it by 54% in a separate China-only trial. [19] Both companies are now running global trials to see whether the results repeat. [19] In June, Gilead and Merck stopped a multi-country lung cancer trial after deciding a survival benefit was unlikely. [19]

Why it matters. Detailed findings from that halted trial were presented at the same meeting. [19] Whether a result found in one country's patients holds elsewhere is the live question over a wave of licensed Chinese drugs. [19]

Arrowhead Pharmaceuticals said on Tuesday that a single dose of Aro-Dimer-Pa lowered two proteins in the blood at once in an early trial: PCSK9 by 72% and APOC3 by 88%. [14] Both proteins are tied to heart disease. [14] The drug uses RNA interference, which silences a gene so its protein is never made. [14] Arrowhead is aiming it at mixed hyperlipidemia, where harmful cholesterol and other blood fats are high at the same time. [14]

Why it matters. This is an early-stage result in a small group, with nothing yet on heart attacks or strokes. A Cantor Fitzgerald analyst called the drug highly differentiated and the market very large. [14]

Corbus Pharmaceuticals said on Monday that CRB-913, a daily pill, helped people with obesity lose up to 5 percentage points more weight than a dummy pill over 12 weeks in an early trial. [14] The drug blocks a cannabinoid receptor that helps regulate appetite, a route that has produced disappointing results before. [14] Between 1.5% and almost 10% of patients, depending on the dose, stopped taking it because of psychiatric side effects. [14] Corbus plans a mid-stage trial next year. [14]

Why it matters. Psychiatric side effects are the known problem with this class of drug. A Jefferies analyst called them the key debate around this one. [14]

The US National Science Foundation, which has funded basic research since 1950, told researchers it is moving to what it calls a portfolio approach to grants. [20] Alongside proposals driven by scientists' own questions, it will fund four named topics: the Future of Intelligence, Frontier of Quantum Phenomena, Advanced Materials and Matter, and Foundations of Biological Design. [20] The four were chosen by the office that advises the US president on science. [20] The letter does not say what they will cost or where the money comes from. [20]

Why it matters. Agency staff, who spoke to Nature anonymously for fear of retaliation, expect the money to come out of curiosity-led research. [20] Some is likely to come from $1bn the agency has been holding centrally instead of awarding. [20]

US Republican lawmakers added $50bn over five years for rural health care while cutting $1 trillion over ten years from Medicaid, the programme that pays for care for people on low incomes. [21] Hospital leaders expected cash to keep their doors open; the fund as built pays for new programmes instead, with spending caps. [21] Announced projects include drone drug deliveries and AI image reading in Alaska, remote monitoring in West Virginia and equipment in North Dakota. [21] Some hospitals have told their states they are not interested. [21]

Why it matters. The fund is 5% of the money coming out of Medicaid. [21] Randy Clark, who leads three hospitals in Maine's Northern Light Health, said it is hard to think about transformation when you are thinking about survival. [21]

The US regulator added a warning on Tuesday to the label of Kybella, an injection approved in 2015 to reduce fat under the chin. [23] Since then the agency has found 129 reports of harm when it was injected elsewhere on the body, which the label has never covered, including blurred or reduced vision, muscle and nerve damage, numbness and difficulty walking. [23] It also found 117 reports of lumps that did not go away, lasting about five months on average, including after injections given as approved. [23]

Why it matters. The drug is only cleared for use under the chin, and people have been getting it elsewhere. The regulator also urges people to avoid fat-dissolving products it has not approved at all. [23]

David Samra of Artisan Partners, a major Novartis shareholder, called for a shake-up of the Swiss drugmaker's board, STAT reported, citing Reuters. [24] Novartis shares had just suffered a record fall after two trial setbacks. [24] A drug for a muscle-wasting disease, acquired in its $12bn takeover of Avidity, failed a late-stage study, and results for its heart drug pelacarsen disappointed investors. [24] Samra said successive chairmen had failed the company on acquisitions and that it must change how it oversees deals. [24]

Why it matters. The complaint is about how the company buys other companies rather than about its science. Both of the failed drugs came in through that buying. [24]

Researchers at UCLA grew cancer-hunting immune cells from blood stem cells taken from donated umbilical cord blood. [26] Engineered T cell treatments are normally made from each patient's own cells, which takes weeks and costs six figures, while cells from another person can attack the patient's healthy tissue. [26] Starting from stem cells means the mature cells all carry one chosen receptor and none of the random ones that cause that attack. [26] In mice with ovarian cancer and melanoma, one dose controlled tumours and extended survival. [26]

Why it matters. The cells carry a second way to spot a tumour, so they can still kill a cancer cell that has hidden the marker they were built to find. [26] This was done in mice, and no person has been given these cells. [26]

Doctors in Italy compared 29 adults with high-risk acute myeloid leukaemia who were given azacitidine after a stem cell transplant from a donor with 21 similar patients who were given nothing. [27] The cancer returning is the main reason such transplants fail. [27] Two years on, 69% of the treated group were alive and 60% were alive with no return of the cancer. [27] Patients had a median of 12 monthly cycles, and 45% stopped early. [27]

Why it matters. This looks back at patients already treated rather than assigning them at random, and an earlier randomised trial of the same drug found no survival benefit overall. [27] It is a signal worth testing, not a result. [27]

Scientists at the US World Trade Center Health Program published a review in JAMA of what 25 years of follow-up has shown after the 2001 attacks. [28] More than 154,000 responders and survivors are enrolled. [28] Many have several conditions at once, including cancer, lung disease, reflux and mental illness, and some appeared years after the exposure. [28] Responders with cancer in the programme died at rates 26% to 64% lower than the New York State population across several cancer types. [28]

Why it matters. The authors say the harms of a large exposure can unfold over decades, so monitoring has to last that long. [28] They also say that care with no cost to the patient appears to improve survival. [28]

The US drug regulator permanently filled four leadership roles this month. [25] Michael Davis, a psychiatrist, becomes director of the Center for Drug Evaluation and Research, which decides on most medicines. [25] Karim Mikhail, formerly of Merck, becomes director of the Center for Biologics Evaluation and Research, which covers vaccines and cell therapies. [25] Bret Koplow, a lawyer at the agency since 2011, takes the tobacco products centre. [25] Heidi Overton was nominated to lead the whole agency late last month. [25]

Why it matters. Davis and Mikhail are the two officials who announced this week's pilot to speed up first human trials. [9][10] Three of the four had already been doing the jobs in an acting capacity. [25]

Frazier Life Sciences, based in Palo Alto, added more than $1.1bn to a fund that buys shares in small and mid-sized listed biotech companies, taking that fund to $2.8bn. [22] The firm now manages more than $6.9bn across its public and venture funds. [22] Several of its largest holdings have been bought by big drugmakers, including Verona Pharma, taken by Merck for $10bn in 2025, and Alpine Immune, bought by Vertex for $4.9bn in 2024. [22]

Why it matters. A fund holding dozens of companies survives the failures that end a single one. Sionna and Contineum each had a drug fail this week. [13][17]

Novo Nordisk said on Monday it will use Novo as its everyday name, keeping Novo Nordisk as the legal one. [14] The Danish company makes Ozempic and Wegovy. [14] It has lost more than two-thirds of its market value since mid-2024 and has given up its lead in obesity treatment to Eli Lilly. [14] The change comes a week before an investor day at which it will set out an updated strategy. [14]

Why it matters. The company is presenting that new strategy to investors next week. The shorter name is the part of it announced first. [14]