Biotech, in brief

Telix Pharmaceuticals, an Australian company that sells radioactive drugs for scanning cancer, is buying its rival ITM in a deal worth at least $1.65 billion. [1][2][3]

ITM Isotope Technologies Munich is a privately owned German company. [2] Its medicines carry a radioactive atom straight to cancer cells. [2] The radioactive metal is what kills the cancer. [1]

Last month the FDA, the US drug regulator, turned down ITM's lead drug, ITM-11. [1] It cited how the drug is made and problems found at a factory run by another company. [1] ITM plans to file again. [1]

ITM-11 is for neuroendocrine tumours, cancers that start in hormone-making cells, here in the gut and pancreas. [1] It would compete directly with Lutathera, a drug Novartis already sells. [3]

Telix could pay up to $700 million more on top. [1] Up to $250 million depends on approvals, and up to $450 million on how well ITM-11 sells. [1]

The first extra $100 million is paid only if the FDA approves ITM-11 for one gut tumour by the end of next year. [1]

The deal cannot close until ITM files with the FDA again, or the two companies agree a way forward. [1]

Telix's chief executive, Christian Behrenbruch, said a rejection is not the end of the road when the drug itself is strong. [1] He pointed to Telix's own brain-scan drug Pixclara as proof. [1]

ITM also makes lutetium-177, the radioactive metal inside Pluvicto, Novartis's big-selling prostate cancer drug. [1] Other companies make it too. [1]

Novartis leads this field, with two such drugs launched in 2018 and 2022. [2] ITM sold $273 million of products in 2025, and Telix expects the joined company to take in $1.3 billion this year. [1]

On 22 September Novartis licensed an early-stage radioactive drug from BoomRay, a company in Suzhou, China, for up to $900 million. [3]

Analysts at the bank William Blair wrote that they were extremely bullish on the joined company. [1]

Why it matters. The next step is ITM filing again with the FDA: the deal closing and the first $100 million both depend on the regulator accepting the fixed factory. [1]

A panel of outside experts that advises the FDA meets on 23 September to review Grail's Galleri. [4] It is a blood test for adults aged 50 and over that looks for signs of many kinds of cancer at once. [4] In Grail's two big studies, it flagged 32% to 35% of the cancers diagnosed within a year. [4] Of the people it flagged, 66% to 77% turned out to have cancer. [4] About one healthy person in 500 got a false alarm. [4] FDA documents posted before the meeting suggest a favourable view, with one oddity, STAT reported. [5]

Why it matters. Galleri is already sold through laboratories, and FDA approval would be expected to help it get paid for by Medicare and other insurers. [5] Grail's own document says people should keep having their usual screening, because the test misses most cancers. [4]

Amgen said its drug dazodalibep improved Sjogren's disease more than a dummy drug in a final-stage trial of 651 people. [6] Sjogren's is a disease in which the immune system attacks the glands that make tears and saliva, and it can affect the whole body. [6] There is no medicine approved in the US specifically for it. [6] The trial measured overall disease activity after 48 weeks, and Amgen has not yet given the scores. [6] A second final-stage trial, in people with milder disease, is still running. [6]

Why it matters. Novartis and Sanofi have each dropped a Sjogren's drug in recent years, and Astellas stopped one in April. [6] Patients now mostly get general anti-inflammatory drugs to ease the symptoms. [6]

The US Treasury is drafting rules that would probably let US drugmakers keep licensing new medicines from Chinese companies, Reuters reported, citing unnamed sources. [11] Deals on germs or biotechnology that could be made into weapons would be excluded. [11] Licensing deals for Chinese biotech drugs were worth $115 billion last year, according to the research firm GlobalData. [11] The rules come under a 2025 US law that screens investment in sensitive technology in certain countries, and a drug industry group is lobbying over them. [5] They are not final, and could change if President Trump weighs in. [11]

Why it matters. Pfizer and other big drugmakers argued that limits would cut US companies off from a fast-growing source of new medicines. [11] Some smaller biotech companies and two Michigan members of Congress want the deals restricted, saying they help China lead in biotechnology. [11]

Ulefnersen, a drug for a rare form of ALS, slowed the disease in a final-stage trial of 89 people, its makers Otsuka and Ionis said. [8] ALS destroys the nerve cells that control muscles. [8] This form, caused by faults in a gene called FUS, is about 0.6% of cases and can kill younger patients within two years. [8] The drug lowers the amount of the harmful FUS protein, and patients did better than on a dummy drug on function and survival after 72 weeks. [8] Otsuka, a Japanese drugmaker, will take it to regulators. [8]

Why it matters. The market is tiny: analysts at Guggenheim expect Ionis to earn only about $14 million a year from it by 2030. [8] The drug was designed for ALS with one genetic cause, not for ALS in general. [8]

Celldex said its antibody barzolvolimab beat a dummy injection in two final-stage trials with 1,939 people who have chronic hives with no known cause. [7] The drug blocks a signal that mast cells, the immune cells behind hives and swelling, need to survive. [7] In people not helped by the standard injection Xolair, 55% on the lower dose were fully clear at 12 weeks in one trial, against 9% on a dummy. [7] In the other trial the figures were 42% against 15%. [7] Two patients had a severe allergic reaction, and 16% left the trials, mostly by choice. [7]

Why it matters. The same drug failed a trial in a different itchy skin disease in July, so this is a recovery for the company. [7] Celldex plans to ask the FDA for approval in 2027. [7]

Novo, the Danish maker of Ozempic and Wegovy, laid out its plans to 2030 to investors on Monday, and its shares fell 8%. [9] The US patent on semaglutide, the ingredient in both drugs, runs out in 2032, which its chief executive called the elephant in the room. [9][10] Novo expects its sales to grow at about the same rate as other big drugmakers until 2030. [10] It aims for more than $23 billion of sales from drugs in its pipeline by 2035, allowing for some to fail, and plans to buy companies as well as invent. [9]

Why it matters. Novo has lost ground to Eli Lilly, whose drug Zepbound beat Wegovy for weight loss in a head-to-head trial. [10] The same day Novo said its newer drug CagriSema beat Zepbound in another trial, though against a lower dose of Lilly's drug. [10]

Iambic Therapeutics, which designs cancer drugs with AI and is backed by the chipmaker Nvidia, filed to list on the US Nasdaq stock market. [12] It has raised about $461.8 million from investors since 2019. [12] Its most advanced drug, for tumours driven by a gene called HER2, is in a first safety trial. [13] Retension Pharmaceuticals and TRex Bio filed on Friday, and ADARx began pitching its share sale to investors. [12] On the same day, AbbVie signed a research deal with Iambic. [12][15]

Why it matters. One analyst said this year's rush of listings has been helped by big drugmakers buying up young companies. [12] Eli Lilly has already said it may buy shares in TRex, whose drug it licensed in 2023. [14]

Roche said its weekly injection enicepatide cut blood sugar in a mid-stage trial in people with type 2 diabetes. [16] HbA1c, a measure of average blood sugar, fell 2.65 points on the top dose after 48 weeks. [16] Weight fell by 15.5%, with no sign yet of levelling off. [16] Roche got the drug by buying Carmot Therapeutics for $2.7 billion, and 2% of patients stopped because of side effects. [16]

Why it matters. Lilly's Mounjaro lowered HbA1c by up to 1.8 points in its own trial, but the trials ran for different lengths, so the numbers cannot be compared directly. [16] Roche wants to become one of the three biggest companies in obesity drugs. [16]

Vertex said its pill inaxaplin cut protein leaking into the urine in a mid-stage trial for kidney disease driven by a gene called APOL1. [17] In 22 people with milder leaking, it fell by 42.7% on average after 13 weeks. [17] In 17 people who also had type 2 diabetes, it fell by 17.3%. [17] A larger trial has finished enrolling, and early results are due next year. [17]

Why it matters. If the larger trial is positive, Vertex could ask the FDA for a faster, conditional approval. [17] Vertex says the group in that larger trial covers about 150,000 people in the US and Europe. [17]

Viking Therapeutics said people kept most of the weight they lost on its drug VK2735 when injections were spaced out. [18] After 21 weeks of weekly doses, people had lost 16% to 19% of their weight. [18] Moved to one injection every two weeks, they kept 90% of that loss on average; moved to monthly injections, 85%. [18] Viking's shares rose sharply on the news. [18]

Why it matters. Weight-loss drugs have to be taken for good to keep weight off, and many people stop taking them. [18] Fewer injections could make staying on the drug easier. [18]

Johnson & Johnson said Caplyta reduced manic symptoms in people with bipolar I disorder after three weeks in a final-stage trial, compared with a dummy drug. [15] Bipolar I is a mental illness defined by manic episodes, sudden swings in mood and behaviour. [15] Some improvements showed up within three days. [15] Caplyta is already approved for schizophrenia, bipolar depression and major depression, but not for mania. [15]

Why it matters. J&J paid nearly $15 billion last year for the company that made Caplyta. [15] A second final-stage trial in mania is still running. [15]

The FDA issued a final rule saying the safety of drugs and biological medicines can be tested with methods other than animals where appropriate. [19] The rule replaces the words 'animal tests' with 'nonclinical tests' in its regulations. [19] It is the second package of this kind announced this year by the US health department. [19] Health secretary Robert F. Kennedy Jr. said it moves research toward human biology. [19]

Why it matters. Kennedy said researchers have been kept dependent on animal testing when better tools exist. [19] The department says it is paying to develop testing methods based on human biology. [19]

The FDA accepted a proposal to use a wearable movement sensor to measure foot drop in people with a nerve disease called CMT1A. [20] Foot drop means being unable to lift the front of the foot, which disturbs walking. [20] The sensor records it during a six-minute walk at clinic visits. [20] It is the first such digital tool for a rare nerve disease accepted into this FDA programme. [20]

Why it matters. CMT1A gets slowly worse over years, so trials need a way to measure small changes. [20] The sensor is meant to track that change as a trial result. [20]

Heidi Overton, nominated to lead the FDA, goes before the US Senate health committee on Thursday for her confirmation hearing. [25] Two former FDA commissioners, Scott Gottlieb and Mark McClellan, wrote in support of her. [25] They noted that more early human trials now begin in China than in the US. [25] In February, the FDA's leaders at the time argued that one large final trial, instead of two, should be the usual standard for approval. [26]

Why it matters. The FDA has since published draft guidance on the one-trial idea, which is not yet final. [26] How much evidence a company must gather before a drug reaches patients depends on it. [26]

The FDA approved Cordis's Selution balloon for heart arteries that narrow again inside a stent, a metal tube put in earlier to hold them open. [21] The balloon coats the artery wall with sirolimus, a drug usually given to stop the body rejecting a transplanted organ. [21] It leaves no new metal behind. [21] About 10% of the roughly one million US artery-opening procedures a year treat this re-narrowing. [21]

Why it matters. Doctors often have to put a second stent inside the first one when this happens. [21] The approval rested on a study of more than 400 patients, and a 960-patient trial in small arteries has finished enrolling. [21]

Boehringer Ingelheim, a German drugmaker, signed a deal worth up to $1 billion with Envisagenics, a New York company. [22] Envisagenics uses AI to screen more than 14 million ways cells edit gene messages. [22] It looks for protein versions made only by tumours, which a drug could hit while sparing healthy tissue. [22] Its chief executive said the money will take its own brain drug into human trials. [22]

Why it matters. Neither company gave the split between money paid now and money paid later. [22] Boehringer can choose targets and license them for itself alone. [22]

A final-stage trial registered in China tested aumolertinib, a pill that blocks a faulty EGFR gene, with radiotherapy for lung cancer that cannot be operated on. [24] It was compared with chemotherapy plus radiotherapy. [24] Half the patients on the pill went 34 months before their cancer grew, against 7.4 months on chemotherapy. [24] The trial stopped early with only 43 of a planned 98 patients, partly because it enrolled slowly. [24]

Why it matters. Chemotherapy with radiation has been the standard for this stage of lung cancer. [24] A trial this small cannot yet show a clear difference in how long people live. [24]

Lexeo Therapeutics bought four drugs for Friedreich's ataxia from Mantle for $8 million. [23] Friedreich's ataxia is an inherited disease that destroys balance and coordination and affects about one American in 50,000. [23] About 80% of patients develop heart muscle disease. [23] The lead drug improved a 93-point disease rating by 6 points in 11 patients in early testing. [23]

Why it matters. Lexeo already has a gene therapy for the same disease in a late-stage trial. [23] It has money to run until 2028. [23]

Scientists at Stanford University built a 'virtual biotech' of up to 37,000 AI agents that split drug-discovery work between them, reported in the journal Science. [27] It studied more than 55,000 past clinical trials. [27] It found that drugs aimed at proteins active in only certain cell types were nearly 50% likelier to reach the market. [27] It also proposed an antibody carrying a cancer drug to a protein called CD276 in lung tumours. [27]

Why it matters. None of its predictions has been tested in a lab, let alone a trial, other scientists pointed out. [27] It is a test of whether AI can speed up the early, slow part of making a drug. [27]

Researchers mapped two serotonin receptors, the docking points that psychedelic drugs act on, in the journal Nature Communications. [28] One, called 2A, is behind the effects doctors want; the other, 2B, causes side effects. [28] Using that map, they designed a compound that switches on 2A and blocks 2B. [28] Images of the molecule in place confirmed how it works. [28]

Why it matters. The authors say avoiding the 2B receptor is critical for safe medicines inspired by psychedelics. [28] This was laboratory work, not a test in people. [28]

Researchers tested blood for 28 sugar-coated protein fragments in 550 people, half with stomach or bowel cancer and half healthy. [29] A computer model using them picked out 77.3% of the cancers and wrongly flagged 18.4% of the healthy people. [29] It did less well on early cancers. [29]

Why it matters. About one healthy person in five was wrongly flagged, against about one in 500 for Galleri. [4][29] It compared known patients with healthy people, which is not the same as screening the general public. [29]

The FDA cleared the latest version of Aptitude's Metrix, a small molecular test for COVID. [30] It was paid for partly by BARDA, the US health department's fund for emergency medical products. [30] It arrives before the US ends emergency approvals for COVID tests on 26 December. [30]

Why it matters. Tests already shipped under the emergency rules can still be used until they expire, the company said. [30] Aptitude is also seeking clearance for a combined COVID and flu test. [30]