Biotech, in brief

Ten women aged 51 to 72 with advanced osteoporosis, the bone-thinning that makes fractures easy, got one infusion of their own bone-marrow cells in a trial in Murcia, Spain. [1]

Before the infusion the women had been breaking bones in their spines, hips and arms every year or two on average, often after little more than a stumble. [1]

Afterwards, fractures from small knocks fell to a rate closer to once a decade, according to the trial report published this week in the journal Cell. [1]

The cells are mesenchymal stromal cells, a kind of stem-like cell from bone marrow that can grow into bone. Injected into the blood, they normally cannot get into bone at all. [1]

In 2008 the same team found that adding a sugar called fucose to the cells' surface changed that in mice. The sugar makes the cells stick to blood-vessel walls, slow down, and squeeze into the marrow. [1]

The added sugar wears off in about two days. No previous trial in any disease had deliberately changed cells to improve where they travel. [1]

Ajit Varki, a physician-scientist at the University of California, San Diego, who was not involved, called it remarkable: almost complete protection sustained for several years, with no side effects. [1]

The trial was small, had no comparison group, and most of the women were taking ordinary osteoporosis drugs before and during it, so the cells' own effect cannot be separated out. [1]

The researchers never tracked the cells inside the women, so nobody knows how many actually reached bone. [1]

Because the cells come from each patient's own marrow, every dose is its own batch. If a batch fails it cannot be replaced from stock, which is what makes this kind of therapy slow and costly to make. [2]

Osteoporosis affects around 200 million women worldwide, mostly after the menopause. The trial began in 2015 and took several years of manufacturing work before it could start. [1]

Why it matters. Watch whether anyone runs a controlled trial of these sugar-coated cells; without one, the drop in fractures cannot be separated from the osteoporosis drugs the women were already taking. [1]

GSK's drug risvutatug rezetecan, or ris-rez, gave a median survival of 18.5 months in a final-stage trial of small-cell lung cancer, against 10.3 months on the chemotherapy drug topotecan. [3] The trial, run by GSK's Chinese partner Hansoh Pharma, enrolled 461 patients at sites in China whose cancer had come back after first-line chemotherapy. [3] Ris-rez is an antibody-drug conjugate: an antibody that finds a protein called B7-H3 on the cancer cell and carries a poison to it. [3] The trial ran only in China; GSK started its own global trial last year, with results expected in 2027. [3]

Why it matters. Small-cell lung cancer has had few options once chemotherapy stops working, and Amgen's rival drug reached 13.6 months in its own trial, so a China-only 18.5 months is a number the whole field now has to read against. [3]

Amgen said on Tuesday that its drug Imdelltra, given with AstraZeneca's immunotherapy Imfinzi, helped people with newly diagnosed extensive small-cell lung cancer live longer than Imfinzi alone in a final-stage trial. [4] Amgen gave no figures; it said a trial monitoring committee saw the survival benefit at an early check, along with slower tumour growth. [4] Imdelltra is currently approved only after chemotherapy has failed, and sold about $627 million last year. [4] Amgen will now discuss the data with regulators. [4]

Why it matters. Moving a drug from second-line to first-line use is what one analyst called a significant expansion of its sales, and it lands in the same week GSK reported its own small-cell result. [4]

Confirmed Ebola cases in the Democratic Republic of Congo passed 7,000 on Friday, with 7,022 cases and 3,398 deaths across seven provinces on government figures. [5] A case was found in South Ubangi province in the northwest, far from the outbreak's centre in the east. [5] The outbreak is caused by the Bundibugyo species of the virus and is now the largest and deadliest in Congo's history, second worldwide only to the 2014 to 2016 West African epidemic that infected more than 28,600 people. [5] The World Health Organization says shortages of trained staff, partners and funding are holding back the response. [5]

Why it matters. A case far from the epicentre means a second front for a response that the WHO already says cannot expand, and more than 3,300 people have died. [5][6]

Jean Kaseya, head of the Africa CDC, the African Union's disease agency, convened the presidents of Congo and South Africa in July to discuss the Ebola epidemic. [6] He told them and the Western donors present that the response is led in Africa by Africans, with support from partners, and not the reverse. [6] The Africa CDC was set up in 2017 after the 2014 West African Ebola outbreak; Kaseya, a Congolese doctor, took office in 2023 as its second head. [6] He is pushing presidents across the continent to take financial and political ownership of public health, which he calls a matter of sovereignty. [6]

Why it matters. Western aid budgets are shrinking, so the money the outbreak needs has to come from somewhere, and Kaseya's answer is African governments; his leadership has split opinion in global health. [6]

Bangladesh has recorded 999 suspected and confirmed measles-related deaths since March, its worst outbreak on record and the largest in the world on a US CDC list. [7] More than 166,000 suspected cases have been reported, 19,835 confirmed in a laboratory, and more than 146,000 people have been hospitalised. [7] The health minister, Sardar Md. Sakhawat Husain, told parliament on Monday that changes to vaccine purchasing under the interim government caused a shortage, after a 2024 campaign was postponed and a 2025 one cancelled. [7] One Dhaka hospital built for 1,350 patients is caring for more than 2,350, with 60 measles beds and patients on the floor. [7]

Why it matters. Bangladesh kept measles vaccination at or above 95% for most of the past decade, and two years of missed shots during political turmoil were enough to bring the disease back. The same hospitals are also filling with dengue, with more than 45,000 admissions this year. [7]

Kenvue, which makes Tylenol, and a group of US retailers asked a federal appeals court in Manhattan on Thursday to reconsider its July decision reviving more than 500 lawsuits claiming the painkiller causes autism. [8] A trial judge, Denise Cote, had thrown the cases out by excluding three doctors' testimony as unreliable; the appeals court overturned that on 13 July. [8] The companies argue that judges, not juries, are the gatekeepers of scientific evidence, and that lay juries are not scientists. [8] There is no firm scientific evidence linking acetaminophen, Tylenol's ingredient, to autism, and doctors consider it the preferred treatment for pain and fever in pregnancy. [8]

Why it matters. The fight is now about who gets to decide whether a scientific claim is solid enough to put to a jury, and the appeals court rarely reconsiders its own decisions. [8]

A large US hospital made a test for congenital cytomegalovirus, or CMV, an automatic nursing order for any newborn who failed a hearing screen, and checked 91,079 babies born between 2016 and 2023. [9] Testing among babies who failed the hearing screen rose from 7.7% before the policy to 95.2% after it, and the share of babies found to carry the infection rose from 0.015% to 0.036%, a difference just short of statistical significance. [9] CMV is the most common infection passed from mother to baby before birth, and it causes hearing loss that often appears only after a baby has passed its first hearing test. [9] The infection can only be confirmed in the first 21 days of life, and an antiviral treatment exists if it is started within 13 weeks. [9]

Why it matters. Nothing about the test changed. What changed was a staff training programme and an order nobody had to remember to place, and that is where the missing 87 percentage points had been. [9]

Researchers at the University of Illinois Chicago put 32 adults with type 1 diabetes and obesity into three groups for six months: eating only between noon and 8pm with no calorie counting, cutting calories by 25%, or changing nothing. [10] The eating-window group's HbA1c, a blood test of average sugar over recent months, fell by about 0.5%, more than the calorie-cutting group. [10] There was no rise in dangerous high or low blood sugar, or in ketoacidosis, the emergency that follows when insulin runs short. [10] The study is published in Diabetes Care, and the team wants larger trials at several sites. [10]

Why it matters. Type 1 diabetes means the body makes little or no insulin, and this is the first trial of a timed eating pattern in that condition; 32 people is a first step, and the lead researcher said so. [10]

Type II interferon is a signal the immune system uses to attack cancer, but kept on for a long time it does the opposite, and a paper in Science on 10 September shows how. [11] Long exposure made cells leak RNA from their mitochondria, the cell's power units, which the cell read as a virus and answered with a second signal, type I interferon. [11] Together the two signals raised a fat-like molecule called prostaglandin E2 that calms the immune attack, and tumours grew. [11] Blocking prostaglandin production in melanoma cells that had stopped responding to immunotherapy made them respond again. [11]

Why it matters. It offers one explanation for why immunotherapy stops working in some patients, and a target to test; the work is laboratory science, not a study in people. [11]

Macrophages are immune cells that swallow dead and dying cells, and inside tumours they are often turned to the cancer's side. [12] A study in Nature Chemical Biology on 8 September used mass spectrometry to show that proteins from the cells a macrophage eats can keep working inside it. [12] Proteins from the eaten cell's surface are even moved onto the macrophage's own surface. [12] The finding comes from a research team whose paper is open access, with a commentary by Jacob Geri of Weill Cornell Medicine. [12]

Why it matters. Other cells identify a macrophage by what is on its surface, so a macrophage carrying a tumour's proteins can be read as something it is not. [12]

Diabetes destroys beta cells, the pancreas cells that make insulin, and researchers have long tried to convert other pancreas cells into replacements. [13] A study in Nature Chemical Biology reports that changing the metabolism of alpha cells, their neighbours, gave them beta-cell-like features, both in a dish and in living animals. [13] An inhibitor opened up the parts of the DNA that beta cells use and closed down sites where a protein called ETV1 binds. [13] The work was highlighted in Nature Reviews Endocrinology on 7 September. [13]

Why it matters. A cell's identity is a set of open and closed pages in its DNA, and this is a chemical way of turning some pages; it is a long way from a treatment. [13]

A study of 3,119 older adults in China, followed for a median of 13.7 years, is published in Nature Medicine. [14] It measured two things over time: the build-up of Alzheimer's proteins in spinal fluid, and how well thinking held up beyond what those proteins predicted. [14] Each step up in the protein score raised the risk of dementia by 2.5 times, and each step up in resilience halved it. [14] The two contributed comparable shares of ten-year risk, and the lowest risk was in people with both low protein levels and high resilience. [14] The result was repeated in an independent US cohort of 597 people. [14]

Why it matters. Current Alzheimer's drugs attack the proteins; this argues that whatever keeps thinking steady in the face of them deserves trials of its own, and the corresponding author reports consulting fees from several drugmakers. [14]

Immune cells that grow in bone marrow can help clear the damage of Alzheimer's disease, but few of them ever reach the brain, and a paper in Nature Neuroscience says why. [15] In a mouse model, a type I interferon signal in the marrow disrupted how monocytes, the cells in question, develop; monocytes from Alzheimer's patients showed the same pattern. [15] Blocking the interferon signal with antibodies, or giving the mice marrow that cannot respond to it, restored normal monocytes, sent more of them to the brain, and reduced the disease's damage. [15]

Why it matters. It puts part of Alzheimer's disease outside the skull, in the bone marrow, and shows a way to fix the supply of cells rather than the brain itself; it is mice for now. [15]

Andy Plump, president of research and development at Takeda, the Japanese drugmaker, will leave in June after twelve years in the job, Fierce Biotech reported on Friday. [16] He joined in 2015 from Sanofi and will stay until a successor is named. [16] The announcement comes as Takeda's sales are dragged by lost exclusivity on its ADHD drug Vyvanse; last month the US regulator approved its narcolepsy drug Orzeyful and a blood cancer drug. [16] The same week, Vaxcyte hired a former Pfizer scientist as medical chief ahead of final-stage vaccine results due by the end of October. [16]

Why it matters. The person who runs research decides what a drugmaker bets on for the next decade, and Takeda is looking for growth as its Vyvanse sales fall. [16]

The US Food and Drug Administration has pushed its decision on Exelixis' colorectal cancer pill zanzalintinib from 3 December 2026 to March 2027, the company said in a securities filing. [17] The agency asked for the extra time while it waits for updated safety and efficacy data. [17] The pill met its main survival goal across all patients in a final-stage trial but missed it in June in a subgroup defined by liver spread. [17] Analysts at William Blair said the risk of rejection remained low and that they could not tell why the agency wanted more data. [17]

Why it matters. Fifteen extra months is what an unexplained request for data costs a company and its patients, and Exelixis has two more final-stage trials of the same pill due to report. [17]

Karyopharm Therapeutics, a US cancer drugmaker short of cash, reached a deal with its lenders over a $15.8 million debt payment that came due on Thursday. [4] The lenders, noteholders and royalty investors will hold off until 15 October, giving the company a little over a month to negotiate, find a buyer or raise money. [4] In return Karyopharm will pay a $20 million fee in newly issued convertible stock. [4] It is waiting to hear whether US regulators will accept its application for its myeloma drug Xpovio in another blood disorder. [4]

Why it matters. Paying $20 million in stock to delay a $15.8 million payment shows how little room the company has left, and the cost lands on existing shareholders. [4]

Crimean-Congo haemorrhagic fever is a tick-borne virus that kills up to 30% of the people it infects in some areas, across Africa, Asia, the Middle East and eastern Europe, and has recently appeared in Spain and Portugal. [18] Researchers built four vaccines using a harmless adenovirus to carry pieces of the virus, and tested them in mice, four per group. [18] After two doses, several versions protected 100% of mice from a lethal dose of the virus, while unvaccinated mice had no protection. [18] The vaccines were made by a process compatible with manufacturing rules, and the next steps are monkeys and a first human trial. [18]

Why it matters. The fever kills up to 30% of the people it infects in some areas, and four mice per group is a first signal, not a result. [18]

Aflatoxins are poisons made by mould on crops such as peanuts, and they endanger food safety and human health. [19] Researchers at the Chinese Academy of Agricultural Sciences in Wuhan built a device that mixes ozone into water as tiny bubbles and used it on peanut meal, the leftover after oil pressing. [19] In two hours it removed 88% of the most dangerous aflatoxin and 78% to 89% of three others, and it kept working over 12 reuse cycles. [19] Tests of the treated meal found it safe, and the work is aimed at animal feed. [19]

Why it matters. Peanut meal goes into animal feed, and the work is aimed at cleaning it at industrial scale, where no efficient method yet exists. [19]

Prostate cancer that spreads to bone is a leading cause of death from the disease, and a study in Oncogene shows one way it does the damage. [20] Tumour cells release a protein called CTHRC1 that pushes bone-eating cells, osteoclasts, to develop, in cell cultures and in mice. [20] The protein binds directly to a receptor called integrin beta 3 on the osteoclast's surface, switching on the signals that turn it into a bone-destroyer. [20] Blocking that contact weakened the bone loss. [20]

Why it matters. A tumour protein docking onto a surface receptor is a target a drug can aim at, and bone spread is where advanced prostate cancer hurts most. [20]

Multiple myeloma is a cancer of the bone marrow, and its risk is usually judged by the genetics of cells taken from the marrow. [21] A study of 631 patients published in Leukemia found cancer cells circulating in the blood in 63.9% of those tested at diagnosis. [21] Above a threshold of 0.38%, those cells independently predicted a faster return of the disease, and within the group already rated genetically high-risk they picked out the worst outcomes: 12.4% of patients. [21] The blood count also tracked outcomes during treatment, alongside marrow testing. [21]

Why it matters. A blood test is far easier on a patient than a marrow biopsy, and this adds a risk reading the marrow genetics alone were missing. [21]

Researchers built a genetic risk score for osteoporosis from up to 10,794 people in Japan's Tohoku Medical Megabank, using a heel ultrasound measure of bone, and tested it in 8,711 others. [22] Added to age and sex it improved prediction only modestly, but people in the lowest-scoring fifth had 1.42 times the rate of new cases over 3.5 years and the highest fifth had 0.70 times. [22] In adults aged 20 to 44, the analysis suggested those at high genetic risk already had lower bone readings around the age of peak bone mass. [22]

Why it matters. The paper calls the gain over age and sex modest, and its point is a reading available decades before bone starts to thin. [22]

The adrenal glands sit on top of the kidneys and make stress and blood-pressure hormones, and a study in Nature Genetics mapped every cell type in adult human and mouse glands. [23] Mice have no equivalent of the zona reticularis, the human layer that makes weak male hormones, and the markers of the cortisol-making layer differ between the species. [23] In humans, cells that renew the gland were scattered through its outer layer; in mice they sit in one zone. [23] The map also found age-linked cell states in humans able to make cortisol directly. [23]

Why it matters. A mouse is the standard stand-in for a person in hormone research, and this shows which parts of the stand-in are not there. [23]

Metabolic fatty liver disease, now called MASLD, is a whole-body condition in which heart disease is a leading cause of death. [24] An expert recommendation in Nature Reviews Gastroenterology and Hepatology says cardiologists often miss the liver disease and liver doctors often underrate the heart risk, so care is split. [24] It proposes a shared pathway: a blood-test score, then a scan of liver stiffness, with scarring rather than fat as the measure that matters. [24] It also lists the drugs that help both organs, including the weight-loss class and a newer liver drug, resmetirom. [24]

Why it matters. Two specialities treating one patient's two organs separately is how early warning gets missed, and this is a published attempt to put them in one room. [24]

A commentary in Nature Aging on 7 September describes an analysis by Ashwin and colleagues that sorts the world's diseases into four groups by the stage of life they strike. [25] Diseases of ageing are now the largest group, and they tend to arrive together and last for years. [25] Most countries still carry all four burdens at once. [25] The authors argue health systems must shift from treating disease to preserving health across every stage of life. [25]

Why it matters. The measles wards in Bangladesh and the osteoporosis trial in Spain sit at opposite ends of a life, and the claim here is that one health system now has to pay for both at once. [25]