Biotech & Longevity · Saturday, 26 September 2026

01 Briefing what happened

The FDA approved Lilly's Onswik, a once-a-week insulin for type 2 diabetes. It matched the daily shots it replaces, and cuts more than 300 injections a year.

Biotech & Longevity 21 sources

The US drug regulator cleared Eli Lilly's weekly insulin on 24 September, six months after Novo Nordisk's. The same week Novo paid up to 1.17 billion euros to make its own shots monthly, three cancer drugs were approved or withdrawn on very different levels of proof, and biotech share sales kept landing.

300+

fewer injections a year than a daily insulin, by Lilly's count

a shot every day is 365 a year; a shot every week is 52 [2]

3,400+

adults with type 2 diabetes in the four trials behind the approval

Onswik was compared with the daily insulins glargine and degludec, not with a dummy shot [1][2]

2nd

weekly insulin approved in the United States

Novo Nordisk's Awiqli was the first, in March 2026 [3]

$27.6bn

sold by Lilly's two big diabetes and weight-loss drugs in six months

Mounjaro and Zepbound, January to June 2026; Novo's equivalents sold about $15.7bn [4]

The lead story — what happened

The US Food and Drug Administration (FDA), the US drug regulator, approved Onswik on Thursday 24 September. It is an insulin for adults with type 2 diabetes, injected under the skin once a week, on the same day each week. [1]
Onswik is made by Eli Lilly, the US drugmaker that also sells the weight-loss shot Zepbound. Lilly says it will be on sale in the United States in the coming months, in a pre-filled pen in two strengths. [2][4]
Type 2 diabetes means the body does not make enough insulin, or does not use it well, so sugar stays in the blood. Insulin moves that sugar out of the blood and into cells. [1]
Many people with type 2 diabetes inject a long-acting insulin every day. Lilly says a weekly shot cuts more than 300 injections a year. [2]
The approval rests on four trials with more than 3,400 adults. Onswik was compared with two daily insulins, glargine and degludec, and patients kept their other diabetes drugs. [1][2]
In all four trials Onswik lowered A1c, a measure of average blood sugar over months, about as much as the daily insulins did. It was tested to be no worse than them, not better. [1][3]
It is not for type 1 diabetes, where the body makes no insulin at all. In those patients the FDA saw a higher risk of severe low blood sugar, which can cause seizures, unconsciousness or death. [1][2]
The most common side effects are low blood sugar, allergic reactions, reactions where the needle goes in, itching, swelling and weight gain. The label also warns about dosing mistakes and accidental overdose. [1]
Onswik is the second weekly insulin approved in the United States. The first, Novo Nordisk's Awiqli, was approved in March. [1][3]
Novo Nordisk, the Danish drugmaker, is Lilly's main rival in diabetes and obesity. Lilly's two big diabetes and weight-loss drugs sold $27.6 billion in the first half of 2026, against about $15.7 billion for Novo's. [4]
Onswik was already approved in Europe, Japan and Mexico before the US decision. [2]
The same week Novo agreed to pay up to 1.17 billion euros for a coating technology to make its own injections monthly or every three months. [6]

365Daily insulin, injections a year

52Onswik, injections a year

One shot a day against one shot a week. Lilly puts the difference at more than 300 injections a year.

Who is involved

Eli Lilly

the US drugmaker behind Onswik and the weight-loss shot Zepbound; it says Onswik goes on sale in the coming months
The FDA

the US drug regulator; it approved Onswik on 24 September on the strength of four trials
Novo Nordisk

the Danish drugmaker that is Lilly's main rival; its Awiqli was the first weekly insulin, approved in March
People with type 2 diabetes on daily insulin

the patients who could swap 365 shots a year for 52; people with type 1 diabetes are excluded

How it unfolded

March 2026 the FDA approves Novo Nordisk's Awiqli, the first weekly insulin [2]
Before this week Onswik is approved in Europe, Japan and Mexico [2]
Thu 24 Sep the FDA approves Onswik for adults with type 2 diabetes [1]
Coming months Lilly puts it on sale in the US as a pre-filled pen, in two strengths [2]

Where this points

Watch what Lilly charges for Onswik and whether US insurers list it beside the daily insulins; the next test is whether patients on it outside a trial keep their blood sugar as steady as those inside one.

What is pushing on the whole day

The bar and the word are our reading of how hard each one is pushing today. The arrow is where it is heading. The evidence is in the stories below.

Fewer injections as the sales pitch Building↑

Lilly says Onswik replaces a daily insulin shot with a weekly one. [2] Novo Nordisk agreed to pay up to 1.17 billion euros for a coating that could make its shots monthly or quarterly. [6] Novo's research chief said three or four once-a-month weight-loss drugs would reach first human tests within a year. [5]

Big drugmakers buying from small ones High↑

Genentech, Roche's US arm, is paying Earendil Labs $55 million up front for cancer antibody designs, in a deal that could reach $1.5 billion. [8] AbbVie agreed to buy up to $100 million of ADARx's shares as it listed. [13] Novo's deal with Nanexa doubled the Swedish company's share price in a morning. [6]

Approvals on partial proof Steady→

Onswik was approved for matching daily insulin, not for beating it. [1] Merck's Welireg was approved with Lenvima because it held kidney cancer back longer than the standard drug, though patients did not live significantly longer. [10] Lyrfigtu was approved for bile-duct cancer on 116 patients and no comparison group. [11]

Drug partnerships ending in court Building↑

A California jury awarded Nektar $90 million from Eli Lilly, far below the nearly $1 billion Nektar asked for. [3] Lepu, a Chinese drugmaker, is suing its former US deal adviser in New York for alleged double-dealing. [18] Indivior handed back every drug from an eight-year partnership with Addex before its merger. [19]

Biotech share sales still landing Steady→

ADARx raised $446 million on Nasdaq, above the $350 million it had planned. [12] Electra Therapeutics, which listed a week earlier, has fallen about 21% since. [13]

The rest of the day

17 more stories on this beat.

Each with its own sources. None of these is a link to the story above.

02
Novo pays up to 1.17bn euros for monthly shots

Novo Nordisk, the Danish maker of Ozempic and Wegovy, agreed on Thursday to license a drug-coating technology from Nanexa, a small Swedish company. [4][6] The coating, PharmaShell, is an ultra-thin layer on each drug particle that dissolves slowly, so the drug enters the blood over a long time. [5][6] Novo may use it on up to five drugs for obesity, type 2 diabetes and related diseases, aiming at shots given monthly or every three months. [5][7] Up to 615 million euros is paid up front and on development goals, and neither company has said what the upfront part is. [5] Nanexa's shares more than doubled on Friday morning. [6]

Why it matters — Novo's weekly shots face Lilly's daily pill and a US patent on its best-seller that ends in 2032, so it is now selling how rarely a patient injects. [4] Lilly signed a similar deal with Camurus last year. [5]
03
Genentech pays Earendil $55m for cancer antibodies

Genentech, the US arm of the Swiss drugmaker Roche, is paying Earendil Labs $55 million up front to design bispecific antibodies for cancer. [8] A bispecific antibody is one molecule that grabs two targets at once, for instance a cancer cell and an immune cell. [8] Earendil designs antibodies with AI software and will run the early trials before handing the drugs to Genentech. [8] Payments could pass $1.5 billion if the drugs keep passing tests. [8] Sanofi has paid Earendil $285 million across two earlier deals, and the company raised $787 million in March. [8]

Why it matters — Sanofi, WuXi XDC and now Genentech have all paid a startup that made its name only in the past 18 months for designs rather than make them in-house. [8] It is the same trade as Novo and Nanexa: the big company pays a little for the idea, and most of the money arrives only if it works.
04
Merck and Daiichi pull a lung cancer application

Merck, the US drugmaker, and Daiichi Sankyo of Japan withdrew their US application for ifinatamab deruxtecan this week. [9] The drug was under review for small-cell lung cancer that had grown back after chemotherapy, an aggressive form of the disease. [9] It is an antibody-drug conjugate: an antibody that finds the cancer cell and carries a dose of chemotherapy into it. [9] It is the second of three drugs in the pair's 2023 partnership, worth up to $22 billion, to have a US application pulled. [9] A larger trial against chemotherapy is nearly fully enrolled and will be the basis for a new application. [9]

Why it matters — People whose small-cell lung cancer has come back have few treatments, and now wait for the larger trial. [9] The first drug pulled from this partnership went after it failed to extend patients' lives. [9]
05
Merck's Welireg cleared for kidney cancer after Keytruda

The FDA approved Merck's Welireg with Eisai's Lenvima on Thursday for advanced clear-cell kidney cancer that has grown despite an immune-boosting drug such as Keytruda. [10][3] In a 747-patient trial the pair held the cancer back for a median 14.6 months, against 10.6 months on cabozantinib, the drug it was compared with. [10] Patients on the pair lived a median 33.7 months against 28.6, but that gap was not statistically significant, so it may be chance. [10] It is the latest of several approved uses for Welireg, which sold $470 million in the first half of 2026. [3]

Welireg with Lenvima14.6 months
Cabozantinib10.6 months
Median months before the kidney cancer grew again, in the 747-patient trial behind the approval.

Why it matters — It gives kidney cancer patients a next step after immunotherapy stops working, on evidence about delaying the cancer rather than about living longer. [10]
06
A pill for a rare bile-duct cancer is approved

The FDA approved lirafugratinib, sold as Lyrfigtu by Elevar Therapeutics, on 23 September. [11] It is for cholangiocarcinoma, a cancer of the bile ducts, in patients whose tumour carries a fault in a gene called FGFR2 and who have already had chemotherapy. [11] The pill blocks the faulty FGFR2 signal that tells those cells to keep growing. [11] In a trial of 116 patients with no comparison group, tumours shrank in 46%, and the shrinkage lasted a median 11.8 months. [11] The label warns of eye damage and high phosphate levels. [11]

Why it matters — For a cancer with few options, the regulator accepted tumour shrinkage in a single-arm trial as proof. Whether patients live longer has not been shown. [11]
07
Nektar wins $90m from Lilly, not $1bn

A federal jury in California sided on Thursday with Nektar Therapeutics in its 2023 lawsuit against Eli Lilly. [3] Nektar said Lilly had undermined an autoimmune drug the two were developing together after Lilly bought a rival treatment in a company takeover. [3] The jury awarded Nektar $90 million, far below the nearly $1 billion it had asked for. [3] Nektar told regulators the verdict is still subject to post-trial proceedings. [3]

Why it matters — Small companies license drugs to big ones and then depend on the big partner's attention. [3] A jury agreed Lilly fell short, and priced the harm at about a tenth of what Nektar asked. [3]
08
ADARx raises $446m in an enlarged share sale

ADARx Pharmaceuticals, a San Diego company with 100 staff, priced its Nasdaq share sale on Thursday night at $17 a share, the top of its range, and sold more shares than planned. [12] It raised $446.3 million, against the $350 million it had expected. [12][13] AbbVie, a large US drugmaker that paid ADARx $335 million last year for options on its drugs, agreed to buy up to $100 million of shares alongside. [12][13] ADARx makes siRNA drugs, which stop a cell making a harmful protein by blocking the instructions for it. [12]

Why it matters — Biotech listings are the strongest part of this year's US share-sale market, though Electra, which listed a week earlier, has fallen about 21% since. [13]
09
A hepatitis B treatment that tags the virus silent

Scientists at nChroma Bio in Boston and institutes in Milan reported on 21 September a one-time therapy that switches off the hepatitis B virus without cutting DNA. [14] Hepatitis B infects more than 250 million people and hides in liver cells as small loops of DNA, so today's daily drugs rarely clear it. [14] The therapy uses a disabled CRISPR enzyme to carry chemical tags onto the virus's genes, which silences them. [14] One injection worked in human liver cells and in mice, with minimal side effects in monkeys. [14] A human trial in Hong Kong and New Zealand gave its first dose in January. [14]

Why it matters — Cutting DNA can raise cancer risk in people with hepatitis B, so a tag instead of a cut matters most for exactly these patients. [14] No human results exist yet.
10
Edited donor cells shield leukaemia patients

Doctors at Washington University in St Louis and 14 other centres used CRISPR to remove a protein called CD33 from donor stem cells. [15] They transplanted the cells into 30 adults with acute myeloid leukaemia or a related bone marrow disease. [15] CD33 sits on those cancer cells but also on healthy blood cells, so drugs aimed at it usually damage both. [15] All 30 transplants took hold by day 28, and patients then got a CD33-targeting drug without the usual collapse in blood counts. [15] Seven patients died, four from the cancer and three from transplant complications. [15]

Why it matters — Vor Biopharma funded the trial, published in Nature Medicine. [15] It clears a path for CAR-T cell therapy against these cancers, which has failed so far because the target sits on healthy cells too. [15]
11
Alzheimer's risk gene scars brain vessels, and mice recovered

Researchers at Mount Sinai in New York reported in two papers, in Cell and Cell Stem Cell, how APOE4, the strongest genetic risk factor for Alzheimer's disease, harms the brain. [16] The gene turns pericytes, cells that keep small blood vessels stable, into scar-forming cells that thicken the vessels and speed the build-up of amyloid, the protein clumps that current drugs target. [16] Blocking a protein called TGF-beta protected the pericytes and reversed the vessel damage in mice. [16]

Why it matters — It suggests blood-vessel damage in Alzheimer's is a cause that could be treated, not a late effect. [16] It has only been shown in mice.
12
Gilead's cell therapy: all 38 myeloma patients responded

A CAR-T therapy from Gilead Sciences, the US drugmaker, produced responses in all 38 patients with hard-to-treat multiple myeloma in a phase 1 trial reported in the New England Journal of Medicine. [16] CAR-T takes a patient's own immune cells, engineers them in a lab to hunt a protein on the cancer, and returns them to the body. [16] Nearly 80% had no detectable cancer; more than half showed no progression two years on, and 65% were alive after three years. [16] Phase 1 trials test safety, so larger trials are under way to confirm this. [16]

Why it matters — Multiple myeloma is a blood cancer that keeps coming back, and a treatment that holds it off for years in most patients would change what those patients can expect. [16] The number is 38 people, and severe side effects were uncommon. [16]
13
A way to spot old cells without killing them

Researchers at MIT found a way to identify senescent cells, old cells that have stopped dividing but refuse to die, without destroying them, in a study in Nature Aging. [16] They used Raman microscopy, which reads a cell's chemistry with light, together with the cell's gene activity, to find a unique signature for these cells. [16] The work was in mouse cells, and the team is now adapting it for human tissue. [16] It is part of a US National Institutes of Health programme on cell senescence. [16]

Why it matters — Senescent cells feed inflammation and age-related disease, and drugs to remove them need a way to find them first. [16]
14
Nine European drugmakers warn of a 'slow agony'

The chairs of nine drugmakers, including AstraZeneca, GSK, Novo Nordisk, Novartis, Roche and Sanofi, published an open letter on Tuesday. [17] It asks the European Union and its governments to spend more on medicines, run trials faster and protect patents. [17] They said Europe's share of global drug research has fallen to 31% from 43% in 1990, and its share of commercial trials has halved to 9% in a decade. [17] About 40% of new therapies never reach European patients, they said. [17] Europe spends about 1% of GDP on medicines, against 2% in the United States. [17]

Why it matters — The companies are asking governments for more spending and faster approvals, and they have already delayed some European launches with an eye on US pricing rules. [17]
15
A Chinese drugmaker sues its US deal adviser

Lepu Biopharma, a Chinese drugmaker, is suing an adviser named Zhang in a New York federal court. It hired her in December 2024 to find Western partners for its obesity and diabetes drugs. [18] Lepu says Zhang and her team worked at the same time for rival Chinese companies while holding Lepu's confidential data. [18] Zhang made her name at Hengrui, where she set up the first 'NewCo' deal: a Chinese company spins early drugs into a new Western firm and keeps a stake. [18] Her lawyers say her firm was an independent contractor and owed Lepu no such duty. [18]

Why it matters — These deals are how Western investors reach Chinese drugs early; Chinese assets have produced 15 of them this year, and this case tests who owes loyalty to whom inside one. [18]
16
Indivior hands drug rights back to Addex

Indivior, the maker of the opioid-addiction treatment Sublocade, has returned all rights from an eight-year partnership with Addex Therapeutics, a Swiss biotech, Addex said. [19] The two had worked since 2018 on drugs that turn up a brain receptor called GABAB to treat addiction; the first candidate failed and Indivior took on a second in 2024. [19] Indivior is cutting research ahead of its planned merger with Supernus. [19] Addex says it now holds one of the broadest portfolios against that receptor and may seek a new partner. [19]

Why it matters — A merger at the big partner ends the small company's programme. [19] Addex gets its drug back, and loses the money that was developing it.
17
Travere's chief hands over to Amicus's former CEO

Eric Dube, chief executive of Travere Therapeutics, is stepping down. Bradley Campbell, who led Amicus Therapeutics through its $4.8 billion sale to BioMarin, takes over on 1 December. [20] Dube renamed the company from Retrophin, which was founded by Martin Shkreli, and won US approval in 2023 for Filspari, a drug for the rare kidney disease IgA nephropathy. [20] He stays on as an adviser until February 2027. [20] Campbell spent 20 years at Amicus and took it from one product to two, with a Pompe disease treatment approved in 2023. [20]

Why it matters — Filspari was approved on an accelerated basis, so Travere still has to confirm it works, and the new chief's job is that and the sales. [20]
18
Anthropic says its AI found a CRISPR-like enzyme

Anthropic, the company behind the Claude AI models, said on 23 September that Claude had discovered a new enzyme system with repeated DNA sequences it called reminiscent of CRISPR. [21] CRISPR is the gene-editing tool copied from a bacterial defence against viruses, and its discoverers won a Nobel prize in 2020. [21] New Scientist's assessment is that the two discoveries are not comparable, and that at best the find is another gene-editing tool. [21]

Why it matters — AI companies are now announcing biology findings directly, and the specialist press is grading them. [21]

02 Lesson why it matters

What a new drug must prove depends on what it is replacing

A regulator judges a new drug against the best treatment already sold, so the bar rises and falls with what came before.

The twist

Onswik needed more than 3,400 patients to show it was no worse than a daily insulin. Lyrfigtu needed 116 patients and no comparison at all, because those patients had nothing else.

The picture

Onswik, weekly insulin3,400 patients

Welireg with Lenvima, kidney cancer747 patients

Lyrfigtu, bile-duct cancer116 patients

Patients in the trials behind three approvals this week. Onswik was tested against daily insulins, Welireg against cabozantinib, and Lyrfigtu against nothing.

How it works

A company asks the regulator to approve a new drug
The regulator asks what patients with that disease get today
If a good drug exists, the new one is tested against it
Matching it can be enough, if the new drug is easier or safer
If nothing exists, a small trial with no comparison can be enough
So the size of the proof follows the disease, not the drug

The same force, elsewhere today

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

Merck's Welireg in kidney cancer

it was tested against cabozantinib, the standard drug, and approved for holding the cancer back longer, even though patients did not live significantly longer
Lyrfigtu for bile-duct cancer

there was no standard drug for these patients, so a 116-patient trial with no comparison group and tumour shrinkage as the measure was enough
Merck and Daiichi pulling their lung cancer application

chemotherapy already exists for relapsed small-cell lung cancer, so the pair will now wait for a larger trial that compares their drug against it

Where you've seen this

Job interviews

a candidate for an empty post only has to be good enough; one replacing a strong colleague has to be at least as good

Phone upgrades

a new model is judged against last year's, not against having no phone

River crossings

where a ferry runs, a bridge must beat it on time or cost; where there is no crossing, any bridge is an improvement

Restaurants

the first pizza place in a town gets customers for existing; the second has to be better or cheaper than the first

The catch

Matching an old drug only counts if the new one is easier or safer, and a small trial with no comparison can approve a drug that later turns out not to help people live longer.

And the whole of it

A person with type 2 diabetes will get a shot tested on thousands of people. A person with bile-duct cancer will get a pill tested on 116, because there was nothing to compare it with. Neither of them chose how much proof sat behind their prescription, and neither can see the other's.

03 Truth what's really going on

What is really going on

Lilly's weekly insulin was approved because it matched a daily one, not because it beat it, and what Lilly is selling is fewer injections. [1][2] The same week Novo Nordisk agreed to pay up to 1.17 billion euros for a coating that could make its own shots monthly. [6] The two biggest diabetes drugmakers are now competing on how rarely a patient has to inject, because the drugs themselves already work. [4][5]

Why it works on us — A figure like 300 fewer injections a year is easy to picture, and it comes from the company that sells the pen. [2]

Who gains

Eli Lilly — It now has a weekly insulin to sell against Novo Nordisk's Awiqli, which reached the US market first in March, and it already outsells Novo almost two to one in the newer diabetes and weight-loss drugs. [2][3][4]
Nanexa's shareholders — The Swedish company's shares more than doubled on Friday morning, to their highest in more than five years, on the Novo deal. [6]
Earendil Labs — It collects $55 million up front from Genentech for antibody designs, its fourth big deal in 18 months, with most of the $1.5 billion due only if the drugs pass tests. [8]
Merck — Welireg gained another approved use, on evidence about delaying kidney cancer rather than extending life; the drug sold $470 million in the first half of 2026. [3][10]
ADARx — It raised $446 million by selling more shares at the top of its price range, about $96 million more than it had planned, with AbbVie buying up to $100 million alongside. [12][13]
Addex — It got back every drug from its Indivior partnership and can now offer them to a new partner. [19]

Who pays

People with type 1 diabetes — Onswik is not for them. The FDA found a higher risk of severe low blood sugar, which can cause seizures or death, so the weekly shot stays a type 2 product. [1][2]
Nektar Therapeutics — It spent three years suing Lilly for nearly $1 billion and a jury gave it $90 million, which it may still have to defend in post-trial proceedings. [3]
Patients with relapsed small-cell lung cancer — The Merck and Daiichi drug they were waiting on has left US review and will not return until a larger trial against chemotherapy reads out. [9]
The seven patients who died in the Vor Biopharma trial — Four died of their leukaemia and three of transplant complications, including kidney failure, liver toxicity and sepsis, in a 30-person study. [15]
Investors who bought Electra's shares at its listing — The stock has fallen about 21% in the week since, while the next biotech listing raised more than it planned. [13]
Patients in Europe — About 40% of new therapies never reach them, according to the nine drugmakers' own letter, which asks European governments to pay more. [17]

What nobody knows yet

Open questions from across today’s stories — ours included.

01

What Onswik will cost, and whether US insurers will pay for it.

Lilly said it goes on sale in the coming months and named two pen strengths, but neither its statement nor the FDA's notice says what a patient or an insurer will pay. [1][2]
02

Whether Welireg with Lenvima helps kidney cancer patients live longer.

Patients on the pair lived a median 33.7 months against 28.6 on cabozantinib, and the FDA says that gap was not statistically significant, so it could be chance. [10]
03

How much Novo Nordisk paid Nanexa on the day.

Both companies say up to 615 million euros comes from the upfront payment and development goals combined, and neither has said how much of that is upfront. [5]
04

Why Merck and Daiichi Sankyo pulled their lung cancer application.

The Reuters report gives the withdrawal and the plan for a larger trial against chemotherapy; the reason for pulling this application is not stated in it. [9]
05

Whether the edited transplant makes CAR-T therapy safer.

The 30-patient trial tested an antibody drug against CD33, not CAR-T cells; the CAR-T result is one patient, reported separately. [15]
06

Whether the hepatitis B tags hold in people.

The results are in human liver cells, mice and monkeys. The human trial in Hong Kong and New Zealand began dosing in January and has reported nothing yet. [14]
07

Whether Nektar keeps its $90 million.

Nektar told regulators the verdict is subject to post-trial proceedings, so the figure can still change. [3]
08

What the enzyme system Anthropic's AI found actually does.

Anthropic called it reminiscent of CRISPR. New Scientist says the two discoveries are not comparable, and that at best it is another gene-editing tool. [21]

04 Hope carry this

People with type 2 diabetes who inject a long-acting insulin every day can now get one shot a week in the United States. Lilly counts that as more than 300 fewer injections a year.

Also true today

All 38 people with hard-to-treat multiple myeloma who got Gilead's cell therapy in a first trial responded to it. Nearly 80% had no detectable cancer afterwards, and 65% were alive three years later.
All 30 leukaemia patients given gene-edited donor cells in a trial had the transplant take hold within 28 days, and they kept their blood counts through a drug that usually knocks them down.
Tumours shrank in 46% of the 116 patients with advanced bile-duct cancer who took lirafugratinib, for a median of 11.8 months. The drug is now approved in the United States.
One injection of a new hepatitis B therapy silenced the virus in human liver cells and in mice, and a human trial gave its first dose in January.