Artiva ended its AlloNK autoimmune cell therapy on October 27, 2026 after two patient deaths, cutting 94% of staff and weighing a sale or wind down.
Last updated: October 28, 2026
Artiva Biotherapeutics discontinued AlloNK, canceled its Phase 3, and cut roughly 94% of staff on October 27, 2026 after two deaths in its autoimmune program and durability that fell short of its own bar. The board is weighing a sale, merger, or liquidation.
This page covers the Artiva decision in full, the 2026 autoimmune cell therapy safety record across three sponsors, the SAFFRON lung cancer data, Novartis third quarter results, the BioNTech and Hansoh ADC readouts, the Novo Nordisk ruling, and the day’s deals and movers.
Why did Artiva discontinue AlloNK?
Artiva stopped all AlloNK development because longer term follow up showed the durability of clinical response fell below the company’s required threshold, and because two participants died. The Phase 3 pivotal trial is canceled and no further patients will be dosed in any study.
AlloNK, also called AB-101, was an allogeneic natural killer cell therapy given with rituximab in rheumatoid arthritis and other B cell driven autoimmune diseases. In its October 27 release, the company said updated data still showed deep B cell depletion and clinical activity consistent with earlier findings. The problem was how long responses lasted, alongside risk benefit considerations, likely enrollment delays, and the program’s development and commercial outlook.
The two deaths were both Grade 5 events in or around the Phase 2a basket study. The first was leukoencephalopathy in a participant who had received cyclophosphamide and fludarabine lymphodepletion. The study investigator and an independent safety review assessed that death as likely related to fludarabine associated neurotoxicity, and Artiva says it plans to publish the case given how widely fludarabine is used in cell therapy conditioning. The second participant died suddenly about two months after starting treatment. The investigator assessed that death as unrelated to AlloNK or any other component of the regimen, and the cause has not been determined. After the first event the FDA asked Artiva to pause enrollment and dosing in the Phase 2a, and Artiva also paused the planned Phase 3.
The corporate consequences are near total. Artiva is cutting approximately 94% of its workforce, has canceled its ACR Convergence data presentation and a November 9 investor webcast, and has retained Leerink Partners to run a strategic review that the company says may end in a merger, business combination, asset sale or licensing, or dissolution and liquidation. Artiva reported approximately $319.1 million in cash, cash equivalents and investments as of September 30, a figure it labels preliminary and unaudited. Shares closed at $4.27, down 30.7% from Monday’s $6.16 close, both settled exchange figures with the percentage our arithmetic.
What happened to the other autoimmune cell therapy programs in 2026?
Artiva is the third sponsor to hit a safety wall in autoimmune cell therapy since August. Novartis and Bristol Myers Squibb paused trials of their autologous CAR T therapies over deaths and inflammatory events, per BioPharma Dive, though both programs continue.
| Program | Design | 2026 events | Status |
|---|---|---|---|
| Artiva AlloNK (AB-101) | Allogeneic NK cells plus rituximab | Two Grade 5 events; one assessed likely related to fludarabine conditioning, one undetermined and assessed unrelated to the regimen | Discontinued October 27; Phase 3 canceled; strategic review |
| Bristol Myers Squibb zola-cel | Autologous CD19 CAR T | Trials paused in August over inflammatory events the company called transient and reversible; one death 44 days after dosing disclosed last week | Development continuing |
| Novartis rap-cel | Autologous CD19 CAR T | Three deaths from a serious immune reaction reported; trials paused in August | Development continuing |
The event descriptions for zola-cel and rap-cel are as reported by BioPharma Dive on October 27. On zola-cel specifically, Bristol Myers Squibb’s interim Breakfree-2 data, described by BioPharm International from the company’s presentations, covered 66 patients with multiple sclerosis: all 60 at the higher dose level were off standard MS therapies at last follow up and 92% of evaluable progressive MS patients had stable or improved disability, while the patient who died developed late onset motor symptoms, received high dose cyclophosphamide, and died of septic pneumonia 44 days after infusion. The company said it has had no other treatment related deaths across 190 patients in its zola-cel autoimmune studies.
The distinction worth keeping is that the Artiva death with an assessed cause traces to conditioning chemotherapy, not the cells. Fludarabine lymphodepletion is standard across much of cell therapy, which is why Artiva’s planned publication matters well beyond one company.
What did the SAFFRON trial show for savolitinib plus Tagrisso?
SAFFRON showed that adding savolitinib to osimertinib after progression on osimertinib cut the risk of progression or death 45% and the risk of death 32% versus platinum doublet chemotherapy in MET driven, EGFR mutated lung cancer. The data ran at ESMO’s Presidential Symposium on October 25 and simultaneously in The New England Journal of Medicine.
The global Phase 3 enrolled 338 patients with EGFR mutated advanced NSCLC whose tumors showed MET overexpression or amplification after first or second line osimertinib, randomized to savolitinib 300 mg twice daily plus osimertinib 80 mg once daily, or to platinum doublet chemotherapy. All figures below are from AstraZeneca’s October 25 release, with a July 6, 2026 data cutoff.
| Endpoint (full analysis set) | Savolitinib + osimertinib | Chemotherapy | Hazard ratio |
|---|---|---|---|
| Median PFS by blinded review | 7.6 months | 5.4 months | 0.55 (0.42 to 0.72), p<0.001 |
| PFS at 12 months | 36% | 10% | |
| Median overall survival | 17.8 months | 12.9 months | 0.68 (0.50 to 0.91), p=0.010 |
| OS at 12 months | 70% | 53% | |
| Grade 3 or higher TEAEs | 64% | 63% |
In the MET overexpressed subset median PFS was 8.5 versus 5.6 months with a 0.51 hazard ratio, and in the MET amplified subset 7.64 versus 4.2 months at 0.487, without formal statistical testing in that set. Median treatment exposure was 8.5 months on the combination versus 3.5 on chemotherapy, which is context for the nearly identical severe adverse event rates. A US NDA is already under FDA review, announced September 29. China has approved the combination on the SACHI trial and Switzerland granted a temporary authorization on SAVANNAH. HUTCHMED estimates about 34% of tumors develop high MET levels after progression on a third generation EGFR inhibitor, and principal investigator Shun Lu framed the result as making MET testing at progression part of standard workup. HCM ADSs closed at $13.80 Tuesday, down 3.3% from Monday, our arithmetic.
How did Novartis perform in the third quarter of 2026?
Novartis grew sales 4% in constant currencies to $14.5 billion, grew core operating income 10%, and raised its full year core operating income guidance, while taking a $1.9 billion impairment on the lead asset from its Avidity acquisition. Figures are from the company’s October 27 results materials.
| Measure | Q3 2026 | Note |
|---|---|---|
| Net sales | $14.5 billion, up 4% cc | |
| Core operating income | $6.0 billion, up 10% cc | 41.6% core margin |
| Core EPS | $2.46, up 8% cc | |
| Reported operating income | Down 59% | Mainly the impairment charge |
| FY26 core operating income guidance | Raised | Now in line to low single digit growth, from a low single digit decline |
Kisqali grew 37% in constant currencies to $1.82 billion, Scemblix 67%, Leqvio 63%, Kesimpta 22% and Pluvicto 20%. The $1.9 billion charge is a partial impairment of del-desiran, the myotonic dystrophy candidate acquired with Avidity, after the Phase 3 HARBOR trial missed its primary endpoint; the asset’s carrying value is now $5.5 billion and the company told investors it stands by the roughly $12 billion acquisition. Novartis also raised Rhapsido peak sales guidance to above $4 billion for chronic urticaria indications and said remibrutinib, which hit in the REMODEL-1 and REMODEL-2 multiple sclerosis trials, goes to the FDA in the fourth quarter with a priority review voucher. NVS ADRs closed at $156.86.
What did BioNTech and DualityBio report from DYNASTY-Breast02?
DYNASTY-Breast02 met its primary endpoint, but BioNTech and DualityBio said the progression free survival improvement was modest, interstitial lung disease came in above prior trials of the drug, and the potential patient benefit in this setting would be limited. No figures were disclosed.
The global Phase 3 randomized 541 patients with chemotherapy naive, hormone receptor positive, HER2 low metastatic breast cancer that had progressed on endocrine therapy to trastuzumab pamirtecan, the HER2 ADC also known as BNT323 or DB-1303, or to single agent chemotherapy of the investigator’s choice. The October 26 release describes a statistically significant yet modest PFS improvement, immature interim overall survival, and interstitial lung disease reported more often than in prior trastuzumab pamirtecan trials, mostly Grade 1 and 2, with severe and fatal events in a small proportion of patients and no counts given. The companies say they will discuss the data and determine next steps, which is unusually cool language for a positive Phase 3 and reflects the setting: AstraZeneca and Daiichi Sankyo’s Enhertu is already approved in HER2 low disease, so a late entrant needs more than a statistical win over chemotherapy. BNTX closed at $87.06 Tuesday, roughly flat in the second session after the release.
How effective was Ris-Rez in relapsed osteosarcoma?
In Hansoh’s China Phase 3 ARTEMIS-011, the B7-H3 ADC Ris-Rez cut the risk of progression or death 67% versus chemotherapy in relapsed osteosarcoma, with median progression free survival of 5.5 versus 1.4 months. Figures are from the ESMO presentation as reported by trade press.
The trial randomized patients whose disease had progressed after at least two prior lines to risvutatug rezetecan, also coded HS-20093, at 12 mg/kg every three weeks, or to gemcitabine plus docetaxel. At a median follow up of 7.1 months the hazard ratio was 0.33 with p below 0.0001, per BioPharm International’s account of the late breaking Presidential Symposium data, with Fierce Biotech reporting the same 67% risk reduction. Overall survival is immature. Grade 3 or higher treatment related adverse events ran 93% versus 83%, mostly hematologic, with discontinuations of 4% versus 2%. The 12 mg/kg dose is higher than the 8 mg/kg used in small cell lung cancer, chosen for bone penetration. Hansoh plans a China filing; GSK, which holds rights outside greater China and has dosed more than 1,000 patients in its EMBOLD program, has FDA breakthrough designation in this setting. The caveats: a China only population, an open label design, and short follow up. For the B7-H3 class it is a second Phase 3 win in a second tumor type, in a disease with no approved therapies after relapse.
What did the court decide in Novo Nordisk’s case against Eli Lilly?
A federal judge denied Novo Nordisk’s request to block Eli Lilly’s head to head obesity drug advertising on Monday, ruling that Novo had not shown the harm needed for a preliminary injunction. The underlying false advertising case continues.
Judge Zahid Quraishi noted that Wegovy prescription volume did not fall while the ads ran from April to July, that Lilly’s ads acknowledged the newer higher dose Wegovy, and that Lilly had already revised its Zepbound ads after a cease and desist letter, per BioPharma Dive. The judge also pointed at Novo’s own Ozempic ads claiming superiority over Trulicity without disclosing dose context. Novo sued in July 2026 over Lilly promotions built on a head to head trial it argues compared unequal doses. Novo said the ruling does not resolve the merits. The practical read: comparative advertising stays on the air through the obesity market’s biggest commercial quarter, and the bar for stopping a rival’s claims in this category is now visibly high.
Which deals crossed: Ferring, Labcorp, and Corxel?
Ferring agreed to buy IVF startup Gameto for up to $1.38 billion, Labcorp disclosed its third lab deal in eight days with Millennium Health, and obesity developer Corxel is going public through a SPAC with a $300 million private placement.
Ferring’s acquisition of Gameto, announced Tuesday, is reported at up to $1.38 billion by Endpoints News, with BioPharma Dive rounding to $1.4 billion; the upfront portion was not disclosed and the balance is development, regulatory and sales milestones. Gameto’s Fertilo uses stem cell derived ovarian support cells to mature eggs outside the body, which the company says can compress IVF and egg freezing cycles to as little as three days with fewer hormone injections. Fertilo is cleared in Australia and parts of Latin America and is in US Phase 3. Gameto had raised $127 million since 2020, so the headline total is roughly ten times its lifetime venture funding.
Labcorp’s purchase of Millennium Health, a specialty toxicology lab doing mass spectrometry urine testing and oral fluid testing, was announced Friday October 23 at undisclosed terms, per MedTech Dive. It follows the BioReference clinical assets deal of October 15, at $722 million cash plus up to $150 million in earnout, and the $210 million Myriad prenatal testing deal days later. Corxel, which licenses the oral GLP-1 CX11 from Vincentage Pharma, said Monday it will list through Amanat Acquisition Corp. with a $300 million private investment; CX11 showed up to 11.5% weight loss at 36 weeks in a US Phase 2 and could start an international Phase 3 as soon as the first quarter of 2027, per BioPharma Dive.
Which stocks moved on Tuesday, October 27?
Cell therapy and the ESMO follow through set the day’s single stock moves inside a strong tape, with the XBI up 2.9%. All closes are settled exchange figures and percentages are our arithmetic against Monday’s closes.
| Ticker | Close | Move | Context |
|---|---|---|---|
| ARTV | $4.27 | −30.7% | AlloNK discontinued, 94% staff cut, strategic review; first reaction session |
| IBRX | $8.63 | +10.5% | No company disclosure found; our checks surfaced only stale 2025 lung data and the May quarterly release |
| SMMT | $15.33 | +5.7% | Bounce after Monday’s HARMONi driven slide; still below its pre readout level |
| LQDA | $45.44 | +3.9% | Fifth straight up session, up 32.3% since October 16, with no disclosure we could find |
| HCM | $13.80 | −3.3% | Second session after SAFFRON’s full data; ADR |
| AMGN | $447.95 | +0.7% | Held Monday’s DeLLphi driven gain through a second session |
Context: XBI 154.84, up 2.9%; SPY 769.66, up 0.6%. Merck added 0.9% to $141.82 ahead of Thursday’s third quarter call.
Frequently asked questions
Why did Artiva stop developing AlloNK?
Durability fell short: responses did not last long enough to meet the company’s own bar for continued development, and two patient deaths added regulatory pauses, likely enrollment delays, and a weakened commercial case.
Were the two deaths caused by AlloNK?
Artiva says no. One was assessed by the investigator and an independent review as likely related to fludarabine conditioning chemotherapy, not the cells. The other is undetermined and was assessed as unrelated to any part of the regimen.
What happens to Artiva now?
A strategic review run with Leerink Partners, with roughly $319.1 million in preliminary, unaudited cash as of September 30. Options named by the company include a merger, asset sale or licensing, or dissolution.
Is this the end of cell therapy in autoimmune disease?
No, but it is a reset. Bristol Myers Squibb and Novartis both continue their paused or restricted CD19 CAR T programs, and the field’s open question is whether conditioning chemotherapy, not the cells, carries much of the risk.
What is fludarabine and why does it matter here?
Fludarabine is a decades old chemotherapy used to deplete a patient’s lymphocytes before cell therapy. A death assessed as likely related to fludarabine neurotoxicity implicates a backbone used across the modality, which is why Artiva plans to publish the case.
What did SAFFRON show?
Savolitinib plus osimertinib beat chemotherapy in MET driven, EGFR mutated lung cancer after osimertinib progression: median overall survival 17.8 versus 12.9 months, hazard ratio 0.68, with progression free survival 7.6 versus 5.4 months at 0.55.
Is savolitinib approved in the United States?
Not yet. AstraZeneca submitted the NDA, announced September 29, 2026. China approved the combination on the SACHI trial and Switzerland granted temporary authorization on SAVANNAH.
Did Novartis raise 2026 guidance?
Yes. Core operating income guidance moved to in line to low single digit growth, from an expected low single digit decline, on a quarter with sales up 4% and core operating income up 10% in constant currencies.
What is the Novartis $1.9 billion impairment?
A partial writedown of del-desiran, the lead asset from the Avidity acquisition, after its Phase 3 HARBOR myotonic dystrophy trial missed in September. The asset still carries $5.5 billion and Novartis says it stands by the deal.
What did DYNASTY-Breast02 show?
The trial met its progression free survival endpoint, but BioNTech and DualityBio called the improvement modest and the potential benefit limited, with interstitial lung disease above prior trials. No figures were released.
Will trastuzumab pamirtecan be filed in HER2 low breast cancer?
Undecided. The companies say they will discuss the data and determine next steps. Enhertu’s incumbency in HER2 low raises the bar for any filing built on a modest chemotherapy comparison.
How strong is the Ris-Rez osteosarcoma result?
A 0.33 hazard ratio for progression free survival is a large effect, reported from a China only, open label trial with 7.1 months median follow up, immature survival data, and 93% Grade 3 or higher treatment related adverse events.
What did the judge decide in Novo versus Lilly?
He denied Novo’s preliminary injunction on Monday, so Lilly’s comparative obesity ads continue. The false advertising case itself is unresolved and continues.
What is Gameto’s Fertilo?
A treatment that uses stem cell derived ovarian support cells to mature eggs outside the body, aiming to compress IVF and egg freezing cycles to as little as three days. It is cleared in Australia and parts of Latin America, with US Phase 3 underway.
What should readers watch next?
Boston Scientific’s Wednesday call with its first cyberattack quantification, Merck’s Thursday third quarter call, Friday’s FDA Expedited IND Pilot deadline, and Summit’s ivonescimab decision on November 14.
Sources
Primary sources
- Artiva Biotherapeutics press release, via Form 8-K exhibit 99.1, SEC, October 27, 2026
- AstraZeneca press release, SAFFRON Phase 3 results, astrazeneca.com, October 25, 2026
- Novartis Q3 2026 results materials, novartis.com, October 27, 2026
- BioNTech and DualityBio press release, DYNASTY-Breast02 topline, October 26, 2026
- Disc Medicine Form 8-K, offering closing, SEC, filed October 22, 2026
- Merck investor notice, Q3 2026 call October 29, merck.com
- Novartis press release, pelacarsen Lp(a)HORIZON topline, September 4, 2026
Trade press and other coverage, October 26 to 27, 2026
- BioPharma Dive: Artiva discontinuation; Novo versus Lilly ruling; Ferring and Gameto; Corxel SPAC
- Endpoints News: Artiva wind down; Ferring and Gameto at up to $1.38 billion; Treasury COINS rule analysis
- BioPharm International: Ris-Rez ARTEMIS-011 figures; zola-cel Breakfree-2 interim data
- Fierce Biotech: Ris-Rez 67% risk reduction
- MedTech Dive: Labcorp and Millennium Health; AdvaMed conference coverage
Market data: settled exchange closes for October 27, 2026; all percentage moves are our arithmetic.
