Artiva Halts Cell Therapy Program After Two Patient Deaths in Autoimmune Trial

For patients with severe autoimmune disease, the promise of cell therapy has felt like a lifeline. These are people whose own immune systems attack their bodies, often after years of cycling through drugs that only partially work. A one-time treatment that could reset a malfunctioning immune system and offer lasting remission is the kind of breakthrough that changes lives. That promise just took a serious hit.

Artiva Biotherapeutics announced Tuesday it is halting development of its lead cell therapy program, AlloNK, and exploring "strategic alternatives" for the company itself. In plain terms, that often means a company is preparing to sell itself, merge, or wind down. The San Diego biotech will lay off about 94% of its staff, a gutting reduction from the 115 full-time employees it reported at the end of June. The company still has a substantial cash cushion, $319.1 million as of the end of the third quarter, but that money is now more likely to fund an exit than a comeback.

The decision follows two patient deaths in a mid-stage trial testing AlloNK for rheumatoid arthritis, along with updated data suggesting the therapy simply isn't working well enough, long enough, to justify moving forward.

Artiva's approach differs from the CAR-T therapies many people associate with cancer treatment. Instead of engineering a patient's own T cells, which is a slow and expensive process, Artiva used natural killer cells, or NK cells, sourced from donated cord blood. Think of NK cells as the immune system's rapid-response unit: they don't need to be trained to recognize a threat the way T cells do. Because the cells come from donors rather than the patient, this "allogeneic" approach is designed to be faster and more scalable to manufacture. The therapeutic goal, however, mirrors the cancer application: wipe out the B cells that are driving the autoimmune attack.

What Went Wrong

Before any cell therapy is infused, patients typically undergo a preconditioning regimen, a short course of chemotherapy meant to clear out existing immune cells. This step prevents the body from rejecting the new cells and gives them a better chance to take hold. It's a bit like clearing weeds from a garden bed before planting something new. But that clearing process carries its own risks, and Artiva's trial illustrates just how serious they can be.

One patient in Artiva's Phase 2a study died from a fatal neurological event after receiving AlloNK. According to the company, both the trial investigator and an independent safety review concluded the complication was likely caused by toxicity from fludarabine, a chemotherapy drug widely used in preconditioning regimens across the cell therapy field, not by AlloNK itself. That fatality was enough to halt enrollment and dosing in the study and pause plans for a Phase 3 trial. Because fludarabine is used so broadly in this field, Artiva said it intends to publish additional details about the case in a scientific journal, a move that could help other companies and researchers understand and manage the risk.

A second patient died roughly two months after receiving AlloNK. In this instance, the investigator did not find a link between the death and either the study drug or the preconditioning regimen. The cause of death, however, remains undetermined. That lingering uncertainty is its own kind of unsettling, a reminder that even careful investigation doesn't always produce clear answers.

Separate from the fatalities, Artiva's own data raised doubts about whether AlloNK was ever going to be a durable solution. Updated Phase 2 results showed the therapy was still depleting B cells and producing clinical activity consistent with earlier findings. But longer-term follow-up told a less encouraging story: the effects weren't strong or lasting enough to support continued development. "Following an assessment of the risk-benefit of the therapy, likely enrollment delays and the overall development and potential commercial outlook, Artiva has decided to halt further development of AlloNK," the company said, adding that no additional patients will be dosed and the planned Phase 3 trial will not move forward.

Artiva isn't alone in confronting these problems. In late August, Novartis paused development of its own cell therapy, rap-cel, after immune-related complications led to three deaths, the company told Fierce Biotech. Just last week, Bristol Myers Squibb disclosed that a patient receiving its experimental therapy zola-cel developed septic pneumonia and died 44 days after treatment. Three companies, three different therapies, the same unsettling pattern of serious safety signals emerging in a field that had generated enormous enthusiasm.

Artiva has brought in Leerink Partners as a financial advisor to help navigate what comes next for the company. Leerink analyst Daina Graybosch, who tracks allogeneic cell therapy developers Nkarta and Fate Therapeutics, said the Artiva setback carries particularly negative implications for Nkarta. The reasoning comes down to the preconditioning regimen: like Artiva, Nkarta's therapy requires fludarabine, and neurotoxicity is a rare but recognized risk tied to that drug. Fate Therapeutics, by contrast, uses a lower-intensity preconditioning approach that skips fludarabine entirely, potentially sidestepping this specific risk.

Graybosch expects the bar for these therapies to rise sharply. Going forward, she said physicians and investors will likely demand evidence of prolonged clinical remission, more than 12 months, before they're willing to back further investment or adoption. Even therapies that clear that bar may end up with a narrower path to market than companies originally hoped. Leerink anticipates use could be limited to academic medical centers equipped to monitor for and manage complications, and possibly reserved for the most severe patients, restrictions that would shrink the commercial opportunity considerably.

Nkarta is scheduled to present Phase 1 data for its own therapy, NKX019, at the American College of Rheumatology conference in Orlando next month. Graybosch believes Nkarta's approach, which engineers NK cells to bind directly to a protein found on disease-causing B cells, could outperform AlloNK's unmodified cells. But she cautioned that proving superiority will require more data over time, and that the safety concerns now swirling around the field will likely make it harder to recruit patients for future trials. "We anticipate elevated attention on these presentations, especially NKX019's safety profile, following the field's recent toxicity setbacks," she said. "We expect to leave ACR with outstanding clinical uncertainty, as more limited patient follow-up for clinical durability will limit a satisfactory assessment of NKX019's benefit-risk profile."

Why It Matters

Behind every one of these trial numbers is a person who volunteered for an experimental treatment, often after conventional options had failed them. That willingness to take on risk is part of what makes clinical research possible, but it also demands rigorous safety oversight and honest public reporting when things go wrong. Artiva's decision to publish details about the fludarabine-linked death, rather than quietly shelving the information, is a small but meaningful act of transparency in a field still working out how to balance ambition with caution. The companies that survive this wave of setbacks will likely be the ones that take both halves of that equation seriously.