Novartis halts 8 trials of a cell therapy for lupus and MS after 3 deaths

Novartis has stopped eight clinical trials of an experimental immune cell therapy after three patients died, the company confirmed in late August.¹ The treatment, called rapcabtagene autoleucel or rap-cel, was being tested in autoimmune and nerve diseases such as lupus, rheumatoid arthritis, vasculitis, multiple sclerosis and myasthenia gravis.² At about the same time, Bristol Myers Squibb paused enrollment in trials of its own, similar therapy, called zola-cel, after seeing short-lived inflammatory side effects.³

This is Biohack’s first article on this subject. The pause is now about two months old, and the companies have not yet published case details or a cause.

What happened

Novartis said it began placing holds on the trials on August 24, and the companies confirmed the pauses on Monday, August 31.³ Novartis found three cases of a rare and possibly life-threatening reaction called immune effector cell-associated hemophagocytic syndrome, or IEC-HS. Complications from those events led to “fatal outcomes,” according to a company spokesperson.³ In plain terms, IEC-HS is a severe, whole-body inflammatory reaction to immune cell therapies.⁴ It is an extreme form of immune overactivity that can lead to organ damage.⁴

The hold covers screening, randomization (assigning patients to study groups) and giving the treatment, across eight studies.⁴ The affected program spans lupus and lupus-related kidney disease, systemic sclerosis (a disease that hardens skin and organs), ANCA-associated vasculitis (inflamed blood vessels), myositis (muscle inflammation), rheumatoid arthritis, Sjögren’s disease, generalized myasthenia gravis and multiple sclerosis.⁴

A Novartis spokesperson told BioPharma Dive:

“The temporary halt will allow for a more comprehensive review of the evolving clinical and safety data across the program.”

³

Two Novartis trials of the same therapy in people with lymphoma and leukemia, both blood cancers, were not included in the pause.²

How the therapy works

CAR-T therapy starts by collecting a patient’s own T cells, a type of immune cell. These are genetically engineered in a lab to recognize a specific target, then infused back into the patient, where they multiply and attack that target.⁴ In cancer, the target is tumor cells. In Novartis’s autoimmune program, the target is B cells carrying a marker called CD19. B cells make antibodies and play a central role in autoimmune attacks.⁴

The idea is different from standard treatment. Instead of suppressing an overactive immune system day after day, the therapy aims to wipe out a population of misbehaving immune cells and let the immune system rebuild, possibly giving a deeper and longer-lasting reset.⁴

One of the paused studies shows what testing looked like. It was a mid-stage trial in adults with active lupus or lupus kidney disease that had not responded well to at least two earlier treatments. Patients got a single infusion after a “lymphodepletion” step, drugs that clear out existing immune cells first. Everyone knew which treatment they got, and there was no comparison group.⁵ Researchers planned to check kidney response and remission at 24 and 52 weeks, with follow-up out to 76 weeks.⁵

What Bristol Myers Squibb did

Bristol Myers voluntarily paused enrollment in its autoimmune trials of zola-cel (zolacabtagene autoleucel).³ It did not stop the studies outright.⁶ A company spokesperson said the move was made “out of an abundance of caution” after routine safety checks found “transient and reversible inflammatory events,” and that the company aims to resume testing “as quickly as possible.”³ Unlike the Novartis cases, the reported Bristol Myers events were not fatal.⁴

Results published in February had already described one case of IEC-HS in an early study of zola-cel.³ The company said the drug’s safety profile “remains consistent with the known profile of CAR-T therapies,” and that it still has confidence in the treatment.³ Bristol Myers had been set to start a zola-cel trial in systemic sclerosis in September.⁶

What is still not known

Much of the picture is missing. Novartis has not said when the deaths happened or described the individual cases.⁴ Everything known so far comes from company statements to reporters, not from published data.

Novartis said this kind of severe immune reaction is a known risk of CAR-T therapy in general.¹ The company is working with independent safety boards to understand what happened, and it is looking at whether dangerous side effects can be caught earlier.¹ It said it is sharing information with health authorities and will keep monitoring patients who have already been treated.¹

One possible explanation is being discussed, though it is not confirmed. Both rap-cel and zola-cel are made with newer technology meant to speed up production compared with earlier CAR-T products.³ William Blair analyst Sami Corwin wrote that this technology “could be driving increased cell expansion and the reported toxicities,” while cautioning that other factors could be involved.³ Other developers, including Cabaletta Bio, Kyverna Therapeutics and Autolus Therapeutics, make their therapies with more traditional processes.⁶

Alfred Kim, director of the Washington University Lupus Center, gave a similar view to Healio:

“Rap-cel and zola-cel are both made on rapid manufacturing platforms, and the working hypothesis is that faster production yields fitter T cells that expand greater in the patient, with immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome (IEC-HS) sitting at the severe end of that inflammation.”

² Kim reports receiving research funding from Bristol Myers Squibb and Novartis, among other companies.²

Why the risk weighs differently here

CAR-T is an established treatment for several blood cancers, where it has produced lasting responses in some patients.⁴ In advanced cancer, patients and doctors may accept large risks because the disease can be quickly life-threatening. People with autoimmune disease can be very sick but are often expected to live for decades with their condition, so the acceptable level of risk is different.⁴

Kim put it this way: people with lupus, myositis, scleroderma or other autoimmune diseases “are often younger, are not facing imminent death and increasingly have other options.”² He also said: “Three deaths have to be taken seriously, and pausing was the only call. This is not a verdict, though.”²

The interest in this approach comes from early results. Case reports and early studies have suggested that one CAR-T treatment can bring long remissions in some people with hard-to-treat autoimmune disease.⁴ At a European rheumatology meeting earlier in 2026, Novartis showed early data from its studies in severe muscle inflammation and systemic sclerosis, reporting improvements and rapid loss of B cells, and described the safety profile as manageable at the time.⁴ These were early, small studies, and the new deaths put those results in a more complicated light.

How it fits the wider field

Novartis and Bristol Myers both already sell CAR-T products for certain cancers, and both are among the leaders trying to turn these therapies against autoimmune disease.³ They are not alone. Cabaletta Bio has a late-stage study in myositis, and Kyverna Therapeutics could submit an approval application in stiff person syndrome by the end of the year.³

Because two companies paused related programs at nearly the same time, regulators, researchers and manufacturers are likely to watch this kind of immune reaction as a possible issue for the whole class of treatments, not just one product.⁵ “This will clearly need to be monitored to determine how it could impact the total market adoption of these products in the future,” Corwin wrote.³

What to watch next

• The Novartis review. Whether the company and independent safety boards publish details of the three cases and a likely cause.¹
• Bristol Myers’s restart. Whether zola-cel enrollment resumes, and whether the planned systemic sclerosis trial goes ahead.⁶
• Changes to trial rules. Observers expect possible tighter patient selection, better early detection of toxic reactions and closer monitoring, which could slow the move of CAR-T into long-term diseases.⁵
• Other developers. How companies using slower, traditional manufacturing report their own safety data.⁶

Biohack will update this thread when the companies or regulators share new information.