Novartis and Bristol Myers Pause Autoimmune CAR-T Trials Over Safety Signals

The excitement around using CAR-T cell therapy for autoimmune disease has hit a note of caution. Novartis and Bristol Myers Squibb have paused autoimmune CAR-T trials after cases of a rare, potentially life-threatening immune reaction.

What was paused

Novartis halted trials of rap-cel across multiple autoimmune indications — including lupus, myasthenia gravis and multiple sclerosis — after identifying three cases of immune effector cell-associated hemophagocytic syndrome (IEC-HS), described as rare and possibly life-threatening. The holds began on August 24, and the company is engaging with regulators.

Bristol Myers Squibb voluntarily paused enrolment in trials of zola-cel after detecting “transient and reversible inflammatory events,” including one prior IEC-HS case, acting “out of an abundance of caution.”

What IEC-HS is

Immune effector cell-associated hemophagocytic syndrome is a severe hyperinflammatory state that can follow cell therapy. The immune system enters a self-amplifying activation loop, producing high fevers, organ dysfunction, blood count abnormalities and, in severe cases, multi-organ failure.

It is related to but distinct from cytokine release syndrome, the more familiar CAR-T toxicity. IEC-HS tends to appear later, is harder to recognise, and responds less predictably to the standard treatments used for cytokine release syndrome. It is rare, and it is genuinely dangerous.

Why the risk calculation is different outside cancer

This is the central point. CAR-T was developed for patients with refractory blood cancers who had exhausted other options and faced death within months. Against that, a meaningful risk of severe toxicity — even mortality — is an acceptable trade.

Autoimmune patients are in a different position. They are frequently young, they have available treatments that control disease imperfectly but adequately, and they are not facing imminent death. A therapy with a small chance of a life-threatening hyperinflammatory syndrome is judged against a very different baseline.

Three cases would be unremarkable in an oncology programme. In an autoimmune programme, they stop it.

What may be driving it

Analysts identified a plausible culprit. The accelerated cell-production technology used in these next-generation therapies “could be driving increased cell expansion and the reported toxicities,” noted William Blair’s Sami Corwin, adding this “will clearly need to be monitored” for its impact on adoption.

The reasoning is coherent. Conventional CAR-T manufacturing takes weeks, and rapid-manufacturing processes were developed to shorten that — a major practical advantage. But cells produced quickly may be less differentiated and more proliferative, expanding more aggressively once infused. More expansion means more potency, and potentially more inflammation than intended.

If that is the mechanism, the problem lies in the manufacturing process rather than in the concept of autoimmune CAR-T — which would be the more tractable of the two possibilities.

What is at stake scientifically

The reason this matters is how striking the efficacy has been. CAR-T in autoimmune disease has produced reports of drug-free remission in rheumatoid arthritis and lupus — patients off all immunosuppression after a single treatment.

That is not incremental improvement. It suggests deep B-cell depletion allows the immune system to reconstitute without the autoreactive population, a genuine reset rather than ongoing suppression. Very little in autoimmune medicine has ever offered that.

The prospect is what drove enormous investment across the sector, including the antibody-based approaches now being pursued as scalable alternatives.

The field has not stopped

Other developers, including Cabaletta Bio and Kyverna, are continuing their programmes. That is informative rather than merely reassuring: it suggests the problem may be specific to particular manufacturing approaches or constructs rather than intrinsic to the modality.

The pauses are also voluntary and precautionary in at least one case, not regulatory shutdowns. Companies pausing enrolment while investigating is the system functioning as intended.

What to watch

Three things will determine how this resolves. Whether the cases cluster around a specific manufacturing process, which would localise the problem. Whether predictive factors can be identified, allowing higher-risk patients to be excluded. And whether the holds are lifted with protocol amendments — monitoring requirements, dose adjustments, patient selection criteria — or extended.

What the pauses mean for the antibody alternatives

These holds land in the middle of a broader industry pivot, and their effect on it is not straightforward.

Large pharmaceutical companies have been investing heavily in off-the-shelf antibody approaches — bispecific and multi-specific engagers designed to achieve deep B-cell depletion without the manufacturing burden of cell therapy. The commercial argument has always been scalability. A safety problem in CAR-T strengthens that argument by adding a second reason to prefer an alternative.

The complication is that IEC-HS is a consequence of intense immune activation, and any therapy achieving comparably deep depletion may provoke comparable inflammation regardless of format. If the toxicity tracks the biology rather than the manufacturing, antibody-based approaches would inherit it. If it tracks how the cells were made, they would not.

Which of those turns out to be true is the most consequential open question in the field, and it will not be settled by the current investigations alone.

The broader lesson is one the field will keep encountering: repurposing a powerful cancer therapy for patients who are not dying means the safety bar moves, and effects tolerated in oncology may be disqualifying elsewhere. Clinical and business news, not medical advice.