A woman with three autoimmune diseases was bedridden and dependent on blood transfusions. Seven days after a single experimental infusion, she got out of bed.

“I just saw her yesterday. She’s perfectly fine,” her physician Fabian Müller of University Hospital Erlangen in Germany said in May — 11 months after the treatment.

The therapy was not designed for autoimmune disease. CAR T cell therapy, one of oncology’s biggest breakthroughs, was built to hunt cancer. Now it is being redirected to strip out the malfunctioning immune cells that drive lupus, multiple sclerosis, stiff person syndrome, and a growing list of autoimmune conditions.

The mechanism is straightforward once you strip away the jargon. B cells normally produce antibodies against viruses and bacteria. In autoimmune disease, they start producing antibodies against the body’s own tissues instead. CAR T therapy engineers a patient’s T cells — the immune system’s hunters — to recognize and destroy those rogue B cells, giving the immune system what amounts to a fresh start.

Early trials, striking results

The FDA approved the first CAR T cancer treatment in 2017. A German team reported the first autoimmune application, in a lupus patient, in 2021. Clinical trials have since multiplied.

Results have been remarkable so far. In a Kyverna Therapeutics trial of 26 patients with stiff person syndrome — a rare, debilitating condition with no approved treatments — a single CAR T infusion produced statistically significant improvements across every measured endpoint at 16 weeks, according to data presented at the American Academy of Neurology in April. Two-thirds of patients who had needed walking aids no longer required them. All 26 remained free of other immunotherapies through their last follow-up.

Cartesian Therapeutics reported that two-thirds of 15 autoimmune patients receiving its mRNA-based CAR T therapy showed improvement, with no serious long-term side effects.

The risk gap

The calculus shifts when the underlying disease is not cancer. CAR T therapy can trigger severe inflammation and, in rare cases, the engineered cells themselves have turned malignant. That is an acceptable trade-off against a lethal blood cancer. Against lupus, it is harder to justify.

“Causing a secondary cancer may be an acceptable risk when treating a life-threatening cancer, but probably not for autoimmunity,” Matt Lunning, medical director for gene and cellular therapy at Nebraska Medicine, told Knowable Magazine.

Patients also face up to a year of vulnerability to infection while their B cell populations rebuild. Durability remains uncertain — some patients have seen disease-driving cells return, requiring additional treatment.

Newer approaches aim to narrow these risks. Cartesian’s mRNA-based therapy uses short-lived genetic instructions rather than permanent DNA modifications, so engineered cells lose their ability to target B cells after a limited period, eliminating the risk that lingering modified cells could turn cancerous.

The price problem

CAR T therapy costs between $200,000 and $600,000 per cancer treatment, driven largely by the need to genetically engineer each patient’s cells individually in a laboratory. Researchers are pursuing “off-the-shelf” versions using donor cells — one donor’s blood could theoretically produce CAR T cells for more than 1,000 patients, according to Bing Du, an immunologist at East China Normal University in Shanghai.

Müller argues the upfront cost may be offset by eliminating a lifetime of immunosuppressive drugs and lost productivity. “It costs a lot to start with, but you save a lot of money in the long run.”

For Jan Janisch-Hanzlik, a Nebraska nurse with multiple sclerosis who received an off-the-shelf CAR T infusion in June 2025, the equation is personal. Nearly a year later, she no longer uses a cane, rarely falls, and recently visited the Grand Canyon. She still has symptoms — right leg weakness, numbness, occasional difficulty finding words. When she asks her doctors whether she will keep improving, their answer is honest: “We don’t know, you’re the first. We’re just going to have to wait and see.”

Sources