How is CAR T therapy for arthritis different from its use in cancer?

The mechanism is the same: engineered T cells target CD19 on disease-driving B cells and eliminate them. In cancers, targets are malignant and numerous; in RA they are normal-origin memory B cells with a damaging specificity. Side effects are proportionally milder.

CAR T erased rheumatoid arthritis in 3 patients who had exhausted all other options

Q: Does CAR T therapy for arthritis destroy the immune system?

No. CAR T cells eliminate CD19-carrying B cells; the immune system then repopulates from naive precursors without disease memory. Protective antibodies from prior vaccinations were preserved in all six Charité trial patients.

Rheumatoid arthritis had defeated every treatment tried on six patients at Berlin’s Charité hospital — four to eight biologics and targeted drugs over ten years, none of them sufficient. Then each received one infusion of cells engineered to hunt down the immune cells driving their disease. Three of them have needed nothing since.

The therapy is mivocabtagene autoleucel (miv-cel), a CD19-targeting CAR T cell product developed for blood cancers. In rheumatoid arthritis, disease-driving memory B cells — immune cells that recognize the body’s own tissues as foreign and keep triggering inflammation — carry the same CD19 surface marker that the therapy was built to track. The cells found them.

All six patients saw marked reductions in disease activity. Three achieved sustained remission without any rheumatoid arthritis medication for up to 12 months. A fourth patient showed initial improvement before disease returned. Autoantibodies — the molecules that signal the immune system attacking the body — declined by more than 90% across the group.

What the therapy actually does

Rheumatoid arthritis is driven by long-lived memory B cells that have accumulated over years in joint tissue, bone marrow, and lymph nodes. These cells remember the mistaken signal that told them to attack the body’s cartilage and joint lining, and they keep repeating that signal. Existing treatments — from methotrexate to biologics that block inflammation signals — suppress the consequences of that mistake. None of them erase the memory.

CD19 is a surface protein found on B cells at most stages of their development. It functions as a kind of name tag that marks them for destruction by CAR T cells, regardless of whether those B cells are healthy or disease-driving. The infusion triggers near-complete depletion of B cells from the blood and tissues.

What happens next is what makes this trial different. Rather than staying depleted, the B-cell system repopulates from naive precursor cells that carry no record of the original disease signal. When the Charité team measured the returning cells, they were predominantly naive. The autoantibodies characteristic of rheumatoid arthritis — anti-citrullinated protein antibodies (ACPA) — were no longer detectable in most patients. And protective antibodies from prior vaccinations against chickenpox and tetanus remained intact, suggesting that long-lived plasma cells providing baseline immunity were spared.

How the trial was designed

The phase 1 COMPARE trial enrolled six patients — three women, three men, aged 31 to 69 — with ACPA-positive rheumatoid arthritis who had each failed between four and eight prior targeted or biologic therapies. All disease-modifying drugs were stopped before treatment. Each patient received a preparatory course of lymphodepletion chemotherapy to clear the existing immune cell population, followed by a single infusion of miv-cel.

Follow-up ran from 36 to 52 weeks. The researchers, led by Prof. David Simon and Prof. Gerhard Krönke at Charité’s Department of Rheumatology and Clinical Immunology, collaborated with the German Rheumatology Research Center (DRFZ).

“Disease activity decreased markedly in all six patients,” Prof. Krönke reported. “Three patients were in sustained remission without any medication.”

Side effects were limited. Every participant experienced cytokine release syndrome — a temporary inflammatory reaction that is the expected response to CAR T infusion — but all six cases were mild to moderate. “We observed only a temporary, mild-to-moderate cytokine release syndrome in all participants,” said Dr. Marie Luise Hütter-Krönke, who oversaw safety assessment. No patient developed severe neurological complications. Infections were rare.

What this compares to

CAR T cell therapy has been used to treat blood cancers since 2017, when the first product received FDA approval. In those patients, cytokine release is frequently severe, sometimes requiring intensive care. The milder response in the Charité trial reflects a key difference: cancer patients often carry enormous numbers of malignant cells that all react simultaneously to the infusion; RA patients carry fewer disease-driving B cells, so the immune system’s response is proportionally less intense.

Rheumatoid arthritis affects roughly 18 million people worldwide. Current first-line treatments suppress disease activity in many patients, but approximately 30 to 40% of those with severe RA fail to achieve adequate control even with multiple biologics. The Charité patients had already failed the available B-cell-depleting therapies — rituximab among them — and were selected specifically because other options had been exhausted.

Three of six entering sustained remission in that population is a result without precedent in rheumatology. No prior intervention has produced medication-free remission in treatment-refractory RA.

What the trial does not settle

Six patients is the evidence base. That sample is appropriate for a phase 1 safety trial but does not establish efficacy. The three patients in remission may represent unusually responsive individuals; the one who relapsed after an initial response illustrates that the therapy does not work for everyone.

The durability of remission beyond 12 months is unknown. RA is a chronic disease, and it is possible that disease-driving memory B cells re-emerge from tissue reservoirs — particularly in joint tissue — that the infusion did not fully clear.

“The B-cell system essentially starts fresh,” Dr. Hütter-Krönke noted. “Whether that clean slate is preserved long-term depends on the disease environment the new cells encounter.”

Miv-cel is an autologous therapy: it is manufactured from each patient’s own T cells, a process that takes several months. The cost and lead time make it inaccessible to most patients globally. The head-to-head comparison with less resource-intensive options has not yet been made.

Common questions about CAR T therapy and arthritis

Is CAR T therapy approved for rheumatoid arthritis? No. Mivocabtagene autoleucel is not approved for any autoimmune disease. The COMPARE trial is a phase 1 study. Regulatory approval would require larger controlled trials showing efficacy and long-term safety.

How is this different from CAR T therapy for cancer? The mechanism is the same: engineered T cells track a surface marker on disease-driving cells and eliminate them. In blood cancers, those cells are malignant and numerous. In RA, they are normal-origin memory B cells with a damaging specificity. The body’s inflammatory response is proportionally milder, which explains the lower rate of severe side effects.

Does the therapy destroy the immune system? No. B cells that carry CD19 are eliminated; the immune system then repopulates from naive precursors without disease memory. Protective antibodies from prior vaccinations were preserved in all six patients.

Could this work for other autoimmune diseases? Possibly. Lupus, inflammatory myositis, and systemic sclerosis share a similar B-cell-driven mechanism. Clinical trials in those conditions are underway at multiple centers. No published results are available yet.

The phase 2 COMPARE trial will enroll ten additional patients and compare miv-cel directly against an approved B-cell-targeting drug. Enrollment is expected to begin before the end of 2026.

Three patients are still medication-free. The question the field is now watching is not whether CAR T can clear treatment-resistant arthritis, but how long the reset lasts.

Reference: Albach FN et al., “CD19 CAR T cell therapy for treatment-refractory seropositive rheumatoid arthritis: a phase 1 trial,” Nature Medicine, 2026. DOI: 10.1038/s41591-026-04603-3

Tags: medicine, autoimmune disease, clinical trial, rheumatoid arthritis, CAR T cell therapy, cell therapy