In 2021, a woman with severe lupus at University Hospital Erlangen in Germany became the first person to receive CAR-T cell therapy for an autoimmune disease. Within weeks, her symptoms receded. She stopped taking immunosuppressive drugs. More than two years later, she remained in remission. That single case launched a wave of clinical research that, by early 2026, has produced results across multiple autoimmune conditions and prompted dozens of new trials worldwide.
CAR-T therapy works by reprogramming a patient’s own immune cells (or, in newer versions, donor cells) to hunt and destroy specific targets. In cancer, those targets are malignant cells. In autoimmune disease, the target is CD19, a protein found on the surface of B cells, the immune cells that, when they malfunction, produce the antibodies that attack healthy tissue in diseases like lupus, inflammatory myositis, and systemic sclerosis. Wipe out the rogue B cells, the theory goes, and the immune system can rebuild itself without the self-destructive programming.
The early evidence suggests the theory holds, at least for some patients. But the data also reveals limits that will determine whether this approach can move from a handful of academic medical centers to widespread clinical use.
The strongest results so far
The most cited evidence comes from a 15-patient case series published in the New England Journal of Medicine by Georg Schett and colleagues at the University of Erlangen. The study enrolled patients with three different autoimmune diseases: systemic lupus erythematosus, idiopathic inflammatory myositis, and systemic sclerosis. All had failed multiple rounds of conventional immunosuppressive therapy. Each received a single infusion of autologous CD19-targeted CAR-T cells, meaning the engineered cells were manufactured from their own blood.
Every one of the 15 patients was able to discontinue immunosuppressive medication. All maintained sustained clinical and serologic responses over the follow-up period. For people who had spent years cycling through drugs like mycophenolate, cyclophosphamide, and rituximab, drugs that suppress the entire immune system and carry risks ranging from serious infections to organ toxicity, stopping treatment while staying in remission was a profound change.
A separate line of research has tested whether the same principle works with off-the-shelf donor cells, which could make the therapy faster and far cheaper to produce. In a trial registered as NCT05859997, three patients with severe refractory lupus in China received allogeneic CD19 CAR-T cells at a dose of 1×10^6 cells per kilogram. After 12 months of follow-up reported in Cell Reports Medicine, none of the three developed graft-versus-host disease, cytokine release syndrome, or immune effector cell-associated neurotoxicity syndrome. These are the three complications clinicians worry about most with donor-derived cell therapies, and their absence in all three patients is a meaningful early safety signal.
A third study, also from the Erlangen group, focused on six patients with severe diffuse systemic sclerosis, a condition in which the immune system drives progressive scarring of the skin, lungs, and other organs that can be fatal. Researchers tracked a broad set of outcomes, including event-free survival, skin thickening scores, lung imaging, and a composite measure called ACR-CRISS probability, as detailed in The Lancet Rheumatology. The choice to measure skin and lung function alongside survival reflects what matters most to patients: not just staying alive, but regaining the ability to move, breathe, and use their hands.
Across these three studies, the consistent finding is that CD19-targeted CAR-T therapy can deplete the autoreactive B cells driving disease and allow the immune system to reconstitute without immediate relapse. A Nature news feature surveying research across lupus, rheumatoid arthritis, and ulcerative colitis has described the collective results as evidence that the field is approaching broader clinical use, while noting that expert definitions of what counts as true remission in autoimmune disease remain unsettled.
Where the limits are showing
The clearest evidence of a limitation comes from a single patient. A person with idiopathic inflammatory myositis who initially responded to CD19 CAR-T therapy later relapsed. The case, published in Nature Medicine, prompted the Erlangen team to try a second, different CAR-T product, this one targeting B-cell maturation antigen (BCMA). Unlike CD19, which sits on the surface of most B cells, BCMA is found on long-lived plasma cells, the factories that churn out antibodies for years or even decades. Some of these plasma cells can survive CD19-directed killing because they no longer express CD19, continuing to produce the autoantibodies that drive disease even after the broader B-cell population has been wiped out.
That relapse raises a question no current study has definitively answered: if a subset of patients harbor CD19-negative plasma cells, will a single-target approach produce lasting remission for everyone? A sequential protocol, hitting CD19 first and BCMA second, could theoretically close the gap. But this hypothesis rests on one case report, and no controlled trial has compared dual-target or sequential strategies against single-target therapy.
Sample size is the other obvious constraint. The published studies involve 3 to 15 patients each. Numbers that small can reveal promising signals but cannot establish how often the therapy fails, which patients are most likely to relapse, or what the safety profile looks like beyond one year. The allogeneic lupus study, for instance, followed patients for 12 months. That is a meaningful window, but autoimmune diseases often wax and wane over decades, and 12 months of remission is not the same as a cure.
Then there is cost. Approved CAR-T products for blood cancers, such as Novartis’s Kymriah and Gilead’s Yescarta, carry list prices ranging from roughly $373,000 to more than $475,000 per infusion in the United States. Manufacturing is complex and time-consuming, requiring weeks of cell processing in specialized facilities. None of the autoimmune studies published so far include cost-effectiveness analyses. The allogeneic (donor-cell) approach could reduce both cost and turnaround time by eliminating the need to manufacture a custom product for each patient, but no published data quantify that advantage yet.
Regulatory status adds another layer of uncertainty. As of spring 2026, no CAR-T therapy has received FDA or European Medicines Agency approval for any autoimmune indication. More than 50 clinical trials are registered globally, according to ClinicalTrials.gov listings, but most are early-phase studies. The path from a 15-patient case series to a licensed therapy typically takes years and requires the kind of large, randomized, controlled trials that have not yet been completed in this space.
What patients and doctors should weigh
For the general-audience reader, the most important thing to understand is the gap between what these results show and what they do not yet prove. The data are real and striking: people with diseases that had resisted every available drug stopped all treatment after a single infusion and stayed well for months. That is not hype. It is documented in peer-reviewed journals by independent research groups.
But the patients in these studies were carefully selected. They had severe, refractory disease yet were stable enough to undergo lymphodepleting chemotherapy (a necessary step before CAR-T infusion) and weeks of close monitoring at specialized academic centers. Outcomes in that controlled setting may not reflect what would happen if the therapy were offered more broadly, in community hospitals or to patients with different disease stages or comorbidities.
Safety, too, deserves a careful read. The absence of cytokine release syndrome, neurotoxicity, and graft-versus-host disease in the allogeneic lupus trial is encouraging, particularly given the history of these complications in oncology. But with only three patients, rare adverse events could easily go undetected. As more patients are treated, clinicians will need to watch for delayed effects: prolonged B-cell depletion that leaves patients vulnerable to infections, or theoretical risks like secondary malignancies, which early studies are not large enough to detect.
Definitions of remission also matter more than they might seem. Clinical scores, blood markers, imaging, and patient-reported symptoms can all point in different directions. The current reports describe “sustained clinical and serologic responses,” but there is no universally accepted standard for what constitutes complete remission versus low disease activity after CAR-T-induced immune reset. Until consensus criteria are applied consistently across trials, comparing results between studies will remain difficult.
What comes next in the research pipeline
The next phase of evidence will need to answer the questions these early studies have raised. Randomized trials comparing CAR-T therapy against current standards of care, such as belimumab for lupus or nintedanib for systemic sclerosis, could clarify which patients benefit most and whether intervening earlier in the disease course changes long-term outcomes. Head-to-head studies of single-target versus dual-target or sequential CAR designs could address the relapse scenarios hinted at in the BCMA case report. And health economic analyses will be essential to determine whether a one-time, high-cost infusion ultimately reduces the cumulative burden of chronic immunosuppression, repeated hospitalizations, and progressive organ damage.
Several of these trials are already underway or in planning stages at centers in Germany, China, and the United States. The pace of enrollment and the rigor of trial design will determine how quickly answers arrive.
For now, the evidence is best understood as a powerful proof of concept. For a small number of patients with devastating autoimmune diseases, CD19-directed CAR-T cells have already rewritten the trajectory of their illness. Whether that transformation can be delivered safely, affordably, and equitably to the millions of people living with these conditions worldwide remains the central unanswered question of this emerging field.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
