LONDON — Cellular therapy continues to advance treatment options for severe, refractory autoimmune diseases, according to newly released data presented at the European Alliance of Associations for Rheumatology (EULAR) 2026 Annual Meeting.

From autologous chimeric antigen receptor (CAR) T-cell therapies to novel off-the-shelf allogeneic natural killer (NK) cell platforms, researchers reported high rates of drug-free clinical remission in patients who had previously failed multiple targeted biological therapies. However, as these living drugs advance into clinical development, experts are also gaining insights into the timelines, characteristics, and management of disease relapse.

As the array of targeted B-cell depletion therapies expands, identifying specific tissue biomarkers and cellular phenotypes through patient biopsies is a research priority, Wolfgang Merkt, MD, Division of Rheumatology, Department of Hematology, Oncology and Rheumatology, Internal Medicine V, Medical Faculty, Heidelberg University, Heidelberg, Germany, told Medscape Medical News.

Obe-cel CAR T-Cell Therapy Induces Rapid Remissions in Severe Systemic Lupus Erythematosus (SLE)

In the phase 1 CARLYSLE study, researchers evaluated obecabtagene autoleucel (obe-cel), a novel autologous CD19-directed CAR T-cell therapy engineered with a fast off-rate binding domain designed to reduce immunotoxicity. The study enrolled patients with severe, refractory SLE across two fixed-dose cohorts: 50 million cells (50M) and 100 million cells (100M).

At baseline, adult patients had active disease (median SLE Disease Activity Index-2000 [SLEDAI-2K] scores of 17.0 and 18.0, respectively) and a history of failing multiple previous lines of therapy, including B-cell targeted agents. All patients in the 50M cohort and two out of three in the 100M cohort had active class III or IV lupus nephritis. At the data cutoff, the median follow-up was 11.4 months for the 50M cohort and 3.3 months for the 100M cohort. The therapy demonstrated a favorable safety profile, with no cases of immune effector cell-associated neurotoxicity syndrome (ICANS) or grade 2 or higher cytokine release syndrome (CRS) observed at either dose.

Clinically meaningful reductions were observed over time in SLEDAI-2K and Physician Global Assessment scores across both cohorts. Laboratory markers also improved, with decreasing anti-double-stranded DNA (dsDNA) antibodies and rising complement 3 levels at 1 month in most patients. Renal parameters turned positive for patients with lupus nephritis, showing decreased median urinary protein-creatinine ratios and stabilized or increased estimated glomerular filtration rates. Five of the six adult patients (83.3%) in the 50M cohort reached Definition of Remission in SLE remission at a median of 5.1 months.

Additionally, three patients (50.0%) achieved a complete renal response within the first month. Robust CAR T-cell expansion and profound B-cell depletion occurred in all patients. Median time to loss of CAR T-cell persistence was 3.0 months in the 50M cohort and 2.0 months in the 100M cohort. The median time to B-cell recovery was approximately 6.0 months for the 50M cohort and 5.8 months for the 100M cohort, with more than 90% of reconstituted B cells in the 50M cohort characterized as transitional or naive at the time of recovery.

“When we’re using B-cell depletion, for example, with a monoclonal antibody like rituximab, patients relapse around 3 months,” said Maria Leandro, MD, a rheumatologist at University College London (UCL) Hospitals and an honorary senior lecturer at UCL in London, England. “This is completely different. There is a qualitative difference in depleting with a lifelong drug that is able to deplete very profoundly in tissues.”

A phase 2 open-label study, the Lumina trial, has commenced and is actively recruiting patients with severe refractory SLE and active lupus nephritis using the 50M dose.

First-in-Rheumatoid Arthritis (RA) CAR T Trial Induces Remission Signals and Seroconversion 

Phase 1 of the prospective, open-label COMPARE trial provided the first clinical and translational data evaluating mivocabtagene autoleucel (miv-cel), an autologous, fully human CD19-directed CAR T-cell therapy, in patients with active, anticitrullinated protein antibody (ACPA)-positive, treatment-refractory RA.

The trial enrolled six adult patients (50% female) with a median age of 51.5 years who had high baseline disease activity (median Disease Activity Score in 28 joints using C-reactive protein [DAS28-CRP] of 5.5) and extensive prior treatment exposure (median of 6.5 biologic/targeted synthetic disease-modifying antirheumatic drugs [DMARDs]), including four patients who had previously failed rituximab. Following the discontinuation of all DMARDs and standard fludarabine-cyclophosphamide lymphodepletion, patients received a single miv-cel infusion. At the time of data cutoff, patients were followed for a median of 24 weeks (range, 16-36 weeks).

The safety profile was favorable. CRS was limited to mild-to-moderate events, with four patients experiencing grade 1 and two patients experiencing grade 2 CRS; all episodes were managed with standard treatment protocols. No cases of ICANS or unexpected toxicities occurred. Transient episodes of neutropenia were observed across all patients.

Following a single infusion, CAR T cells expanded rapidly in the blood, peaking within 3 weeks before gradually declining. B cells were profoundly depleted not only in peripheral blood but also in tissue compartments, as demonstrated by synovial and bone marrow biopsies.

Disease activity decreased across all patients, yielding a median DAS28-CRP reduction of 41% from screening to the latest follow-up, and half of the patients achieved full DAS28-CRP clinical remission. This clinical response was mirrored by impressive reductions in joint swelling counts and local inflammation seen on Doppler ultrasound and PET-CT joint imaging. In addition, autoantibody titers dropped markedly: four of six patients achieved sustained seroconversion to normal values for ACPA, and five of six achieved seroconversion to normal values for rheumatoid factor immunoglobulin M. Conversely, protective vaccine titers, such as tetanus, remained largely stable.

During follow-up, profound B-cell depletion was sustained in most patients, with reconstitution observed in two individuals. Translational tracking noted that repopulating B cells were predominantly naive or transitional phenotypes, whereas antigen-experienced, autoreactive cyclic citrullinated peptide-reactive B cells remained absent. Except for one patient who experienced a moderate flare requiring glucocorticoid reintroduction following steroid withdrawal, all patients remained completely off background immunosuppressive therapies during the follow-up period.

Phase 2 of the COMPARE trial is ongoing, evaluating a randomized design where five patients receive rituximab and five receive CAR T-cell therapy, with an integrated crossover option for rituximab nonresponders at 6 months.

Rese-cel CAR T-Cell Therapy in Idiopathic Inflammatory Myopathies

The ongoing, open-label RESET-Myositis trial that evaluated resecabtagene autoleucel (rese-cel; formerly called CABA-201), a fully human, autologous CD19-directed CAR T-cell therapy designed to deeply and transiently deplete CD19-positive cells to enable an immune system reset.

The trial enrolled adults and young patients across four independent cohorts: dermatomyositis (DM), antisynthetase syndrome (ASyS), immune-mediated necrotizing myopathy (IMNM), and juvenile idiopathic inflammatory myopathy (JIIM). Eligible patients had refractory, active disease on two or more core set measures and a positive myositis-specific autoantibody. All nonglucocorticoid immunomodulatory therapies and DMARDs were discontinued prior to weight-based preconditioning lymphodepletion and a single flat infusion of rese-cel.

As of the data cutoff, 13 patients had completed at least 28 days of safety follow-up. Rese-cel was well tolerated across the entire cohort. CRS was limited to grade 1 fever only in four patients, and no cases of grade 2 or higher CRS, ICANS, or serious infections occurred. The single juvenile patient with JIIM experienced a transient episode of grade 2 febrile neutropenia.

Efficacy and durability outcomes varied substantially across the disease subsets:

• DM and AsyS. Among patients with at least 16 weeks of follow-up, 80% (three of three with DM and one of two with ASyS) achieved at least a moderate Total Improvement Score (TIS ≥ 40) at week 16 while remaining entirely off immunomodulatory drugs and on no more than low-dose steroids. All patients with DM successfully maintained their clinical responses through week 52 or their latest follow-up. However, the two patients with ASyS experienced a recurrence of disease activity and ultimately had to resume other myositis treatments.
• IMNM. Clinical responses were more modest in this cohort; only two of four patients with at least 24 weeks of follow-up achieved a minimal TIS response (TIS ≥ 20) off immunomodulatory medications and low-dose steroids by week 24.

Serologically, clinical responders demonstrated distinct downward trends in their pathogenic myositis-specific autoantibodies. Conversely, persistence or recurrence of baseline autoantibodies was observed in patients with ASyS or IMNM who either lost their initial clinical response or failed to achieve any minimal clinical improvement.

Pharmacokinetic and cell-kinetic data from 12 adult patients confirmed robust cellular expansion, with rese-cel peaking at a median of 13 days post-infusion. Peripheral B cells reached a minimum at a median of 11 days, with an average B-cell depletion duration of 49 days among patients who showed B-cell repopulation at the data cutoff. Consistent with an immunologic reset, repopulating cell populations were characterized predominantly as transitional naive B cells.

“The vast majority of first-antisynthetase patients went into drug-free remission, flared, and then had to be rescued again with immunomodulatory therapy,” said Erin Wilfong, MD, an assistant professor and rheumatologist at Vanderbilt University Medical Center in Nashville, Tennessee. “We’ve actually been rescuing with rituximab, which shouldn’t work, but is working.”

Based on these findings, researchers concluded that long-lived, CD19-negative plasma cells not cleared by the therapy may continue to secrete pathogenic autoantibodies and drive subsequent flares in certain myositis phenotypes. An expanded phase 2b study has initiated enrollment to further evaluate the efficacy of rese-cel specifically in patients with DM and ASyS.

Relapses Following CD19 CAR T-Cell Therapy 

A retrospective study analyzed the frequency of relapses and subsequent management strategies in an institutional cohort of 50 patients with refractory B cell-driven autoimmune diseases treated with zorpocabtagene autoleucel (zorpo-cel).

Over the follow-up period, six of 50 treated patients experienced a return of disease activity, representing an overall relapse rate of 12%. The probability of disease recurrence was highly dependent on the underlying autoimmune indication: Relapses were rare in SLE, affecting just one of 28 patients (3.6%), intermediate in systemic sclerosis (SSc) at 14.3% (two of 14 patients), and highest in idiopathic inflammatory myositis (IIM), where three of eight patients (37.5%) experienced a flare. The median time to relapse after the initial cell therapy was 13 months, with a range spanning 9-24 months.

The researchers detailed the heterogeneous clinical presentations and tailored therapeutic rescue strategies utilized for each of the six relapsing patients:

• Patient 1 was a 46-year-old woman with Jo1-positive ASyS who relapsed with active myositis 9 months after zorpo-cel. She was successfully rescued with B-cell maturation antigen (BCMA)-directed CAR T-cell therapy (idecabtagene vicleucel), achieving a sustained, drug-free American College of Rheumatology (ACR) Major Clinical Response that has persisted through 18 months of follow-up.
• Patient 2 was a 44-year-old woman with PL7-positive ASyS who presented with progressive interstitial lung disease (ILD) and hiker’s feet 18 months post-infusion. Combined treatment with the anti-CD20 monoclonal antibody obinutuzumab and the antifibrotic nintedanib effectively improved her skin manifestations, lung function, and radiographic signs of ILD through 10 months of follow-up.
• Patient 3 was a 57-year-old woman with Jo1-positive ASyS. She relapsed with myositis and arthritis after 10 months. Induction with the JAK inhibitor tofacitinib provided rapid but temporary clinical relief; following a recurrence of disease activity 3 months later, she was transitioned to BCMA-targeted CAR T-cell therapy under the CARAMBA study.
• Patient 4 was a 26-year-old woman with dsDNA-positive SLE who experienced a severe relapse at 9 months characterized by new-onset transverse myelitis, arthritis, and cutaneous lesions. Treatment with BCMA-targeted CAR T cells (idecabtagene vicleucel) led to extensive neurologic improvement, leaving only residual bladder dysfunction. Skin lesions that persisted despite dsDNA antibody seroconversion were biopsied, confirming interface dermatitis; the addition of the type I interferon receptor antagonist anifrolumab resolved the cutaneous lesions completely over a 9-month follow-up.
• Patient 5 was a 58-year-old man with Scl70-positive SSc who relapsed at 24 months with progressive skin thickening, digital ulcers, ILD, and new-onset, hemodynamically relevant atrial flutter. Following electrical and pharmacologic cardioversion for his arrhythmia, he was treated with the BCMA-targeted T-cell engager teclistamab, resulting in improved skin and pulmonary symptoms at 4 weeks.
• Patient 6 was a 24-year-old woman with Scl70-positive SSc. She relapsed at 16 months with progressive skin thickening, declining lung function, respiratory symptoms, muscular impairment, and a reduced left ventricular ejection fraction. She received teclistamab, which led to prompt improvement in cardiac, pulmonary, and cutaneous symptoms over a 4-week follow-up period.

The clinical courses observed in this cohort indicate that repeating identical lines of therapy may be limited by cellular immunogenicity. “The administration of the same cryopreserved CAR-T cell product will possibly fail due to antigenicity, as we’ve seen twice in our patients,” said Andreas Wirsching, MD, a rheumatologist at the Department of Internal Medicine at Universitätsklinikum Erlangen in Erlangen, Germany.

While relapses after CD19 CAR T-cell therapy remain infrequent overall, patients with myositis face a higher recurrence risk. Data suggest that sequentially targeting the plasma cell niche via BCMA-directed approaches — using either T-cell engagers or secondary CAR T-cell infusions — represents a highly promising tool to clear refractory disease tissue and successfully restore deep, drug-free clinical remissions.

Allogeneic Off-the-Shelf NK Cell Therapy Provides Outpatient Option 

Researchers presented data from a phase 2a basket study evaluating AB-101, also known as AlloNK, an allogeneic, non-genetically modified, off-the-shelf, cryopreserved NK cell therapy derived from cord blood units. The therapy aims to address the logistical, manufacturing, and accessibility challenges inherent to autologous CAR T-cell therapy platforms.

Administered in an outpatient clinic setting in combination with rituximab, the therapy aims to achieve deep B-cell depletion comparable to CAR T-cell therapies without the associated safety risks or logistical complexities.

The study reported on 31 patients with treatment-refractory, highly active RA, Sjögren disease (SjD), or SSc with at least 3 months of follow-up.

Patients underwent a 3-day low-dose conditioning regimen of cyclophosphamide and fludarabine, followed by three doses of AB-101 administered 7 days apart, and two doses of rituximab administered 2 weeks apart. The treatment generated rapid, uniform, and consistent peripheral B-cell depletion that persisted for at least 6 months, with expanding B cells demonstrating favorable naive/transitional phenotypes.

Robust clinical responses through 6 months were reported across all three cohorts:

• RA (n = 15): Despite a long mean disease duration (13 years) and high baseline disease activity, all patients showed profound improvements at 3 months. For seven patients with follow-up to 6 months, five (71%) achieved 50% improvement in ACR response criteria, with mean changes from baseline of -73% in swollen joint counts and -67% in tender joint counts. The mean reductions were -36.8 for Clinical Disease Activity Index and -2.6 for DAS28 using erythrocyte sedimentation rate, exceeding the established minimal clinically important improvement criteria for every patient.
• SjD (n = 11): At 6 months, 86% of patients were clinical responders on the clinical EULAR Sjögren Syndrome Disease Activity Index (improvement ≥ 4 points), and 57% responded to EULAR Sjögren Syndrome Patient Reported Index (improvement ≥ 1 point). Stimulated salivary flow collected at 6 months showed a substantial mean improvement of 0.76 mL/min from baseline, accompanied by prominent reductions in chronic fatigue based on Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) scores.
• SSc (n = 5): In patients with high baseline skin disease activity, the mean change in four patients from baseline to 6 months in the modified Rodnan Skin Score was -9.5. At this time point, 100% of these patients achieved 25% improvement in revised Composite Response Index in Systemic Sclerosis (rCRISS25), and 50% achieved an rCRISS50 response.

Across all three indications, standard patient-reported outcome measures revealed a mean 52% reduction in Patient Global Assessment scores and a mean increase of 7.2 points in FACIT-F scores.

The therapy was exceptionally well tolerated. No cases of CRS, ICANS, or hypogammaglobulinemia were observed in any patient, and the overall safety profile was entirely consistent with what is expected from the conditioning regimen and rituximab alone. Also, no patients discontinued the study due to adverse events or a lack of efficacy, and none required new immunomodulatory drugs.

“AB-101 generates robust clinical efficacy responses comparable with the responses seen with autologous CAR T therapy without typical CAR T toxicities,” said Norman Gaylis, MD, a rheumatologist based in Aventura, Florida.

The phase 1 CARLYSLE trial of obe-cel in patients with SLE was sponsored by Autolus Therapeutics. The COMPARE trial of miv-cel in patients with RA was sponsored by Charite Universitätsmedizin Berlin. The RESET-Myositis trial of rese-cel in patients with idiopathic inflammatory myopathies was sponsored by Cabaletta Bio. The phase 2a basket study of AB-101 (AlloNK) was sponsored by Artiva Biotherapeutics. Merkt, Wirsching, and Gaylis had no relevant financial disclosures. Leandro reported having financial relationships with Roche, UCB, and GSK; and serving as primary investigator for clinical trials on B cell-targeting therapies with Artiva, Autolus, Bristol Myers Squibb, Fate Therapeutics, and Roche. Wilfong reported having financial relationships with AstraZeneca, Boehringer Ingelheim, Allogene, Arcellx, and Merck. 

Manuela Callari is a freelance science journalist specializing in human and planetary health. Her work has been published in The Medical Republic, Rare Disease Advisor, New Scientist, The Guardian, MIT Technology Review, and others.