RA research has spent a decade optimizing therapy selection among approved agents — which biologic first, how to sequence after failure, when to add a JAK inhibitor. That work is valuable and ongoing, but the most interesting 2026 development sits outside that conversation entirely. Five patients with severe refractory autoimmune RA received CD19-directed CAR-T cells at the University of Erlangen. All five achieved drug-free remission. This is an extraordinarily small series and caution is warranted — but if it replicates, it represents something categorically different from anything in the current RA treatment landscape. The rest of the 2026 pipeline — TYK2 inhibitors, drug-free remission trials, cardiovascular safety monitoring — is important clinical work. The CAR-T data is a possible paradigm shift.
This article is for informational purposes only and does not constitute medical advice. Clinical trial eligibility and availability vary. Always consult a qualified healthcare professional before making any medical decisions or considering participation in a clinical trial.
Summary
Rheumatoid arthritis affects approximately 1% of the global population and remains a significant cause of disability despite eleven approved biological drug classes. The 2026 trial landscape is focused on three distinct problems: achieving drug-free remission in patients whose disease is controlled, treating the roughly 30–40% who fail multiple biologics and JAK inhibitors (refractory RA), and characterizing the long-term cardiovascular risks associated with JAK inhibitor therapy in the population most likely to receive it. CD19 CAR-T cell therapy — producing drug-free remission in small series of severe autoimmune RA — is the most scientifically novel development in the field.
ClinicalMetric Analysis
- ORAL Surveillance's cardiovascular findings for tofacitinib generated FDA class-wide JAK inhibitor labeling changes — applied to more selective agents without direct comparative safety data — and the class warning may overstate risk for JAK1-selective inhibitors in lower-CV-risk populations. ORAL Surveillance enrolled RA patients aged ≥50 with at least one additional cardiovascular risk factor — a higher-risk population than younger RA patients or dermatology patients receiving JAK inhibitors for atopic dermatitis. Upadacitinib and abrocitinib are more JAK1-selective than tofacitinib, which may translate to meaningfully different cardiovascular and VTE risk profiles in lower-risk patients. Prescribers should make individualized CV risk assessments when choosing among JAK inhibitors rather than treating the uniform class label as a blanket contraindication across all patients and indications.
- Subclinical synovial ultrasound activity — not clinical DAS28 score alone — is the most sensitive predictor of which RA patients in sustained remission will flare on biologic tapering, and most community practices haven't integrated this into tapering decisions. Drug-free remission trials (RETRO, DRESS, AWARE) established that a substantial fraction of RA patients in sustained remission can successfully taper and stop treatment — but predictors of successful tapering are not reliably captured by composite clinical scores. Joints that appear clinically quiet can have active synovial power Doppler signal on ultrasound, predicting relapse during tapering even when the clinical exam is normal. Musculoskeletal ultrasound is available in most academic rheumatology centers; the implementation gap is in community practice where the assessment is not routinely integrated into remission monitoring workflows.
- CAR-T for refractory RA (Erlangen series: all 5 patients achieved drug-free remission) is compelling proof-of-concept — but the biologic rationale is strongest in seropositive RA, and the multi-center trials must stratify on seropositivity to determine who actually benefits. B cell depletion by CD19-targeting CAR-T makes most mechanistic sense in seropositive RA (RF+, anti-CCP+), where autoantibodies are part of the pathological cascade. Seronegative RA has a different immunopathological basis — T cell and cytokine-driven rather than autoantibody-driven — and may not respond equivalently. The Erlangen cohort was seropositive; applying those results broadly to seronegative RA without pre-specified subgroup data would be an overextension. Phase 2 multi-center trials should stratify on seropositivity as a primary subgroup analysis before drawing conclusions about who benefits from this expensive and intensive intervention.
The Treatment Gap That Trials Are Trying to Close
Standard RA management progresses from methotrexate through biologics (TNF inhibitors, IL-6 inhibitors, abatacept, rituximab) to JAK inhibitors when biologics fail. The EULAR and ACR guidelines have converged on treat-to-target strategies, and the targets — low disease activity (DAS28 ≤3.2) or remission (DAS28 <2.6) — are achievable in the majority of patients who cycle through available options. That's the good news.
The less comfortable reality: approximately 30–40% of patients don't reach adequate disease control despite trying multiple agents. Virtually all patients flare when treatment is stopped, suggesting current therapies suppress disease activity rather than modifying the underlying autoimmune process. And the cardiovascular and thrombotic safety signals from JAK inhibitors — formalized in the ORAL Surveillance trial — have created genuine prescribing uncertainty, particularly for the most effective agents in the class. Trials are working all three of these problems simultaneously.
CAR-T Therapy for Refractory RA: Small Data, Large Implications
The Erlangen series published in 2023 treated five patients with severe, refractory RA — having failed multiple biologics and JAK inhibitors — with autologous CD19-directed CAR-T cells. The same approach used in B-cell malignancies. The rationale: RA pathophysiology depends heavily on aberrant B cell activity, particularly autoantibody-producing plasmablasts and the B cell-T cell costimulation that sustains synovial inflammation. Eliminating the CD19+ B cell population resets this process.
All five patients achieved drug-free remission. Follow-up at the time of publication ranged from 6 to 24 months — several patients remained in remission without any disease-modifying therapy for over a year. DAS28 scores fell from active disease range (>3.2) to below 2.6 in all evaluable patients. No patients required retreatment during the observation period. The absence of serious adverse events is also notable — the cytokine release syndrome and neurotoxicity that complicate CAR-T in oncology were not observed at the lower lymphodepletion intensities used.
Phase 1/2 trials are now enrolling to expand this approach. Eligibility typically requires: failure of at least two biologics with different mechanisms, failure of at least one JAK inhibitor, active disease at screening (DAS28 >3.2 with elevated CRP or ESR), no prior malignancy, and adequate organ function. The manufacturing logistics — apheresis, CAR-T production, lymphodepletion, infusion — currently limit access to academic centers. If results hold in larger series, the access question will dominate this field.
TYK2 Inhibitors: A Safer Path Through the JAK Landscape
Deucravacitinib (Sotyktu, Bristol Myers Squibb), approved for psoriasis in September 2022 based on POETYK PSO-1 and PSO-2 data, is now in Phase 3 trials for RA. The mechanism is the clinically interesting part. Tyrosine kinase 2 (TYK2) participates in IL-12, IL-23, and type I interferon signaling — pathways relevant to autoimmune inflammation — but does not directly participate in the JAK1/2/3 signaling that drives the erythropoiesis, thrombopoiesis, and lymphopoiesis changes underlying the cardiovascular and thrombotic risks associated with pan-JAK inhibitors.
In the psoriasis program, deucravacitinib outperformed apremilast on PASI 75 response (53.6% vs. 35.1% at 16 weeks in POETYK PSO-1) and showed a clean safety profile without the laboratory changes or cardiovascular signals observed with tofacitinib and baricitinib. Whether this translates to RA — where JAK1 inhibition drives much of the anti-inflammatory efficacy — remains the key question. The POETYK RA Phase 3 trials are actively recruiting patients who have failed at least one biologic. Results are expected 2025–2026.
Drug-Free Remission: What the ARCTIC REWIND Data Actually Shows
The ambition of drug-free remission — stopping treatment entirely without disease flare — is clinically meaningful and increasingly tractable. The ARCTIC REWIND trial (NCT02463136) randomized RA patients in sustained remission (DAS28 < 2.6 for at least a year) to continue TNF inhibitors or step down to methotrexate monotherapy, then attempt full withdrawal. At 52 weeks, 43% of the step-down group maintained remission without any biologic therapy, compared to 83% in the continuation group. The rate is lower than we'd like, but 43% is not zero — and it identifies a meaningful subpopulation that can successfully stop treatment.
The research question has shifted to identifying which patients those are before they try to stop. Candidate predictors include: RF and anti-CCP negativity (seronegative patients flare less often on withdrawal), ultrasound power Doppler activity at the time of tapering (detectable subclinical synovitis predicts flare), and emerging gene expression signatures in peripheral blood that differ between patients who successfully wean and those who don't. Trials in 2026 are prospectively testing biomarker-guided tapering algorithms — essentially, a decision support tool built around objective measures of residual inflammatory activity to guide the timing of dose reduction.
JAK Inhibitor Cardiovascular Safety: What ORAL Surveillance Changed
The ORAL Surveillance trial (NCT02092467) randomized over 4,300 RA patients over 50 with cardiovascular risk factors to tofacitinib (5 mg or 10 mg twice daily) or TNF inhibitor (etanercept or adalimumab). The 10 mg tofacitinib dose was associated with increased rates of major adverse cardiovascular events (MACE) and cancer — leading the FDA to add a black box warning to all JAK inhibitors approved for immune-mediated conditions, including upadacitinib and baricitinib, by class effect.
The clinical implications are nuanced. The excess risk was concentrated in patients over 65 with cardiovascular risk factors or a history of smoking. Younger patients without cardiovascular risk factors showed no significant MACE signal. Post-marketing studies using real-world data (Medicare, commercial insurance) are now generating observational comparisons across the JAK inhibitor class and against biologics, with results expected to sharpen prescribing guidance. Trials specifically enrolling low-cardiovascular-risk patients with refractory RA on JAK inhibitors are also underway, attempting to define the safety profile in the population where these drugs provide the clearest benefit.
Finding an RA Trial and Assessing Eligibility
Common eligibility requirements across RA trials include: confirmed diagnosis by ACR/EULAR 2010 criteria, documented active disease at screening (specific swollen and tender joint count thresholds, elevated CRP or ESR), defined prior treatment history (typically failed methotrexate, often failed at least one biologic), and washout periods from current medications. Active serious infection and pregnancy are universal exclusions. Knowing your DAS28 score, your RF and anti-CCP antibody status, and your complete treatment history with documented failure dates makes the eligibility screening process much faster.
Academic rheumatology centers with dedicated autoimmune research programs — Mayo Clinic, UCSF, Johns Hopkins, Hospital for Special Surgery, University of Birmingham in the UK — run the most concurrent RA trials. They also frequently have access to expanded access programs and pre-Phase 3 studies that are not yet listed on ClinicalTrials.gov. Asking your rheumatologist directly whether their institution is running any unreported industry studies is often the most efficient path.