EULAR 2025: rheumatoid arthritis
International EULAR Current
EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2025 update
Smolen JS, et al. Ann Rheum Dis. 2026;85:991–1009
DOI 10.1016/j.ard.2026.01.023 · Summary checked against the source: 2026-11
5 overarching principles and 9 recommendations, down from 11 in 2022. Citations give the level of evidence and strength of recommendation as written in the guideline.
At a glance
- Start a DMARD as soon as RA is diagnosed. [1a/A · p. 995]
- Treat to target: sustained remission, or low disease activity. [1a/A · p. 995]
- Methotrexate first, with short-term glucocorticoids tapered and stopped quickly. If MTX is contraindicated: leflunomide or sulfasalazine. [1a/A · p. 995]
- Review: improvement by 3 months, target by 6 months, or change therapy. [2b/B · p. 995]
- Biggest change: if the MTX strategy fails, add a biologic for everyone. Prognostic stratification is gone, and a second csDMARD round is no longer the standard next step. JAK inhibitors only after a risk assessment. [1a/A · p. 995, 999]
- Sustained remission off steroids: continue DMARDs, but dose reduction may be considered. [1b/A · p. 995]
Overarching principles
- A. Best care through shared decision-making between patient and rheumatologist.
- B. Decisions rest on disease activity, safety, comorbidities and structural damage.
- C. Rheumatologists should primarily care for people with RA.
- D. Patients need access to drugs with different modes of action, and may need several over a lifetime.
- E. RA has high individual, medical and societal costs, which should be considered.
Adapted from EULAR 2025 Table 1 [p. 995]
Recommendations
| # | Recommendation | LoE/SoR |
|---|---|---|
| 1 | Start DMARDs as soon as RA is diagnosed | 1a/A |
| 2 | Target sustained remission or low disease activity in every patient | 1a/A |
| 3 | Monitor every 1–3 months in active disease. No improvement by 3 months or target not reached by 6 months: adjust therapy | 2b/B; 5/D |
| 4 | MTX in the first strategy. Contraindication or early intolerance: leflunomide or sulfasalazine | 1a/A |
| 5 | Short-term glucocorticoids when starting or changing csDMARDs, tapered and stopped as fast as possible | 1a/A |
| 6 | Target not reached on csDMARDs: add a bDMARD; JAK inhibitors may be considered after weighing risk factors | 1a/A (efficacy), 2a/B (safety) |
| 7 | Combine b/tsDMARDs with a csDMARD. If no csDMARD can be used, IL-6 inhibitors and JAK inhibitors have some advantages | 1a/A |
| 8 | b/tsDMARD failure: another bDMARD or tsDMARD. After one TNF or IL-6R inhibitor fails, use another mode of action or a second agent of the same class | 1a/A; 5/D; 3/C |
| 9 | Sustained remission after stopping steroids: continue DMARDs; dose reduction (or longer intervals) may be considered | 1b/A |
Adapted from EULAR 2025 Table 1 [p. 995]
Treatment algorithm
flowchart TD
A["Clinical diagnosis of RA<br/>(2010 ACR/EULAR criteria support early diagnosis)"] -->|No contraindication| M["Start methotrexate"]
A -->|Contraindication to MTX| L["Start leflunomide or sulfasalazine"]
M --> G["Plus short-term glucocorticoids<br/>(reduce and stop as rapidly as possible)"]
L --> G
G --> Q1{"Improved at 3 months and<br/>target reached at 6 months?"}
Q1 -->|Yes| C1["Continue → dose reduction<br/>in sustained remission"]
Q1 -->|No| P2["Phase II: add a bDMARD<br/>JAK inhibitor only after risk assessment"]
P2 --> Q2{"Improved at 3 months and<br/>target reached at 6 months?"}
Q2 -->|Yes| C2["Continue → dose reduction or<br/>interval increase in sustained remission"]
Q2 -->|No| P3["Phase III: change the bDMARD<br/>or JAK inhibitor"]
P3 --> Q3{"Improved at 3 months and<br/>target reached at 6 months?"}
Q3 -->|Yes| C3["Continue → dose reduction or<br/>interval increase in sustained remission"]
Q3 -->|No| P3
Adapted from EULAR 2025 algorithm figure [p. 1001]
Notes from the figure
- Target: clinical remission by ACR/EULAR definitions; if unlikely, at least low disease activity. Expect ≥50% improvement by 3 months. Sustained remission = ≥6 months in ACR/EULAR index-based (CDAI/SDAI) or Boolean remission.
- MTX combinations: starting with MTX doesn't exclude combining it with other csDMARDs (the usual combination is MTX + sulfasalazine + hydroxychloroquine), but they add adverse events without added benefit. Leflunomide or sulfasalazine may be an interim step if MTX isn't enough before adding a bDMARD or JAKi.
- bDMARDs: TNF inhibitors (adalimumab, certolizumab, etanercept, golimumab, infliximab), abatacept, IL-6R inhibitors, or rituximab (under certain conditions), and EMA/FDA-approved biosimilars.
- Risk factors to weigh before a JAK inhibitor: age over 65, current or past smoking, other CV risk factors (diabetes, obesity, hypertension), malignancy risk (current or previous cancer other than treated non-melanoma skin cancer), VTE risk (previous MI or heart failure, cancer, inherited clotting disorder or previous clots, combined hormonal contraception or HRT, major surgery, immobility).
- Tapering: reduce dose or lengthen intervals for bDMARDs and JAKi; stopping completely brings a high flare rate. Stop glucocorticoids before tapering DMARDs.
Adapted from EULAR 2025 algorithm footnotes [p. 1001]
Clinical pearls
- Anchor drug: methotrexate, with folic acid. Leflunomide or sulfasalazine if it can't be used.
- Treat to target, review every 1–3 months; change if no response at 3 months or not at target by 6.
- After MTX failure, add a biologic (2025 change). JAK inhibitors carry MACE, malignancy and VTE warnings (ORAL Surveillance).
- Monotherapy without a csDMARD: IL-6 inhibitors (tocilizumab, sarilumab) or JAK inhibitors work best.
- Taper in remission, steroids first; don't stop DMARDs completely.
Classes and levels of evidence
How EULAR 2025 grades its recommendations. Adapted from EULAR 2025 Table 1 (Oxford levels) [p. 995]. Hover over any class or level cited on this page to see what it means.
| LoE | Meaning | Wording used |
|---|---|---|
| LoE 1a | Systematic review of RCTs | |
| LoE 1b | Individual RCT | |
| LoE 2a | Systematic review of cohort studies | |
| LoE 2b | Individual cohort study, or a low-quality RCT | |
| LoE 3a | Systematic review of case-control studies | |
| LoE 3b | Individual case-control study | |
| LoE 4 | Case series, or poor-quality cohort or case-control studies | |
| LoE 5 | Expert opinion |
| Level | Meaning |
|---|---|
| GoR A | Based on consistent level 1 studies |
| GoR B | Based on consistent level 2 or 3 studies, or extrapolated from level 1 |
| GoR C | Based on level 4 studies, or extrapolated from level 2 or 3 |
| GoR D | Based on level 5 evidence, or troublingly inconsistent or inconclusive studies |
EULAR tags each recommendation with a level of evidence (LoE, 1a–5) and a grade of recommendation (GoR, A–D) from the Oxford Centre for Evidence-Based Medicine, written as "1b/A". Overarching principles carry no LoE. Each statement also has a level of agreement: the panel's mean score from 0 to 10.