Article Text
PDF
Abstract
Objectives To update the evidence on efficacy of DMARDs (disease-modifying antirheumatic drugs) and inform the taskforce of the 2022 update of the European Alliance of Associations for Rheumatology (EULAR) recommendations for management of rheumatoid arthritis (RA).
Methods This systematic literature review (SLR) investigated the efficacy of conventional synthetic (cs), biological (b), biosimilar and targeted synthetic (ts)DMARDs in patients with RA. Medline, EMBASE, Cochrane CENTRAL and Web of Science were used to identify all relevant articles published since the previous update in 2019 to 14 January 2022.
Results Of 8969 search results, 169 articles were selected for detailed review and 47 were finally included. Trials investigated the efficacy of csDMARDs, bDMARDs and tsDMARDs, DMARD switching, tapering and trials investigating different treatment strategies. The compounds investigated were csDMARDs (methotrexate (MTX), leflunomide, sulfasalazine, hydroxychloroquine), bDMARDs (abatacept, adalimumab, certolizumab-pegol, denosumab, etanercept, infliximab, levilimab, olokizumab, opineracept, rituximab, sarilumab, tocilizumab) and tsDMARDs (baricitinib, filgotinib, tofacitinib, upadacitinib). The efficacy of csDMARDs+ short-term glucocorticoids in early RA was confirmed and similar to bDMARD+MTX combination therapy. Interleukin-6 pathway inhibition was effective in trials on olokizumab and levilimab. Janus kinase inhibitor (JAKi) was efficacious in different patient populations. After insufficient response to JAKi, patients could respond to TNFi treatment. Tapering of DMARDs was feasible for a proportion of patients, who were able to taper therapy while remaining in low disease activity or remission.
Conclusion The results of this SLR, together with one SLR on safety of DMARD and one on glucocorticoids, informed the taskforce of the 2022 update of the EULAR recommendations for pharmacological management of RA.
- rheumatoid arthritis
- DMARDs (synthetic)
- anti-TNF
- outcomes research
Data availability statement
Data are available in a public, open access repository. All data relevant to the study are included in the article or uploaded as online supplemental information. Detailed data of all included studies are also available in the public domain.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
The 2019 update of the European Alliance of Associations for Rheumatology (EULAR) recommendations for the management of rheumatoid arthritis (RA) provided guidance for the treatment of patients with RA using conventional synthetic, biological and targeted synthetic disease-modifying antirheumatic drugs (DMARDs) and glucocorticoids. To update these recommendations, this systematic literature review (SLR) informed the EULAR task force 2022 with the evolved evidence on the efficacy of DMARD therapies since 2019.
WHAT THIS STUDY ADDS
In patients with insufficient response to upadacitinib, patients could respond to TNFi treatment after immediate switch to adalimumab.
Interleukin-6 pathway inhibition was effective in trials investigating olokizumab and levilimab.
Tapering of DMARDs was feasible for a patients, who were able to taper therapy while remaining in low disease activity or remission.
Synovial biopsy driven histological stratification to tocilizumab or rituximab treatment did not lead to improved response rates in patients with previous insufficient response to TNFi.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This work, together with two further SLRs, one on safety and one on glucocorticoid treatment informed the 2022 EULAR RA management recommendations task force with the available evidence published since 2019.
Introduction
Rheumatoid arthritis (RA) is a chronic autoimmune disease leading to a symmetric polyarthritis with substantial consequences on quality of life, daily living and social participation due to joint swelling and pain, stiffness, fatigue and potential long-term joint damage. Pharmacological management of RA aims to relieve these signs and symptoms, preventing structural damage, improving health-related quality of life and normalising physical function.1 Modern treatment strategies include established pharmacological agents termed disease-modifying antirheumatic drugs (DMARDs), including monoclonal antibodies as well as small molecules targeting specific inflammatory pathways. These are used alongside well known conventional synthetic (cs) DMARDs as well as short-term glucocorticoids.2 Weighing up efficacy and safety remains a constant challenge as data from clinical trials arise continually since decades, providing new insights into the efficacy of drugs established in routine clinical care, as well as the clinical application of molecules targeting novel treatment targets. The availability of many highly effective therapies, combined with a strategy of treating RA patients to target, facilitates the achievement of clinical remission and/or low disease activity (LDA) in a large proportion of patients. Subsequently, many questions on how to taper established therapies arise: who should be tapered, how should treatment be tapered, which treatment should be tapered first and how quickly should it be withdrawn? Many trials investigating tapering have therefore been conducted in recent years.
To address this stream of constantly evolving evidence, the European Alliance of Associations for Rheumatology (EULAR) management recommendations for pharmacological treatment of RA were updated in 2022.3 Three systematic literature reviews (SLRs) were conducted covering the areas of (1) efficacy, (2) safety of DMARD treatment and (3) utilisation of glucocorticoids in RA and are published separately.4 5
This particular SLR focuses on the evidence for efficacy of the pharmacological interventions of DMARDs, published since the previous update in 2019.6
Methods
This SLR was conducted according to the EULAR standard operating procedures published in 2014 and based on a protocol developed and approved by the task force.7
Similar to previous SLRs dealing with efficacy of DMARDs in RA, only randomised controlled trials (RCTs) investigating csDMARD, biological (b) or targeted synthetic (ts) DMARDs in adult patients classified as having RA were eligible for inclusion.8 9 To further update the previous SLR, articles published between 1 January 2019 and 14 January 2022 with no language restriction were searched by an experienced librarian (JSS) using Medline, EMBASE, The Cochrane CENTRAL Register of Controlled Trials (Central), Web of Science and the EULAR/American College of Rheumatology (ACR) abstract archives as information sources. Studies presented as conference abstracts at the EULAR and ACR annual meetings from 2020 to 2021 were also eligible for inclusion.
All detailed search strategies are provided in online supplemental text S1.1-S1.4.
Supplemental material
In total 20 research questions were defined during the first virtual steering committee meeting. The main research questions involved the efficacy of csDMARD, bDMARD and tsDMARDs, efficacy differences between different DMARDs; differences between combination therapy versus monotherapy, switching between bDMARDs and tsDMARDs, evidence on different treatment strategies as well as DMARD dose reduction and treatment discontinuation in patients with ongoing therapy. Patient populations were defined as follows: DMARD-naïve patients, patients with insufficient response (IR) to csDMARD, patients who were bDMARD-IR and/or tsDMARD-IR. Comparator arms were required to include patients, receiving placebo or active therapy. These research questions covered the areas of the efficacy of csDMARDs, csDMARD combination therapies, bDMARDs and tsDMARDs (with and without concomitant csDMARDs), head-to-head comparisons of different bDMARDs and tsDMARDs, DMARD switching as well as DMARD tapering and stopping, and studies on biosimilars. All interventions of interest are listed in online supplemental table 1.5. The detailed research questions and population, intervention, control, outcome are shown in online supplemental table 1.6.
Ten per cent of all titles and abstracts were screened by two separate researchers (AK, SAB) with an agreement of 93%. The remaining studies were screened and assessed by one researcher (AK). Uncertainties were discussed with the senior EULAR methodologist (RBML). After title and abstract screening, selected articles were assessed in detail for eligibility and data of finally eligible articles were extracted using standardised spreadsheet forms. Variables of interest were predefined in the review protocol, including outcomes on signs and symptoms of RA, commonly used composite measures, respective core-set variables, measures of physical function, patient-reported outcomes and outcomes on structural damage. Extensive evidence of the bioequivalence of various biosimilars (bsDMARDs) when compared with their respective biooriginator (boDMARDs) was already shown in a previous SLR, which led to the decision to only include systematic reviews on biosimilars in this SLR. RoB was assessed using the Cochrane Collaborations Risk of Bias tool for RCTs. Conference abstracts were not assessed for RoB. No meta-analytical methods were applied due to the heterogeneity of the available studies; therefore, results are reported descriptively.
Results
In total 8969 search results were obtained, with 5071 unique references (after deduplication) remaining for title and abstract screening. A total of 169 references were selected to be assessed in the detailed article review, resulting in 47 articles describing 38 unique trials were eligible for final inclusion in the SLR. A detailed flow chart is depicted in figure 1. Details of all studies included are shown in online supplemental table 2.1.
PRISMA flow chart describing the study selection process.
RoB assessment resulted in 17/38 (44.7%) trials rated with a low RoB and 17 of 38 trials (44.7%) had a high RoB, primarily due to open-label or single-blinded trial designs. One trial (1/38, 2.6%) had unclear RoB due to insufficient reporting of randomization sequence generation and allocation concealment methods. Three trials (3/38, 7.9%) were published as conference abstracts only and therefore not assessed for RoB. Detailed RoB results of all studies are provided in online supplemental table 2.2.
A summary of included trials and therapies investigated is shown in table 1. Baseline characteristics of all articles included are presented in online supplemental table 2.3 & online supplemental table 2.4 as well as detailed efficacy results, shown in online supplemental tables 3.1.2.
Interventions and therapeutic compounds of trials included for review
Efficacy of csDMARDs
CareRA was an open-label trial in csDMARD-naïve RA patients, who were stratified to different csDMARD combination regimens based on factors of poor prognosis (erosive disease, high disease activity, high titres of rheumatoid factor or anti-citrullinated protein antibodies antibodies). High-risk patients received either methotrexate (MTX)+sulfasalazine (SSZ)+60 mg of (protocolised tapered) prednisone or MTX+leflunomide (LEF)+30 mg (tapered protocolised) prednisone or MTX+30 mg prednisone (tapered protocolised) alone. At year 2, no differences in response rates were observed with about 90% of patients achieving LDA (Disease Activity Score 28-C reactive protein (DAS28-CRP) ≤3.2 86/98 (88%) vs 86/98 (88%) vs 85/93 (91%), p=0.65) and about one-fifth achieving ACR-EULAR Boolean remission across all treatment arms (21/98 (21%), 20/98 (20%), 21/93 (23%), p=1.0). Low-risk patients received either MTX+30 mg of prednisone (step-down) or MTX (with protocolised step-up to 25 mg weekly) alone. Patients with MTX+GC combination therapy showed similar responses in achieving LDA states after 2 years, but MTX+GC treated patients did show faster responses and higher remission rates across several outcomes including Clinical Disease Activity Index (CDAI), Simplified Disease Activity Index (SDAI) and Boolean remission (CDAI≤2.8: 21/43 (49%) vs 12/47 (28%), p=0.04). Radiographic progression was low and comparable across all treatment arms.10
An Indian single-centre open-label RCT (high RoB) investigated the non-inferiority of MTX+LEF+hydroxychloroquine (HCQ) vs MTX+SSZ+HCQ in 136 patients who had an IR (defined as DAS28 ≥3.2) to stable therapy with MTX 25 mg weekly. After 12 weeks, non-inferiority was demonstrated as 40/68 (59%) in the LEF vs 37/68 (54%) in the SSZ arm achieved a EULAR good response (difference 4.4%, 95% CI −12 to 20).11 Another Indian investigator-blinded trial (high RoB) compared two different MTX dosage escalation strategies in MTX-naïve patients—starting with 15 mg once weekly, MTX was either increased by 5 mg every 2 weeks or by 5 mg every 4 weeks. No efficacy differences in EULAR good response rates were observed at week 16 (22.5% vs 28.1%; p=0.39; for every 2 weeks and every 4 weeks MTX escalation, respectively).12
Efficacy of bDMARDs
In total eight trials investigating bDMARDs with or without concomitant csDMARD were included (six with low RoB, one with high RoB, one conference abstract). Primary results are summarised in table 2.13–21
Primary efficacy outcomes of trials comparing biological DMARDs with or without background csDMARD therapy to placebo
Four trials (three with low RoB, one conference abstract) investigated agents targeting the interleukin-6 (IL-6) pathway: olokizumab (OKZ; anti-IL-6 cytokine) in combination with MTX was superior to placebo (+MTX) treatment in CREDO 1 (MTX-IR) and CREDO 3 (TNFi-IR patients)15 16 levilimab (LEV; human anti-IL-6 receptor)+MTX also showed superior efficacy compared with placebo+MTX at week 12 in patients with previous IR to MTX.19 HARUKA investigated Japanese patients receiving sarilumab (SAR; anti-IL6 receptor) in combination with non-MTX csDMARDs or sarilumab monotherapy and did show similar response rates across the treatment arms.18
Two trials (both with low RoB) investigated the efficacy of denosumab (DEM; receptor activator of nuclear factor kappa-Β ligand (RANKL) inhibitor) on structural changes in csDMARD treated patients with presence of erosions at baseline. Patients with or without osteoporosis were eligible for inclusion. Although damage progression was low overall, significant differences between the arms in mean modified total Sharp Scores (mTSS) at 12 months (primary endpoint) were observed in DESIRABLE (placebo: 1.49 (95% CI 0.99 to 1.99) vs DEM 60 mg Q6M: 0.99 (95% CI 0.49 to 1.49; p=0.023) vs DEM 60 mg Q3M: 0.72 (95% CI 0.41 to 1.03; p=0.006)).20 Another study investigated the effect of DEM on healing of erosions (using peripheral high-resolution CT scans) in patients receiving csDMARD treatment who had stable disease and erosions at the heads of metacarpal bones II–IV. The primary endpoint (healing of erosions at 12 months) was not met (13% vs 18%, p=0.45). However, a higher proportion of patients showed healing of erosions in the DEM group compared with placebo at month 24 (10/50 (20%) vs 3/48 (6%); p=0.045).21
AMARA was an RCT investigating rituximab (RTX) combination therapy with LEF in patients who showed IR to LEF alone. Although numerical results were in favour of RTX+LEF, the primary endpoint (ACR 50 response at week 24) was not met (RTX+LEF: 25/68 (27%) vs PBO+LEF: 7/40 (15%); p=0.081).14
Efficacy of tsDMARDs (Janus kinase inhibitors)
Seven articles on five different trials (four low RoB, one unclear RoB) investigating Janus kinase inhibitors (JAKi) were included. Key outcomes are also provided in table 2.
Baricitinib (BARI; 4 mg once daily) in combination with MTX was more effective in reducing signs and symptoms of RA and improving health-related quality of life when compared with placebo+MTX in MTX-IR patients from Brazil, Argentina and China (RA-BALANCE, unclear RoB).22 23
The efficacy of upadacitinib (UPA; 15 mg or 30 mg once daily) monotherapy was investigated in early patients with RA who had no or limited MTX exposure (SELECT-EARLY, low RoB). UPA was superior to MTX monotherapy regarding clinical, patient-reported and radiographic outcomes.24 25 In SELECT-SUNRISE (low RoB) UPA (7.5 mg, 15 mg or 30 mg once daily) + csDMARDs demonstrated clinical and functional superiority to placebo+csDMARD treatment in Japanese csDMARD-IR patients.26
Filgotinib (FIL; 200 mg once daily and 100 mg once daily) in combination with csDMARDs was superior to placebo in bDMARD-IR patients (FINCH2, low RoB).27 In early RA patients with limited or no previous exposure to MTX (FINCH3, low RoB), FIL (200 mg once daily and 100 mg once daily) + MTX was superior to MTX monotherapy. However, superiority of FIL 200 mg once daily monotherapy over MTX monotherapy at week 24 was not demonstrated (ACR20: FIL200 OD: 164/210 (78.1%) vs MTX monotherapy: 297/416 (71.4%); p=0.058).28
Head-to-head studies
Three bDMARD head-to-head trials were included (two open-label studies with high RoB; one study with low RoB; table 3):
IL-6 cytokine inhibition with OKZ (+MTX) was non-inferior compared with adalimumab (ADA) in combination with MTX in patients with previous IR to MTX in the CREDO-2 study (low RoB).29 30
Head-to-head studies comparing biological DMARDs to other biological DMARDs
The R4RA open-label trial (high RoB) stratified patients with IR to TNFi to either RTX or tocilizumab (TCZ) treatment based on synovial histology (B-cell poor vs B-cell rich). The hypothesis that TCZ is superior vs RTX in patients with B-cell poor histology was investigated using CDAI50% changes at week 16 as a primary endpoint. The primary endpoint was not met (CDAI50% change: RTX: 17/38 (45%) vs TCZ: 23/41 (56%); p=0.31). In exploratory analyses, RNA sequencing-based stratification (B-cell poor/B-cell rich); however, showed significant differences between the groups and higher response rates for TCZ treated patients in the B-cell poor population (RTX: 12/33 (36%) vs TCZ: 20/32 (63%); p=0.035). No differences were observed in the B-cell rich stratified patient population.31
NORD-STAR was a single-blinded head-to-head trial (high RoB) in csDMARD-naïve early RA patients. All patients started MTX (escalated within 4 weeks to 25 mg/week) and received additionally (1) conventional therapy (csDMARD combination: SSZ+HCQ+intraarticular glucocorticoids or 20 mg oral prednisone tapered to 5 mg in 9 weeks); (2) TNFi therapy (certolizumab-pegol (CZP), 200 mg every 2 weeks); 3) CD80/86 inhibition (abatacept (ABA) 125 mg weekly) or 4) anti-IL-6R treatment with TCZ (8 mg/kg iv every 4 weeks or 162 mg s.c. once a week). The primary endpoint was the superiority of CZP+MTX or ABA+MTX or TCZ+MTX vs. conventional therapy with respect to adjusted CDAI ≤2.8 rates at week 24 and radiographic damage progression at week 48. After 24 weeks, no clinically meaningful differences between the treatments were observed. Prespecified non-inferiority (margin 15%) analyses on CDAI remission rates showed that conventional therapy (estimated CDAI remission rate: 42.7%) was non-inferior to CZP+MTX (46.5%) and TCZ+MTX (42.1%), however, not to ABA+MTX (52.0%). ABA+MTX was statistically superior compared with conventional therapy, CZP+MTX and TCZ+MTX were not. No difference in radiographic progression was observed after week 48 between the groups (ΔmTSS from baseline to week 48: conventional therapy: 0.45 vs ABA+MTX 0.62 vs CZP+MTX 0.47 vs TCZ+MTX 0.5).32 33
In total four manuscripts on head-to-head trials comparing tsDMARDs to bDMARDs (all with low RoB) were included (table 4).
UPA (15 mg once daily) in combination with MTX showed superiority over ADA (40 mg every 2 weeks) + MTX and over placebo+MTX in MTX-IR patients (SELECT-COMPARE, low RoB). UPA treated patients had significantly less radiographic damage progression compared with placebo, while results were comparable versus ADA at week 26.34
Efficacy outcomes of head-to-head studies comparing JAK inhibitors to biological DMARDs
In SELECT-CHOICE (high RoB), bDMARD-IR patients were either randomised to ABA (125 mg weekly) or UPA 15 mg once daily (both in combination with csDMARDs). UPA+csDMARDs was superior to ABA+csDMARDs in change from baseline to week 12 in DAS28-CRP. This was mainly driven by changes in CRP. Numerically higher response rates, but no clinically meaningful difference in outcomes without acute phase reactants (CDAI ≤10, ACR/EULAR Boolean remission) were observed.35
Another head-to-head trial (FINCH1, low RoB) investigating FIL 200 mg once daily in combination with MTX demonstrated non-inferiority (NI) in achieving DAS28-CRP≤3.2 responses at week 12 when compared with ADA 40 mg every 2 weeks (+MTX) and superiority when compared with placebo+MTX in MTX-IR patients. NI was not shown for FIL 100 mg OD+MTX.36
Biosimilars
Evidence on the bioequivalence of biosimilars was provided through two identified SLRs, both in line with the available evidence from the previous update.6 37 38 Tanaka et al conducted a systematic review on biosimilar treatment to inform the 2020 update of the Japanese College of Rheumatology clinical practice guidelines for the management of RA. In 20 included RCTs on biosimilars of infliximab (INF), etanercept (ETA) and ADA efficacy was similar when comparing the respective bsDMARDs and boDMARDs (risk ratio for ACR50 at week 12: 1.03 (95% CI 0.93 to 1.13); risk ratio for ACR50 at week 24: 1.04 (95% CI 0.98 to 1.10)).37 Another SLR investigated the bioequivalence as well as switching from (and to) the biooriginator of ADA bsDMARDs (FKB327, ABP 501, BI695501, GP2017, MSB11022, PF-06410293 and SB5), and did not find differences in efficacy.38
Strategy trials
In very early patients with RA, the VEDERA trial (open-label; high RoB) did not show statistical superiority of induction therapy with ETA+MTX compared with a treat-to-target strategy using MTX and glucocorticoids in achieving DAS28-CRP<2.6 at week 48 (62/120, 52% vs 46/120, 38%, p=0.211).39
Therapeutic drug monitoring (using INF serum drug levels) did not provide benefits compared with fixed INF dosing (3 mg/kg Q8W) in treatment responses of patients with RA included in the NOR-DRUM trial (open-label; high RoB). DAS28-ESR <2.6 rates for therapeutic drug monitoring vs fixed INF dosing were 55.3% vs 50% (adjusted difference: –8.3 (95% CI –30.4 to 13.8)) in the induction phase (week 0 until week 30) and 76.9% vs 73.7% in the maintenance phase (week 30 – week 52). Rates of sustained disease control without disease worsening were numerically higher for patients randomised to therapeutic drug monitoring (69.2% vs 55%, adjusted difference: 13.6 (95% CI −7.4 to 34.6)).40 41
A 1-year, open-label trial (high RoB) randomised 108 patients with DAS28-CRP ≤3.2, but subclinical synovitis on ultrasound, to DMARD treatment intensification or maintenance. Relapse rate of patients with intensified therapy was significantly lower than with maintenance of therapy (5/54 (9.1%) vs 13/54 (24.1%); p=0.039)).42
Switching from tsDMARDs to bDMARDs (or vice versa)
In SELECT-COMPARE, patients with IR (<20% improvement from baseline in tender or swollen joint counts (SJCs) between week 14 and 22, or CDAI >10 at week 26) to UPA, ADA or placebo received blinded rescue treatment. Patients receiving placebo (305/651; 47%) or ADA (159/327; 49%) were switched to UPA, while patients receiving UPA were switched to ADA (252/651; 39%), all continuing stable background MTX. Three and 6 months after blinded switching, clinical improvements in patients with IR to UPA receiving ADA were observed (CDAI <10 at week 12: 74/242 (30.6%, as observed) and CDAI <10 at week 24: 95/234 (40.6%, as observed)). After switching from ADA to UPA 58/148 (39.2%, as observed) and 77/146 (52.7%, as observed) patients with IR to ADA achieved CDAI LDA at 3 and 6 months, respectively.43
Dose reduction and stopping csDMARDs
Open-label randomisation to half-dose versus stable-dose csDMARD therapy in patients with stable csDMARD therapy (≥1 year) in stable remission (DAS44<1.6 + no swollen joint ≥1 year) in ARCTIC REWIND (high RoB) led to higher flare rates (flare defined as: DAS greater than 1.6, an increase in DAS of 0.6 units or more from the previous visit, and at least two swollen joints on examination of 44 joints) during the 1-year follow-up in the half-dose compared with the stable-dose arm (19/77 (25%) vs 5/78 (6%); risk difference 18% (95% CI 7% to 29%); p=0.003); 85% in the half-dose group vs 92% in the stable-dose had a DAS <1.6 after 1 year, and 47/75 (63%) vs 58/73 (79%), respectively, had no radiographic progression.44
An open-label non-inferiority trial (high RoB) investigated discontinuation of csDMARD treatment in 125 patients who had CZP added due to inadequate response to csDMARDs. Patients with a change in DAS28 of ≥1.2 after 3 or 6 months of CZP treatment were randomised 1:1 to continue their treatment or discontinue csDMARDs. After 18 months, maintenance of change from baseline in DAS28 ≥1.2 and/or DAS28 LDA achievement was 72%, and 69% for continuation vs discontinuation of csDMARDs. Although clinically comparable, these results were not non-inferior (absolute risk difference 2.6%; upper limit of 90% CI 19.1%; one-sided p=0.402).45
ORAL Shift (low RoB) investigated blinded MTX withdrawal of patients achieving CDAI LDA 24 weeks after initiation of tofacitinib (TOFA) modified-release 11 mg once daily treatment: 24 weeks after randomisation, non-inferiority was demonstrated using DAS-28-ESR (difference 0.3, 95% CI 0.12 to 0.48). Numerically better results were consistently obtained for most secondary outcomes (DAS28, CDAI, SDAI, ACR-EULAR Boolean remission) on LDA and remission in patients continuing concomitant MTX.46
bDMARD dose reduction and/or stopping with ongoing csDMARD therapy
In the open-label TapERA trial (high RoB) 66 patients on weekly ETA treatment, with or without background csDMARD therapy, and with sustained DAS28-CRP <2.6 were randomised 1:1 to continue ETA weekly or ETA every other week (Q2W). After 6 months, non-inferiority was not met as 26/34 (76%) in the ETA QW arm vs 19/32 (59%) in the ETA Q2W arm maintained DAS28-CRP<2.6 (difference 17%; (95% CI −5% to 41%); NI margin 10%). Although numerically better results were observed in secondary outcomes for ETA QW, after 12 months 78% of patients could maintain SDAI LDA in the ETA EOW arm.47
Patients with IR to DMARDs received TCZ 162 mg s.c. weekly for 24 weeks in the ToSpace open-label RCT and those achieving DAS28-ESR <2.6 (at week 20 and week 24) were randomised to continue TCZ 162 mg weekly or switch to TCZ 162 mg every 2 weeks for 24 weeks. Seventy-three per cent of patients switched to TCZ Q2W could maintain DAS28-ESR <2.6 at week 48 vs 90% continuing weekly TCZ treatment (p=0.004).48
RTX dose reduction from 1000 mg to 500 mg or 200 mg in patients who previously responded to RTX 1000 mg was investigated in the REDO trial (low RoB). Patients with ongoing RTX treatment and stable disease activity for 6 months (DAS28-CRP <2.9 or DAS28-CRP <3.5 at screening and judgement of LDA by a rheumatologist) were randomised 1:2:2 to receive one infusion of RTX 1000 mg, 500 mg or 200 mg in a double-blind manner. At 3 months, RTX 500 mg and 200 mg had similar response rates compared with 1000 mg (ΔDAS28-CRP: −0.07 (95% CI −0.41 to 0.27) and 0.03 (95% CI −0.32 to 0.38), respectively). At 6 months only RTX 200 mg showed similar responses compared with RTX 1000 mg (ΔDAS28-CRP: −0.02 (95% CI −0.39 to 0.35)), while RTX 500 mg did not (ΔDAS28-CRP: 0.29 (95% CI −0.08 to 0.65)). Due to the predefined hierarchical testing procedure, non-inferiority of RTX 500 mg and 200 mg vs 1000 mg could not be claimed.49
ARCTIC REWIND (conference abstract) also investigated TNFi tapering (half-dosage for 4 months) and discontinuation of TNFi therapy (at month 4) compared with TNFi continuation in patients with stable TNFi therapy (≥1 year) being in stable DAS remission (DAS44<1.6 + 0 SJ) for more than 1 year and concomitant csDMARD therapy. Non-inferiority was not demonstrated as 37.2% (tapering/stopping) vs 95.1% (continuation) maintained DAS44 remission (risk difference: 58% (95% CI 43% to 74%)) after 1 year. In the tapering arm 19% of patients experienced radiographic progression of joint damage, compared with 10% in the TNFi continuation arm.50
In PREDICTRA (low RoB), double-blind dose reduction of ADA 40 mg (every 2 weeks) to every 3 weeks was compared with treatment discontinuation of ADA while csDMARDs were continued in patients with stable DAS28(CRP)<2.6: 36% of patients in the ADA 40 mg every 3 weeks (dose reduction) arm experienced a flare by week 40, compared with 45% in the ADA discontinuation arm. Baseline MRI inflammation was not predictive for occurrence of flares.51
In the RRRR trial (open-label, high RoB), programmed INF dosing Q8W based on TNF-alpha levels vs fixed INF dosing (3 mg/kg) every 8 weeks for 52 weeks followed by withdrawal of INF if patients achieved SDAI remission, did not increase sustained remission rates 1 year after INF withdrawal (22% vs 24%, p=0.631).52
One SLR published by Verhoef et al was eligible for inclusion. RCTs investigating TNFi dose reduction or withdrawal published until March 2018 were included. Dose reduction of TNFi in patients achieving sustained LDA or remission was comparable to continuation of the standard dosing in sustaining persistent remission (absolute risk difference: 1% (95% CI −13% to 18%)). TNFi discontinuation without disease activity guidance increased the risk for disease activity worsening and flares when compared with continuation ((absolute risk difference: 14% (95% CI 9% to 18%)). However, disease activity-guided tapering was similar to treatment continuation (absolute risk difference for risk of disease worsening: 9%–95% CI −20% to 5%).53
Tapering and stopping of csDMARDs or bDMARDs
The single-blinded TARA trial (high RoB) included patients with TNFi and csDMARD combination therapy and controlled disease activity (DAS44 ≤2.4 and SJC ≤1) for more than three consecutive months. Patients were randomised to gradual tapering (and discontinuation) of csDMARDs in year 1, followed by tapering of TNFi treatment in year 2 or vice versa. After 1 year, flare rates (flare defined as DAS >2.4 and/or SJC>1) were numerically higher in the TNFi tapering group compared with the csDMARD tapering group (33% vs 43%; p=0.17).54 After 2 years of follow-up, the flare rate was similar between the groups: 61% (95% CI 50% to 71%) in the group who tapered csDMARDs first and 62% (95% CI 52% to 72%; p=0.84) in the group who tapered TNFi first. Mean disease activity, physical function and radiographic progression were similar.55 A 2-year cost-effectiveness analysis, although formally not included in this SLR, showed similar costs for both tapering strategies: the total costs were €38 833 (±€39 616) for tapering csDMARDs first, and €39 442 (±€47 271) for tapering the TNF-inhibitor first (p=0.88). Healthcare costs were significantly lower if TNFi were tapered first compared with patients who tapered the csDMARD first (€13 616 ± €9162 vs €22 484 ± €8069). However, absenteeism and presenteeism were higher leading to higher productivity costs (€25 826 ± €46 289 vs €16 349 ± €38 277 for tapering TNFi first vs tapering csDMARDs first).55 56
A double-blind RCT (SEAM-RA, low RoB) investigated continuation or withdrawal of either ETA 50 mg weekly or MTX in patients with ETA+MTX combination therapy achieving SDAI≤3.3 after a 24-week open-label run-in phase. After a 48-week double-blind period, half of the patients who stopped MTX and continued ETA monotherapy (49.5%) did not experience a worsening of disease, similar to the treatment continuation group (ETA 50 mg weekly+MTX, 52.9%). More patients who stopped MTX and received ETA monotherapy had sustained SDAI remission compared with patients stopping ETA receiving MTX monotherapy (50/101 (49.5%) vs 29/101 (28.7%): p=0.004). Patients who lost a state of SDAI LDA could recapture SDAI LDA/remission (REM) by the end of the study (LDA: 100%, 92% and 96%; REM: 71%, 73% and 81%; for ETA+MTX combination, ETA mono and MTX mono, respectively).57
AVERT-2 (conference abstract) investigated early RA patients who received ABA 125 mg weekly+MTX for 52 weeks and achieved SDAI remission (≤3.3) at week 40 and 52. Patients were then randomised to continuation of ABA+MTX; or ABA dose reduction (125 mg Q2W) + MTX for 24 weeks followed by ABA withdrawal (placebo treatment) + MTX for another 24 weeks; or ABA 125 mg weekly and MTX stopping (without withdrawal). In the combination group, SDAI remission was maintained in about 80% of patients. After 24 weeks, the SDAI remission rates were 78% in the ABA QW+MTX continuation group, 74% for ABA EOW+MTX, and 64% for ABA QW+PBO, respectively. After 52 weeks, SDAI remission was maintained in 48% of patients who were able to withdraw ABA (ie, on MTX monotherapy), and in 57% of those who discontinued MTX (ie, who were receiving ABA QW monotherapy).58
The open-label RETRO study (high RoB) randomised patients on csDMARD and bDMARD combination therapy, who had a DAS28-ESR <2.6 for 6 months to either continue their treatment regimen, reduce the of csDMARD and bDMARD by 50%, or completely stop b- and csDMARD treatment. Relapse-free DAS28-ESR remission at 12 months was achieved by 83% of patients in the continuation arm, compared with 57% of patients of patients who had their DMARD treatment reduced by 50%, and 45% who stopped DMARD treatment completely.59
Primary results of all trials investigating DMARD tapering and/or stopping are shown in table 5.
Primary outcomes of studies investigating DMARD dose reduction and withdrawal
Discussion
This SLR was conducted to provide an update of the evolving evidence from 2019 to January 2022 on efficacy of DMARDs in RA.
The efficacy of MTX with short-term glucocorticoids in patients with early RA was confirmed. Response rates of patients receiving MTX combined with other csDMARDs and glucocorticoids were similar in the high-risk treatment arm and superior to MTX without glucocorticoids in the low-risk arm in early RA patients in the CareRA trial.10 Moreover, treatment induction with conventional therapy was similarly efficacious and did not show major differences when compared with induction therapy with bDMARDs (anti-TNF, anti-CD80/86, anti-IL6R) combined with MTX in early RA patients participating in the NORD-STAR trial.32 33
Trials on OKZ (anti-IL6 cytokine) and LEV (anti-IL6 receptor) confirmed the efficacy of inhibiting the IL6-pathway in RA. Both molecules showed superior efficacy compared with placebo.15–17 19 Further, olokizumab treatment was non-inferior to ADA (both in combination with MTX) in MTX-IR patients.29 30
Efficacy of different JAKi was further confirmed in trials investigating baricitinib (JAK1/2i),22 23 UPA (JAK1/2i) and filgotinib (JAK1i).24–26 Filgotinib monotherapy treatment induction did not show statistically significant differences compared with MTX monotherapy in early RA patients.28 Head-to-head trials comparing filgotinib and baricitinibto ADA in MTX-IR patients did not show clinically meaningful differences, especially in outcomes not including acute-phase reactants.34 36 SELECT-COMPARE was the first trial to provide evidence on good treatment responses to TNFi after IR to JAKi.43
Several trials in patients with refractory disease were published. SELECT-CHOICE was the first head-to-head trial comparing a JAKi to a non-TNFi bDMARD (ABA) in patients with previous IR to bDMARDs, demonstrating statistical superiority of UPA vs ABA in change of DAS28. Only minor differences between ABA and UPA were observed if outcomes not including acute phase reactants were used.35
Trial data do not yet support the use of biopsy driven treatment allocation as a first trial of treatment allocation based on synovial B-cell counts did not result in superior outcomes in patients allocated to either TCZ or RTX although an exploratory analysis hints that an RNAseq based allocation may improve outcomes.31 No clear benefit for RA patients was observed when comparing therapeutic drug monitoring of INF to standard dosing.40 41
Several trials investigated dose reduction and/or discontinuation of csDMARDs, bDMARDs or both. Although endpoints, populations and tapering strategies differed across most trials, the evidence of the feasibility of tapering was confirmed: many patients could lower their DMARD dosage or discontinue treatment without experiencing worsening of disease activity. The TARA trial showed similar flare rates when comparing gradual disease activity-guided tapering of csDMARD or bDMARDs first (or vice versa) in patients on combination therapy. A cost-effectiveness analysis of the trial showed that tapering either the TNFi or the csDMARD first is equally cost-effective—while medication costs were significantly lower in patients who tapered their TNFi first, indirect costs were higher due to more productivity loss.55 56 Evidence on recapturing LDA or remission after treatment reinduction was also confirmed in patients who experienced a flare.57 Overall, disease activity-guided tapering appeared beneficial compared with stopping DMARDs abruptly.53
The results of this SLR, together with one SLR on the safety of DMARDs and one on glucocorticoids, informed the taskforce of the 2022 update of the EULAR recommendations for pharmacological management of RA.
Data availability statement
Data are available in a public, open access repository. All data relevant to the study are included in the article or uploaded as online supplemental information. Detailed data of all included studies are also available in the public domain.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Handling editor David S Pisetsky
Twitter @AlexSepriano, @Janetbirdope
Contributors All authors contributed and finally approved the current manuscript. AK accepts full responsibility for the work and the conduct of the study, had access to the data, and controlled the decision to publish together with the senior author (RL) and the convenor of the taskforce (JSS).
Funding The European Alliance of Associations for Rheumatology.
Competing interests AK: Speakers bureau, Consultancy: AbbVie, Amgen, Bristol-Myers Squibb, Eli Lilly, Gilead, Janssen, Merck Sharp and Dohme, Novartis, UCB and Pfizer. AS: received speaker/consulting fees from UCB, Novartis and Lilly. SAB: received an ASPIRE grant from Pfizer. JSS received grants from Abbvie, Astra-Zeneca, Janssen, Lilly, Novartis, Roche and honoraria from Abbvie, Amgen, Astra-Zeneca, Astro, BMS, Celgene, Celltrion, Chugai, Gilead, ILTOO, Janssen, Lilly, MSD, Novartis-Sandoz, Pfizer, Roche, Samsung, Sanofi, UCB. DvdH: received consulting fees from AbbVie, Amgen, Astellas, AstraZeneca, BMS, Boehringer Ingelheim, Celgene, Daiichi, Eli-Lilly, Galapagos, Gilead, Glaxo-Smith-Kline, Janssen, Merck, Novartis, Pfizer, Regeneron, Roche, Sanofi, Takeda, UCB Pharma and is Director of Imaging Rheumatology bv. RC received consulting fees from Abbvie, BMS, Amgen, Celltrion, Eli-Lilly, Fresenius-Kabi, Galapagos, Janssen, MSD, Novartis, Pfizer, UCB Pharma. CJE has received fees from Abbvie, Astra Zeneca, BMS, Celgene, Celltrion, Fresenius Kabi, Gilead, Galapagos, GSK, Janssen, Eli Lilly, Pfizer, Roche, Sanofi for advisory boards, speakers bureau and research support. PV received consulting fees from Abbvie, BMS, Celltrion, Eli Lilly, Galapagos, Gilead, Nordic Pharma, Pfizer, Sidekick Health, UCB and speaker's fees from Eli Lilly, Galapagos, MSD and Roularta and holds the Pfizer Chair Early RA Management at KU Leuven. JEP received research grants from AbbVie, Frensenius Kabi, Mallinckrodt Pharmaceuticals, Pfizer, Seattle Genetics and consulting fees from AbbVie, Amgen, Astra Zeneca, BI, BMS, Celltrion, EMERALD, Frensenius Kabi, Galapagos, GSK, Janssen, Lilly, Mallinckrodt Pharmaceuticals, Medexus, Mitsubishi Tanabe Pharma, Novartis, Pfizer, Roche, Sandoz, Samsung, Sobi, Viatris. TT received grants from AbbVie, Asahikasei, AYUMI, Chugai, Eli Lilly Japan, Mitsubishi-Tanabe, Ono, and honoraria from AbbVie, Asahikasei, Astellas, Bristol-Myers Squibb, Chugai., Daiichi Sankyo., Eisai, Eli Lilly Japan, Gilead Sciences, Janssen K.K., Mitsubishi-Tanabe, Pfizer Japan. KLH received speaker’s fees from Abbvie and research grants from Pfizer and BMS. KLW received research grants from Pfizer, BMS, and GSK, and scientific consulting fees from UCB, Abbvie, BMS, Galapagos, Gilead, Lilly, GSK, Roche, and Novartis. DA received grants, speaker fees, or consultancy fees from Abbvie, Amgen, Lilly, Janssen, Merck, Novartis, Pfizer, Roche, Sandoz, Sanofi, and Sobi. TAS has received grant/research support from AbbVie and Roche, has been a consultant for AbbVie and Sanofi Genzyme, and has been a paid speaker for AbbVie, Novartis, Roche, Sanofi, and Takeda. RBML: received consulting fees from AbbVie, BMS, Celgene, Eli-Lilly, Galapagos, Gilead, Glaxo-Smith-Kline, Janssen, Merck, Novartis, Pfizer, Roche, UCB and is Director of Rheumatology Consultancy bv
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.