Overview of included studies.
| Study | Study Design | Disease Duration | Patients | Criteria Used to Initiate Tapering/discontinuation | Medication Tapered/stopped | Comedication | No. Patients Tapered/stopped | Flare Definition | Flare, % (n)/followup | Median/mean Time to Flare | Time to Remission After Flare | Radiological Progression | Study Limitations, Items# |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| bDMARD: TNFi | |||||||||||||
| Smolen, et al17 (PRESERVE) | RCT | Mean 6.9 yrs | RA, 18–70 yrs old; ETN + MTX 36 weeks | DAS28 ≤ 3.2 for 24 weeks | ETN 50 mg/week + MTX, randomized 1:1:1 to A) ETN 50 mg/week + MTX, B) ETN 25 mg/week +MTX, C) PBO + MTX | MTX ± GCS | 202 full-dose ETN, 202 half-dose ETN, 200 PBO | DAS28 > 3.2 at 52 weeks | 50 mg: 17.4 (35), 25 mg: 20.9 (42), PBO: 57.4 (113)/1 yr | — | — | Group A: −0.06 u/yr, B: 0.05 u/yr, C: 0.60 u/yr; A vs C was significant | 9, 11, 12, 26; partly: 1 |
| van der Maas, et al21 | Single-arm trial | Median 12 yrs | RA, 1987 ACR | DAS28 < 3.2 for 6 mos | IFX, down titration 3 mg/kg every 8–12 weeks | ± sDMARD | 51 | Reversed EULAR response criteria* | 54 (28)/1 yr | Median 200 days | — | — | 9, 12,15, 19, 26 |
| Heimans, et al5 (IMPROVED) | Single-arm trial | 8 mos | Early RA, ACR 2010; or undifferentiated arthritis | DAS44 < 1.6 for 4 mos | ADA 40 mg/2 weeks, MTX 25 mg/week, tapered† to MTX monotherapy | MTX | 26 | DAS44 > 1.6 4 mos | 35 (9)/ | — | — | — | 9, 12, 14, 19, 27 |
| Maneiro, et al13 | Retrospective observational study | Median 10.6 yrs | Early and established RA, ≤ and > 2 yrs of diagnosis, respectively | Early RA: sustained‡ DAS28 < 2.6, established RA: sustained‡ DAS28 < 3.2 | IFX 5 to 3 mg/kg and/or 6 to 8 weeks, ETN 7 to 10 days, ADA 2 to 3 weeks, CTZ 2 to 3 weeks | ± sDMARD ± GCS | 54: ADA 9, CTZ 7, ETN 28, IFX 10 | DAS28 increase > 20% or increase in dose or frequency of bDMARD, sDMARD, or GCS | All 19.1 (ADA 30.8, CTZ 50.0, ETN 11.1, IFX 25.0)/1 yr | ADA 19 mos, ETN 15.5 mos, IFX 16.5 mos, CTZ not reported | — | — | 1, 5, 11–15, 19, 27 |
| Tanaka, et al22 (RRR) | Single-arm trial | Mean 5.9 yrs | RA, 1987 ACR | DAS28 < 3.2 for > 24 weeks, PRED < 5 mg/day | IFX, stop | MTX | 114 | IFX restarted within 1 yr, DAS28 ≥ 3.2 at Yr 1 | 40 (46)/1 yr | Mean 6.4 mos | Majority within 24 weeks | — | 11, 12, 19, 26; partly: 3, 9, 15 |
| van den Broek, et al20 (BeSt) | Single-arm trial | Median 23 mos | RA, 1987 ACR | DAS44 < 2.4 for 6 mos | IFX, stop | MTX | 104 | DAS44 > 2.4 | 20 (21) /1 yr | Median 17 mos | — | — | 9, 12,15,19, 26; partly: 1 |
| Brocq, et al8 | Single-arm trial | Mean 11.3 yrs | Inflammatory joint disease, 304/442 fulfilling 1987ACR criteria | DAS28 < 2.6 for 6 mos, DMARD stable for 6 mos, no NSAID, PRED < 5 mg | TNF blocker (IFX, ETN, ADA), stop | sDMARD | 24 | DAS28 > 3.2 | 63 (15)/1 yr | Mean 14.7 weeks | Mean 5.6 weeks | — | 9, 12, 15, 19, 26, 27; partly: 11 |
| Harigai, et al10 (BRIGHT) | Retrospective cohort study | Mean 10.3 yrs | RA | DAS28-CRP ≤ 2.7 | ADA, stop | MTX ± GCS | 22 | DAS28-CRP > 2.7 or restart of bDMARD | 54 (12)/1 yr | — | — | — | 8, 9, 12, 15, 19, 23, 26, 27; partly: 5 |
| Tanaka, et al19 (HONOR) | Observational cohort with control group | Mean 7.5 yrs, SD 10.2 yrs | RA, 1987 ACR; inadequate response to MTX; and/or sDMARD | DAS28 < 2.6 for 6 mos, stable MTX dose ≥ 12 weeks, no GCS, no NSAID | ADA 40 mg/2 weeks, stop | MTX | A) 52, B) 23 control | DAS28 ≥ 3.2 | A) 40 (21), B) 9 (2)/1 yr | — | Restart ADA ± MTX: 90% LDA within 6 mos, 100% LDA within 9 mos | — | 9,14, 15, 19, 23, 27 |
| Smolen, et al16 (OPTIMA) | RCT | < 1 yr | Early RA, 1987 ACR | DAS28-CRP < 3.2 at weeks 22 and 26 | ADA 40 mg/2 weeks: A) stop, B) continue | MTX 20 mg/week ± NSAID ± GCS | A) 102, B) 105 | DAS28-CRP ≥ 3.2 | A) 19 (19), B) 9 (9)/1 yr | — | — | Radiographic non-progression, ΔTSS ≤ 0.05, from baseline to Week 78: A) 81%, B) 89%, p = 0.06 | 9, 12, 19 |
| Iwamoto, et al11 | Observational cohort | 8.2 yrs | RA, 1987 ACR OR 2010 ACR/EULAR | DAS28 < 2.6 | TNFi (IFX, ETN, ADA, GOL, CTZ), stop | ± MTX ± GCS | 32 | DAS28 > 3.2 and escalation of antirheumatic treatment | 38 (12)/6 mos | Mean 14.8 weeks | — | — | 9, 11, 12, 14, 19, 26, 27; partly: 5, 15 |
| Emery, et al9 | RCT | 6.8 mos | Early active disease; RA, 1987 ACR; MTX + biological-naive; ETN + MTX for 52 weeks | DAS28 ≤ 3.2 at Week 39 and DAS28 < 2.6 at Week 52 | ETN 50 mg/week + MTX 10–25 mg/week, randomized to: A) ETN 25 mg/week + MTX, B) MTX + PBO, C) PBO + PBO for 39 weeks. hereafter if DAS28 ≤ 3.2, all treatment was withdrawn | ± GCS | A) 63, B) 65, C) 65 | DAS28 ≥ 2.6 | A) 21 (13), B) 46 (30), C) 62 (40)/39 weeks | — | — | ΔmTSS, mean ± SE: A) 0.1 ± 0.1, B) −0.0 ± 0.2, C) 0.4 ±0.2/39 weeks; p A vs B = 0.79, p A vs C = 0.48, p B vs C = 0.34 | 9, 17C, 19 |
| Marks, et al14 | Prospective cohort | 129.5 mos | RA, 2010 ACR/EULAR; + TNFi >1 yr | DAS28 ≤ 2.6 PDUS = 0 > 6 mos, no oral GCS | TNFi, tapered 1/3 (increased interval) | ± sDMARD | 69 | DAS28 ≥ 2.6 or PDUS ≥ 1 or according to patient | 63 (43) /9 mos | Median 6–9 mos | — | — | 9, 12, 14, 15, 17A, 18, 19, 26, 27; partly: 3 |
| Raffeiner, et al15 | RCT | 14.3 yrs | RA, 1987 ACR; failure traditional DMARD; ETN 25 mg 2 ×/week | DAS28 < 2.6 for ≥ 12 weeks | A) ETN 25 mg/week, B) ETN 25 mg 2×/week | ± sDMARD ± NSAID ± GCS | A) 159, B) 164 | DAS28 > 2.6 | A) 11 (18), B) not reported/1 yr | — | — | ΔTSS = 0, > 0, ≥ 5 at 1 yr: A) 82%, 18%, 1%; B) 82%, 18%, 1%. At 2 yrs: A) 85%, 16%, 1%; B) 80%, 20%, 1% | 12, 15, 19, 9, 24, 26 |
| Kavanaugh, et al12 | Observational cohort | Median 8 yrs | RA, discontinued first TNFi, no other previous bDMARD | CDAI ≤ 10 | TNFi, stop | ±sDMARD ± GCS | 717 | CDAI > 10 or bDMARD initiation or sDMARD initiation/dose escalation or GCS initiation/dose escalation | 26.6 (191)/1 yr | Median ≥ 20 mos | — | — | 8, 9, 15, 19, 26 |
| bDMARD: TCZ | |||||||||||||
| Nishimoto, et al24,26 (DREAM/RESTORE) | Single-arm trial | Median 7.8 yrs | RA, 1987 ACR; ≥ 20 yrs old | DAS28 ≤ 3.2 at 2–3 consecutive timepoints | TCZ | ± NSAIDs ± oral GCS | 187 | DAS28 > 3.2 at 2 consecutive observations | 86.6 (162)/1 yr | — | 139 of 157 (88.5%) retreated with TCZ achieved DAS28 < 2.6 within 12 weeks | — | 12–15, 19 |
| Aguilar, et al23 | Prospective cohort | Mean 13.7 yrs | RA, MTX + TCZ for 5 yrs | DAS28 < 2.6 and SJC = 0 | TCZ 8 mg/kg/4 weeks, stop | MTX | 45 | SJC ≥ 1 | 55 (25)/1 yr | Median 3 mos | — | — | 9, 12–15, 26, 27; partly: 3, 5 |
| van Herwaarden, et al25 | Retrospective cohort | Median 10 yrs | RA, 1987 ACR or 2010 ACR/EULAR or clinical diagnosis | DAS28 < 3.2 or rheumatologist’s judgement | TCZ 8 mg/kg/4 weeks to 4 mg/kg/4 weeks | ± MTX ± sDMARD ± GCS | 22 | DAS28 > 3.2 or rheumatologist’s judgement | 41 (9)/6 mos | 7/9 (78%) within first 16 weeks | After dose-escalation, 8/9 achieved LDA (clinical judgement) within 6 mos; 1/9 LDA after 6 mos | 9, 12, 15, 26, 27 | |
| bDMARD: ABA | |||||||||||||
| Emery, et al27 (AVERT) | Single-arm trial | ≤ 2 yrs, symptoms | Clinical synovitis ≥ 2 joints for ≥ 8 weeks; ACPA-positive; MTX-naive; ≥ 18 yrs | DAS28-CRP < 3.2 at Mo 12 | A) ABA 125 mg/week + MTX 15–20 mg/week, B) ABA 125 mg/week, C) MTX 15-20 mg/week, ABA stopped immediately, MTX + steroids tapered over 1 mo | ± GCS | A) 84, B) 66, C) 73 | DAS28-CRP ≥ 3.2 | A) 75 (55), B) 72 (36), C) 83 (44)/6 mos | — | — | — | 9, 12, 14, 15, 17C, 19, 23, 26, 27; partly: 5, 8 |
| Westhovens, et al28 (AGREE) | RCT | ≤ 2 yrs | Early RA, seropositive, erosive | DAS28-ESR < 2.6 at 1 yr | ABA 10 mg/kg IV: A) 5 mg/kg IV, B) 10 mg/kg IV | ± sDMARD ± GCS | A) 50, B) 58 | DAS28-CRP ≥ 3.2 at 2 visits or additional DMARD required or ABA 10 mg required or ≥ 2 courses of GCS | A) 34 (17), B) 31 (18) /1 yr | — | Restart ABA 10 mg/kg: 3/4 remission within 1 yr | — | 9, 27 |
| Takeuchi, et al18 | Prospective cohort with control | A) 9.6 yrs, B) 15.3 yrs | RA, 1987 ACR; age ≥ 20 yrs; ABA > 2 yrs | DAS28-CRP < 2.3 | ABA 10 mg/kg/4 weeks: A) stop, B) continue | ± sDMARD ± NSAID ± GCS | A) 34, B) 17 | DAS28-CRP > 2.7 | A) 41 (14), B) 6 (1)/1 yr | — | — | ΔmTSS: A) 0.80/yr, B) 0.32/yr, p = 0.37 | 9, 12, 14, 15, 23, 27 |
| sDMARD | |||||||||||||
| Fleischmann, et al4 (iRAMT) | Single-arm trial | Mean 10.4 yrs | RA, 1987 ACR | 40% reduction in TJC + SJC | MTX, tapering 5 mg/8 weeks to minimum of 5 mg/week | IFX ± GCS | 159 | Loss of response; response defined as 40% reduction in TJC + SJC compared with baseline | 42 (67)/32 weeks | — | — | — | 9, 12, 15, 19; partly: 1, 5 |
| Heimans, et al5 (IMPROVED) | Single-arm trial | 8 mos | Early RA, ACR 2010; or undifferentiated arthritis | DAS44 < 1.6 for 4 mos | PRED 7.5 mg/day, SSZ 2000 mg/day, HCQ 400 mg/day, MTX 25 mg/week; tapered in above order to MTX monotherapy | MTX | 30 | DAS44 > 1.6 | 63 (19)/4 mos | — | — | — | 9, 12, 14, 19, 27 |
| Luis, et al6 | RCT | Mean 2.8 yrs | RA, 1987 ACR; functional class I or II; disease duration < 15 yrs | Clinical remission ACR criteria ≥ 6 mos, stable dose weekly MTX ≥ 9 mos | MTX, weekly to 2-weekly | ± HCQ ± GCS | 25 | Loss of remission; clinical criteria | 8 (2)/24 weeks | — | — | — | 2, 12, 14, 19, 27 ; partly: 3, 5, 24 |
| ten Wolde, et al7 | RCT | Median 9 yrs | RA, 1987 ACR; age 18–85 yrs | Good therapeutic response ARA criteria (5/6), stable disease for 1 yr, RX second-line drugs for 2 yrs, no previous unsuccessful attempt to discontinue second line drugs | sDMARD (CHL, HCQ, PG, DPEN, SSZ, AZA, MTX), stop | ± NSAID | 143 | SJC ≥ 3 and ≥ 2 additional criteria; clear clinical recurrence of synovitis MTX | Overall: 37 (53), HCQ/CHL: 33 (26), PG: 33 (11), SSZ: 47 (8), PEN: 40 (4), AZA: 67 (2), 100 (2)/1 yr | — | 24/51 (47%) patients retreated with same cDMARD achieved ACR20 response within 3 mos | — | 12, 19; partly: 3 |
↵* DAS28 increase ≥ 1.2 compared with baseline at 2 consecutive visits with at least 2 weeks in between or DAS28 increase ≥ 0.6 if DAS28 > 3.2.
↵† Mode of tapering was not described.
↵‡ Remission duration was not further specified.
↵# Study limitations (Supplementary Data 3, available online at jrheum.org): reporting items 1–9, external validity items 11–13, internal validity/bias items 14–20, internal validity/confounding items 21–26, and power item 27. DMARD: disease-modifying antirheumatic drugs; bDMARD: biologic DMARD; sDMARD: synthetic DMARD; TNFi: tumor necrosis factor inhibitor; TCZ: tocilizumab; ABA: abatacept; RCT: randomized controlled trial; RA: rheumatoid arthritis; ETN: etanercept; MTX: methotrexate; ACR: American College of Rheumatology; EULAR: European League Against Rheumatism; ACPA: anticitrullinated protein antibodies; DAS28: Disease Activity Score at 28 joints; DAS44: DAS at 44 joints; PRED: prednisone; NSAID: nonsteroidal antiinflammatory drug; CRP: C-reactive protein; GCS: glucocorticoids; PDUS: power Doppler ultrasound; SJC: swollen joint count; TJC: tender joint count; ESR: erythrocyte sedimentation rate; ARA: American Rheumatism Association; RX: treatment; PBO: placebo; IFX: infliximab; ADA: adalimumab; CTZ: certolizumab pegol; GOL: golimumab; IV: intravenous; SSZ: sulfasalazine; HCQ: hydroxychloroquine; CHL: chloroquine; PG: parenteral gold; DPEN: d-penicillamine; AZA: azathioprine; CDAI: Clinical Disease Activity Index; PEN: penicillamine; LDA: low disease activity; cDMARD: conventional DMARD; TSS: Total Sharp score; mTSS: modified TSS; PRESERVE: a randomized, double-blind study comparing the safety and efficacy of once-weekly ETN 50 mg, ETN 25 mg, and placebo in combination with MTX in subjects with active RA; IMPROVED: remission induction therapy with methotrexate and prednisone in patients with early rheumatoid and undifferentiated arthritis; RRR: Remission induction by Remicade in RA; BeSt: Behandel Strategieën, i.e., Treatment Strategies Study; BRIGHT: Biologics-free remission and low disease activity after stopping adalimumab in Japanese patients with rheumatoid arthritis; HONOR: Humira discontinuation without functional and radiographic damage progression following sustained remission; OPTIMA: Optimal Protocol for Treatment Initiation with MTX and ADA; DREAM: Drug-free REmission/low disease activity after cessation of TCZ (Actemra) Monotherapy; RESTORE: Retreatment Efficacy and Safety of TOcilizumab in patients with RA in Recurrence; AVERT: Assessing Very Early Rheumatoid arthritis Treatment; AGREE: ABA trial to Gauge Remission and joint damage progression in MTX-naive patients with Early Erosive RA; iRAMT: IFX RA MTX Tapering.