Advertisement

Tofacitinib Versus Biologic Treatments in Patients With Active Rheumatoid Arthritis Who Have Had an Inadequate Response to Tumor Necrosis Factor Inhibitors: Results From a Network Meta-analysis

Open AccessPublished:November 24, 2016DOI:https://doi.org/10.1016/j.clinthera.2016.11.004

      Abstract

      Purpose

      Tofacitinib is an oral Janus kinase inhibitor for the treatment of rheumatoid arthritis (RA). This analysis compared the efficacy and safety of tofacitinib with biologic disease-modifying antirheumatic drugs in patients with RA and a prior inadequate response (IR) to tumor necrosis factor inhibitors (TNFi).

      Methods

      A systematic literature review identified 5 randomized placebo-controlled trials that evaluated tofacitinib or biologic disease-modifying antirheumatic drugs (bDMARDs) against placebo in patient populations with RA with a prior IR to TNFi. The definition of TNFi-IR varied across studies, and included patients with an IR or who had failed treatment with TNFi for any reason. A network meta-analysis was conducted comparing study data with regard to American College of Rheumatology response rates and Health Assessment Questionnaire-Disability Index improvement at weeks 12 and 24, rates of treatment withdrawal due to all causes; adverse events (AEs) and lack of efficacy; and rates of AEs, serious AEs, and serious infections.

      Findings

      The 5 trials included a total of 2136 patients. Tofacitinib 5 mg twice daily combined with methotrexate was found to have relative risk estimates of American College of Rheumatology responses and change from baseline in Health Assessment Questionnaire-Disability Index score comparable with abatacept, golimumab, rituximab, and tocilizumab combined with conventional synthetic disease-modifying antirheumatic drugs. Withdrawal rates from trials due to all causes and AEs were comparable between treatments, and tofacitinib had a lower rate of withdrawals due to lack of efficacy. Rates of AEs and HAQ-DI were comparable between tofacitinib, other active treatments, and placebo. No serious infections were reported with tofacitinib during the placebo-controlled period (up to week 12) in this study population; rates of serious infection with other active treatments were generally low and similar to placebo.

      Implications

      During a 24-week period, tofacitinib had efficacy and rates of AEs comparable with currently available bDMARDs in the treatment of patients with RA who had a prior IR to TNFi. ClinicalTrials.gov identifiers: ORAL Step, NCT00960440; ATTAIN, NCT00124982; GO-AFTER, NCT00299546; RADIATE, NCT00106522; REFLEX, NCT00462345.

      Key words

      Introduction

      Rheumatoid arthritis (RA) is a chronic and disabling autoimmune disease that leads to inflammation and destruction of the joints and surrounding tissues. The ultimate goal of treatment is to achieve remission or to slow disease progression if remission is not possible, with low disease activity recognized as an acceptable therapeutic goal.
      • Singh J.A.
      • Saag K.G.
      • Bridges Jr, S.L.
      • et al.
      2015 American College of Rheumatology Guideline for the treatment of rheumatoid arthritis.
      National Institute for Health and Clinical Excellence (NICE)
      National Collaborating Centre for Chronic Conditions. Rheumatoid arthritis: the management of rheumatoid arthritis in adults.
      • Ledingham J.
      • Deighton C.
      Update on the British Society for Rheumatology guidelines for prescribing TNFalpha blockers in adults with rheumatoid arthritis (update of previous guidelines of April 2001).
      Current RA management guidelines
      • Singh J.A.
      • Saag K.G.
      • Bridges Jr, S.L.
      • et al.
      2015 American College of Rheumatology Guideline for the treatment of rheumatoid arthritis.
      National Institute for Health and Clinical Excellence (NICE)
      National Collaborating Centre for Chronic Conditions. Rheumatoid arthritis: the management of rheumatoid arthritis in adults.
      • Ledingham J.
      • Deighton C.
      Update on the British Society for Rheumatology guidelines for prescribing TNFalpha blockers in adults with rheumatoid arthritis (update of previous guidelines of April 2001).
      • Smolen J.S.
      • Landewé R.
      • Breedveld F.C.
      • et al.
      EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2013 update.
      recommend initial treatment with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), such as methotrexate. Patients with an inadequate response (IR) or intolerance to csDMARDs are generally prescribed biologic disease-modifying antirheumatic drugs (bDMARDs),
      • Kiely P.D.
      • Deighton C.
      • Dixey J.
      • Ostor A.J.
      Biologic agents for rheumatoid arthritis--negotiating the NICE technology appraisals.
      usually in combination with methotrexate. Patients who do not respond adequately to treatment with tumor necrosis factor inhibitors (TNFi)—the largest class of bDMARDs (etanercept, infliximab, adalimumab, certolizumab pegol, and golimumab)—are generally prescribed another drug of the same class or a bDMARD with an alternative mechanism of action.
      • Rubbert-Roth A.
      • Finckh A.
      Treatment options in patients with rheumatoid arthritis failing initial TNF inhibitor therapy: a critical review.
      Other bDMARDs recommended by the American College of Rheumatology (ACR), European League Against Rheumatism, and National Institute for Health and Care Excellence, include monoclonal antibodies against B cells (rituximab), blockers of T-cell activation (abatacept), and the interleukin-6 receptor antagonist tocilizumab.
      Tofacitinib is an oral Janus kinase (JAK) inhibitor for the treatment of RA. Tofacitinib preferentially inhibits signaling by receptors associated with JAK1 and JAK3, with functional selectivity over JAK2.
      • Ghoreschi K.
      • Jesson M.I.
      • Li X.
      • et al.
      Modulation of innate and adaptive immune responses by tofacitinib (CP-690,550).

      Meyer D, Head R, Thompson J, et al. Mechanism of action of the JAK inhibitor, CP-690550, in rheumatoid arthritis. Abstract presented at: 8th Cytokines and Inflammation Conference San Diego, California, 2010

      By interfering with signaling pathways, tofacitinib disrupts the inflammatory process and leads to improvement in disease activity.
      • Kontzias A.
      • Laurence A.
      • Gadina M.
      • O’Shea J.J.
      Kinase inhibitors in the treatment of immune-mediated disease.
      Six Phase III randomized controlled trials (RCTs) and 2 long-term extension studies have reported the efficacy and safety of tofacitinib 5 and 10 mg BID, as monotherapy or in combination with csDMARDs, in patients with an IR to csDMARDs or bDMARDs, and in methotrexate-naïve patients with active RA.
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      • Fleischmann R.
      • Kremer J.
      • Cush J.
      • et al.
      Placebo-controlled trial of tofacitinib monotherapy in rheumatoid arthritis.
      • Kremer J.M.
      • Cohen S.
      • Wilkinson B.E.
      • et al.
      A phase IIb dose-ranging study of the oral JAK inhibitor tofacitinib (CP-690,550) versus placebo in combination with background methotrexate in patients with active rheumatoid arthritis and an inadequate response to methotrexate alone.
      • van der Heijde D.
      • Tanaka Y.
      • Fleischmann R.
      • et al.
      Tofacitinib (CP-690,550) in patients with rheumatoid arthritis receiving methotrexate: twelve-month data from a twenty-four-month phase III randomized radiographic study.
      • van Vollenhoven R.F.
      • Wallenstein G.
      • Lee E.B.
      • et al.
      Effects of tofacitinib (CP-690,550), an oral janus kinase lnhibitor, or adalimumab on patient reported outcomes in a phase 3 study of active rheumatoid arthritis.
      • Lee E.B.
      • Fleischmann R.
      • Hall S.
      • et al.
      Tofacitinib versus methotrexate in rheumatoid arthritis.
      • Wollenhaupt J.
      • Silverfield J.
      • Lee E.B.
      • et al.
      Safety and efficacy of tofacitinib, an oral janus kinase inhibitor, for the treatment of rheumatoid arthritis in open-label, longterm extension studies.
      The ORAL Step trial (ClinicalTrials.gov identifier: NCT00960440) investigated the effectiveness of tofacitinib in patients with an IR to TNFi (TNFi-IR population).
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      This 6-month randomized, double-blind, Phase III trial assessed the efficacy safety of tofacitinib 5 and 10 mg BID combined with methotrexate in TNFi-IR patients with moderate to severe RA. Patients treated with tofacitinib 5 and 10 mg BID demonstrated rapid improvements in RA symptoms and physical function compared with placebo, with a safety profile consistent with previous studies of tofacitinib for RA.
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      The available evidence for the efficacy of bDMARDs in TNFi-IR patients is limited to a small number of RCTs, none of which directly compared bDMARD treatments. In the absence of an RCT providing direct comparison of all treatments of interest, a network meta-analysis can be utilized to combine data from multiple RCTs to allow inferences on treatment comparisons not directly available.
      • Caldwell D.M.
      • Ades A.E.
      • Higgins J.P.
      Simultaneous comparison of multiple treatments: combining direct and indirect evidence.
      • Hoaglin D.C.
      • Hawkins N.
      • Jansen J.P.
      • et al.
      Conducting indirect-treatment-comparison and network-meta-analysis studies: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2.
      • Jansen J.P.
      • Fleurence R.
      • Devine B.
      • et al.
      Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1.
      • Lu G.
      • Ades A.E.
      Combination of direct and indirect evidence in mixed treatment comparisons.
      The objective of the present analysis was to compare the efficacy and safety of tofacitinib 5 mg BID relative with bDMARDs for the treatment of RA in TNFi-IR patients by means of a network meta-analysis based on the available evidence from RCTs. The tofacitinib 5 mg BID dose was chosen for comparison, rather than 10 mg BID, as this is the recommended dose in the majority of countries in which tofacitinib has been approved for the treatment of RA. Published literature is available comparing efficacy and safety of bDMARDs in TNFi-IR patients using network meta-analyses. Therefore, the aim of this analysis was to specifically compare bDMARDs with the more recently approved drug, tofacitinib, and not with one another.
      • Schoels M.
      • Aletaha D.
      • Smolen J.S.
      • Wong J.B.
      Comparative effectiveness and safety of biological treatment options after tumour necrosis factor alpha inhibitor failure in rheumatoid arthritis: systematic review and indirect pairwise meta-analysis.
      • Salliot C.
      • Finckh A.
      • Katchamart W.
      • et al.
      Indirect comparisons of the efficacy of biological antirheumatic agents in rheumatoid arthritis in patients with an inadequate response to conventional disease-modifying antirheumatic drugs or to an anti-tumour necrosis factor agent: a meta-analysis.

      Methods

      Retrieval of Published Studies

      A systematic literature search was performed to identify studies evaluating the efficacy and safety of bDMARDs as monotherapy or in combination with csDMARDs in TNFi-IR patients, published from January 1990 to June 2013. The Ovid (comprising MEDLINE and Embase) and Cochrane databases, and abstracts from the ACR 2012 conference and the European League Against Rheumatism 2012 and 2013 conferences, were searched using a predefined search strategy with terms related to RA, tofacitinib, bDMARDs, and RCTs (Supplemental Material).

      Inclusion and Exclusion Criteria

      The analysis included RCTs of Phase II or beyond that fulfilled the criteria described here. Each identified study was assessed for inclusion by 2 independent reviewers.
      Each trial must have studied an adult patient population with moderate to severe RA with IR or failed treatment with TNFi, as defined in each trial. The definition of TNFi-IR could include IR to, or intolerance of, TNFi therapy, patients discontinued primarily due to lack of efficacy and patients who had been treated with more than 1 dose of TNFi therapy and could have discontinued for any reason. Each study must have assessed tofacitinib or 1 or more of the following bDMARDs: abatacept, adalimumab, anakinra, etanercept, golimumab, infliximab, rituximab, and tocilizumab, either as monotherapy or in combination with other csDMARDs. The drugs selected were all currently approved for patients with moderate to severe RA. Certolizumab pegol and anakinra were not included, as there were no data available in the TNFi-IR population. Within each trial, the bDMARD must have been presented in comparison with either placebo or 1 of the other active interventions. Comparisons were excluded if they were limited to different doses or routes of administration of the active agent. In addition, only studies published in English were included, and studies on exclusively Asian populations were excluded.

      Quality Assessment

      Each of the trials identified by the systematic literature review were evaluated for validity by applying 2 quality-assessment instruments, the Jadad Criteria
      • Jadad A.R.
      • Moore R.A.
      • Carroll D.
      • et al.
      Assessing the quality of reports of randomized clinical trials: is blinding necessary?.
      and the Quality Assessment of Studies according to Centre for Reviews and Dissemination.
      University of York, Centre for Reviews and Dissemination
      Study quality was not a criterion for exclusion, and so no minimum score was adopted a priori. The Quality Assessment of Studies tool comprises 7 detailed questions about studies. Quality assessments were conducted by 1 rater and 1 checker, and were not explicitly used for analysis but provided additional information to determine quality of evidence to support interpretation of the results. All studies were randomized and double-blind, and included a description of withdrawals.

      Data Extraction

      For each study meeting the inclusion criteria, data were extracted on study design, baseline population characteristics, interventions, and the following efficacy and safety outcomes: improvement in ACR criteria of 20%, 50%, and 70% from baseline (ACR20, ACR50, and ACR70 response) at weeks 12 and 24; change from baseline in Health Assessment Questionnaire-Disability Index (HAQ-DI) score at weeks 12 and 24; withdrawals due to all causes, AEs, and lack of efficacy; and AEs, serious adverse events (SAEs), and infections.

      Statistical Analysis

      Data from the included studies were combined and indirectly compared by means of a network meta-analysis.
      • Caldwell D.M.
      • Ades A.E.
      • Higgins J.P.
      Simultaneous comparison of multiple treatments: combining direct and indirect evidence.
      • Hoaglin D.C.
      • Hawkins N.
      • Jansen J.P.
      • et al.
      Conducting indirect-treatment-comparison and network-meta-analysis studies: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2.
      • Jansen J.P.
      • Fleurence R.
      • Devine B.
      • et al.
      Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1.
      • Lu G.
      • Ades A.E.
      Combination of direct and indirect evidence in mixed treatment comparisons.
      The network meta-analysis was performed within a Bayesian framework and involved data, a likelihood distribution, a model with parameters, and prior distributions for these parameters.
      • Sutton A.J.
      • Abrams K.R.
      Bayesian methods in meta-analysis and evidence synthesis.
      The ACR response rate data were analyzed separately for ACR20, ACR50, and ACR70 responses with a binomial likelihood function and logit link. An alternative model with a multinomial likelihood function and probit link was employed to analyze the ACR20, ACR50, and ACR70 response criteria simultaneously.

      Dias S, Welton MJ, Sutton AJ, Ades AE. NICE DSU technical support document 2: a generalised linear modelling framework for pairwise and network meta-analysis of randomised controlled trials, 2011. http://www.nicedsu.org.uk. Accessed April 1, 2015.

      This model assumes an underlying continuous ACR distribution, which had been categorized by specifying different cutoffs (≥20%, ≥50% and ≥70% improvement), at which point an individual moves from one category to the next and assumes that the treatment effect is the same regardless of the cutoff. The treatment effects for ACR20, ACR50, and ACR70 estimated with the 2 models might not be consistent if not all studies provide data for all categories.
      Continuous data for change from baseline in HAQ-DI score were analyzed using a normal likelihood function and identity link. The data on withdrawals and AEs were transformed into rates and analyzed with a log-linear model with a Poisson likelihood. Using the models, the relative treatment effect of each intervention compared with placebo was estimated, and these basic parameters were used to subsequently obtain estimates of the relative efficacy between tofacitinib and each of the comparator interventions.
      To avoid the influence of the prior distributions required for the Bayesian analyses, non-informative prior distributions were used for all model parameters; prior distributions were normal with a mean of 0 and a variance of 104. The parameters of the different models were estimated using a Markov chain Monte Carlo method as implemented in the OpenBUGS software.
      • Spiegelhalter D
      • Thomas A
      • Best N
      • Lunn D
      Relative risk estimates were calculated along with 95% credible intervals (CrIs; when using Bayesian probability distributions, CrI are analogous to confidence intervals) obtained from the posterior distributions.

      Results

      Study Selection

      The literature search of the Ovid and Cochrane databases returned 2737 potentially relevant publications (Figure 1). Review of these abstracts led to exclusion of 2680 records, primarily due to duplication (n = 1130) and failure to meet the selection criteria for study population (n = 357) or design (n = 589). Review of the remaining 57 full-text publications led to exclusion of a further 49 records that did not meet the selection criteria for study population (n = 23), study design (n = 19), intervention (n = 1), comparator (n = 1), study outcomes (n = 4), or were duplicates (n = 1). The remaining 8 publications, relating to 5 separate RCTs, were selected for the analysis.
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      • Cohen S.B.
      • Emery P.
      • Greenwald M.W.
      • et al.
      Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: Results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks.
      • Emery P.
      • Keystone E.
      • Tony H.P.
      • et al.
      IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial.
      • Genovese M.C.
      • Becker J.C.
      • Schiff M.
      • et al.
      Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
      • Keystone E.
      • Burmester G.R.
      • Furie R.
      • et al.
      Improvement in patient-reported outcomes in a rituximab trial in patients with severe rheumatoid arthritis refractory to anti-tumor necrosis factor therapy.
      • Smolen J.S.
      • Kay J.
      • Doyle M.K.
      • et al.
      Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
      • Strand V.
      • Burmester G.R.
      • Ogale S.
      • et al.
      Improvements in health-related quality of life after treatment with tocilizumab in patients with rheumatoid arthritis refractory to tumour necrosis factor inhibitors: results from the 24-week randomized controlled RADIATE study.
      • Westhovens R.
      • Cole J.C.
      • Li T.
      • et al.
      Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
      Figure 1.
      Figure 1Flow diagram of study identification and selection process. (A) Ovid (MEDLINE and Embase) and Cochrane searches; and (B) conference abstract search. ACR = American College of Rheumatology; EULAR = European League Against Rheumatism; RCT = randomized controlled trial; RA = rheumatoid arthritis.
      A search of conference abstracts returned 1053 abstracts (Figure 1). After review of abstract titles, 818 abstracts were excluded. After a full review, all other abstracts were excluded because they did not meet the selection criteria for the study population (n = 98), study design (n = 55), intervention (n = 6), comparator (n = 9), study outcomes (n = 24), or they were duplicates (n = 43).

      Study Characteristics

      The characteristics of the 5 RCTs identified for inclusion in this analysis are summarized in the Table. The included studies investigated the efficacy and safety of tofacitinib (ORAL Step, ClinicalTrials.gov identifier: NCT00960440),
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      abatacept (ATTAIN trial, NCT00124982),
      • Genovese M.C.
      • Becker J.C.
      • Schiff M.
      • et al.
      Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
      • Westhovens R.
      • Cole J.C.
      • Li T.
      • et al.
      Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
      golimumab (GO-AFTER trial, NCT00299546),
      • Smolen J.S.
      • Kay J.
      • Doyle M.K.
      • et al.
      Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
      tocilizumab (RADIATE trial, NCT00106522),
      • Emery P.
      • Keystone E.
      • Tony H.P.
      • et al.
      IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial.
      • Strand V.
      • Burmester G.R.
      • Ogale S.
      • et al.
      Improvements in health-related quality of life after treatment with tocilizumab in patients with rheumatoid arthritis refractory to tumour necrosis factor inhibitors: results from the 24-week randomized controlled RADIATE study.
      and rituximab (REFLEX trial, NCT00462345)
      • Cohen S.B.
      • Emery P.
      • Greenwald M.W.
      • et al.
      Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: Results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks.
      • Keystone E.
      • Burmester G.R.
      • Furie R.
      • et al.
      Improvement in patient-reported outcomes in a rituximab trial in patients with severe rheumatoid arthritis refractory to anti-tumor necrosis factor therapy.
      (Figure 2).
      TableSummary of study characteristics, patient demographics and baseline characteristics (all studies were multicenter, randomized, double-blinded, Phase III trials).
      Study CharacteristicsTofacitinibAbataceptGolimumabTocilizumabRituximab
      First author, yearBurmester, 2013Genovese, 2005 Westhovens, 2006Smolen, 2009Emery, 2008 Strand, 2012Cohen, 2006 Keystone, 2008
      Trial nameORAL StepATTAINGO-AFTERRADIATEREFLEX
      Dosage5 mg BID10 mg/kg, days 1, 15, and 29, then q4wk50 mg, q4wk8 mg/kg, q4wk1000 mg, 2 wk apart
      Concomitant DMARD treatmentMethotrexateOral DMARD,
      Oral DMARDs included methotrexate, azathioprine, penicillamine, gold, hydroxychloroquine, chloroquine, leflunomide, sulfasalazine, NSAIDs, and corticosteroids.
      anakinra
      Methotrexate; sulfasalazine; hydroxychloroquineMethotrexateMethotrexate
      Trial duration6 mo24 wk24 wk24 wk2 y
      Centers/countries82 centers; North America, Europe and Latin America89 centers82 centers; Europe, Australasia, USANorth America and Western Europe114 centers; USA, Canada, Europe, and Israel
      Definition of TNFi-IRTreated with ≥1 TNFi, withdrawn due to IR or intoleranceTNFi-IR after ≥3 mo treatmentTreated with ≥1 dose of a TNFi, ≥8–12 wk before the study drug, withdrawn due to IR or intoleranceTreated with ≥1 TNFi within the past year, withdrawn due to IR or intoleranceTNFi-IR after ≥3 mo or intolerant to ≥1 dose
      Reported outcomesWk 12 ACR20, ACR50, and ACR70 response rates; Wk 24 ACR20, ACR50, and ACR70 response rates
      No placebo data available.
      ; week 12 HAQ-DI; withdrawals AEs, SAEs, infections
      Wk 12 ACR20, ACR50, and ACR70 response rates; wk 24 ACR20, ACR50, and ACR70 response rates; withdrawals AEs, SAEs, infectionsWk 12 ACR20, ACR50, and ACR70 response rates; wk 24 ACR20, ACR50, and ACR70 response rates; wk 12 HAQ-DI; withdrawals; AEs, SAEs, infectionsWk 12 ACR20, ACR50, and ACR70 response rates; wk 24 ACR20, ACR50, and ACR70 response rates; wk 12 HAQ-DI; withdrawals; AEs, SAEs, infectionsWk 12 ACR20, ACR50, and ACR70 response rates; wk 24 ACR20, ACR50, and ACR70 response rates; wk HAQ-DI; withdrawals; AEs, SAEs, infections
      Jadad score Patient54533
      Demographics and Baseline CharacteristicsPlaceboTofacitinibPlaceboAbataceptPlaceboGolimumabPlaceboTocilizumabPlaceboRituximab
      Patients, n132
      Randomly assigned.
      133
      Randomly assigned.
      133
      Randomly assigned and treated.
      258
      Randomly assigned and treated.
      155
      Randomly assigned.
      153
      Randomly assigned.
      158
      With baseline characteristics.
      170
      With baseline characteristics.
      209
      Randomly assigned.
      308
      Randomly assigned.
      Age, y, mean (SD)54.4 (11.3)55.4 (11.5)52.7 (11.3)53.4 (12.4)54.0
      Median.
      (46.0–64.0)
      Range.
      55.0
      Median.
      (46.0–63.0)
      Range.
      53.4 (13.3)53.9 (12.7)52.8 (12.6)52.2 (12.2)
      Female, %80858077857479848181
      Disease duration, y, mean (SD)11.3 (0.4–47.0)
      Range.
      13.0 (1.2–55.0)
      Range.
      11.4 (8.9)12.2 (8.5)9.8
      Median.
      (4.9–17.6)
      Range.
      9.6
      Median.
      (5.6–17.2)
      Range.
      11.4 (9.2)12.6 (9.3)11.7 (7.7)12.1 (8.3)
      HAQ-DI score, mean (SD)1.6 (0.7)1.6 (0.7)1.8 (0.6)1.8 (0.6)1.8
      Median.
      (1.3–2.1)
      Range.
      1.6
      Median.
      (1.1–2.0)
      Range.
      1.7 (0.6)1.7 (0.6)1.9 (0.5)1.9 (0.6)
      SJC, mean (SD)17.2 (10.7)16.2 (10.1)22.0 (10.0)22.3 (10.2)14.0
      Median.
      (9.0–23.0)
      Range.
      14.0
      Median.
      (9.0–25.0)
      Range.
      18.9 (11.1)18.9 (10.9)22.9 (12.7)23.4 (11.9)
      TJC, mean (SD)28.2 (16.7)28.4 (18.3)32.8 (13.4)31.2 (13.0)26
      Median.
      (15.0–43.0)
      Range.
      27
      Median.
      (16.0–42.0)
      Range.
      30.4 (16.8)31.7 (15.4)32.9 (15.6)33.9 (15.2)
      ACR20, ACR50, ACR70 = improvement in American College of Rheumatology criteria of 20%, 50%, and 70%, respectively; AE = adverse event; DMARD = disease-modifying antirheumatic drug; HAQ-DI = Health Assessment Questionnaire-Disability Index; IR = inadequate response; SAE = serious adverse event; SJC = swollen joint count; TNFi = tumor necrosis factor inhibitor; TJC = tender joint count.
      low asterisk Oral DMARDs included methotrexate, azathioprine, penicillamine, gold, hydroxychloroquine, chloroquine, leflunomide, sulfasalazine, NSAIDs, and corticosteroids.
      No placebo data available.
      Randomly assigned.
      § Randomly assigned and treated.
      || With baseline characteristics.
      Median.
      # Range.
      Figure 2.
      Figure 2Network diagrams of randomized controlled trials selected for inclusion in the meta-analysis. Diagrams evaluating each outcome measure for rheumatoid arthritis patients with an inadequate response to tumor necrosis factor inhibitors. All trials reported ACR response at weeks 12 and 24, change from baseline in HAQ-DI score at week 12 compared with placebo, and withdrawals and adverse events, unless otherwise stated. *ORAL Step had no placebo arm at week 24. Comparisons for ACR responses and HAQ-DI score at week 24 used the data last observed at week 12 for the placebo arm and the data actually observed at week 24 for the active treatment arm; **ATTAIN did not report change from baseline in HAQ-DI score at week 12. ACR = American College of Rheumatology; DMARD = disease-modifying antirheumatic drug; HAQ-DI = Health Assessment Questionnaire Disability-Index; MTX = methotrexate.
      All were multicenter, randomized, double-blind, and international Phase III trials. In all trials except 1, the treatment drug was administered in combination with csDMARDs: tofacitinib, tocilizumab, and rituximab were administered in combination with methotrexate; and golimumab was administered with methotrexate, sulfasalazine, or hydroxychloroquine. Abatacept was administered with either an oral csDMARD or the bDMARD anakinra.
      The definition of TNFi-IR varied across studies; IR to, or intolerance of, TNFi therapy was the most common definition (tofacitinib, tocilizumab, and rituximab).
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      • Cohen S.B.
      • Emery P.
      • Greenwald M.W.
      • et al.
      Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: Results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks.
      • Emery P.
      • Keystone E.
      • Tony H.P.
      • et al.
      IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial.
      • Keystone E.
      • Burmester G.R.
      • Furie R.
      • et al.
      Improvement in patient-reported outcomes in a rituximab trial in patients with severe rheumatoid arthritis refractory to anti-tumor necrosis factor therapy.
      • Strand V.
      • Burmester G.R.
      • Ogale S.
      • et al.
      Improvements in health-related quality of life after treatment with tocilizumab in patients with rheumatoid arthritis refractory to tumour necrosis factor inhibitors: results from the 24-week randomized controlled RADIATE study.
      In the abatacept study, patients classed as TNFi-IR were discontinued primarily due to lack of efficacy.
      • Genovese M.C.
      • Becker J.C.
      • Schiff M.
      • et al.
      Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
      • Westhovens R.
      • Cole J.C.
      • Li T.
      • et al.
      Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
      Patients enrolled in the golimumab trial were only required to have been treated with more than 1 dose of TNFi therapy and could have discontinued for any reason before being termed TNFi-IR.
      • Smolen J.S.
      • Kay J.
      • Doyle M.K.
      • et al.
      Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
      Patient characteristics are summarized in the Table. The studies randomly assigned between 132 and 308 patients per treatment arm. The percentage of females ranged from 74% to 85% and mean age and disease duration were consistent across studies, ranging from 52.2 to 55.4 years and 9.6 to 13.0 years, respectively. Mean baseline HAQ-DI scores of included patients were comparable across trials, ranging from 1.6 to 1.9.

      Efficacy Outcomes

      ACR Response Rates at Week 12

      All included RCTs reported ACR response rates around week 12. The abatacept trial
      • Genovese M.C.
      • Becker J.C.
      • Schiff M.
      • et al.
      Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
      • Westhovens R.
      • Cole J.C.
      • Li T.
      • et al.
      Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
      reported data for ACR20 response rates only, with data collected during week 12. The golimumab trial
      • Smolen J.S.
      • Kay J.
      • Doyle M.K.
      • et al.
      Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
      reported ACR response rates at week 14.
      All studies reported a greater proportion of patients achieving ACR20, ACR50, or ACR70 response for the active treatment group compared with each study’s respective placebo at week 12 (Supplemental Table I).
      Higher ACR response rates were reported with tofacitinib 5 mg BID than placebo (Figure 3) (relative risk = 2.10; 95% CrI, 1.50–2.83; relative risk = 2.86; 95% CrI, 1.77–4.44; and relative risk = 3.80; 95% CrI, 2.05–6.80, for ACR20, ACR50, and ACR70, respectively). Risk estimates of ACR response rates for tofacitinib 5 mg BID were comparable with abatacept, golimumab, rituximab, and tocilizumab at week 12 (Figure 3), demonstrating comparable efficacy between these active treatments.
      Figure 3.
      Figure 3Forest plot for ACR20, ACR50, and ACR70 response criteria at weeks 12 and 24. Data were obtained using fixed-effects multinomial likelihood models. Diamonds show the relative risk estimate of each biologic DMARD combined with conventional synthetic DMARDs compared with tofacitinib 5 mg BID combined with methotrexate. Arms show the 95% CrI associated with each relative risk estimate. ACR20, ACR50, and ACR70 = improvement in American College of Rheumatology criteria of 20%, 50%, and 70%, respectively; CrI = credible interval; DMARD = disease-modifying antirheumatic drug.

      ACR Response Rates at Week 24

      All included studies reported ACR20, ACR50, and ACR70 response rates at week 24. However, due to the reassignment design of the tofacitinib trial, the analyses at week 24 used the data last observed at week 12 for the placebo arm and the data actually observed at week 24 for the active arm. In addition, the golimumab, rituximab, and tocilizumab trials offered rescue therapy to nonresponders at or after week 16. For the tocilizumab and rituximab trials, patients who received rescue therapy were classified as nonresponders for week 24 analyses; for the golimumab trial, week 16 efficacy data were carried forward for week 24 analyses.
      All studies reported a greater proportion of patients achieving ACR20, ACR50, and ACR70 response for the active treatment group compared with each study’s respective placebo at week 24 (Supplemental Table II). The relative risk for tofacitinib 5 mg BID versus placebo at week 24 was 2.69 (95% CrI, 1.98–3.53) for ACR20, 3.92 (95% CrI, 2.54–5.84) for ACR50, and 5.77 (95% CrI, 3.26–9.84) for ACR70 (Figure 3). Relative risks for bDMARDs versus tofacitinib at week 24 ranged from 0.74 (95% CrI, 0.53–1.01) to 1.24 (95% CrI, 0.83–1.86) for ACR20, 0.63 (95% CrI, 0.38–1.01) to 1.36 (95% CrI, 0.76–2.44) for ACR50, and from 0.53 (95% CrI, 0.27–1.02) to 1.50 (95% CrI, 0.70–3.25) for ACR70.

      HAQ-DI Outcomes at Week 12

      HAQ-DI data for week 12 were reported in only 4 studies; the abatacept trial did not report HAQ-DI data at this time point. The change from baseline in mean HAQ-DI score at week 12 ranged from −0.20 to 0.02 in the placebo groups and from −0.44 to −0.21 in the active treatment groups, across studies (Supplemental Table III).
      Comparison of changes from baseline in HAQ-DI score suggested that tofacitinib 5 mg BID was more efficacious than placebo, and had efficacy comparable with other interventions in terms of HAQ-DI (Figure 4). The difference in change from baseline for tofacitinib 5 mg BID versus placebo was −0.25 (95% CrI, −0.36 to −0.14). Difference in change from baseline versus bDMARDs ranged from −0.04 (95% CrI, −0.19 to 0.11) to 0.10 (95% CrI, −0.05 to 0.25).
      Figure 4.
      Figure 4Forest plot for change from baseline in HAQ-DI score at weeks 12 and 24. Data were obtained using fixed-effects normal likelihood models. Diamonds show the difference in change from baseline of each biologic DMARD combined with conventional synthetic DMARDs compared with tofacitinib 5 mg BID combined with methotrexate. Arms show the 95% CrI associated with each treatment difference. CrI = credible interval; DMARD = disease-modifying antirheumatic drug; HAQ-DI = Health Assessment Questionnaire-Disability Index.

      HAQ-DI Outcomes at Week 24

      All included studies reported HAQ-DI data at week 24. In the tofacitinib trial, the analyses at week 24 used the data last observed at week 12 for the placebo arm and the data actually observed at week 24 for the active arm.
      Consistent with the data at week 12, tofacitinib 5 mg BID was more efficacious than placebo, and its efficacy was comparable with other interventions (Figure 4). For tofacitinib 5 mg BID versus placebo, the difference in change from baseline was −0.33 (95% CrI, −0.46 to −0.20). Difference in change from baseline versus bDMARDs ranged from −0.12 (95% CrI, −0.28 to 0.04) to 0.04 (95% CrI, −0.11 to 0.19).

      Safety Outcomes

      Withdrawals

      Safety analyses were conducted by comparing data reported on withdrawals due to all causes, AEs, and lack of efficacy (the rituximab trial did not report withdrawals due to lack of efficacy). Because all trials but the abatacept trial imposed reassignment and rescue schemes between weeks 12 and 24, withdrawals due to all causes were defined as any patient who withdrew before reassignment or rescue therapy, and analyses were adjusted based on treatment exposure.
      The percentage of patients withdrawing due to all causes, AEs, and lack of efficacy in placebo and treatment groups is summarized in Supplemental Table IV. The percentage of patients withdrawing due to all causes varied widely across studies, from 5% to 65% with placebo and from 2% to 33% with active treatments; the highest rate of withdrawals was reported with golimumab and the lowest with tofacitinib.
      The rate of withdrawal due to all causes was similar for tofacitinib, bDMARDs, and placebo (Figure 5); the relative risk estimate for tofacitinib versus placebo was 0.46 (95% CrI, 0.09–1.84) and, compared with bDMARDs, ranged from 0.92 (95% CrI, 0.17–3.98) to 1.40 (95% CrI, 0.27–5.91). Similarly, withdrawal rates due to AEs were comparable for tofacitinib, bDMARDs, and placebo, with relative risk estimates versus tofacitinib ranging from 0.56 (95% CrI, 0.06–3.61) to 4.18 (95% CrI, 0.87–23.31). Tofacitinib 5 mg BID treatment was associated with reduced withdrawals due to lack of efficacy compared with placebo and comparator interventions, with the relative risk estimates comparing tofacitinib with placebo, abatacept, golimumab, and tocilizumab equal to zero in each case.
      Figure 5.
      Figure 5Forest plot for withdrawal rates (all causes, adverse events and lack of efficacy) and adverse events (all adverse events, serious adverse events and serious infections). *Risk estimates for serious infections are presented compared with placebo. Data were obtained using fixed-effects Poisson likelihood models. Diamonds show the relative risk estimate of each biologic DMARD combined with conventional synthetic DMARDs compared with tofacitinib 5 mg BID combined with methotrexate. Arms show the 95% CrI associated with each relative risk estimate. CrI = credible interval; DMARD = disease-modifying antirheumatic drug.

      Adverse Events

      Analyses were conducted for AEs, SAEs, and serious infections. All included studies reported data on these outcomes: at week 24 for abatacept, tocilizumab, and rituximab; at week 16 for golimumab; and at week 12 for tofacitinib. As with withdrawal analyses, each analysis took into account the different rescue therapy and reassignment schemes for each study and was adjusted for treatment exposure.
      The percentages of patients experiencing AEs, SAEs, and serious infections are summarized in Supplemental Table V.
      Tofacitinib 5 mg BID was comparable with placebo and all other interventions in terms of AEs and SAEs (Figure 5). Network meta-analysis estimates found that the risk ratios (95% CrI) of AEs for active treatment compared with placebo ranged from 0.81 (95% CrI, 0.67–0.99) (rituximab) to 1.04 (95% CrI, 0.82–1.33) (abatacept). The risk ratio for tofacitinib 5 mg BID versus placebo was 0.94 (95% CrI, 0.68–1.29). For SAEs, risk ratios versus placebo ranged from 0.29 (95% CrI, 0.04–1.36) (tofacitinib) to 0.87 (95% CrI, 0.48–1.70) (abatacept). There were no reported occurrences of serious infection up to week 12 in the trial assessing tofacitinib and, therefore, the risk ratio versus placebo could not be calculated; after week 12, 2 cases of serious infection were reported for tofacitinib 5 mg BID. For the other 4 interventions, risk ratios versus placebo of serious infections ranged from 0.98 (95% CrI, 0.17–5.76) (golimumab) to 1.45 (95% CrI, 0.47–4.85) tocilizumab) (Figure 5).

      Discussion

      The objective of this network meta-analysis was to compare the efficacy and safety of tofacitinib 5 mg BID relative to bDMARDs for the treatment of patients with RA and a prior IR to TNFi. Indirect comparisons were made between RCTs reporting ACR response rates, HAQ-DI improvement and rates of withdrawals, AEs, SAEs, and serious infections. Five RCTs were identified that assessed tofacitinib 5 mg BID and the comparators abatacept, golimumab, tocilizumab, and rituximab in combination with DMARD therapy.
      The results of this network meta-analysis of tofacitinib and bDMARDs suggest that tofacitinib 5 mg BID was more efficacious than placebo and comparable with the bDMARDs in terms of efficacy, measured by ACR response rates at weeks 12 and 24, and HAQ-DI improvement at week 12. Safety of tofacitinib 5 mg BID, assessed by rate of withdrawals due to all causes, withdrawals due to AEs and risk of AEs and SAEs, was comparable with the bDMARDs assessed in this study. No serious infections were reported with tofacitinib over 12 weeks. These data were reported at an earlier time point in the tofacitinib study than in the comparator studies, however, the analyses were adjusted for treatment exposure. Tofacitinib 5 mg BID had a lower relative risk of withdrawals due to lack of efficacy than placebo and other comparator interventions.
      Network meta-analyses are an accepted technique used to combine results of multiple RCTs, to allow treatment comparisons not directly made within trials. In recent years, several network meta-analyses of bDMARD treatments for RA have been published, including an analysis of TNFi-IR patients,
      • Salliot C.
      • Finckh A.
      • Katchamart W.
      • et al.
      Indirect comparisons of the efficacy of biological antirheumatic agents in rheumatoid arthritis in patients with an inadequate response to conventional disease-modifying antirheumatic drugs or to an anti-tumour necrosis factor agent: a meta-analysis.
      and give comparable findings despite differences in methodology.
      • Salliot C.
      • Finckh A.
      • Katchamart W.
      • et al.
      Indirect comparisons of the efficacy of biological antirheumatic agents in rheumatoid arthritis in patients with an inadequate response to conventional disease-modifying antirheumatic drugs or to an anti-tumour necrosis factor agent: a meta-analysis.
      • Bergman G.J.
      • Hochberg M.C.
      • Boers M.
      • et al.
      Indirect comparison of tocilizumab and other biologic agents in patients with rheumatoid arthritis and inadequate response to disease-modifying antirheumatic drugs.
      • Devine E.B.
      • Alfonso-Cristancho R.
      • Sullivan S.D.
      Effectiveness of biologic therapies for rheumatoid arthritis: an indirect comparisons approach.
      • Launois R.
      • Avouac B.
      • Berenbaum F.
      • et al.
      Comparison of certolizumab pegol with other anticytokine agents for treatment of rheumatoid arthritis: a multiple-treatment Bayesian metaanalysis.
      • Schmitz S.
      • Adams R.
      • Walsh C.D.
      • et al.
      A mixed treatment comparison of the efficacy of anti-TNF agents in rheumatoid arthritis for methotrexate non-responders demonstrates differences between treatments: a Bayesian approach.
      • Singh J.A.
      • Christensen R.
      • Wells G.A.
      • et al.
      A network meta-analysis of randomized controlled trials of biologics for rheumatoid arthritis: a Cochrane overview.
      • Turkstra E.
      • Ng S.K.
      • Scuffham P.A.
      A mixed treatment comparison of the short-term efficacy of biologic disease modifying anti-rheumatic drugs in established rheumatoid arthritis.
      However, when conducting network meta-analyses, randomization only holds within each trial, not across trials; therefore, only relative treatment effects can be assessed. This is because differences regarding study and patient characteristics may modify the treatment effects and introduce bias to the indirect comparisons made by the network meta-analyses.
      One notable difference across the network of studies, but not a likely effect modifier, was that although in all studies bDMARDs were administered in combination with csDMARDs at stable doses, the type of csDMARD varied across studies. Most administered methotrexate therapy only; however, the golimumab
      • Smolen J.S.
      • Kay J.
      • Doyle M.K.
      • et al.
      Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
      and abatacept
      • Genovese M.C.
      • Becker J.C.
      • Schiff M.
      • et al.
      Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
      • Westhovens R.
      • Cole J.C.
      • Li T.
      • et al.
      Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
      studies allowed concomitant treatment with a variety of different csDMARDs.
      A potential source of bias is that the definition of TNFi-IR varied across studies. IR to, or intolerance of, TNFi therapy was the most common definition (tocilizumab,
      • Emery P.
      • Keystone E.
      • Tony H.P.
      • et al.
      IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial.
      • Strand V.
      • Burmester G.R.
      • Ogale S.
      • et al.
      Improvements in health-related quality of life after treatment with tocilizumab in patients with rheumatoid arthritis refractory to tumour necrosis factor inhibitors: results from the 24-week randomized controlled RADIATE study.
      rituximab,
      • Cohen S.B.
      • Emery P.
      • Greenwald M.W.
      • et al.
      Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: Results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks.
      • Keystone E.
      • Burmester G.R.
      • Furie R.
      • et al.
      Improvement in patient-reported outcomes in a rituximab trial in patients with severe rheumatoid arthritis refractory to anti-tumor necrosis factor therapy.
      and tofacitinib
      • Burmester G.R.
      • Blanco R.
      • Charles-Schoeman C.
      • et al.
      Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
      studies). However, in the abatacept study, patients classed as TNFi-IR were discontinued primarily due to lack of efficacy,
      • Genovese M.C.
      • Becker J.C.
      • Schiff M.
      • et al.
      Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
      • Westhovens R.
      • Cole J.C.
      • Li T.
      • et al.
      Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
      and the golimumab study included patients who had discontinued TNFi treatment for any reason ≥8 weeks before the first dose of study drug,
      • Smolen J.S.
      • Kay J.
      • Doyle M.K.
      • et al.
      Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
      which may indicate a less severely affected trial population. A further potential source of bias was the different rescue schemes for nonresponders after week 12 between studies, and reassignment of the placebo arm in the tofacitinib study. This reflects the lack of consensus at the time these studies were conducted around acceptable study designs to provide rescue of nonresponders and patients assigned to placebo. In the tofacitinib study, this resulted in comparison of tofacitinib data at week 24 with placebo data from week 12, which may have magnified the treatment effect.
      Due to a scarcity of studies (only 1 study for each comparison intervention relative to placebo), it was not possible to determine the extent of bias and the level of effect these differences may have had in modifying the observed treatment effects. All studies were given a validity rating of ≥3 based on the Jadad Criteria (scores can range from 0 to 5, with higher scores indicating greater study validity)
      • Jadad A.R.
      • Moore R.A.
      • Carroll D.
      • et al.
      Assessing the quality of reports of randomized clinical trials: is blinding necessary?.
      and reported the method of randomization, blinding, and a description of withdrawals. However, when interpreting the results of this network meta-analysis, the scarcity of studies must be taken into account. In addition, given the limited number of studies in this analysis, it was not possible to adjust for any differences across studies within a modeling framework. The impact of study and patient characteristics on the outcomes are reflected in the baseline risk. Extending network meta-analytic models with treatment by covariate interactions would allow heterogeneity to be explored and the impact of bias due to similarity or consistency violations to be reduced.
      • Hoaglin D.C.
      • Hawkins N.
      • Jansen J.P.
      • et al.
      Conducting indirect-treatment-comparison and network-meta-analysis studies: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2.
      • Jansen J.P.
      • Fleurence R.
      • Devine B.
      • et al.
      Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1.
      • Lu G.
      • Ades A.E.
      Combination of direct and indirect evidence in mixed treatment comparisons.
      • Cooper N.J.
      • Sutton A.J.
      • Morris D.
      • et al.
      Addressing between-study heterogeneity and inconsistency in mixed treatment comparisons: Application to stroke prevention treatments in individuals with non-rheumatic atrial fibrillation.
      In addition, the high level of patient withdrawals observed in some studies, particularly in the placebo group, could potentially have affected the efficacy estimates of active treatments versus placebo. However, the variation in discontinuations in the placebo arm across studies does not automatically invalidate comparisons, as it is not clear if there would be any change in the treatment effects if rates of discontinuation were different in each trial.
      The efficacy and safety of tofacitinib as a second-line therapy option in patients without IR to TNFi, but with an IR to csDMARDs (including methotrexate), have been described.
      • Fleischmann R.
      • Kremer J.
      • Cush J.
      • et al.
      Placebo-controlled trial of tofacitinib monotherapy in rheumatoid arthritis.
      • Kremer J.M.
      • Cohen S.
      • Wilkinson B.E.
      • et al.
      A phase IIb dose-ranging study of the oral JAK inhibitor tofacitinib (CP-690,550) versus placebo in combination with background methotrexate in patients with active rheumatoid arthritis and an inadequate response to methotrexate alone.
      • van Vollenhoven R.F.
      • Wallenstein G.
      • Lee E.B.
      • et al.
      Effects of tofacitinib (CP-690,550), an oral janus kinase lnhibitor, or adalimumab on patient reported outcomes in a phase 3 study of active rheumatoid arthritis.
      These studies reported generally higher efficacy at week 12 with tofacitinib 5 mg BID (ACR20 response rate: 52% to 60%; change from baseline in HAQ-DI: −0.44 to −0.55), compared with that reported in the TNFi-IR population analyzed here (ACR20 response rate: 42%; change from baseline in HAQ-DI: −0.43), with similar rates of AEs.
      Tofacitinib is included in the 2015 ACR guidelines for the treatment of RA, which recommend tofacitinib as a second-line option (after DMARD monotherapy), administered as either monotherapy or combination therapy.
      • Singh J.A.
      • Saag K.G.
      • Bridges Jr, S.L.
      • et al.
      2015 American College of Rheumatology Guideline for the treatment of rheumatoid arthritis.
      The ACR guidelines state that for patients with established RA and moderate or high disease activity despite DMARD monotherapy (including methotrexate), the use of combination DMARDs or the addition of a TNFi or non-TNFi biologic or tofacitinib (all choices with or without methotrexate) is strongly recommended, with no particular order of preference. Moreover, tofacitinib may also be used as monotherapy or in combination with csDMARDs.
      For third-line use, ACR recommends that if disease activity remains moderate or high despite use of a single TNFi or non-TNFi biologic, another non-TNFi biologic should be used over tofacitinib (both with or without methotrexate). However, the ACR acknowledges that evidence for this guidance is “very low.” This study provides additional evidence to support the use of tofacitinib as a well-tolerated and effective therapy for patients with moderate to high disease activity who have failed at least 1 bDMARD.

      Conclusions

      In the absence of head-to-head comparisons of bDMARDs with tofacitinib in patients with RA and an IR to TNFi, a network meta-analysis was performed based on currently available evidence from RCTs. This analysis concluded that oral tofacitinib 5 mg BID has efficacy and rates of AEs comparable with currently available bDMARDs during a 24-week period. However, longer-term follow-up data are required to fully understand the benefit−risk profile of tofacitinib compared with bDMARDs for the treatment of RA.

      Conflicts of Interest

      This study was sponsored by Pfizer Inc. Employees of the sponsor were involved in study conception, design, and conduct, and in data collection and analysis. All authors critically reviewed and revised the manuscript, and read and approved the final manuscript for submission. Medical writing support was provided by Alice Palmer of Complete Medical Communications and was funded by Pfizer Inc. Maria-Cecilia Vieira, Brielan Smiechowski, and Jeroen P. Jansen were consultants for Pfizer Inc. Gene V. Wallenstein, Dean Spurden, and Samuel H. Zwillich are shareholders and employees of Pfizer Inc.

      Acknowledgements

      The authors would like to thank the investigators, staff, patients, and the study teams involved in the tofacitinib Phase III studies.
      Maria-Cecilia Vieira, Jeroen P. Jansen, Brielan Smiechowski, and Dean Spurden were involved in conducting the systematic review and performed the data and statistical analyses. All authors were involved in data interpretation and manuscript drafting, reviewing, and development. All authors read and approved the manuscript.

      Appendix A. Supplementary material

      Additional Table IPercentage of patients achieving ACR20, ACR50 and ACR70 responses at week 12.
      TreatmentNACR20ACR50ACR70
      PlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive Treatment
      Abatacept13325618%
      Calculated using data extracted from publications.
      46%
      Calculated using data extracted from publications.
      Golimumab
      Golimumab ACR responses were measured at week 14.
      15515318%35%6%16%2%10%
      Tocilizumab16017515%43%3%23%0%8%
      Rituximab20129820%48%7%22%4%7%
      Tofacitinib13113224%
      Calculated using data extracted from publications.
      42%
      Calculated using data extracted from publications.
      8%
      Calculated using data extracted from publications.
      27%
      Calculated using data extracted from publications.
      2%
      Calculated using data extracted from publications.
      14%
      Calculated using data extracted from publications.
      N, number of patients; ACR20, ACR50 and ACR70, improvement in American College of Rheumatology criteria of 20%, 50% and 70%.
      low asterisk Calculated using data extracted from publications.
      Golimumab ACR responses were measured at week 14.
      Additional Table IIPercentage of patients achieving ACR20, ACR50 and ACR70 responses at week 24.
      TreatmentNACR20ACR50ACR70
      PlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive Treatment
      Abatacept13325620%50%4%20%2%10%
      Golimumab15515317%34%5%18%3%12%
      Tocilizumab16017510%50%4%29%1%12%
      Rituximab20129818%51%5%27%1%12%
      Tofacitinib131
      Results carried from week 12.
      13224%
      Calculated using data extracted from publications.
      52%
      Calculated using data extracted from publications.
      8%
      Calculated using data extracted from publications.
      37%
      Calculated using data extracted from publications.
      2%
      Calculated using data extracted from publications.
      16%
      Calculated using data extracted from publications.
      N, number of patients; ACR20, ACR50 and ACR70, improvement in American College of Rheumatology criteria of 20%, 50% and 70%.
      low asterisk Results carried from week 12.
      Calculated using data extracted from publications.
      Additional Table IIIMean change from baseline in HAQ-DI score at weeks 12 and 24.
      Week 12Week 24
      TreatmentNHAQ-DI score Mean CFB (SE)NHAQ-DI score Mean CFB (SE)
      PlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive Treatment
      Abatacept
      The abatacept trial did not report HAQ-DI data at week 12.
      133256–0.11 (0.04
      Calculated using data extracted from publications.
      )
      –0.45 (0.04
      Calculated using data extracted from publications.
      )
      Golimumab1551530.00 (0.03
      Calculated using data extracted from publications.
      )
      –0.21
      Calculated using data extracted from publications.
      (0.04
      Calculated using data extracted from publications.
      )
      1551530.00
      Calculated using data extracted from publications.
      (0.03
      Calculated using data extracted from publications.
      )
      –0.21
      Calculated using data extracted from publications.
      (0.04
      Calculated using data extracted from publications.
      )
      Tocilizumab1391510.02 (0.03
      Calculated using data extracted from publications.
      )
      –0.33 (0.04
      Calculated using data extracted from publications.
      )
      62130–0.05 (0.04
      Calculated using data extracted from publications.
      )
      –0.39 (0.04
      Calculated using data extracted from publications.
      )
      Rituximab201298–0.20 (0.03
      Calculated using data extracted from publications.
      )
      –0.44 (0.04
      Calculated using data extracted from publications.
      )
      201298–0.07 (0.03
      Calculated using data extracted from publications.
      )
      –0.44 (0.00
      Calculated using data extracted from publications.
      )
      Tofacitinib118117–0.18 (0.04)–0.43 (0.04)118
      Results carried from week 12.
      103–0.18 (0.04)–0.51 (0.05)
      N, number of patients; HAQ-DI, Health Assessment Questionnaire-Disability Index; CFB, change from baseline; SE, standard error of the mean.
      low asterisk Calculated using data extracted from publications.
      Results carried from week 12.
      The abatacept trial did not report HAQ-DI data at week 12.
      Additional Table IVPercentage of patients withdrawing due to all causes, adverse events and lack of efficacy.
      All causesAdverse eventsLack of efficacy
      PlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive Treatment
      TreatmentNs%Ns%Ns%Ns%Ns%Ns%
      Abatacept133262581413342583133202585
      Golimumab15565152338311111483111113
      Tocilizumab1606017525949155694201553
      Rituximab
      The rituximab trial did not report lack of efficacy data.
      209463081812923073
      Tofacitinib132513321322133213221330
      All withdrawals extracted at week 24, except for the tofacitinib trial which was extracted at Week 12.
      Ns, starting population.
      low asterisk The rituximab trial did not report lack of efficacy data.
      Additional Table VPercentage of patients experiencing adverse events, serious adverse events and serious infections.
      Adverse eventsSerious adverse eventsSerious infections
      PlaceboActive TreatmentPlaceboActive TreatmentPlaceboActive Treatment
      TreatmentNs%Ns%Ns%Ns%Ns%Ns%
      Abatacept1337125880133112581113322582
      Golimumab15570152611557152515521522
      Tocilizumab160811758416011175616031755
      Rituximab209883088520910308720913082
      Tofacitinib13257133531325133213201330
      Adverse events, serious adverse events and serious infections were extracted at week 24 for the abatacept, tocilizumab and rituximab trials, at week 16 for the golimumab trial and at week 12 for the tofacitinib trial.
      Ns, starting population.
      Table A1Ovid search. Database: MEDLINE and Embase. Date of search: June 14, 2013.
      IDSearchHits
      1“randomized controlled trial”.pt.366764
      2(random$ or placebo$ or single blind$ or double blind$ or triple blind$).ti,ab.1598709
      3(retraction of publication or retracted publication).pt.5751
      41 or 2 or 31681204
      5(animals not humans).sh.3813858
      6((comment or editorial or meta-analysis or practice-guideline or review or letter or journal correspondence) not “randomized controlled trial”).pt.5972363
      7(random sampl$ or random digit$ or random effect$ or random survey or random regression).ti,ab. not “randomized controlled trial”.pt.94407
      85 or 6 or 79690101
      94 not 81341846
      10(random$ or placebo$ or single blind$ or double blind$ or triple blind$).ti,ab.1598709
      11RETRACTED ARTICLE/6157
      1210 or 111604733
      13(animal$ not human$).sh,hw.6324971
      14(book or conference paper or editorial or letter or review).pt. not exp randomized controlled trial/6513296
      15(random sampl$ or random digit$ or random effect$ or random survey or random regression).ti,ab. not exp randomized controlled trial/91171
      1613 or 14 or 1512577612
      1712 not 161200550
      189 or 171375691
      19Arthritis, Rheumatoid/172412
      20rheumatoid arthritis.ti,ab.151700
      2119 or 20206726
      22(adalimumab or Humira).ti,ab.9077
      23(etanercept or Enbrel).ti,ab.10932
      24(infliximab or Remicade).ti,ab.18450
      25(golimumab or Simponi or CNTO 148).ti,ab.825
      26(tocilizumab or Actemra or RoActemra).ti,ab.1963
      27(abatacept or Orencia or CTLA-4Ig).ti,ab.2178
      28(tofacitinib or tasocitinib or CP-690550).ti,ab.279
      29(rituximab or Rituxan or Mabthera).ti,ab.26169
      30(tumor necrosis factor or tumor necrosis factor inhibitor or tumor necrosis factor blocker or tumor necrosis factor receptor or anti-tumor necrosis factor or TNF or anti-TNF).ti,ab.305310
      31(biologic or biological).ti,ab.976386
      3222 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 311301483
      3318 and 21 and 322575
      34limit 33 to human2158
      35limit 34 to humans2158
      36limit 35 to english language2051
      37limit 36 to yr=“1990 - Current”2039
      38remove duplicates from 371342
      Table A2Cochrane search. Database: Cochrane Library. Date of search: June 14, 2013.
      IDSearchHits
      #1MeSH descriptor: [Arthritis, Rheumatoid] explode all trees3982
      #2rheumatoid and arthritis:ti,ab,kw5565
      #3#1 or #25691
      #4adalimumab or Humira:ti,ab,kw373
      #5etanercept or Enbrel:ti,ab,kw586
      #6infliximab or Remicade:ti,ab,kw706
      #7golimumab or Simponi or CNTO148:ti,ab,kw66
      #8tocilizumab or Actemra or RoActemra:ti,ab,kw57
      #9abatacept or Orencia or CTLA-4Ig:ti,ab,kw110
      #10tofacitinib or tasocitinib or CP-690550:ti,ab,kw13
      #11rituximab or Rituxan or Mabthera:ti,ab,kw762
      #12tumor necrosis factor or tumor necrosis factor inhibitor or tumor necrosis factor blocker or tumor necrosis factor receptor or anti-tumor necrosis factor or TNF or anti-TNF:ti,ab,kw4713
      #13biologic or biological:ti,ab,kw19154
      #14#4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #1324343
      #15#3 and #14986
      #16#15 from 1990967
      #17#16 in trials698

      Additional Appendix:

      Search strategy and results

      References

        • Singh J.A.
        • Saag K.G.
        • Bridges Jr, S.L.
        • et al.
        2015 American College of Rheumatology Guideline for the treatment of rheumatoid arthritis.
        Arthritis Rheumatol. 2016; 68: 1-26
        • National Institute for Health and Clinical Excellence (NICE)
        National Collaborating Centre for Chronic Conditions. Rheumatoid arthritis: the management of rheumatoid arthritis in adults.
        NICE Clin Guide, London (UK)2009: 79
        • Ledingham J.
        • Deighton C.
        Update on the British Society for Rheumatology guidelines for prescribing TNFalpha blockers in adults with rheumatoid arthritis (update of previous guidelines of April 2001).
        Rheumatology (Oxford). 2005; 44: 157-163
        • Smolen J.S.
        • Landewé R.
        • Breedveld F.C.
        • et al.
        EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2013 update.
        Ann Rheum Dis. 2014; 73: 492-509
        • Kiely P.D.
        • Deighton C.
        • Dixey J.
        • Ostor A.J.
        Biologic agents for rheumatoid arthritis--negotiating the NICE technology appraisals.
        Rheumatology (Oxford). 2012; 51: 24-31
        • Rubbert-Roth A.
        • Finckh A.
        Treatment options in patients with rheumatoid arthritis failing initial TNF inhibitor therapy: a critical review.
        Arthritis Res Ther. 2009; 11: S1
        • Ghoreschi K.
        • Jesson M.I.
        • Li X.
        • et al.
        Modulation of innate and adaptive immune responses by tofacitinib (CP-690,550).
        J Immunol. 2011; 186: 4234-4243
      1. Meyer D, Head R, Thompson J, et al. Mechanism of action of the JAK inhibitor, CP-690550, in rheumatoid arthritis. Abstract presented at: 8th Cytokines and Inflammation Conference San Diego, California, 2010

        • Kontzias A.
        • Laurence A.
        • Gadina M.
        • O’Shea J.J.
        Kinase inhibitors in the treatment of immune-mediated disease.
        F1000 Med Rep. 2012; 4: 5
        • Burmester G.R.
        • Blanco R.
        • Charles-Schoeman C.
        • et al.
        Tofacitinib (CP-690,550) in combination with methotrexate in patients with active rheumatoid arthritis with an inadequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial.
        Lancet. 2013; 381: 451-460
        • Fleischmann R.
        • Kremer J.
        • Cush J.
        • et al.
        Placebo-controlled trial of tofacitinib monotherapy in rheumatoid arthritis.
        N Engl J Med. 2012; 367: 495-507
        • Kremer J.M.
        • Cohen S.
        • Wilkinson B.E.
        • et al.
        A phase IIb dose-ranging study of the oral JAK inhibitor tofacitinib (CP-690,550) versus placebo in combination with background methotrexate in patients with active rheumatoid arthritis and an inadequate response to methotrexate alone.
        Arthritis Rheum. 2012; 64: 970-981
        • van der Heijde D.
        • Tanaka Y.
        • Fleischmann R.
        • et al.
        Tofacitinib (CP-690,550) in patients with rheumatoid arthritis receiving methotrexate: twelve-month data from a twenty-four-month phase III randomized radiographic study.
        Arthritis Rheum. 2013; 65: 559-570
        • van Vollenhoven R.F.
        • Wallenstein G.
        • Lee E.B.
        • et al.
        Effects of tofacitinib (CP-690,550), an oral janus kinase lnhibitor, or adalimumab on patient reported outcomes in a phase 3 study of active rheumatoid arthritis.
        Ann Rheum Dis. 2012; 71 (Abstract): 206
        • Lee E.B.
        • Fleischmann R.
        • Hall S.
        • et al.
        Tofacitinib versus methotrexate in rheumatoid arthritis.
        N Engl J Med. 2014; 370: 2377-2386
        • Wollenhaupt J.
        • Silverfield J.
        • Lee E.B.
        • et al.
        Safety and efficacy of tofacitinib, an oral janus kinase inhibitor, for the treatment of rheumatoid arthritis in open-label, longterm extension studies.
        J Rheumatol. 2014; 41: 837-852
        • Caldwell D.M.
        • Ades A.E.
        • Higgins J.P.
        Simultaneous comparison of multiple treatments: combining direct and indirect evidence.
        BMJ. 2005; 331: 897-900
        • Hoaglin D.C.
        • Hawkins N.
        • Jansen J.P.
        • et al.
        Conducting indirect-treatment-comparison and network-meta-analysis studies: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2.
        Value Health. 2011; 14: 429-437
        • Jansen J.P.
        • Fleurence R.
        • Devine B.
        • et al.
        Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1.
        Value Health. 2011; 14: 417-428
        • Lu G.
        • Ades A.E.
        Combination of direct and indirect evidence in mixed treatment comparisons.
        Stat Med. 2004; 23: 3105-3124
        • Schoels M.
        • Aletaha D.
        • Smolen J.S.
        • Wong J.B.
        Comparative effectiveness and safety of biological treatment options after tumour necrosis factor alpha inhibitor failure in rheumatoid arthritis: systematic review and indirect pairwise meta-analysis.
        Ann Rheum Dis. 2012; 71: 1303-1308
        • Salliot C.
        • Finckh A.
        • Katchamart W.
        • et al.
        Indirect comparisons of the efficacy of biological antirheumatic agents in rheumatoid arthritis in patients with an inadequate response to conventional disease-modifying antirheumatic drugs or to an anti-tumour necrosis factor agent: a meta-analysis.
        Ann Rheum Dis. 2011; 70: 266-271
        • Jadad A.R.
        • Moore R.A.
        • Carroll D.
        • et al.
        Assessing the quality of reports of randomized clinical trials: is blinding necessary?.
        Control Clin Trials. 1996; 17: 1-12
        • University of York, Centre for Reviews and Dissemination
        Systematic Reviews: CRD’s Guidance for Undertaking Review in Health Care. Centre for Reviews and Dissemination, University of York, York, UK2008
        • Sutton A.J.
        • Abrams K.R.
        Bayesian methods in meta-analysis and evidence synthesis.
        Stat Methods Med Res. 2001; 10: 277-303
      2. Dias S, Welton MJ, Sutton AJ, Ades AE. NICE DSU technical support document 2: a generalised linear modelling framework for pairwise and network meta-analysis of randomised controlled trials, 2011. http://www.nicedsu.org.uk. Accessed April 1, 2015.

        • Spiegelhalter D
        • Thomas A
        • Best N
        • Lunn D
        WinBUGS User Manual: Version 1.4. MRC Biostatistics Unit, Cambridge, UK2003
        • Cohen S.B.
        • Emery P.
        • Greenwald M.W.
        • et al.
        Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: Results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks.
        Arthritis Rheum. 2006; 54: 2793-2806
        • Emery P.
        • Keystone E.
        • Tony H.P.
        • et al.
        IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial.
        Ann Rheum Dis. 2008; 67: 1516-1523
        • Genovese M.C.
        • Becker J.C.
        • Schiff M.
        • et al.
        Abatacept for rheumatoid arthritis refractory to tumor necrosis factor alpha inhibition.
        N Engl J Med. 2005; 353: 1114-1123
        • Keystone E.
        • Burmester G.R.
        • Furie R.
        • et al.
        Improvement in patient-reported outcomes in a rituximab trial in patients with severe rheumatoid arthritis refractory to anti-tumor necrosis factor therapy.
        Arthritis Rheum. 2008; 59: 785-793
        • Smolen J.S.
        • Kay J.
        • Doyle M.K.
        • et al.
        Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial.
        Lancet. 2009; 374: 210-221
        • Strand V.
        • Burmester G.R.
        • Ogale S.
        • et al.
        Improvements in health-related quality of life after treatment with tocilizumab in patients with rheumatoid arthritis refractory to tumour necrosis factor inhibitors: results from the 24-week randomized controlled RADIATE study.
        Rheumatology (Oxford). 2012; 51: 1860-1869
        • Westhovens R.
        • Cole J.C.
        • Li T.
        • et al.
        Improved health-related quality of life for rheumatoid arthritis patients treated with abatacept who have inadequate response to anti-TNF therapy in a double-blind, placebo-controlled, multicentre randomized clinical trial.
        Rheumatology (Oxford). 2006; 45: 1238-1246
        • Bergman G.J.
        • Hochberg M.C.
        • Boers M.
        • et al.
        Indirect comparison of tocilizumab and other biologic agents in patients with rheumatoid arthritis and inadequate response to disease-modifying antirheumatic drugs.
        Semin Arthritis Rheum. 2010; 39: 425-441
        • Devine E.B.
        • Alfonso-Cristancho R.
        • Sullivan S.D.
        Effectiveness of biologic therapies for rheumatoid arthritis: an indirect comparisons approach.
        Pharmacotherapy. 2011; 31: 39-51
        • Launois R.
        • Avouac B.
        • Berenbaum F.
        • et al.
        Comparison of certolizumab pegol with other anticytokine agents for treatment of rheumatoid arthritis: a multiple-treatment Bayesian metaanalysis.
        J Rheumatol. 2011; 38: 835-845
        • Schmitz S.
        • Adams R.
        • Walsh C.D.
        • et al.
        A mixed treatment comparison of the efficacy of anti-TNF agents in rheumatoid arthritis for methotrexate non-responders demonstrates differences between treatments: a Bayesian approach.
        Ann Rheum Dis. 2012; 71: 225-230
        • Singh J.A.
        • Christensen R.
        • Wells G.A.
        • et al.
        A network meta-analysis of randomized controlled trials of biologics for rheumatoid arthritis: a Cochrane overview.
        CMAJ. 2009; 181: 787-796
        • Turkstra E.
        • Ng S.K.
        • Scuffham P.A.
        A mixed treatment comparison of the short-term efficacy of biologic disease modifying anti-rheumatic drugs in established rheumatoid arthritis.
        Curr Med Res Opin. 2011; 27: 1885-1897
        • Cooper N.J.
        • Sutton A.J.
        • Morris D.
        • et al.
        Addressing between-study heterogeneity and inconsistency in mixed treatment comparisons: Application to stroke prevention treatments in individuals with non-rheumatic atrial fibrillation.
        Stat Med. 2009; 28: 1861-1881