Advertisement

Real-world Effectiveness of Biologic Disease-modifying Antirheumatic Drugs for the Treatment of Rheumatoid Arthritis After Etanercept Discontinuation in the United Kingdom, France, and Germany

Open AccessPublished:July 17, 2017DOI:https://doi.org/10.1016/j.clinthera.2017.06.009

      Abstract

      Purpose

      The purpose of this study was to assess the real-world effectiveness of patients with rheumatoid arthritis (RA) who discontinued etanercept treatment and subsequently received another tumor necrosis factor α (TNF-α) inhibitor or a non–TNF-α biologic in the United Kingdom, France, and Germany.

      Methods

      Medical record data of patients with RA were collected from a panel of rheumatologists in the United Kingdom, France, and Germany. Patients were required to have a diagnosis of RA, be ≥18 years old, and have initiated use of another TNF-α inhibitor (adalimumab, certolizumab pegol, golimumab, or infliximab) or a non–TNF-α biologic (abatacept or tocilizumab) between January 2014 and May 2015 after discontinuing use of etanercept. Reasons for discontinuing use of etanercept and selecting a second biologic disease-modifying antirheumatic drug (DMARD) were described. Study outcomes included European League Against Rheumatism (EULAR) response and change in Clinical Disease Activity Index (CDAI) score. The study outcomes were compared among treatment groups (ie, TNF-α inhibitors and non–TNF-α biologics) using descriptive and multivariable-adjusted analyses. As a secondary analysis, the study outcomes were also descriptively compared between each of the TNF-α inhibitors. Because adalimumab is one of the most commonly used TNF-α inhibitor to treat RA, a secondary analysis was conducted to compare the outcomes among adalimumab, other TNF-α inhibitors, and non–TNF-α inhibitors.

      Findings

      Patient characteristics before initiating treatment with a second DMARD were similar across treatment groups (all TNF-α inhibitors [n = 296] and non–TNF-α biologics [n = 276]). The most common reasons for discontinuing etanercept treatment were inadequate response, adverse effects, and patient preference. After etanercept, TNF-α inhibitors overall were associated with a significantly lower EULAR good response rate (56.0% vs. 64.4%, P < 0.05) and smaller CDAI score change (−6.3 vs −7.3, P = .06) relative to non–TNF-α biologics. However, the proportion of patients achieving an EULAR good response was numerically higher for adalimumab versus other TNF-α inhibitors (61.1% vs 51.6%, P = 0.11) and comparable versus non–TNF-α biologics (61.1% vs 64.4%, P = 0.52). Adalimumab was also associated with a CDAI score change significantly greater than that of other TNF-α inhibitors (−7.1 vs −5.8, P < 0.05) and comparable to that of non–TNF-α biologics (−7.1 vs −7.3, P = 0.79). The results were consistent in the multivariable-adjusted analysis and secondary analysis.

      Implications

      In this retrospective analysis of patients with RA in the United Kingdom, France, and Germany, after discontinuation of etanercept treatment, TNF-α inhibitors as a class were overall less effective as second biologic DMARDs relative to non–TNF-α biologics; however, adalimumab was more or as effective as other TNF-α inhibitors and non–TNF-α biologics.

      Key words

      Introduction

      Rheumatoid arthritis (RA) is a chronic autoimmune disorder characterized by inflammation, pain, stiffness, and swelling of the joints.
      • Smolen J.S.
      • Aletaha D.
      • McInnes I.B.
      Rheumatoid arthritis.
      It is estimated to affect approximately 0.5% to 1% of the world population, with women being 2 to 3 times more likely to be affected than men.
      • Clements J.N.
      Treatment of rheumatoid arthritis: a review of recommendations and emerging therapy.
      • Gibofsky A.
      Epidemiology, pathophysiology, and diagnosis of rheumatoid arthritis: A Synopsis.
      • Helmick C.G.
      • et al.
      Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part I.
      As a consequence of progressive joint damage, synovitis, and often intense pain, RA is associated with both physical and mental disability, substantially impairing the quality of life of patients.
      • Smolen J.S.
      • Aletaha D.
      • McInnes I.B.
      Rheumatoid arthritis.
      • Matcham F.
      • et al.
      The impact of rheumatoid arthritis on quality-of-life assessed using the SF-36: A systematic review and meta-analysis. Semin Arthritis.
      • Pollard L.
      • Choy E.H.
      • Scott D.L.
      The consequences of rheumatoid arthritis: quality of life measures in the individual patient.
      • Brooks P.M.
      The burden of musculoskeletal disease--a global perspective.
      The treatment of RA includes primarily disease-modifying antirheumatic drugs (DMARDs), which can be grouped into conventional synthetic DMARDs, such as methotrexate, and biologic DMARDs, such as tumor necrosis factor alpha (TNF-α) inhibitors.
      • Kim G.
      • et al.
      Factors associated with the initiation of biologic disease-modifying antirheumatic drugs in Texas Medicaid patients with rheumatoid arthritis.
      • Keystone E.C.
      • Smolen J.
      • van Riel P.
      Developing an effective treatment algorithm for rheumatoid arthritis.
      The most recent European League Against Rheumatism (EULAR) guidelines recommend therapy with conventional synthetic DMARDs be started as soon as a diagnosis of RA is established and a first biologic DMARD be added if inadequate or no response is observed.
      • Smolen J.S.
      • et al.
      EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update.
      However, approximately 30% of patients treated with TNF-α inhibitors fail to achieve an improvement of 20% in the American College of Rheumatology criteria (ACR20),
      • Lipsky P.E.
      • et al.
      Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group
      Infliximab and methotrexate in the treatment of rheumatoid arthritis.
      • Weinblatt M.E.
      • et al.
      A trial of etanercept, a recombinant tumor necrosis factor receptor:Fc fusion protein, in patients with rheumatoid arthritis receiving methotrexate.
      • Weinblatt M.E.
      • et al.
      Adalimumab, a fully human anti-tumor necrosis factor alpha monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial.
      • Singh J.A.
      • et al.
      Biologics for rheumatoid arthritis: an overview of Cochrane reviews.
      which typically defines a threshold for response in patients with RA.
      • Felson D.T.
      • LaValley M.P.
      The ACR20 and defining a threshold for response in rheumatic diseases: too much of a good thing.
      Although switching to a second biologic DMARD after failure of a first one is generally advocated, no consensus exists on whether patients should switch to a biologic DMARD of the same or different mechanism of action (ie, a TNF-α inhibitor or a non–TNF-α biologic).
      • Villeneuve E.
      • Haraoui B.
      To switch or to change class-the biologic dilemma in rheumatoid arthritis.
      • Reynolds A.
      • et al.
      When is switching warranted among biologic therapies in rheumatoid arthritis?.
      Indeed, the EULAR guidelines consider both switching approaches acceptable.
      • Smolen J.S.
      • et al.
      EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update.
      In practice, because TNF-α inhibitors are the most frequently prescribed first biologic DMARDs,
      • Villeneuve E.
      • Haraoui B.
      To switch or to change class-the biologic dilemma in rheumatoid arthritis.
      • Reynolds A.
      • et al.
      When is switching warranted among biologic therapies in rheumatoid arthritis?.
      • Rashid N.
      • et al.
      Rates, factors, reasons, and economic impact associated with switching in rheumatoid arthritis patients newly initiated on biologic disease modifying anti-rheumatic drugs in an integrated healthcare system.
      a therapeutic dilemma that a rheumatologist may face is to whether switch patients from a first TNF-α inhibitor to another TNF-α inhibitor or to a non–TNF-α biologic.
      • Villeneuve E.
      • Haraoui B.
      To switch or to change class-the biologic dilemma in rheumatoid arthritis.

      Stetka BS, van Vollenhoven VF. New Rheumatoid Arthritis Management Guidelines: A Quick and Easy Guide. http://www.medscape.com/viewarticle/821404_10. Medscape Rheumatology, 2014.

      Because limited information is available on the most appropriate approach after failure and subsequent discontinuations of use of a first TNF-α inhibitor, more real-world data are needed to optimize treatment strategies in clinical practice.
      • Zhou Z.Y.
      • Griffith J.
      • Du E.X.
      • Chin D.
      • et al.
      Economic burden of switching to a non-tumor necrosis factor inhibitor versus a tumor necrosis factor inhibitor biologic therapy among patients with rheumatoid arthritis.
      Accordingly, this study aimed to assess the real-world effectiveness outcomes of patients with RA who discontinued etanercept treatment and subsequently received another TNF-α inhibitor (adalimumab, certolizumab pegol, golimumab, or infliximab) or a non–TNF-α biologic (abatacept or tocilizumab) as a second biologic DMARD in the United Kingdom, France, and Germany. An additional objective of this study was to compare the effectiveness of switching from etanercept to adalimumab (one of the most commonly used biologic DMARDs
      • Rashid N.
      • et al.
      Rates, factors, reasons, and economic impact associated with switching in rheumatoid arthritis patients newly initiated on biologic disease modifying anti-rheumatic drugs in an integrated healthcare system.
      ), another TNF-α inhibitor (certolizumab pegol, golimumab, or infliximab), or a non–TNF-α biologic. Rituximab was not included in the analysis because of unfavorable reimbursement decisions issued by health technology assessment authorities in the United Kingdom and France.
      National Institute for Health and Care Excellence (NICE)
      Haute Authorite De Sante (HAS)

      Patients and Methods

      Patient Selection

      Medical record data of patients with RA in the United Kingdom, France, and Germany were collected from a panel of rheumatologists. To contribute data to this study, rheumatologists were required to have treated at least 5 adult patients with RA who fulfilled the inclusion criteria described below. These patients were required to have a diagnosis of RA in their medical records, be at least 18 years of age, not be a participant in a clinical trial, and have initiated use of a TNF-α inhibitor (adalimumab, certolizumab pegol, golimumab, or infliximab) or a non–TNF-α biologic (abatacept or tocilizumab) between January 1, 2014, and May 31, 2015, after discontinuation of etanercept therapy. Patients were required to have been biologic naïve before beginning use of etanercept. Patients were required to have at least 1 year of continuous treatment between the initiation of use of their second biologic DMARD and their last available visit while still using this drug. Finally, patients were required to have medical records, including Disease Activity Score 28 (DAS28), available for at least one RA-related visit they had immediately before initiating use of their second biologic DMARD and at least 1 year after initiating use of their second biologic DMARD. Eligibility criteria were verified in the process of data abstraction.

      Medical Record Data Collection

      A case report form was developed in English and translated to French and German to standardize and facilitate data collection. After verifying the patient inclusion criteria, each participating rheumatologist was asked to randomly select up to 5 patients with RA to reduce physician fatigue and ensure data quality. Up to one medical record for each biologic DMARD was allowed per physician to ensure a representative sample. The entire case report form was pilot tested by 3 rheumatologists (one each from the United Kingdom, France, and Germany) to ensure that every question was clear and easy to interpret in each language. Stratified sampling was used and a quota was implemented to ensure a sufficient sample size to compare the treatment groups.
      The study sponsor and authors were not involved in the selection of physicians or eligible medical records, and the physicians who participated in the study were masked to the identity of the study sponsor. All the medical record data collected in this study were anonymous and nonidentifiable. The study was approved by the New England Institutional Review Board.

      Study Measures

      Patient characteristics and treatment considerations (listed below) were collected at the time of initiation of use of the second biologic DMARD. Disease characteristics, including physician- and patient-reported outcomes, were collected at the visit patients had immediately before initiating use of their second biologic DMARD (index visit) and at the last available visit while still taking the second biologic DMARD (follow-up visit).
      Patient characteristics included age, sex, duration of RA, duration of follow-up, the Charlson comorbidity index (CCI),
      • Quan H.
      • et al.
      Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data.
      autoimmune comorbidities (ankylosing spondylitis, celiac disease, Crohn’s disease, hidradenitis suppurativa, psoriasis, psoriatic arthritis, and other autoimmune comorbidities), DAS28 score, Clinical Disease Activity Index (CDAI) score, and nonbiologic DMARD use before the index visit. Treatment considerations included reasons for discontinuing use of etanercept and reasons for selecting the second biologic DMARD. On the basis of the current treatment guidelines,
      • Smolen J.S.
      • et al.
      EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update.
      EULAR response (good, moderate, none) and change in CDAI scores between the follow-up visit and index visit were included as study outcomes.

      Statistical Analyses

      In the primary analyses, baseline patient characteristics were summarized for the two treatment groups (TNF-α and non–TNF-α biologics) and compared using t tests for continuous variables and χ2 tests for categorical variables. Reasons for discontinuing etanercept treatment and selecting the second biologic DMARD were described. The study outcomes (including EULAR response and change in CDAI scores) were compared among treatment groups using descriptive and multivariable-adjusted analyses. The proportion of patients achieving an EULAR response corresponding to good was evaluated using a multivariable-adjusted logistic regression model that controlled for age, sex, country, CCI score, duration of RA at the index visit, and whether the patient had any autoimmune comorbidity. The mean change in CDAI score was assessed using a multivariable-adjusted linear regression model that adjusted for the same factors listed above.
      In the secondary analyses, baseline patient characteristics, EULAR response, and change in CDAI scores were descriptively compared between adalimumab and (1) each other TNF-α inhibitors (ie, certolizumab pegol, golimumab, and infliximab), (2) other TNF-α inhibitors as a class, and (3) non–TNF-α inhibitor biologics.
      All the analyses were conducted using SAS software, version 9.3 (SAS Institute, Cary, North Carolina). P ≤ 0.05 was considered statistically significant for all comparisons.

      Results

      Patient Characteristics

      Patient characteristics at the index visit were similar across treatment groups (all TNF-α inhibitors [n = 296], including adalimumab [n = 137] and other TNF-α inhibitors [n = 159], and non–TNF-α biologics [n = 276]). The mean age ranged from 47.6 to 50.7 years, most patients were women (TNF-α inhibitors, 68.9%; non–TNF-α biologics, 70.7%), and the mean duration of RA was 4.5 years for the TNF-α inhibitors and 5.0 years for the non–TNF-α biologics (Table I). The mean duration of follow-up between the index visit and the follow-up visit was 1.6 years for the TNF-α inhibitors and 1.5 years for the other TNF-α inhibitors. In addition, in all the treatment groups, the mean CCI score was 0.4, and autoimmune comorbidities were uncommon. Baseline RA severity was also comparable across treatment groups: the mean DAS28 was 4.8 for the TNF-α inhibitors and 5.0 for non–TNF-α biologics; the mean CDAI score was 13.3 for the TNF-α inhibitors and 14.1 for the non–TNF-α biologics. The most commonly used nonbiologic DMARD before the index visit was methotrexate (85.1% for TNF-α inhibitors and 83.7% for non–TNF-α biologics). On the index visit, most patients did not initiate use of a steroid (24.7% for TNF-α inhibitors and 30.1% for non–TNF-α biologics).
      Table IPatient characteristics at the index visit.
      Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      TNF-α InhibitorsNon–TNF-α Biologics
      Included abatacept and tocilizumab.
      (n = 276)
      CharacteristicAll TNF-α Inhibitors
      Included adalimumab, certolizumab pegol, golimumab, and infliximab.
      (n = 296)
      Adalimumab (n = 137)Other TNF-α Inhibitors
      Included certolizumab pegol, golimumab, and infliximab.
      (n = 159)
      Age at index date, y
       Mean (SD)48.3 (13.2)47.6 (12.8)48.9 (13.6)50.7 (13.1)
       Median47.745.649.750.9
      Male92 (31.1)44 (32.1)48 (30.2)81 (29.3)
      Duration of RA at index date, y
       Mean (SD)4.5 (4.4)4.4 (4.3)4.5 (4.4)5.0 (4.7)
       Median3.12.93.13.6
      Duration of follow-up from index date, y
       Mean (SD)1.6 (0.4)1.6 (0.4)1.5 (0.4)1.6 (0.4)
       Median1.51.51.41.4
      Charlson comorbidity index, mean (SD)0.4 (0.8)0.4 (0.9)0.4 (0.7)0.4 (0.9)
      Autoimmune
       Ankylosing spondylitis3 (1.0)3 (2.2)0 (0.0)0 (0.0)
       Celiac disease3 (1.0)3 (2.2)0 (0.0)4 (1.5)
       Crohn’s disease2 (0.7)1 (0.8)1 (0.6)2 (0.7)
       Hidradenitis suppurativa1 (0.3)1 (0.7)0 (0.0)1 (0.4)
       Psoriasis13 (4.4)4 (3.0)9 (5.7)6 (2.2)
       Psoriatic arthritis1 (0.3)1 (0.8)0 (0.0)1 (0.4)
      Other autoimmune comorbidities0 (0.0)0 (0.0)0 (0.0)1 (0.4)
      DAS28 score
       Mean (SD)4.8 (1.3)4.9 (1.3)4.6 (1.3)5.0 (1.2)
       Median4.954.75.1
      CDAI score
       Mean (SD)13.3 (3.8)14.0 (3.4)12.8 (4.0)14.1 (3.7)
       Median14141415
      Nonbiologic DMARD used before index date
       Azathioprine6 (2.0)3 (2.2)3 (1.9)3 (1.1)
       Cyclosporine4 (1.4)2 (1.5)2 (1.3)3 (1.1)
       Hydroxychloroquine41 (13.9)22 (16.1)19 (11.9)55 (19.9)
       Leflunomide39 (13.2)21 (15.3)18 (11.3)50 (18.1)
       Methotrexate252 (85.1)115 (83.9)137 (86.2)231 (83.7)
       Sulfasalazine41 (13.9)23 (16.8)18 (11.3)55 (19.9)
       None23 (7.8)11 (8.0)12 (7.5)27 (9.8)
      Steroid treatment initiated on index date73 (24.7)32 (23.4)41 (25.8)83 (30.1)
      CCI = Charlson comorbidity index; CDAI = Clinical Disease Activity Index; DAS28 = Disease Activity Score 28; DMARD = disease-modifying antirheumatic drug; RA = rheumatoid arthritis; TNF-α = tumor necrosis factor α.
      low asterisk Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      Included adalimumab, certolizumab pegol, golimumab, and infliximab.
      Included certolizumab pegol, golimumab, and infliximab.
      § Included abatacept and tocilizumab.

      Treatment Considerations

      Across the treatment groups, the most common reasons for discontinuing etanercept treatment were inadequate response (TNF-α inhibitors, 72.0%; non–TNF-α biologics, 76.8%), adverse effects (TNF-α inhibitors, 21.6%; non–TNF-α biologics, 22.5%), and patient preference (TNF-α inhibitors, 15.9%; non–TNF-α biologics, 22.5%) (Table II). The 3 most common reasons for selecting any of the TNF-α inhibitors were physician experience (64.2%), efficacy (48.3%), and patient preference (30.4%), whereas those for selecting non–TNF-α biologics were different mechanism of action (57.2%), efficacy (51.8%), and physician experience (49.3%).
      Table IIReasons for discontinuing etanercept treatment and choosing a second biologic DMARD.
      Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      ReasonTNF-α InhibitorsNon–TNF-α Biologics
      Included abatacept and tocilizumab.
      (n = 276)
      All TNF-α Inhibitors
      Included adalimumab, certolizumab pegol, golimumab, and infliximab.
      (n = 296)
      Adalimumab (n = 137)Other TNF-α Inhibitors
      Included certolizumab pegol, golimumab, and infliximab.
      (n = 159)
      Reason for discontinuation of etanercept
      Rates do not sum to 100% because rheumatologists were asked to select all the appropriate response options.
       Economic considerations3 (1.0)2 (1.5)1 (0.6)2 (0.7)
       Inadequate response213 (72.0)105 (76.6)108 (67.9)212 (76.8)
       Patient preference47 (15.9)15 (10.9)32 (20.1)33 (12.0)
       Physician preference23 (7.8)10 (7.3)13 (8.2)14 (5.1)
       Adverse effects64 (21.6)33 (24.1)31 (19.5)62 (22.5)
       Other2 (0.7)0 (0.0)2 (1.3)3 (1.1)
      Reasons for choosing second biologic DMARD after etanercept failure
      Rates do not sum to 100% because rheumatologists were asked to select all the appropriate response options.
       Patient comorbidities15 (5.1)8 (5.8)7 (4.4)36 (13.0)
       Cost8 (2.7)1 (0.7)7 (4.4)3 (1.1)
       Efficacy143 (48.3)74 (54.0)69 (43.4)143 (51.8)
       Different mechanism of action65 (22.0)26 (19.0)39 (24.5)158 (57.2)
       Physician experience190 (64.2)93 (67.9)97 (61.0)136 (49.3)
       Patient preference90 (30.4)43 (31.4)47 (29.6)38 (13.8)
       Peer experience50 (16.9)24 (17.5)26 (16.4)43 (15.6)
       Route of administration70 (23.6)22 (16.1)48 (30.2)38 (13.8)
       Tolerability of drug45 (15.2)20 (14.6)25 (15.7)62 (22.5)
       Treatment guidelines61 (20.6)32 (23.4)29 (18.2)65 (23.6)
       Other2 (0.7)1 (0.7)1 (0.6)2 (0.7)
      DMARD = disease-modifying antirheumatic drug; TNF-α: tumor necrosis factor α.
      low asterisk Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      Included adalimumab, certolizumab pegol, golimumab, and infliximab.
      Included certolizumab pegol, golimumab, and infliximab.
      § Included abatacept and tocilizumab.
      || Rates do not sum to 100% because rheumatologists were asked to select all the appropriate response options.

      Study Outcomes

      The proportions of patients achieving various EULAR responses between the index and follow-up visits were generally similar across treatment groups (Table III and Figure). After etanercept, TNF-α inhibitors overall were associated with significantly lower EULAR good response rate (56.0% vs 64.4%, P < 0.05) relative to non–TNF-α biologics. However, in the secondary analysis, the proportion of patients achieving an EULAR good response was comparable for adalimumab versus non–TNF-α biologics (61.1% vs 64.4%, P = 0.52) and adalimumab versus other TNF-α inhibitors (61.1% vs 51.6%, P = 0.11). In multivariable analyses, relative to all the TNF-α inhibitors, the adjusted odds ratio (OR) for achieving an EULAR good response was 1.42 (95% CI, 1.01–1.99; P = 0.04) for non–TNF-α biologics; relative to adalimumab in the secondary analysis, the adjusted ORs for achieving an EULAR good response were 0.68 (95% CI, 0.43–1.08; P = 0.10) for other TNF-α inhibitors and 1.15 (95% CI, 0.75–1.76; P = 0.52) for non–TNF-α biologics, respectively.
      Table IIIEULAR response and CDAI score change from index date to follow-up visit.
      Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      TNF-α InhibitorsNon–TNF-α Inhibitors
      Included abatacept and tocilizumab.
      (n = 264)
      VariableAll TNF-α Inhibitors
      Included adalimumab, certolizumab pegol, golimumab, and infliximab.
      (n = 296)
      Adalimumab (n = 131)Other TNF-α Inhibitors
      Included certolizumab pegol, golimumab, and infliximab.
      (n = 153)
      EULAR response
       Good response
      The Disease Activity Score 28 (DAS28) at the follow-up visit was ≤3.2; the DAS28 change was <−1.2.
      159 (56.0)80 (61.1)79 (51.6)170 (64.4)
      P < 0.05 for comparison with all TNF-α inhibitors.
       Moderate response
      The DAS28 at the follow-up visit was ≤5.1 and >3.2; the DAS28 change was ≥−1.2 and <−0.6.
      88 (31.0)36 (27.5)52 (34.0)69 (26.1)
       No response
      The DAS28 score at the follow-up visit >5.1; the DAS28 change was ≥−0.6.
      37 (13.0)15 (11.5)22 (14.4)25 (9.5)
      CDAI score change, mean (SD)−6.3 (4.9)−7.1 (5.0)−5.8 (4.8)
      P < .05 for comparison with adalimumab.
      −7.3 (5.0)
      CDAI = Clinical Disease Activity Index; EULAR = European League Against Rheumatism; TNF-α = tumor necrosis factor α.
      low asterisk Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      Included adalimumab, certolizumab pegol, golimumab, and infliximab.
      Included certolizumab pegol, golimumab, and infliximab.
      § Included abatacept and tocilizumab.
      || The Disease Activity Score 28 (DAS28) at the follow-up visit was ≤3.2; the DAS28 change was <−1.2.
      P < 0.05 for comparison with all TNF-α inhibitors.
      # The DAS28 at the follow-up visit was ≤5.1 and >3.2; the DAS28 change was ≥−1.2 and <−0.6.
      low asterisklow asterisk The DAS28 score at the follow-up visit >5.1; the DAS28 change was ≥−0.6.
      †† P < .05 for comparison with adalimumab.
      Figure
      FigureEuropean League Against Rheumatism response categories by treatment group. TNF-α = tumor necrosis factor α.
      After etanercept, TNF-α inhibitors overall were associated with a smaller CDAI score change (−6.3 vs −7.3, P = 0.06) relative to non–TNF-α biologics. However, the adalimumab treatment group experienced significantly greater change in CDAI score relative to the other TNF-α treatment group (−7.1 vs −5.8, P < 0.05) but comparable change relative to the non–TNF-α biologic treatment group (−7.1 vs −7.3, P = 0.79) (Table III). In multivariable analyses, relative to TNF-α inhibitors, the adjusted mean difference for CDAI score change was −0.94 (95% CI, −1.88 to 0.00; P = 0.05) for non–TNF-α biologics; relative to adalimumab, the adjusted mean differences for CDAI score change were 1.19 (95% CI, −0.12 to 2.50; P = 0.07) for other TNF-α inhibitors and −0.26 (95% CI, −1.46 to 0.94; P = 0.67) for non–TNF-α biologics, respectively.
      In the secondary analyses, compared with each other TNF-α inhibitor, adalimumab exhibited a higher rate of EULAR good response (61.1% vs 52.0% for certolizumab, 40.0% for golimumab, and 58.7% for infliximab) and greater change in CDAI score (−7.1 vs −6.0 for certolizumab, −5.6 for golimumab, and −5.6 for infliximab) (Table IV).
      Table IVEULAR response and CDAI change for each TNF-α inhibitor from index date to follow-up visit.
      Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      VariableAdalimumab (n = 131)Certolizumab (n = 50)Golimumab (n = 40)Infliximab (n = 63)
      EULAR response
       Good response
      The Disease Activity Score 28 (DAS28) at the follow-up visit was ≤3.2; the DAS28 change was <−1.2.
      80 (61.1%)26 (52.0%)16 (40.0%)
      P < 0.05 for comparison with adalimumab.
      37 (58.7%)
       Moderate response
      The DAS28 at the follow-up visit was ≤5.1 and >3.2; the DAS28 change was ≥−1.2 and <−0.6.
      36 (27.5%)17 (34.0%)19 (47.5%)
      P < 0.05 for comparison with adalimumab.
      16 (25.4%)
       No response
      The DAS28 score at the follow-up visit >5.1; the DAS28 change was ≥−0.6.
      15 (11.5%)7 (14.0%)5 (12.5%)10 (15.9%)
      CDAI score change, mean (SD)−7.1 ± 5.0−6.0 ± 4.9−5.6 ± 4.7−5.6 ± 4.8
      CDAI = Clinical Disease Activity Index; EULAR = European League Against Rheumatism; TNF-α = tumor necrosis factor α.
      low asterisk Data are presented as number (percentage) of patients unless otherwise indicated. Data are from 572 medical records (163 from the United Kingdom, 251 from France, and 158 from Germany).
      The Disease Activity Score 28 (DAS28) at the follow-up visit was ≤3.2; the DAS28 change was <−1.2.
      P < 0.05 for comparison with adalimumab.
      § The DAS28 at the follow-up visit was ≤5.1 and >3.2; the DAS28 change was ≥−1.2 and <−0.6.
      || The DAS28 score at the follow-up visit >5.1; the DAS28 change was ≥−0.6.
      Overall, at the country level, the results related to the EULAR response and CDAI score were directionally similar to those presented above, but no statistical significance was observed because the study was not powered for country-level comparisons.

      Discussion

      This medical record review study assessed real-world effectiveness associated with the use of several TNF-α inhibitors and non–TNF-α biologics after discontinuation of etanercept treatment in patients with RA in the United Kingdom, France, and Germany. Patients receiving TNF-α inhibitors (adalimumab, certolizumab pegol, golimumab, and infliximab) and non–TNF-α biologics (abatacept and tocilizumab) as second biologic DMARDs were found to have similar baseline characteristics and disease severity. The most common reasons for discontinuing etanercept treatment were inadequate response and the presence of adverse effects. After discontinuation of etanercept treatment, the real-world effectiveness, as measured by EULAR response and CDAI score change, was worse among TNF-α inhibitors than non–TNF-α biologics; however, when singling out one of the most commonly used biologics (ie, adalimumab
      • Rashid N.
      • et al.
      Rates, factors, reasons, and economic impact associated with switching in rheumatoid arthritis patients newly initiated on biologic disease modifying anti-rheumatic drugs in an integrated healthcare system.
      ) from the remaining TNF-α inhibitors, the effectiveness of adalimumab was comparable to, if not better than, that associated with the other TNF-α inhibitors and non–TNF-α biologics. In particular, patients treated with adalimumab as a second biologic DMARD experienced a significantly greater change in CDAI score compared with patients treated with all the other TNF-α inhibitors.
      Switching biologic DMARDs is common in clinical practice.
      • Reynolds A.
      • et al.
      When is switching warranted among biologic therapies in rheumatoid arthritis?.
      Several US studies using different methods estimated that 7.8% to 15.4% of patients switched to a second biologic DMARD within 1 year after initiating use of their first biologic DMARD.
      • Rashid N.
      • et al.
      Rates, factors, reasons, and economic impact associated with switching in rheumatoid arthritis patients newly initiated on biologic disease modifying anti-rheumatic drugs in an integrated healthcare system.
      • Meissner B.
      • et al.
      Switching of biologic disease modifying anti-rheumatic drugs in patients with rheumatoid arthritis in a real world setting.
      • Harnett J.
      • et al.
      Real-world evaluation of TNF-inhibitor utilization in rheumatoid arthritis.
      • Bonafede M.M.
      • et al.
      Factors associated with the initiation of disease-modifying antirheumatic drugs in newly diagnosed rheumatoid arthritis: a retrospective claims database study.
      • Ogale S.
      • Hitraya E.
      • Henk H.J.
      Patterns of biologic agent utilization among patients with rheumatoid arthritis: a retrospective cohort study.
      As the number of biologic DMARDs approved to treat RA increases, the selection of subsequent biologic DMARDs becomes increasingly challenging.
      • Husni M.E.
      • Betts K.A.
      • Griffith J.
      • et al.
      For instance, the 2013 EULAR guidelines recommend a second biologic DMARD be started if no improvement is observed within 3 months or the treatment goal is not achieved within 6 months after initiating use of the first biologic DMARD.
      • Smolen J.S.
      • et al.
      EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update.
      However, no guidance is provided regarding which drug and/or drug class should be used as first or second biologic DMARD.
      • Smolen J.S.
      • et al.
      EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update.
      The existing literature on the comparative efficacy or effectiveness of TNF-α inhibitors versus non–TNF-α biologics as second biologic DMARDs has been limited and sometimes contradictory. Several studies supported the beneficial effects of switching to a second TNF-α inhibitor after failure of a first one.
      • Nikas S.N.
      • et al.
      Efficacy and safety of switching from infliximab to adalimumab: a comparative controlled study.
      • Gomez-Reino J.J.
      • Carmona L B.G.
      The, Switching TNF antagonists in patients with chronic arthritis: an observational study of 488 patients over a four-year period. Arthritis Res Ther.
      • Hjardem E.
      • et al.
      Do rheumatoid arthritis patients in clinical practice benefit from switching from infliximab to a second tumor necrosis factor alpha inhibitor?.
      • van Vollenhoven R.
      • et al.
      Treatment with infliximab (Remicade) when etanercept (Enbrel) has failed or vice versa: data from the STURE registry showing that switching tumour necrosis factor alpha blockers can make sense.
      For example, one study found that 75% of patients who switched from infliximab to adalimumab achieved a 20% response in the American College of Rheumatology criteria after 12 months.
      • Nikas S.N.
      • et al.
      Efficacy and safety of switching from infliximab to adalimumab: a comparative controlled study.
      In contrast, other studies have reported superior clinical benefits associated with the switch from a TNF-α inhibitor to a non–TNF-α biologic versus the switch from one TNF-α inhibitor to another.
      • Chatzidionysiou K.
      • van Vollenhoven R.F.
      Rituximab versus anti-TNF in patients who previously failed one TNF inhibitor in an observational cohort.
      • Emery P.
      • et al.
      Rituximab versus an alternative TNF inhibitor in patients with rheumatoid arthritis who failed to respond to a single previous TNF inhibitor: SWITCH-RA, a global, observational, comparative effectiveness study.
      • Finckh A.
      • et al.
      B cell depletion may be more effective than switching to an alternative anti-tumor necrosis factor agent in rheumatoid arthritis patients with inadequate response to anti-tumor necrosis factor agents.
      • Harrold L.R.
      • et al.
      The comparative effectiveness of abatacept versus anti-tumour necrosis factor switching for rheumatoid arthritis patients previously treated with an anti-tumour necrosis factor.
      The findings of the present study indicate that, after etanercept, TNF-α inhibitors overall were less effective relative to non–TNF-α biologics; however, adalimumab was found to have similar or more effectiveness than the other TNF-α inhibitors and non–TNF-α biologics.
      Given the conflicting results regarding the benefits of continued TNF-α inhibition versus a non–TNF-α biologic after discontinuing use of a first TNF-α inhibitor, this study provides health care decision makers with important evidence on the real-world comparative effectiveness of TNF-α inhibitors and non–TNF-α biologics after discontinuation of treatment with etanercept, a commonly prescribed biologic DMARDs. Because switching from one TNF-α inhibitor to another is so common, the results reported here are particularly important to help physicians optimize their treatment strategies. More real-world studies that compare different biologic agents used as first and second biologic DMARDs are needed to optimize the treatment algorithm for RA and maximize clinical outcomes, as well as inform policy decisions and clinical guidelines in the United Kingdom, France, and Germany. In particular, future studies should evaluate and compare the clinical outcomes associated with the switch from other commonly used first-line TNF-α inhibitors, such as adalimumab, to another TNF-α inhibitor or a non–TNF-α biologic.
      The results of this study should be interpreted in light of some limitations. First, medical record data abstraction might be subject to errors and/or omissions. However, to the extent that all treatment groups were affected in a similar way, no significant bias was expected to affect the current comparative findings. Second, unobserved confounding factors and selection bias could result from the nonrandomization of patients to treatment groups. To minimize this problem, multivariable analyses controlled for patient characteristics commonly found in medical record data and known to be prognostic for outcomes in patients with RA. Third, the EULAR response rates reported here were found to be higher than previous studies.
      • Hetland M.L.
      • et al.
      Direct comparison of treatment responses, remission rates, and drug adherence in patients with rheumatoid arthritis treated with adalimumab, etanercept, or infliximab: results from eight years of surveillance of clinical practice in the nationwide Danish DANBIO registry.
      This is most likely because, in this study, patients were required to have received a second biologic DMARD for at least 1 year in an effort to select only long-term responders (ie, patients in whom a second biologic DMARD failed within a year of treatment initiation were not included in the analysis). However, comparisons among treatment groups were not expected to have been affected because this selection criterion was equally applied to all treatment groups.

      Conclusions

      In this retrospective analysis of patients with RA in the United Kingdom, France, and Germany, after discontinuation of etanercept treatment, TNF-α inhibitors as a class were overall less effective as second biologic DMARDs than non–TNF-α biologics; however, adalimumab was more or as effective as other TNF-α inhibitors and non–TNF-α biologics.

      Funding Sources

      Funding for this study was provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the manuscript.

      Conflict of Interest

      K.A. Betts is an employee of Analysis Group, which received payment from AbbVie for participation in this research. N. Li and A.J. Messali were employees of Analysis Group at the time of this research. V. Garg and M. Skup are employees of AbbVie and may hold stocks or stock options in AbbVie. The authors have indicated that they have no other conflicts of interest regarding the content of this article.

      Acknowledgments

      Medical writing assistance was provided by Cinzia Metallo, PhD, an employee of Analysis Group Inc.

      References

        • Smolen J.S.
        • Aletaha D.
        • McInnes I.B.
        Rheumatoid arthritis.
        The Lancet. 2016; 388: 2023-2038
        • Clements J.N.
        Treatment of rheumatoid arthritis: a review of recommendations and emerging therapy.
        Formulary J. 2011; 46: 532-545
        • Gibofsky A.
        Epidemiology, pathophysiology, and diagnosis of rheumatoid arthritis: A Synopsis.
        Am J Manag Care. 2014; 20: S128-S135
        • Helmick C.G.
        • et al.
        Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part I.
        Arthritis Rheum. 2008; 58: 15-25
        • Matcham F.
        • et al.
        The impact of rheumatoid arthritis on quality-of-life assessed using the SF-36: A systematic review and meta-analysis. Semin Arthritis.
        Rheumatism. 2014; 44: 123-130
        • Pollard L.
        • Choy E.H.
        • Scott D.L.
        The consequences of rheumatoid arthritis: quality of life measures in the individual patient.
        Clin Exp Rheumatol. 2005; 23: S43-S52
        • Brooks P.M.
        The burden of musculoskeletal disease--a global perspective.
        Clin Rheumatol. 2006; 25: 778-781
        • Kim G.
        • et al.
        Factors associated with the initiation of biologic disease-modifying antirheumatic drugs in Texas Medicaid patients with rheumatoid arthritis.
        J Manag Care Spec Pharm. 2015; 21: 401-407
        • Keystone E.C.
        • Smolen J.
        • van Riel P.
        Developing an effective treatment algorithm for rheumatoid arthritis.
        Rheumatology. 2012; 51: v48-v54
        • Smolen J.S.
        • et al.
        EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update.
        Ann Rheum Dis. 2017; 76: 960-977
        • Lipsky P.E.
        • et al.
        • Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group
        Infliximab and methotrexate in the treatment of rheumatoid arthritis.
        N Engl J Med. 2000; 343: 1594-1602
        • Weinblatt M.E.
        • et al.
        A trial of etanercept, a recombinant tumor necrosis factor receptor:Fc fusion protein, in patients with rheumatoid arthritis receiving methotrexate.
        N Engl J Med. 1999; 340: 253-259
        • Weinblatt M.E.
        • et al.
        Adalimumab, a fully human anti-tumor necrosis factor alpha monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial.
        Arthritis Rheum. 2003; 48: 35-45
        • Singh J.A.
        • et al.
        Biologics for rheumatoid arthritis: an overview of Cochrane reviews.
        Cochrane Database Syst Rev. 2009; : Cd007848
        • Felson D.T.
        • LaValley M.P.
        The ACR20 and defining a threshold for response in rheumatic diseases: too much of a good thing.
        Arthritis Res Ther. 2014; 16 (101–101)
        • Villeneuve E.
        • Haraoui B.
        To switch or to change class-the biologic dilemma in rheumatoid arthritis.
        Nat Rev Rheumatol. 2010; 6: 301-305
        • Reynolds A.
        • et al.
        When is switching warranted among biologic therapies in rheumatoid arthritis?.
        Expert Rev Pharmacoecon Outcomes Res. 2012; 12: 319-333
        • Rashid N.
        • et al.
        Rates, factors, reasons, and economic impact associated with switching in rheumatoid arthritis patients newly initiated on biologic disease modifying anti-rheumatic drugs in an integrated healthcare system.
        J Med Econ. 2016; 19: 568-575
      1. Stetka BS, van Vollenhoven VF. New Rheumatoid Arthritis Management Guidelines: A Quick and Easy Guide. http://www.medscape.com/viewarticle/821404_10. Medscape Rheumatology, 2014.

        • Zhou Z.Y.
        • Griffith J.
        • Du E.X.
        • Chin D.
        • et al.
        Economic burden of switching to a non-tumor necrosis factor inhibitor versus a tumor necrosis factor inhibitor biologic therapy among patients with rheumatoid arthritis.
        Adv Ther. 2016; 33 (http://dx.doi.org/10.1007/s12325-016-0318-5): 807-823
        • National Institute for Health and Care Excellence (NICE)
        Adalimumab, etanercept, infliximab, rituximab and abatacept for the treatment of rheumatoid arthritis after the failure of a TNF inhibitor. 2010; ([cited 2016 December])
        • Haute Authorite De Sante (HAS)
        Transparency Committee Opinion - Mabthera. 2012; ([cited 2016 December])
        • Quan H.
        • et al.
        Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data.
        Med Care. 2005; 43: 1130-1139
        • Meissner B.
        • et al.
        Switching of biologic disease modifying anti-rheumatic drugs in patients with rheumatoid arthritis in a real world setting.
        J Med Econ. 2014; 17: 259-265
        • Harnett J.
        • et al.
        Real-world evaluation of TNF-inhibitor utilization in rheumatoid arthritis.
        J Med Econ. 2016; 19: 101-112
        • Bonafede M.M.
        • et al.
        Factors associated with the initiation of disease-modifying antirheumatic drugs in newly diagnosed rheumatoid arthritis: a retrospective claims database study.
        Clin Ther. 2012; 34: 457-467
        • Ogale S.
        • Hitraya E.
        • Henk H.J.
        Patterns of biologic agent utilization among patients with rheumatoid arthritis: a retrospective cohort study.
        BMC Musculoskel Disord. 2011; 12 (204–204)
        • Husni M.E.
        • Betts K.A.
        • Griffith J.
        • et al.
        Rheumatol Int. 2017; (http://dx.doi.org/10.1007/s00296-017-3760-z)
        • Nikas S.N.
        • et al.
        Efficacy and safety of switching from infliximab to adalimumab: a comparative controlled study.
        Ann Rheum Dis. 2006; 65: 257-260
        • Gomez-Reino J.J.
        • Carmona L B.G.
        The, Switching TNF antagonists in patients with chronic arthritis: an observational study of 488 patients over a four-year period. Arthritis Res Ther.
        . 2006; 8 (R29–R29)
        • Hjardem E.
        • et al.
        Do rheumatoid arthritis patients in clinical practice benefit from switching from infliximab to a second tumor necrosis factor alpha inhibitor?.
        Ann Rheum Dis. 2007; 66: 1184-1189
        • van Vollenhoven R.
        • et al.
        Treatment with infliximab (Remicade) when etanercept (Enbrel) has failed or vice versa: data from the STURE registry showing that switching tumour necrosis factor alpha blockers can make sense.
        Ann Rheum Dis. 2003; 62: 1195-1198
        • Chatzidionysiou K.
        • van Vollenhoven R.F.
        Rituximab versus anti-TNF in patients who previously failed one TNF inhibitor in an observational cohort.
        Scand J Rheumatol. 2013; 42: 190-195
        • Emery P.
        • et al.
        Rituximab versus an alternative TNF inhibitor in patients with rheumatoid arthritis who failed to respond to a single previous TNF inhibitor: SWITCH-RA, a global, observational, comparative effectiveness study.
        Ann Rheum Dis. 2015; 74: 979-984
        • Finckh A.
        • et al.
        B cell depletion may be more effective than switching to an alternative anti-tumor necrosis factor agent in rheumatoid arthritis patients with inadequate response to anti-tumor necrosis factor agents.
        Arthritis Rheum. 2007; 56: 1417-1423
        • Harrold L.R.
        • et al.
        The comparative effectiveness of abatacept versus anti-tumour necrosis factor switching for rheumatoid arthritis patients previously treated with an anti-tumour necrosis factor.
        Ann Rheum Dis. 2015; 74: 430-436
        • Hetland M.L.
        • et al.
        Direct comparison of treatment responses, remission rates, and drug adherence in patients with rheumatoid arthritis treated with adalimumab, etanercept, or infliximab: results from eight years of surveillance of clinical practice in the nationwide Danish DANBIO registry.
        Arthritis Rheum. 2010; 62: 22-32