TYPE Original Research PUBLISHED 03 October 2022 DOI 10.3389/fphar.2022.980832 Real-world effectiveness and OPEN ACCESS persistence of reference EDITED BY Elham Rahme, McGill University, Canada etanercept versus biosimilar REVIEWED BY Juliana De Oliveira Costa, etanercept GP2015 among University of New South Wales, Australia Marc Henri De Longueville, UCB Pharma, Belgium rheumatoid arthritis patients: A *CORRESPONDENCE Nuria Carballo, cohort study ncarballo@psmar.cat Santiago Grau, sgrau@psmar.cat Nuria Carballo 1,2*, Carolina Pérez García 3, Santiago Grau 1,2*, SPECIALTY SECTION Jordi Monfort 2,3, Xavier Durán-Jordà 4, Daniel Echeverría-Esnal 1 This article was submitted to Drugs and Olivia Ferrández 1,2 Outcomes Research and Policies, a section of the journal 1 Pharmacy Department, Hospital del Mar—Parc de Salut Mar, Barcelona, Spain, 2Universitat Autònoma Frontiers in Pharmacology de Barcelona, Barcelona, Spain, 3Rheumatology Department, Hospital del Mar—Parc de Salut Mar, Barcelona, Spain, 4Methodology and Biostatistics Support Unit, Institute Hospital del Mar for Medical RECEIVED 29 June 2022 Research (IMIM), Barcelona, Spain ACCEPTED 25 August 2022 PUBLISHED 03 October 2022 CITATION Carballo N, Pérez García C, Grau S, Although several randomized clinical trials have confirmed that there is no Monfort J, Durán-Jordà X, difference in efficacy between etanercept and its biosimilar versions in the Echeverría-Esnal D and Ferrández O (2022), Real-world effectiveness and treatment of rheumatoid arthritis (RA), limited real-world evidence is persistence of reference etanercept available. We conducted a cohort study to compare the effectiveness and versus biosimilar etanercept treatment persistence between the reference etanercept (ETN) and the GP2015 among rheumatoid arthritis patients: A cohort study. biosimilar GP2015 in RA patients in a real-life setting. Adults with a Front. Pharmacol. 13:980832. diagnosis of RA who initiated treatment with ETN or GP2015, between doi: 10.3389/fphar.2022.980832 January 2007 and December 2019, were included. The follow-up period COPYRIGHT was 52 weeks. The primary outcome was the mean of change in the DAS28- © 2022 Carballo, Pérez García, Grau, Monfort, Durán-Jordà, Echeverría- CRP values and the adjusted mean difference from baseline to 52 weeks Esnal and Ferrández. This is an open- between ETN and GP2015. Other effectiveness endpoints assessed were the access article distributed under the terms of the Creative Commons rate of patients who achieved remission or low disease activity (LDA) at week Attribution License (CC BY). The use, 52, who showed a reduction of DAS28-CRP value greater than or equal to distribution or reproduction in other 1.2 from baseline to week 52 and rate of good responder patients (those forums is permitted, provided the original author(s) and the copyright meeting both effectiveness measures) at week 52. Treatment effectiveness owner(s) are credited and that the over time (baseline, 26 and 52 weeks) was compared between the ETN and original publication in this journal is GP2015 groups using mixed effects models. Treatment persistence cited, in accordance with accepted academic practice. No use, distribution (probability of maintaining the same treatment over time) was also or reproduction is permitted which does evaluated and shown using Kaplan–Meier survival curves. A total of not comply with these terms. 115 RA patients were included (ETN, n = 90; GP2015, n = 25). No differences were observed in the primary outcome: DAS28-CRP score decreased from baseline to week 52 [5.1 to 2.7 (mean of change -2.37) in ETN group and 5.0 to 2.2 (mean of change -2.84) in GP2015 group, p-value = 0.372] and the adjusted mean difference was −0.37 (−1.03 to 0.29). No differences were also observed in the other effectiveness endpoints assessed among patients treated with ETN or GP2015: rate of patients who achieved remission (54.1% vs. 66.7%, p-value = 0.303) and LDA Frontiers in Pharmacology 01 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 (71.6% vs. 80.9%, p-value = 0.391) at week 52, reduction of DAS28-CRP value greater than or equal to 1.2 from baseline to week 52 (75.6% vs. 80.9%, p-value = 0.613) and rate of good responder patients (58.1% vs. 76.1%, p-value = 0.202). Drug survival was 82% and 80% for ETN and GP2015, respectively (log-rank p-value = 0.804). Etanercept and its biosimilar GP2015 show similar effectiveness and treatment persistence in RA patients in a real-life setting. KEYWORDS GP2015, etanercept (enbrel), biosimilar agents, effectivenes, rheumatoid anhritis, drug survival real-life data 1 Introduction development of biosimilar drugs, which substantially reduce the economic impact of treatments and improve drug accessibility Rheumatoid arthritis (RA) is a chronic, autoimmune, and (Azevedo et al., 2015). inflammatory disease that can lead to accumulating joint damage A biosimilar is a biological agent that contains a version of and disability. This pathology can cause significant morbidity the active substance of an already authorized original biological and decreased quality of life. RA is a heterogeneous disease, with medicinal product (reference or originator biologic) and whose variable clinical presentation and pathogenic mechanisms in similarity to the reference medicinal in terms of quality which both genetic and environmental factors may be characteristics, biological activity, safety and efficacy has been involved (Smolen et al., 2018). proved (European Medicines Agency, 2014). Its prevalence in the adult population ranges around 0.5–1% GP2015 (Erelzi®) is the second etanercept biosimilar worldwide (Alamanos et al., 2006), being of approximately 0.5% approved for all indications included in the reference product in Spain (Silva-Fernández et al., 2020). However, there is a label (Deeks, 2017; Fitton et al., 2018). Bioequivalence between considerable variation among the different populations. GP2015 and etanercept originator Enbrel® (ETN) was firstly New management recommendations, such as early diagnosis demonstrated in healthy patients (von Richter et al., and the implementation of prompt initiation of effective therapy 20152017). Through the EGALITY study, in which bio in a treat-to-target approach, have considerably improved RA similarity between GP2015 and ETN in patients with plaque outcomes (Grigor et al., 2004; Smolen et al., 2016). According to psoriasis was demonstrated (Griffiths et al., 2017), GP2015 was the recommendations of the European League against approved by the Food & Drug Administration (FDA) and the Rheumatism (EULAR; Felson et al., 2011), the current European Medicines Agency (EMA) in 2016 and 2017, treatment goal is to achieve disease remission or, at least, a respectively. Efficacy equivalence of both drugs in RA patients low disease activity (Aletaha et al., 2005). was later confirmed following the EQUIRA study (Matucci- The emergence of biological disease modifying anti- Cerinic et al., 2018). However, these randomized controlled rheumatic drugs (DMARDs), such as tumor necrosis factor trials (RCTs) usually have high internal but low external alpha inhibitors (TNFi), has revolutionized the treatment of validity (Haugeberg et al., 2021). RA. Five TNFi drugs are currently approved for the treatment The retention rate of treatment with biological therapy, drug of RA: Etanercept, infliximab, adalimumab, golimumab, and survival or the probability of treatment persistence with the same certolizumab pegol. biological drug (i.e. the probability of maintaining the treatment Etanercept was the first TNFi to obtain approval for the over time) provides an index of overall drug effectiveness, patient treatment of RA in 1998 and currently remains a first-line satisfaction and treatment compliance (Wolfe, 1995) (21). However, biological DMARD therapy for RA worldwide (Chadwick the evidence in terms of treatment persistence generated in RCTs et al., 2018). Etanercept is a dimeric fusion protein consisting presents many limitations, mainly due to inherent biases associated of the extracellular ligand-binding portion of the human with its measurement under optimal conditions. Therefore, it may 75 kilodalton (p75) TNF receptor linked to the Fc portion of not be translated into the assessment of persistence in real-world human IgG1. By competitively inhibiting TNF binding to its cell settings (Hernandez et al., 2019). surface receptors, etanercept prevents TNF-mediated In this context, it is necessary to generate real-world evidence inflammatory cellular responses (Burness and Duggan, 2016). comparing the originator and biosimilar products of etanercept Despite the significant benefits of these biologicals, the high to guide clinicians in daily practice. acquisition cost of these drugs places a significant economic The aim of this study was to compare the effectiveness and burden on healthcare resources (Curtis and Singh, 2011; Putrik treatment persistence between ETN and GP2015 in a real-life et al., 2014). The expiration of patents protection has led to the cohort of RA patients. Frontiers in Pharmacology 02 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 2 Materials and methods antibodies (ACPA), serum levels of C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) at etanercept 2.1 Participants introduction; medication data (concomitant DMARDs and corticosteroids, prior treatments [both DMARDs and other Patients with RA aged 18 years and older, who initiated biologic agents] and number of previous treatments). Adverse treatment with ETN between January 2007 and December events (AE) during etanercept therapy that were the reason for 2019 and with GP2015 between January 2018 and December interruption of treatment were reflected. 2019, were included in this study. Patients ever treated with DAS28-CRP, DAS28-ESR, TJC, and SJC, serum CRP and etanercept that had undergone a wash-up period of at least ESR levels, use of concomitant DMARDs and corticosteroids 24 weeks were also included. Patients with a diagnosis of a were assessed at the baseline, at 26 and 52 weeks after the rheumatologic disease other than RA, and patients who had initiation of ETN or GP2015. previously received treatment with another biosimilar of All treatment decisions were at the discretion of the treating etanercept different than GP2015 were excluded. physician as part of routine clinical practice and clinical All patients were identified through pharmacy registry and evaluations were performed by the same clinician for all patients. an expert in RA recorded and assessed these data. 2.3 Study outcome 2.2 Study design and procedures 2.3.1 Effectiveness measurement This was a retrospective real-life observational cohort study The primary outcome of the study was the mean of change in carried out in a tertiary university hospital that serves as a the DAS28-CRP values and the adjusted mean difference from reference hospital for the population residing in one of the baseline to 52 weeks between ETN and GP2015. four health areas into which the Barcelona public health Other effectiveness endpoints assessed were the rate of system divides the territory, and which encompasses around patients who achieved remission (defined as DAS28-CRP ≤ 300,000 residents (Informació, 2022). Prevalence of rheumatoid 2.6) or low disease activity (LDA; defined as DAS28-CRP ≤ arthritis in Catalonia is estimated in 0.42% (Fina-Aviles et al., 3.2) at week 52 and patients who showed a reduction of DAS28- 2016). All patients were visited in the Rheumatology Department CRP value greater than or equal to 1.2 from baseline to week 52. and diagnosed with RA according to the 2010 American College The rate of patients meeting both effectiveness measures are of Rheumatology/European League Against Rheumatism classified as good responders according to the European EULAR collaborative initiative RA classification criteria (Aletaha et al., response criteria (Fransen and van Riel, 2009). 2010). The DAS28-CRP was used as the primary effectiveness The decision to initiate treatment with ETN or GP2015 was measure and was calculated as it is a validated instrument to made at the physician’s discretion and considering the fact that determine disease activity of RA at any given time point during GP2015 was introduced at the center in 2018. All treatments are the disease course (Wells et al., 2009). provided free of charge to patients and paid for by the publicly Treatment effectiveness was also measured through a financed National Health Care Service. longitudinal data analysis. The evolution profile of DAS28- The follow-up period was 52 weeks. Study ethical approval CRP, DAS28-ESR, SJC, TJC, serum CRP, and ESR levels and was obtained from the Clinical Research Ethics Committee of the concomitant use of DMARDs and corticosteroids were evaluated center (reference number: CEIm 2021/10022) and the at baseline, 26 and 52 weeks after the initiation of ETN or Strengthening The Reporting of Observational Studies in GP2015 and compared between both groups. Epidemiology (STROBE) statement guidelines were followed (Von Elm et al., 2007). 2.3.2 Treatment persistence measurement Data collected were extracted from the computerized medical Overall persistence was determined according to the ISPOR record system: demographic data (age, gender, ethnicity, body Medication Compliance and Persistence Work Group definition mass index (BMI)); smoking status; comorbid conditions; disease (i.e., “duration of time from initiation to discontinuation of status (disease duration [defined as the time from RA diagnosis therapy”; Cramer et al., 2008) and was estimated as the by the rheumatologist to the start of etanercept], clinical disease duration of time from etanercept therapy initiation to activity at the start of therapy [measured by the Disease Activity discontinuation during a follow-up period of 52 weeks. Score-28 using C-reactive protein (DAS28-CRP) and using Patients who discontinued treatment during the follow-up erythrocyte sedimentation rate (DAS28-ESR)], number of period were considered non-persistent. The start of follow-up swollen and tender joints (SJC and TJC respectively), presence was defined as the date on which the patient first received of bone erosions at study baseline, positive findings for etanercept, and the end of follow-up was defined as the date rheumatoid factor (RF) and anti-citrullinated protein of treatment interruption (last day of supply) for those who Frontiers in Pharmacology 03 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 discontinued or censored at 52 weeks of follow-up. Temporary between drugs were tested using the log-rank test. The rate of treatment interruptions (i.e., due to infections, surgery or poor treatment persistence at 52 weeks was also derived from these adherence) of less than 12 weeks’ duration were allowed in Kaplan–Meier survival curves and its confidence intervals (CI) accordance with other published studies of persistence with were based on beta product confidence procedure (BPCP). TNFi in RA patients where drug-free intervals of 30 days Results were considered as statistically significant at (Yazici et al., 2009), 60 days (Wu et al., 2008) or 90 days p-value<0.05. In addition, p-values for baseline characteristics (Grijalva et al., 2007) have been used. were adjusted for Benjamini and Hochberg correction. Statistical Persistence was retrospectively calculated through the analyses were performed using STATA 15.1 and bpcp electronic pharmacy dispensing record and etanercept survival package in R. rates were estimated by the Kaplan-Meier analysis. Mean time of treatment and reasons for etanercept discontinuation in those patients with treatment interruption 3 Results (lack of effectiveness, AEs or pregnancy) were recorded and compared between the two groups. A total of 186 RA patients initiated etanercept treatment during the study period, of whom 184 were treated for the first time with etanercept and only two had been treated previously. 2.4 Sample size calculation Seventy-one patients were excluded (38.1%) as absence of data from the first year of treatment (42, 22.6%), loss of follow-up (15, Accepting an alpha risk of 0.05 and a beta risk of 0.2 in a two- 8.1%) and treatment with another biosimilar different than sided test, a sample size of 87 subjects in the reference group and GP2015 (14, 7.5%). Finally, 115 (61.9%) patients were 29 in the biosimilar group was deemed necessary to recognize as included in the study, 90 patients in treatment with ETN and statistically significant a difference greater than or equal to 1 unit 25 patients in treatment with GP2015. Figure 1 summarized the in DAS28-CRP score. The SD is assumed to be 1.66. A patient study flow chart. allocation ratio of 3:1 was anticipated considering the low All patients started etanercept therapy at a once-weekly number of patients that received GP2015 over the study period. 50 mg subcutaneous injection. Clinical and demographic characteristics of the study population are summarized in Table 1. There were no clinically relevant differences in 2.5 Statistical analysis baseline and demographics characteristics between ETN and GP2015 groups, apart from differences in the ethnicity, a Categorical variables were presented as frequencies (number higher proportion of patients with erosive disease, and and percentage) and were compared using Chi squared and treatment with concomitant corticosteroids at baseline. Fisher exact tests, as appropriate. Continuous variables were described as the mean and standard deviation (SD) and were analyzed using Mann-Whitney U test. In addition, for 3.1 Effectiveness measurement parameters assessed at baseline, 26 and 52 weeks, differences between groups in each phase were tested separately using Mann- The DAS28-CRP score for the ETN-treated patients Whitney U test. The non-parametric choice was due because decreased significantly from 5.1 (1.5) at baseline to 2.7 (1.2) at most of the continuous variables did not fulfill the normality 52 weeks; the mean (SD) of change was −2.37 (1.66). The assumption and considering the low number of patients in the GP2015-treated patients also significantly improved DAS28- smaller group. CRP from 5.0 (1.1) at baseline to 2.2 (1.3) at 52 weeks; the Effectiveness evaluation was performed through longitudinal mean (SD) of change was −2.84 (1.91). There was no statistically data analysis using linear mixed models, introducing a time-per- significant difference (p-value = 0.372) between both groups. The arm interaction term to test the differences between both groups. adjusted mean difference for DAS28-CRP at week 52 between Main effect, drug and time, and their interaction were variables ETN and GP2015 was −0.37 (95% CI: −1.03 to 0.29). taken into account in the model. These analyses were adjusted for The rate of patients who achieved remission and LDA at potential confounders including baseline imbalances (ethnicity, week 52 and the rate of patients with good response are shown in erosive disease, concomitant corticosteroids) and use of Figure 2 and Figure 3, respectively. concomitant DMARDs using bivariate screening/stepwise Fifty-six (75.7%) patients and 17 (80.9%) in the ETN and selection. These variables were introduced based on clinical GP2015 group, respectively showed a reduction of DAS28-CRP criteria and taking into account their statistical significance in value greater than or equal to 1.2 from baseline to week 52, the bivariate analysis. p-value = 0.613. Drug survival analysis comparing ETN and GP2015 groups The evolution profile of DAS28-CRP and DAS28-ESR are was shown using Kaplan–Meier survival curves. Differences presented in Table 2 and evolution profile of SJC, TJC, serum Frontiers in Pharmacology 04 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 FIGURE 1 The study flow chart in line with the STROBE: 186 patients were initially assessed for eligibility, of which 71 were excluded. Finally, 115 patients were analyzed, 90 in the group ETN-treated patients and 25 in the group GP2015-treated patients. All patients completed the follow up. STROBE: (Strengthening the Reporting of Observational Studies in Epidemiology). CRP and ESR levels and concomitant use of DMARDs and Reasons for etanercept discontinuation comparing the ETN corticosteroids are presented as additional files. and GP2015 groups were lack of effectiveness in 7 (7.8%) patients Figure 4 shows the changes over time in DAS28-CRP and and in 4 (16%) patients, AEs in 7 (7.8%) and in 1 (4%) patient, DAS28-VSG values and the comparison between ETN and and pregnancy in 2 (2.2%) and in 0 (0.0%) patients, respectively. GP2015 during the 52-weeks study period, without statistical No statistical differences between the two groups (p-value = differences. There were also no differences in the time course of 0.301) were observed. Reported AEs are described in a table as SJC, TJC, serum CRP and ESR levels (figures provided as additional. No other patients experienced any AE during follow- additional data). up period. 3.2 Treatment persistence measurement 4 Discussion Drug survival rates estimated by the Kaplan-Meier analysis In the present real-world analysis of RA patients treated with are shown in Figure 5. The rate of treatment persistence at etanercept, no differences in effectiveness and treatment 52 weeks was 0.82 (95% CI: 0.72–0.89) for ETN and 0.80 persistence for ETN compared to its biosimilar (95% CI: 0.59–0.93) for GP 2015. GP2015 during the first year of treatment were observed. Mean (SD) time of treatment was 46.74 (12.97) weeks and Baseline demographics and clinical characteristics were 46.51 (13.54) weeks for ETN and GP2015, respectively similar in both groups. However, differences in ethnicity were (p-value = 0.686). In non-persistence patients (patients observed, with significantly more Caucasian patients in the who discontinued treatment during follow-up) this value ETN group. This difference may be explained by the recent was 22.35 (15.03) and 24.53 (18.60) weeks, respectively increase in the number of non-Caucasian population in (p-value = 1.000). Spanish society (related to the increase in migratory flows Frontiers in Pharmacology 05 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 TABLE 1 Baseline demographic and clinical characteristics of RA patients treated with etanercept originator or its biosimilar GP2015. ETN GP2015 Total p-value p-valuea (n = 90) (n = 25) (n = 115) Demographics Age (years) at ETN introduction 53.1 (12.4) 55.9 (10.5) 53.7 (12.1) 0.362 0.766 Female 66 (73.3) 23 (92.0) 89 (77.4) 0.059 0.341 Weight (kg) 68.3 (14.8) 67.9 (17.5) 68.2 (15.3) 0.646 0.865 BMI (kg/m2) 25.5 (4.7) 26.4 (6.3) 25.7 (5.1) 0.793 0.920 Ethnicity 0.002 0.017 Caucasian 78 (86.7) 16 (64.0) 94 (81.7) Hispanic (American Indian or Alaska Native) 12 (13.3) 5 (20.0) 17 (14.8) Asian 0 (0.0) 3 (12.0) 3 (2.6) African American 0 (0.0) 1 (4.0) 1 (0.9) Smoking history Current smoker 13 (15.1) 4 (16.0) 17 (15.3) 1.000 1.000 Ex-smoker 14 (16.3) 5 (20.0) 19 (17.1) 0.763 0.920 Comorbidities Cardiovascular disease 26 (28.9) 11 (44.0) 37 (32.2) 0.255 0.564 Osteoporosis 31 (34.4) 7 (28.0) 38 (33.0) 0.635 0.865 Lung disease 14 (15.6) 3 (12.0) 17 (14.8) 1.000 1.000 Depression/anxiety 21 (23.3) 9 (36.0) 30 (26.1) 0.209 0.564 Diabetes 6 (6.7) 3 (12.0) 9 (7.8) 0.406 0.773 Glaucoma 3 (3.3) 3 (12.0) 6 (5.2) 0.116 0.443 Thyroid disorder 18 (20.0) 10 (40.0) 28 (24.3) 0.063 0.341 Disease status Duration of RA (years) 9.4 (6.9) 8.2 (7.1) 9.2 (6.9) 0.416 0.773 DAS28-CRP 5.1 (1.5) 5.0 (1.1) 5.1 (1.4) 0.655 0.865 DAS2-ESR 5.3 (1.6) 5.3 (1.2) 5.3 (1.5) 0.673 0.865 Erosive disease 79 (91.9) 16 (64.0) 95 (85.6) 0.002 0.017 RF positive 63 (73.3) 20 (80.0) 83 (74.8) 0.606 0.865 ACPA positive 57 (66.3) 21 (84.0) 78 (70.3) 0.135 0.443 CRP serum levels (mg/dl) 1.7 (2.5) 0.8 (0.7) 1.5 (2.2) 0.236 0.564 ESR serum levels (mm/h) 26.7 (20.6) 25.4 (21.6) 26.4 (20.7) 0.626 0.865 Prior therapy Number of previous DMARDs 2.2 (1.0) 1.9 (1.0) 2.2 (1.0) 0.154 0.443 Previous DMARDs 0.199 0.865 1 24 (26.7) 10 (40) 34 (29.6) 2 33 (36.7) 10 (40) 43 (37.4) 3 24 (26.7) 2 (8) 26 (22.6) ≥4 9 (10.0) 3 (12.0) 12 (10.4) (Continued on following page) Frontiers in Pharmacology 06 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 TABLE 1 (Continued) Baseline demographic and clinical characteristics of RA patients treated with etanercept originator or its biosimilar GP2015. ETN GP2015 Total p-value p-valuea (n = 90) (n = 25) (n = 115) Number of previous biologic agents 0.3 (0.6) 0.2 (0.5) 0.3 (0.6) 0.430 0.654 Biologic-naïve 66 (73.3) 21 (84.0) 87 (75.7) 0.429 0.773 Previous biologic agents 0.627 0.564 1 20 (22.2) 3 (12.0) 23 (20.0) 2 3 (3.3) 1 (4.0) 4 (3.5) 3 1 (1.1) 0 (0.0) 1 (0.9) Concomitant therapy Concomitant DMARDs Yes 69 (78.4) 19 (76.0) 88 (77.9) 0.789 0.920 MTX 37 (41.1) 9 (36.0) 46 (40.0) 0.818 0.920 MTX dose (mg/week) 19.29 (4.8) 20 (6.1) 19.43 (5.1) 0.503 0.835 Leflunomide 21 (23.3) 10 (40) 31 (27.0) 0.126 0.443 Leflunomide dose (mg/day) 17.6 (4.4) 19 (3.2) 18.1 (4.0) 0.540 0.865 Other 12 (13.3) 0 (0.0) 12 (10.4) 0.066 0.341 Concomitant corticosteroids Yes 65 (72.2) 7 (28.0) 72 (62.6) <0.001 0.004 Prednisone equivalent dose ≥10 mg/day 35 (55.6) 4 (57.1) 39 (55.7) 1.000 1.000 <10 mg/day 28 (44.4) 3 (42.9) 31 (44.3) a Benjamini & Hochberg correction. Data presented as mean (SD) and n (%). ETN etanercept originator (Enbrel®), GP2015 etanercept biosimilar (Erelzi®), RA rheumatoid arthritis, BMI body mass index, DAS28-CRP disease activity score 28 using c-reactive protein, DAS28-ESR disease activity score 28 using erythrocyte sedimentation rate, RF rheumatoid factor, ACPA anti-citrullinated protein antibody; CRP c-reactive protein, ESR erythrocyte sedimentation rate, MTX methotrexate, DMARDs disease modifying antirheumatic drugs. Statistically significant. FIGURE 2 Proportion of patients achieving remission defined as DAS28-CRP ≤2.6 and low-disease activity score (LDAS) defined as DASR28-CRP ≤3.2. The rate of patients who achieved remission at week 52 were 40 (54.6%) and 14 (66.7%), p-value = 0.303, in the ETN and GP2015 group, respectively. Fifty-three (71.6%) patients in the ETN group and 17 (80.9%) in the GP2015 group (p value = 0.391) achieved LDAS. Frontiers in Pharmacology 07 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 demonstrated at 24 weeks (Matucci-Cerinic et al., 2018). In this study the mean (SD) DAS28-CRP change from baseline to week 24 was similar between the GP2015 [−2.78 (1.1)] and ETN groups [−2.78 (1.0)]. Efficacy was assessed using the DAS28-CRP score, in the same way as we proceeded in our study. Treatment effectiveness has generally been determined by comparing group means of changes in disease activity variables. However, a significant difference between groups does not reflect the number of patients who responded to treatment (Fransen and van Riel, 2009). The present study therefore assessed the rate of patients who achieved remission or LDA and showed a reduction of DAS28-CRP value greater than or equal to FIGURE 3 1.2 from baseline to week 52 and confirmed that no Rate of good responder patients: who achieved low disease differences were found between both groups (p-value = activity (DAS28-CRP ≤ 3.2) at week 52 and show a significant 0.202). Although no statistically significant differences were change of 1.2 in DAS28-CRP from baseline to 52 weeks. It was reached in 43 (58.1%) and 16 (76.1%) patients in the ETN and observed, we are aware that 58.1% of ETN-patients versus GP2015 group respectively, p-value = 0.202. 76.1% of GP2015-patients achieved LDA and showed a significant reduction in DAS28-CRP at 52 weeks (≥1.2 from baseline). This could be in line with previously mentioned in recent years (Ministerio de Inclusión Seguridad Social y about updates in treatment strategy in recent years with a migraciones, 2020)) together with the fact that ETN was possible prompt therapy initiation in GP2015 patients. marketed almost 20 years before GP2015. A significantly A fundamental limitation of clinical trials is that the higher proportion of patients with erosive disease and in included patients generally differ from those attended in treatment with corticosteroids therapy at baseline were also daily practice, and patients in a normal clinical setting observed in the ETN group. Bone erosions are generally would be ineligible for such trials (Ugalde et al., 2021). This associated with disease severity and consistent with advanced fact could account for some differences between the EQUIRA disease (Fuchs et al., 1989) and these findings could be justified and the present study, such a shorter evaluation period by the update in RA treatment strategies, which lately employ early (24 weeks in front of 52 weeks in the present study) and initiation of DMARDs and concomitant short-term inclusion of patients with active disease in the first one. corticosteroids to suppress inflammation, especially in early RA Although RCTs are considered the gold standard for assessing (Crossfield et al., 2021; Smolen et al., 2020). efficacy, their results are often limited because of their rigorous The results of the present study are consistent with those of inclusion criteria. Furthermore, the generalizability of their clinical trials, as the EQUIRA study, in which efficacy findings is often limited because of the lack of extrapolation of equivalence between GP2015 and ETN in patients with RA was real-life setting (Zink et al., 2006; Aaltonen et al., 2017). TABLE 2 Evolution of DAS28 at baseline, at 26 and 52 weeks after the initiation and comparison between ETN and GP2015 groups. ETN (n = 90) GP2015 (n = 25) Total (n = 115) p-value DAS28-CRP Baseline 5.1 (1.5) 5.0 (1.1) 5.1 (1.4) 0.897 At 26 weeks 2.6 (1.1) 2.3 (1.5) 2.6 (1.2) 0.243 At 52 weeks 2.7 (1.2) 2.2 (1.3) 2.6 (1.2) 0.103 DAS28-ESR Baseline 5.3 (1.6) 5.3 (1.2) 5.3 (1.5) 0.425 At 26 weeks 2.6 (1.2) 2.3 (1.7) 2.5 (1.3) 0.992 At 52 weeks 2.6 (1.3) 2.1 (1.5) 2.5 (1.3) 0.377 Data presented as mean (SD) and n (%). ETN etanercept originator (Enbrel®), GP2015 etanercept biosimilar (Erelzi®), DAS28-CRP disease activity score 28 using c-reactive protein, DAS28-ESR disease activity score 28 using erythrocyte sedimentation rate. The differences observed in the baseline p-values compared to Table 1 are derived from the longitudinal statistical analysis used. Frontiers in Pharmacology 08 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 FIGURE 4 Mean change from baseline over 52 weeks in DAS28-CRP and DAS28-ESR scores of ETN and GP2015 groups. DAS28-CRP disease activity ® score 28 using c-reactive protein, DAS28-ESR disease activity score 28 using erythrocyte sedimentation rate, ETN etanercept originator (Enbrel ), ® GP2015 etanercept biosimilar (Erelzi ). DAS28-CRP (p-value = 0.263). DAS28-ESR (p-value = 0.293) Considering previous published real-life data, Atzeni et al. Furthermore, both studies compare the effectiveness and (2021) evaluated in a retrospective double center study safety of etanercept originator and its biosimilar SB4 and our effectiveness and safety of biosimilar SB4 (Benepali®) and study compares effectiveness with a different biosimilar product, ETN in RA real-life patients. At 6 months the DAS28-CRP GP2015. To our knowledge, this is the first real-world study score was not different between the 11 patients receiving first- evaluating effectiveness between ETN and GP2015, as all other line biosimilar etanercept and 51 patients receiving the originator (2.3 ± 1.2 versus 2.7 ± 1.3; p > 0.05). Other scores as DAS28-ESR, Clinical Disease Activity Index (CDAI) and Simple Disease Activity Index (SDAI) were evaluated, without significant differences. Codreanu et al. also demonstrated similar findings. Similar effectiveness and safety were shown between biosimilar SB4 and ETN after the first 6 months of treatment in a real-life national cohort of RA patients. After this period the DAS28-CRP score was 3.3 ± 1.3 in both 123 patients treated with etanercept and 119 patients treated with SB4 with a p = 0.829 (Codreanu et al., 2019). DAS28 remission and Boolean remission were achieved by 18.7% in ETN group and 17.6% in SB4 group (p = 0.823), and by 11.4% in ETN group and 11.8% in SB4 group (p = 0.926), respectively. Similar to the study of Atzeni et al., DAS28-ESR, CDAI and SDAI scores were also evaluated without significant differences. Both real-life studies evaluated a 6 months-treatment period. FIGURE 5 However, in our study a period of 12 months was considered, Overall treatment persistence with ETN and GP2015 in RA patients shown as the fraction (between 1 and 0) of patients since assessments of disease activity should be made at remaining on therapy during 52 weeks after therapy initiation 12 months after start of treatment in maintenance trials (long-rank p = 0.804). (Committee for Medicinal Products for Human Use, 2017). Frontiers in Pharmacology 09 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 published studies evaluated the effectiveness of other etanercept marketed. Another limitation is that some baseline biosimilar products, mainly SB4. differences are observed between the two study populations Real-world treatment persistence has been considered a because of the difference in time between both treatment proxy measure not only for treatment safety and effectiveness, strategies; however, this is a common occurrence in but also for patient satisfaction (Luttropp et al., 2019). Evaluation retrospective analyses. Pain and fatigue assessments, key of survival times and discontinuation rates associated with patient data, were not available also due to the retrospective treatment is necessary when evaluating their real-world nature of the study. Finally, we acknowledge that a sensitivity effectiveness, however there are many reasons associated with analysis with all included patients would have been of interest. treatment modification in addition to lack of effectiveness such as However, we have waived such analysis for the following AE, pregnancy, or death. RCTs are relatively short compared with reasons: the lack of primary outcome measure in patients the chronic course of RA and therefore real-world studies with with loss to follow-up and the use of other biosimilars, which longer patient follow-up can be very useful. would have increased heterogeneity. In the present study, survival probability curves were highly The strengths of our study are based on data collection similar for the ETN and GP2015. Although previous studies from a real-life clinical setting and a longer follow-up time on comparing TNFi persistence in patients with RA reported the measure of effectiveness than other real-life studies. conflicting results, our persistence rates at 52 weeks (82% for Another strength of the study is the scarcity of missing ETN and 80% for GP2015) appear to be higher than those data which contributes to reduce selection bias and increase reported in other publications, considering the inclusion of statistical power. both biological- naïve and non-naïve patients. A recent To the authors’ knowledge, no previous analysis systematic literature review and meta-analysis reported 12- comparing drug survival between etanercept originator and months survival rates in the first-line and second-line its biosimilar GP2015 in patients with RA in real-life setting treatment setting for etanercept originator of 71% and 61%, had been carried out, so these findings are the first reported in respectively (Emery et al., 2020). In other study involving this field. 6,153 biological-naïve RA patients, persistence rates of 82% In summary, ETN and its biosimilar GP2015 showed similar for etanercept up 52 weeks were reported (Sruamsiri et al., effectiveness and treatment persistence in RA patients in a real- 2018). Among non-persistent patients, there was no difference life setting. These results should be confirmed in other in mean treatment duration. observational studies with larger sample sizes. As we wanted to show the global persistence, all reasons for treatment discontinuation have been evaluated. The most prevalent reasons for discontinuation are lack of effectiveness Data availability statement and AEs in both groups. Although no statistically significant differences were observed in the reasons for drug The original contributions presented in the study are discontinuation, discontinuation due to a lack of included in the article/Supplementary Material, further effectiveness was higher in patients treated with the inquiries can be directed to the corresponding authors. biosimilar and it may be due to an earlier switch of therapy comparing to previous years due to an increase in the currently available biological drugs alternatives (Tak and Kalden, 2011; Ethics statement Conti et al., 2013). The study carried out presents several limitations, mainly The studies involving human participants were reviewed and related to its retrospective nature, which may lead to missing approved by the Clinical Research Ethics Committee of Hospital or incomplete data. It consists of a non-randomized single- del Mar (reference number: CEIm 2021/10022). Written center observational study. The prescription bias could be informed consent for participation was not required for this present, although no significant differences were found at study in accordance with the national legislation and the baseline characteristics in terms of disease activity. A institutional requirements. limitation which stands out above the rest is the difference in sample size between the study groups, with a particularly small sample size for the GP2015 group. Moreover, 25 patients Author contributions were included out of the 29 expected for the GP2015 group which represented a loss of 13,8% with respect the calculated NC, CPG, and OF have contributed to the design of the sample size. It was limited by the number of RA patients study, registration, and patient follow-up and writing of the treated in a tertiary hospital, the rise of variety of treating RA study. DE-E, SG, and JM have contributed to manuscript drugs other than etanercept in recent years, and the fact that at revision. XD-J has developed statistical analysis. All authors the time the study was initiated, GP2015 had just been approved the final version of the manuscript. Frontiers in Pharmacology 10 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 Acknowledgments Publisher’s note This study is part of a PhD program in Medicine of the All claims expressed in this article are solely those of the authors Universitat Autònoma de Barcelona. This study was presented, in and do not necessarily represent those of their affiliated organizations, part, as a poster at the “SEFH Virtual Congress,” 18–21 October or those of the publisher, the editors and the reviewers. Any product 2021 in La Coruña, Spain. that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Conflict of interest Supplementary material The authors declare that the research was conducted in The Supplementary Material for this article can be found the absence of any commercial or financial relationships that online at: https://www.frontiersin.org/articles/10.3389/fphar. could be construed as a potential conflict of interest. 2022.980832/full#supplementary-material References Aaltonen, K. J., Ylikylä, S., Joensuu, J. T., Isomäki, P., Pirilä, L., Kauppi, M., Deeks, E. D. (2017). GP2015: An etanercept biosimilar. BioDrugs 31 (6), 555–558. et al. (2017). Efficacy and effectiveness of tumour necrosis factor inhibitors doi:10.1007/s40259-017-0246-1 in the treatment of rheumatoid arthritis in randomized controlled trials and Emery, P., Vlahos, B., Szczypa, P., Thakur, M., Jones, H. E., Woolcott, J., et al. routine clinical practice. Rheumatology 56 (5), 725–735. doi:10.1093/ (2020). Longterm drug survival of tumor necrosis factor inhibitors in patients rheumatology/kew467 with rheumatoid arthritis. J. Rheumatol. 47 (4), 493–501. doi:10.3899/jrheum. Alamanos, Y., Voulgari, P. V., and Drosos, A. A. (2006). Incidence and prevalence 181398 of rheumatoid arthritis, based on the 1987 American college of rheumatology European Medicines Agency (2014). Biosimilar Guidel. 44 (18), 1–16. doi:10. criteria: A systematic review. Semin. Arthritis Rheum. 36 (3), 182–188. doi:10.1016/ 1089/blr.2011.9970 j.semarthrit.2006.08.006 Felson, D. T., Smolen, J. S., Wells, G., Zhang, B., Van Tuyl, L. H. D., Funovits, J., Aletaha, D., Neogi, T., Silman, A. J., Funovits, J., Felson, D. T., Bingham, C. O., et al. (2011). American college of rheumatology/European league against et al. (2010). 2010 rheumatoid arthritis classification criteria: An American college rheumatism provisional definition of remission in rheumatoid arthritis for of rheumatology/European league against rheumatism collaborative initiative. Ann. clinical trials. Arthritis Rheum. 63 (3), 573–586. doi:10.1002/art.30129 Rheum. Dis. 69 (9), 1580–1588. doi:10.1136/ard.2010.138461 Fina-Aviles, F., Medina-Peralta, M., Mendez-Boo, L., Hermosilla, E., Elorza, J. M., Aletaha, D., Ward, M. M., Machold, K. P., Nell, V. P. K., Stamm, T., and Smolen, Garcia-Gil, M., et al. (2016). The descriptive epidemiology of rheumatoid arthritis J. S. (2005). Remission and active disease in rheumatoid arthritis: Defining criteria in Catalonia: A retrospective study using routinely collected data. Clin. Rheumatol. for disease activity states. Arthritis Rheum. 52 (9), 2625–2636. doi:10.1002/art.21235 35 (3), 751–757. doi:10.1007/s10067-014-2801-1 Atzeni, F., Gerratana, E., Bongiovanni, S., Talotta, R., Miceli, G., Salaffi, F., et al. Fitton, J., Giollo, A., and Buch, M. H. (2018). GP2015 as a promising therapy for (2021). Efficacy and safety of biosimilar and originator etanercept in rheumatoid rheumatoid arthritis. Expert Opin. Biol. Ther. 18 (4), 477–481. doi:10.1080/ arthritis patients: Real-life data. Isr. Med. Assoc. J. 23 (7), 344–349. 14712598.2018.1452909 Azevedo, V. F., Galli, N., Kleinfelder, A., D’Ippolito, J., and Urbano, P. C. M. Fransen, J., and van Riel, P. L. C. M. (2009). The disease activity score and the (2015). Etanercept biosimilars. Rheumatol. Int. 35 (2), 197–209. doi:10.1007/ EULAR response criteria. Rheum. Dis. Clin. North Am. 35 (4), 745–757. doi:10. s00296-014-3080-5 1016/j.rdc.2009.10.001 Burness, C. B., and Duggan, S. T. (2016). Etanercept (SB4): A review in Fuchs, H. A., Kaye, J. J., Callahan, L. F., Nance, E. P., and Pincus, T. (1989). autoimmune inflammatory diseases. BioDrugs 30 (4), 371–378. doi:10.1007/ Evidence of significant radiographic damage in rheumatoid arthritis within the first s40259-016-0188-z 2 years of disease. J. Rheumatol. 16 (5), 585–591. Chadwick, L., Zhao, S., Mysler, E., and Moots, R. J. (2018). Review of biosimilar Griffiths, C. E. M., Thaçi, D., Gerdes, S., Arenberger, P., Pulka, G., Kingo, K., et al. trials and data on etanercept in rheumatoid arthritis. Curr. Rheumatol. Rep. 20 (12), (2017). The EGALITY study: A confirmatory, randomized, double-blind study 84. doi:10.1007/s11926-018-0799-0 comparing the efficacy, safety and immunogenicity of GP2015, a proposed Codreanu, C., Popescu, C. C., Mogoșan, C., Enache, L., Daia, S., Ionescu, R., et al. etanercept biosimilar, vs. the originator product in patients with moderate-to- (2019). Efficacy and safety of original and biosimilar etanercept (SB4) in active severe chronic plaque-type psoriasis. Br. J. Dermatol. 176 (4), 928–938. doi:10.1111/ rheumatoid arthritis – a comparison in a real-world national cohort: Original and bjd.15152 biosimilar etanercept. Ra. Biol. 62 (5), 27–32. doi:10.1016/j.biologicals.2019.10.009 Grigor, C., Capell, H., Stirling, A., McMahon, A. D., Lock, P., Vallance, R., et al. Committee for Medicinal Products for Human Use (2017). Guideline on clinical (2004). Effect of a treatment strategy of tight control for rheumatoid arthritis (the investigation of medicinal products for the treatment of rheumatoid arthritis. Eur. TICORA study): A single-blind randomised controlled trial. Lancet 364 (9430), Med. Agency 44 (2017), 2–16. 263–269. doi:10.1016/S0140-6736(04)16676-2 Conti, F., Ceccarelli, F., Massaro, L., Cipriano, E., Di Franco, M., Alessandri, C., Grijalva, C. G., Chung, C. P., Arbogast, P. G., Stein, C. M., Mitchel, E. F., and et al. (2013). Biological therapies in rheumatic diseases. Clin. Ter. 164 (Suppl. 5), Griffin, M. R. (2007). Assessment of adherence to and persistence on disease- e413–e428. doi:10.7417/CT.2013.1622 modifying antirheumatic drugs (DMARDS) in patients with rheumatoid arthritis. Med. Care 45 (10 Suppl. 2), 66–76. doi:10.1097/mlr.0b013e318041384c Cramer, J. A., Roy, A., Burrell, A., Fairchild, C. J., Fuldeore, M. J., Ollendorf, D. A., et al. (2008). Medication compliance and persistence: Terminology and definitions. Haugeberg, G., Bakland, G., Rødevand, E., Widding Hansen, I. J., Diamantopoulos, A., and Pripp, A. H. (2021). Pos0612 exploring equivalence Value Health 11 (1), 44–47. doi:10.1111/j.1524-4733.2007.00213.x between biosimilar Sb4 and reference etanercept by assessing effectiveness in Crossfield, S. S. R., Buch, M. H., Baxter, P., Kingsbury, S. R., Pujades-Rodriguez, rheumatoid arthritis patients treated in ordinary clinical practice. Ann. Rheum. M., and Conaghan, P. G. (2021). Changes in the pharmacological management of Dis. 80 (1), 543.1–1543. doi:10.1136/annrheumdis-2021-eular.1604 rheumatoid arthritis over two decades. Rheumatology 60 (9), 4141–4151. doi:10. Hernandez, M. V., Sanchez-Piedra, C., Garcia-Magallon, B., Cuende, E., Manero, 1093/rheumatology/keaa892 J., Campos-Fernandez, C., et al. (2019). Factors associated with long-term retention Curtis, J. R., and Singh, J. A. (2011). Use of biologics in rheumatoid arthritis: of treatment with golimumab in a real-world setting: An analysis of the Spanish Current and emerging paradigms of Care. Clin. Ther. 33 (6), 679–707. doi:10.1016/j. BIOBADASER registry. Rheumatol. Int. 39 (3), 509–515. doi:10.1007/s00296-018- clinthera.2011.05.044 4177-z Frontiers in Pharmacology 11 frontiersin.org Carballo et al. 10.3389/fphar.2022.980832 Informació (2022). Informació, tràmits i serveis de la Generalitat de Catalunya. Tak, P. P., and Kalden, J. R. (2011). Advances in rheumatology: New targeted Available at: https://salutintegralbcn.gencat.cat/ca/Arees_Integrals_de_Salut/. therapeutics. Arthritis Res. Ther. 13 (Suppl. 1), S5–S14. doi:10.1186/1478-6354-13- S1-S5 Luttropp, K., Dozier, M., Justo, N., Cornillie, F., Kachroo, S., Govoni, M., et al. (2019). Real-world treatment persistence of golimumab in the management of Ugalde, A., Kiss, N., Livingston, P. M., and Rankin, N. (2021). Commentary on immune-mediated rheumatic diseases in europe: A systematic literature review. ‘exclusion rates in randomized trials of treatments for physical conditions: A BMJ Open 9 (5), e027456. doi:10.1136/bmjopen-2018-027456 systematic review. Trials 22 (1), 76–11. doi:10.1186/s13063-021-05019-9 Matucci-Cerinic, M., Allanore, Y., Kavanaugh, A., Buch, M. H., Schulze-Koops, Von Elm, E., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., H., Kucharz, E. J., et al. (2018). Efficacy, safety and immunogenicity of GP2015, an Vandenbrouckef, J. P., et al. (2007). The strengthening the reporting of etanercept biosimilar, compared with the reference etanercept in patients with observational studies in epidemiology (STROBE) statement: Guidelines for moderate-To-severe rheumatoid arthritis: 24-week results from the comparative reporting observational studies. Bull. World Health Organ. 85 (11), 867–872. phase III, randomised, double-blind EQUIRA study. RMD Open 4 (2), e000757–9. doi:10.2471/BLT.07.045120 doi:10.1136/rmdopen-2018-000757 von Richter, O., Skerjanec, A., Afonso, M., Sanguino Heinrich, S., Poetzl, J., Ministerio de Inclusión Seguridad Social y migraciones (2020). Estadísticas de Woehling, H., et al. (20152017). GP2015, a proposed etanercept biosimilar: residentes extranjeros en España. Madrid: Observatorio Permanente de la Pharmacokinetic similarity to its reference product and comparison of its Inmigración, 1–19. autoinjector device with prefilled syringes. Br. J. Clin. Pharmacol. 83 (4), 732–741. doi:10.1111/bcp.13170 Putrik, P., Ramiro, S., Kvien, T. K., Sokka, T., Pavlova, M., Uhlig, T., et al. (2014). Inequities in access to biologic and synthetic DMARDs across 46 European countries. Wells, G., Becker, J. C., Teng, J., Dougados, M., Schiff, M., Smolen, J., et al. (2009). Ann. Rheum. Dis. 73 (1), 198–206. doi:10.1136/annrheumdis-2012-202603 Validation of the 28-joint Disease Activity Score (DAS28) and European League against Rheumatism response criteria based on C-reactive protein against disease Silva-Fernández, L., Macía-Villa, C., Seoane-Mato, D., Cortés-Verdú, R., Romero- progression in patients with rheumatoid arthritis, and comparison with the Pérez, A., Quevedo-Vila, V., et al. (2020). The prevalence of rheumatoid arthritis in DAS28 based on erythrocyte sedimentation rate. Ann. Rheum. Dis. 68 (6), Spain. Sci. Rep. 10 (1), 21551–21559. doi:10.1038/s41598-020-76511-6 954–960. doi:10.1136/ard.2007.084459 Smolen, J. S., Aletaha, D., Barton, A., Burmester, G. R., Emery, P., Firestein, G. S., Wolfe, F. (1995). The epidemiology of drug treatment failure in rheumatoid et al. (2018). Rheumatoid arthritis. Nat. Rev. Dis. Prim. 4, 18001–18023. doi:10. arthritis. Baillieres Clin. Rheumatol. 9 (4), 619–632. doi:10.1016/S0950-3579(05) 1038/nrdp.2018.1 80305-X Smolen, J. S., Breedveld, F. C., Burmester, G. R., Bykerk, V., Dougados, M., Emery, Wu, E., Chen, L., Birnbaum, H., Yang, E., and Cifaldi, M. (2008). Retrospective P., et al. (2016). Treating rheumatoid arthritis to target: 2014 update of the claims data analysis of dosage adjustment patterns of TNF antagonists among recommendations of an international task force. Ann. Rheum. Dis. 75 (1), 3–15. patients with rheumatoid arthritis. Curr. Med. Res. Opin. 24 (8), 2229–2240. doi:10. doi:10.1136/annrheumdis-2015-207524 1185/03007990802229548 Smolen, J. S., Landewé, R. B. M., Bijlsma, J. W. J., Burmester, G. R., Dougados, M., Yazici, Y., Krasnokutsky, S., Barnes, J. P., Hines, P. L., Wang, J., and Rosenblatt, L. Kerschbaumer, A., et al. (2020). EULAR recommendations for the management of (2009). Changing patterns of tumor necrosis factor inhibitor use in 9074 patients with rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic rheumatoid arthritis. J. Rheumatol. 36 (5), 907–913. doi:10.3899/jrheum.080592 drugs: 2019 update. Ann. Rheum. Dis. 79 (6), S685–S699. doi:10.1136/ annrheumdis-2019-216655 Zink, A., Strangfeld, A., Schneider, M., Herzer, P., Hierse, F., Stoyanova-Scholz, Sruamsiri, R., Kameda, H., and Mahlich, J. (2018). Persistence with biological M., et al. (2006). Effectiveness of tumor necrosis factor inhibitors in rheumatoid disease-modifying antirheumatic drugs and its associated resource utilization arthritis in an observational cohort study: Comparison of patients according to their and costs. Drugs Real World Outcomes 5 (3), 169–179. doi:10.1007/s40801-018- eligibility for major randomized clinical trials. Arthritis Rheum. 54 (11), 3399–3407. 0139-8 doi:10.1002/art.22193 Frontiers in Pharmacology 12 frontiersin.org