Abstract
Objectives
Etanercept biosimilars show comparable efficacy to their originators among biologic-naïve patients with RA in randomized controlled trials. Nationwide guidelines have obligated prescribing of etanercept biosimilars from 2016, resulting in significant cost savings. This analysis aimed to compare the effectiveness of etanercept originator vs etanercept biosimilar amongst biologic-naïve RA patients treated in routine clinical practice in the UK.
Methods
Biologic-naïve RA patients starting etanercept in the British Society for Rheumatology Biologics Register in Rhematoid Arthritis (BSRBR-RA) cohort study from 2010 were included. Data collected at start of therapy includes patient demographics and disease activity. Follow-up data includes changes in disease activity and anti-rheumatic therapy. Six- and 12-month primary outcomes include DAS for 28-joints (DAS28) remission, EULAR response and minimal clinically important difference in function. Etanercept drug survival was assessed using Kaplan–Meier and Cox regression, including reasons for treatment withdrawal. Multiple imputation accounted for missing data. Propensity-decile adjustment was used to account for confounding by indication.
Results
A total of 1806 biologic-naïve RA patients started etanercept: 1009 originator, 797 biosimilar. At 6 and 12 months, the proportion of patients achieving DAS28 remission and EULAR response were similar between treatments. During follow-up, 19% of originator patients switched onto etanercept biosimilar. Patients were censored at time of switch. Patients on originator were no more likely to stop therapy vs biosimilar; 71% of originator and 76% of biosimilar patients remained on therapy at 1 year.
Conclusions
In one of the largest analyses of patients with RA, biologic-naïve RA patients treated with etanercept originator showed similar outcomes vs biosimilar using real-world data. Drug survival, and disease activity after 6 and 12 months of therapy, was similar between cohorts.
Keywords: RA, biologic therapy, originator, biosimilar, disease activity, outcomes, epidemiology
Graphical Abstract
Rheumatology key messages.
Etanercept biosimilars show similar outcomes to originator amongst biologic-naïve RA patients using real-world data.
Drug-survival and disease activity after 6 and 12 months of therapy was similar between therapies.
Data are reassuring to patients/clinical teams when considering starting etanercept (originator/biosimilar) as a first biologic.
Introduction
One of the first biologic therapies to be licenced for use in patients with RA was etanercept, a TNF inhibitor, in the early 2000s [1]. Since then, it has been one of the main treatment options to treat and successfully control disease activity in these patients [2, 3]. However, biologic therapies have high treatment costs with historical annual costs of etanercept in the UK estimated at £150 million [4]. In 2015, the patent for etanercept originator expired in Europe and consequently biosimilars were introduced [5]. Biosimilar therapies demonstrate equivalent clinical efficacy in phase III clinical trials [6–9], but due to the complicated manufacturing process, products cannot be identical [10]. From 2016, etanercept biosimilars have been available in the UK, and due to the reduced costs of these therapies, the National Health Service commissioning framework set guidelines for at least 90% of new patients to be prescribed the biosimilar product by 2019 [4].
Previous observational studies have shown outcomes in over 900 patients with RA defined with stable disease activity switching from etanercept originator to biosimilar [11]. However, real-world evidence comparing etanercept originator and etanercept biosimilar is limited. A Romanian study of 123 etanercept originator and 119 etanercept biosimilar patients showed no difference in DAS28 remission or EULAR response in a combined analysis of both biologic-naïve and biologic-experienced patients [12]. A recent analysis in the Portuguese cohort study Rheuma.pt identified similar proportions of patients achieving remission and a good EULAR response, as well as similar treatment survival between originator and biosimilars [13].
However, the UK has a different healthcare system and it is important to replicate analyses in larger datasets. The aim of this analysis was to compare outcomes after 6 and 12 months of etanercept treatment between biologic-naïve patients with RA starting etanercept originator and etanercept biosimilar as their first biologic, including disease activity and drug survival, treated in routine clinical practice in the UK.
Methods
Study setting
The British Society for Rheumatology Biologics Register for RA (BSRBR-RA) is an observational prospective cohort study established in 2001 with the aim of monitoring the long-term safety of biologic therapies nationally in routine clinical practice. Patients with RA were registered at the point they started biologic therapy, and data were collected on patient characteristics (age, gender), disease duration, disease activity using core outcome variables [swollen joint count (SJC), tender joint count (TJC), patient global visual analogue scale (100 mm) (PGA), ESR and/or CRP and the 28-joint DAS (DAS28)], functional ability using the HAQ, and current or previous anti-rheumatic therapies. Follow-up data was collected every 6 months for 3 years, and annually thereafter, including changes to disease activity and anti-rheumatic therapy. The BSRBR-RA methods have been previously published [14]. The register was approved by the UK North West Multicentre Research Ethics Committee in December 2000 (MREC 00/8/53), and all participants provided written informed consent upon registration.
Patient inclusion criteria
All patient were included in this analysis if they had RA and were starting etanercept therapy as their first biologic (i.e. biologic naïve) from 1 January 2010 until 16 June 2022. All patients had to have at least 1 year of follow-up data available.
Statistical analysis
Disease activity
Patient baseline characteristics (at the point of starting biologic therapy) were presented in patients starting etanercept originator and etanercept biosimilar. Disease activity, including core outcome variables, DAS28 and HAQ (function), were presented [median and interquartile range (IQR)] at baseline, 6 months and 1 year, including change in DAS28 and change in HAQ. Values were reported regardless of whether patients remained on therapy (intention to treat).
At 6 months and 1 year, the proportion of patients who achieved DAS28 remission (DAS28 <2.6) [15], EULAR response (good: 6-month DAS28 ≤3.2 and improvement >1.2 units; no response: 6-month DAS28 >5.1 and improvement ≤1.2 units, or 6-month DAS28 ≤5.1 and >3.2 with improvement ≤0.6 units; moderate: all others) [16], and minimal clinically important difference (MCID) in HAQ (reduction in HAQ score ≥0.22 units) [17] were calculated. As recommended by recent EULAR points to consider, patients who stopped biologic therapy prior to the DAS28 assessment were identified as ‘non-responders’ (DAS28 remission, no EULAR response) unless (i) they stopped for remission and where thus identified as ‘responders’ [18], or (ii) they were switched to another etanercept therapy and therefore the DAS28 value was used (to avoid bias in more patients switching from originator to biosimilar vs the other way around). The same alterations were applied to the MCID in HAQ.
Primary outcome measures were DAS28 remission, any EULAR response (moderate and good vs no response), good EULAR response (vs moderate/no response) and MCID in HAQ. Logistic regression was used to identify whether patients on etanercept originator were more likely to achieve the outcome after 6 months compared with patients on etanercept biosimilar. Models were adjusted by propensity deciles. Propensity scores were generated using baseline variables: age, gender, ethnicity (white vs non-white), disease duration, smoking status (past, current, never), BMI, number of comorbidities (0, 1, 2, 3+; based on hypertension, angina/myocardial infarction, stroke, epilepsy, asthma/chronic obstructive pulmonary disorder, peptic ulcer, liver disease, renal disease, tuberculosis, demyelination, diabetes, hyperthyroidism, depression, cancer), history of joint surgery history, anti-CCP positivity, RF positivity, concomitant MTX use, concomitant steroid use, concomitant conventional-synthetic DMARD (csDMARD) use, DAS28 score, SJC, TJC, CRP, ESR, PGA and HAQ score.
Multiple imputation using chained equations (20 datasets) was used to account for missing data. Imputed variables included ethnicity (white, non-white), joint surgery history (yes or no), anti-CCP positivity, RF positivity, smoking status (past, current, never), disease duration, body mass index, as well as baseline, 6-month and 1-year SJC, TJC, CRP, ESR, PGA, DAS28 score and HAQ score. Complete variables included in the imputation model were age, gender, comorbidities, csDMARDs, steroids, MTX and registration year (<2016, ≥2016). In addition, change in DAS28, DAS28 remission, change in HAQ, MCID in HAQ and EULAR response were all included as passive variables and calculated from imputed values. As mentioned above, patients who stopped biologic therapy prior to the DAS28 assessment were identified as non-responders (no DAS28 remission, no EULAR response) unless they stopped for remission (responders) or switched to a another etanercept therapy (used DAS28).
Drug survival
All patients were followed until their last completed follow-up form within the study. To assess drug survival, patients entered the analysis at the start of etanercept, and were censored if they stopped therapy, or at their date of final follow-up form, date of death or cut-off date or this analysis (if they remained on therapy), whichever was first. The number of patients who discontinued therapy during follow-up was calculated, including time to discontinuation and reason for treatment discontinuation [physician reported: ineffectiveness, adverse event (including death), remission, switch to biosimilar (non-medical switch), other or missing]. Kaplan–Meier survival curves were plotted for time on etanercept therapy between cohorts, adjusted by propensity deciles. Cox proportional hazards model was used to assess whether more patients on etanercept originator stopped therapy over time compared with those on biosimilar, adjusted by propensity deciles.
As part of a sensitivity analysis, this analysis was repeated to censor patients stopping etanercept therapy to switch to a biosimilar, rather than identifying these as a ‘stop’. The reason for this was to avoid bias between the cohorts as a large proportion of patients on the originator switched to a biosimilar, whereas there were no patients in the biosimilar cohort who switched to another etanercept product.
All analyses were completed in Stata version 14 [19]. Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this analysis.
Results
There were 1806 patients registered in the BSRBR-RA starting etanercept for their first biologic with at least 1 year of follow-up data available; 1009 starting the originator, and 797 starting the biosimilar (783 Benepali and 14 Erelzi; see Fig. 1 and Supplementary Table S1, available at Rheumatology online). Patient characteristics were similar with 74% female, median age at start of therapy of 59 years and median disease duration before starting etanercept therapy of 5 years (Table 1). The majority of patients (94%) were white, with only 19% currently smoking, 60% on concomitant MTX and 24% on concomitant steroids. The main difference was that the majority of patients starting etanercept originator were registered before 2016 (75%), whereas all biosimilar patients were registered from 2016 onwards. Patient characteristics of those starting etanercept before and after 2016 are broadly similar (see Supplementary Table S2, available at Rheumatology online).
Figure 1.
Patient inclusion into analysis
Table 1.
Baseline characteristics of patients starting etanercept originator or biosimilar as their first biologic therapy (unimputed)
| Baseline characteristics | All patients | Etanercept originator | Etanercept biosimilar |
|---|---|---|---|
| Total, n | 1806 | 1009 | 797 |
| Female, n (%) | 1332 (74) | 734 (73) | 598 (75) |
| Age, years | 59 (51, 67) | 59 (51, 68) | 59 (50, 67) |
| BMI, kg/m2 | 28 (24, 32); N = 1353 | 28 (24, 32); N = 752 | 27 (24, 32); N = 601 |
| Smoking status, n (%) | N = 1650 | N = 927 | N = 723 |
| Current smoker | 309 (19) | 177 (19) | 132 (18) |
| Ex-smoker | 669 (41) | 380 (41) | 289 (40) |
| Never-smoked | 672 (41) | 370 (40) | 302 (42) |
| Ethnicity, n (%) | N = 1282 | N = 773 | N = 509 |
| White | 1211 (94) | 732 (95) | 479 (94) |
| Other | 71 (6) | 41 (5) | 30 (6) |
| Number of previous csDMARDs, n (%) | N = 1795 | N = 1004 | N = 791 |
| 1 | 57 (3) | 29 (3) | 28 (4) |
| 2 | 666 (37) | 366 (36) | 300 (38) |
| ≥3 | 1072 (60) | 609 (61) | 463 (59) |
| Any concurrent csDMARD, n (%) | 1423 (79) | 782 (78) | 641 (80) |
| Current MTX, n (%) | 1077 (60) | 601 (60) | 476 (59) |
| Current steroid treatment, n (%) | 428 (24) | 240 (24) | 188 (24) |
| Comorbiditiesa, n (%) | |||
| None | 693 (38) | 357 (35) | 336 (42) |
| 1 | 567 (31) | 325 (32) | 242 (30) |
| 2 | 354 (20) | 211 (21) | 143 (18) |
| ≥3 | 192 (11) | 116 (12) | 76 (10) |
| Disease duration until registration date, years | 5 (2, 11); N = 1753 | 5 (2, 12); N = 984 | 5 (2, 10); N = 769 |
| Treatment within 2-years of diagnosis, n (%) | 513 (29); N = 1753 | 279 (28); N = 984 | 234 (30); N = 769 |
| DAS28 score | 5.8 (5.3, 6.5); N = 1801 | 6.0 (5.4, 6.7); N = 1007 | 5.6 (5.2, 6.3); N = 794 |
| 28 TJC | 13 (8, 19); N = 1691 | 14 (9, 20); N = 964 | 11 (8, 17); N = 727 |
| 28 SJC | 8 (5, 11); N = 1692 | 8 (5, 12); N = 963 | 7 (4, 10); N = 729 |
| CRP, mg/L | 12 (5, 28); N = 1038 | 12 (5, 30); N = 562 | 11 (5, 26); N = 476 |
| ESR, mm/h | 25 (11, 42); N = 1190 | 26 (12, 44); N = 743 | 24 (9, 38); N = 447 |
| PGA (0–100 mm) | 80 (64, 90); N = 1622 | 80 (65, 90); N = 932 | 80 (60, 90); N = 690 |
| HAQ score (0–3) | 1.6 (1.0, 2.1); N = 1420 | 1.6 (1.0, 2.1); N = 811 | 1.5 (0.8, 2.0); N = 609 |
| Anti-CCP positivity, n (%) | 709 (72); N = 986 | 268 (70); N = 381 | 441 (73); N = 605 |
| RF positivity, n (%) | 948 (64); N = 1482 | 546 (62); N = 874 | 402 (66); N = 608 |
| Joint surgery history, n (%) | 306 (19); N = 1619 | 202 (20); N = 989 | 104 (17); N = 630 |
| Year of registration, n (%) | |||
| 2011 | 2 (<1) | 2 (<1) | 0 |
| 2012 | 139 (8) | 139 (14) | 0 |
| 2013 | 234 (13) | 234 (23) | 0 |
| 2014 | 217 (12) | 217 (22) | 0 |
| 2015 | 164 (9) | 164 (16) | 0 |
| 2016 | 294 (16) | 181 (18) | 113 (14) |
| 2017 | 523 (29) | 69 (7) | 454 (57) |
| 2018 | 217 (12) | 3 (<1) | 214 (27) |
| 2019 | 16 (<1) | 0 | 16 (2) |
| Follow-up in BSRBR-RA since registration, years | 4.0 (2.5, 5.1) | 5.0 (3.4, 6.2) | 2.7 (1.8, 4.0) |
All continuous variables are reported as median (interquartile ranges), unless stated otherwise.
Comorbidities include hypertension, angina/myocardial infarction, stroke, epilepsy, asthma/chronic obstructive pulmonary disorder, peptic ulcer, liver disease, renal disease, tuberculosis, demyelination, diabetes, hyperthyroidism, depression and cancer. csDMARD: conventional synthetic DMARD; DAS28: DAS of 28 joints; TJC: tender joint count; SJC: swollen joint count; PGA: patient’s global assessment; anti-CCP: anti-cyclic citrullinated peptide; BSRBR-RA: British Society for Rheumatology Biologics Register for RA.
Both cohorts had improved disease activity; median improvement of DAS28 from baseline to 6 months was 2.4, and from baseline to 1 year it was improved by a median of 2.5 units (Table 2). At 6 months, DAS28 remission was achieved by 26% of the originator cohort, and 31% of the biosimilar cohort, although there was no statistical difference when adjusted for patient characteristics [odds ratio (OR) 0.88, 95% CI 0.70, 1.11]. At 6 months, a good EULAR response was achieved by 38% of the originator cohort and 45% of the biosimilar cohort, with the odds of achieving a good EULAR response 20% higher for biosimilar cohort compared with the originator cohort [adjusted OR for originator cohort 0.80 (95% CI 0.66, 0.98)]. However, this association did not remain at 1 year [OR 0.86 (95% CI 0.69, 1.07)]. At 6 months, MCID in HAQ was achieved in 41% of the originator cohort and 47% of the biosimilar cohort, with patients on the originator less likely to achieve response [propensity adjusted OR 0.74 (95% CI 0.57, 0.97)]. However, after 1 year, fewer patients achieved MCID in HAQ from baseline, with no significant difference between cohorts (35% in both cohorts).
Table 2.
Six-month and 1-year outcomes of patients on etanercept originator or biosimilar as their first-biologic (imputed)
| All patients | Etanercept originator | Etanercept biosimilar | |
|---|---|---|---|
| Total, n | 1806 | 1009 | 797 |
| Disease activity | |||
| Baseline DAS28 score | 5.8 (5.3, 6.5) | 6.0 (5.4, 6.7) | 5.6 (5.2, 6.3) |
| Six months | 3.4 (2.4, 4.6) | 3.5 (2.5, 4.7) | 3.2 (2.2, 4.4) |
| ΔDAS28 from baseline | –2.4 (–3.5, –1.2) | –2.4 (–3.5, –1.2) | –2.4 (–3.5, –1.2) |
| One year | 3.3 (2.2, 4.5) | 3.4 (2.3, 4.5) | 3.1 (2.1, 4.5) |
| ΔDAS28 from baseline | –2.5 (–3.7, –1.2) | –2.5 (–3.7, –1.3) | –2.5 (–3.6, –1.1) |
| DAS28 remission (DAS28 <2.6)b | |||
| Six month (%) | 29 | 26 | 31 |
| Unadjusted (OR) | 0.78 (0.63, 0.97)* | [ref] | |
| Propensity decile adjusteda (OR) | 0.88 (0.70, 1.11) | [ref] | |
| One year (%) | 30 | 27 | 34 |
| Unadjusted (OR) | 0.75 (0.60, 0.93)* | [ref] | |
| Propensity decile adjusteda (OR) | 0.83 (0.66, 1.05) | [ref] | |
| EULAR responseb | |||
| Six month (%) | |||
| No response | 27 | 29 | 25 |
| Moderate response | 32 | 34 | 31 |
| Good response | 41 | 38 | 45 |
| Any response (moderate/good vs no response) | |||
| Unadjusted (OR) | 0.81 (0.65, 1.01) | [ref] | |
| Propensity decile adjusteda (OR) | 0.80 (0.63, 1.00) | [ref] | |
| Good EULAR response (vs moderate/no response) | |||
| Unadjusted (OR) | 0.74 (0.61, 0.91)* | [ref] | |
| Propensity decile adjusteda (OR) | 0.80 (0.66, 0.98)* | [ref] | |
| One year (%) | |||
| No response | 35 | 36 | 34 |
| Moderate response | 24 | 26 | 22 |
| Good response | 41 | 39 | 44 |
| Any response (moderate/good vs no response) | |||
| Unadjusted (OR) | 0.93 (0.76, 1.15) | [ref] | |
| Propensity decile adjusteda (OR) | 0.94 (0.76, 1.17) | [ref] | |
| Good EULAR response (vs moderate/no response) | |||
| Unadjusted (OR) | 0.79 (0.64, 0.97)* | [ref] | |
| Propensity decile adjusteda (OR) | 0.86 (0.69, 1.07) | [ref] | |
| Physical function outcomes | |||
| Baseline HAQ (0–3) | 1.6 (1.0, 2.1) | 1.6 (1.0, 2.1) | 1.5 (0.8, 2.0) |
| Six months | 1.4 (0.6, 1.9) | 1.5 (0.7, 2.0) | 1.2 (0.4, 1.8) |
| ΔHAQ from baseline | –0.16 (–0.50, 0.13) | –0.13 (–0.50, 0.13) | –0.23 (–0.54, 0.06) |
| MCID in HAQb (reduction in HAQ score ≥0.22 units) (%) | 44 | 41 | 47 |
| Unadjusted (OR) | 0.79 (0.62, 1.02) | [base] | |
| Propensity decile adjusteda (OR) | 0.74 (0.57, 0.97)* | [base] | |
| One year | 1.4 (0.6, 2.0) | 1.5 (0.7, 2.0) | 1.3 (0.5, 2.0) |
| ΔHAQ from baseline | –0.13 (–0.50, 0.18) | –0.13 (–0.50, 0.14) | –0.12 (–0.49, 0.21) |
| MCID in HAQb (reduction in HAQ score ≥0.22 units) (%) | 35 | 35 | 35 |
| Unadjusted (OR) | 1.00 (0.76, 1.31) | [base] | |
| Propensity decile adjustedb (OR) | 0.97 (0.73,1.28) | [base] |
All continuous variables are reported as median (interquartile range), unless stated otherwise.
Adjusted by propensity deciles—baseline (start of therapy) variables in the propensity score included: age, gender, ethnicity, disease duration, smoking status, BMI, number of comorbidities, joint surgery history, anti-CCP positivity, RF positivity, concomitant MTX use, concomitant steroid use, concomitant conventional synthetic DMARD use, DAS28 score, swollen joint count, tender joint count, CRP, ESR, patient’s global assessment and HAQ score.
Outcomes were updated if patients had stopped therapy prior to assessment (‘non-responder’), unless it was for remission (‘responder’) or to switch to an etanercept therapy (value was used).
P <0.05. DAS28: DAS of 28-joints; OR: odds ratio; MCID: minimal clinically important difference.
During the course of study, the etanercept originator cohort had a median of 5.0 years of follow-up (IQR 3.4, 6.2), while the biosimilar cohort had a median of 2.7 years (IQR 1.8, 4.0) (Table 3). During this time, 81% of originator patients had stopped therapy after a median of 18 months (IQR 7, 41) on treatment (Fig. 2a); 19% were switched to a biosimilar product, 21% stopped for ineffectiveness and 22% for an adverse event. Of the biosimilar patients, 41% had stopped therapy after a median of 10 months (IQR 5, 18); 20% stopped for ineffectiveness and 14% stopped for an adverse event. In the adjusted Cox proportional hazards model, etanercept originator patients were 62% more likely to stop etanercept therapy compared with biosimilar patients [hazard ratio (HR) 1.62 (95% CI 1.41, 1.85)]. When originator patients were censored when they switched to a biosimilar therapy (Fig. 2b), 53% discontinued etanercept therapy after a median of 11 months (5, 26). In the propensity decile–adjusted Cox proportional hazards model, etanercept originator patients were no more likely to stop etanercept therapy compared with biosimilar patients [HR 1.15 (95% CI 0.99, 1.33)].
Table 3.
Frequency and reasons for discontinuing etanercept therapy, including the sensitivity analysis results
| Etanercept originator | Sensitivity analysisa | Etanercept biosimilar | |
|---|---|---|---|
| Total | 1009 | 1009 | 797 |
| Follow-up in BSRBR-RA since registration, years | 5.0 (3.4–6.2) | 5.0 (3.4–6.2) | 2.7 (1.8–4.0) |
| One-year Kaplan–Meier drug survival (%) (95% CI) | 68 (65, 71) | 71 (68, 74) | 76 (73, 79) |
| Discontinued treatment, n (%) | 819 (81) | 530 (53) | 330 (41) |
| Time to discontinuation, months | 18 (7–41) | 11 (5–26) | 10 (5–18) |
| Reasons for treatment discontinuation, n (% of whole cohort) | |||
| Ineffectiveness | 208 (21) | 208 (21) | 158 (20) |
| Adverse events including deaths | 219 (22) | 219 (22) | 108 (14) |
| Remission | 3 (<1) | 3 (<1) | 6 (1) |
| Switched to a biosimilar | 189 (19) | – | – |
| Otherb | 65 (6) | 65 (6) | 39 (5) |
| Missing | 35 (3) | 35 (3) | 19 (2) |
| Discontinuation | |||
| Unadjusted [HR (95% CI)] | 1.69 (1.48, 1.92)* | 1.21 (1.05, 1.39)* | [base] |
| Propensity decile adjustedc [HR (95% CI)] | 1.62 (1.41, 1.85)* | 1.15 (0.99, 1.33) | [base] |
All values are n (%), or median (interquartile range), unless stated otherwise.
Sensitivity analysis for the etanercept originator cohort to censor patients who switched to a biosimilar.
Other reasons for treatment discontinuations were unique to the patient.
Adjusted by propensity deciles—baseline variables in the propensity score included age, gender, ethnicity, disease duration, smoking status, BMI, number of comorbidities, joint surgery history, anti-CCP positivity, RF positivity, baseline MTX use, baseline steroid use, baseline conventional synthetic DMARD use, number of previous conventional synthetic DMARDs, DAS of 28-joints score, swollen joint count, tender joint count, CRP, ESR, patient’s global assessment and HAQ score.
P < 0.05. BSRBR-RA: British Society for Rheumatology Biologics Register for RA; HR: hazard ratio.
Figure 2.
Kaplan–Meier survival curves. Time on originator (dashed) and biosimilar (solid) [imputed data, propensity deciles adjusted]. (a) etanercept drug survival, including drug discontinuation for originator patients switching to a biosimilar, and (b) Sensitivity analysis censoring etanercept originator patients who were switched to a biosimilar
Discussion
This large analysis included over 1800 patients with RA starting etanercept as their first biologic from 2010 in a national cohort study, comparing outcomes in patients on originator vs biosimilar. After 1 year of therapy, disease activity was similar between cohorts, with over one in five achieving DAS28 remission, over two in five achieving a good EULAR response and one-third achieving a minimally clinical important difference in functional ability. However, one in five of the originator patient subsequently switched to an etanercept biosimilar, likely due to national guidelines for non-medical switching. Taking this switch into account (i.e. censoring patients rather than assuming they have stopped originator therapy) etanercept originator patients were no more likely to stop etanercept therapy compared with biosimilar patients, with 71% on originator after 1 year vs 76% on biosimilar.
Data from a randomized control trial comparing first-line treatment with etanercept originator vs biosimilar in patients with RA has shown similar efficacy, with no difference in the adjusted difference rate of ACR 20, 50 or 70 response between treatments after 52 weeks [6, 7]. Improvement in DAS28 after 24 weeks (5.5 months) was similar to the current study for 6-month outcomes, with no differences found between treatments (mean 2.5 and 2.6 in originator and biosimilar, compared with median 2.4 in both in the current analysis). However, the proportion of patients achieving a good EULAR response or DAS28 remission at 6 months was higher in the current analysis; 38% and 45% good EULAR response, and 29% and 26% DAS28 remission in originator and biosimilar patients, respectively, compared with 30% and 32% good EULAR response, and 16% and 17% DAS28 remission in the trial. This could be explained by the baseline DAS28 being lower in the current study (median 6.0 and 5.6 in originator and biosimilar respectively) compared with the randomized controlled trial (mean 6.5), meaning those with a higher starting DAS28 would need to achieve greater numerical reduction to meet the response criteria. In addition, the current analysis used a more strict definition of DAS28 remission (DAS28 <2.6) as published by Fransen et al. [15] compared with the randomized controlled trial (DAS ≤2.6). However, when the trial was extended to 52 weeks, the same authors, using the same definition as the current analysis (DAS <2.6) identified a higher proportion achieving remission; 35% and 42% in originator and biosimilar, respectively, compared with 30% and 27% at 12 months in the current analysis. In addition to differences in initial DAS28 between the trial and this real-world cohort study, patients in the trial were younger (mean 52 years vs median 59 years), more were female (86% vs 74%), with longer disease duration (mean 6.1 vs median 5), although the populations were similar with respect to ethnicity (93% white in trial vs 94% in current analysis) and BMI (mean 27 vs median 28). Approximately two-third of population in the current analysis had prior exposure to at least three csDMARDs, whereas the trial only considered patients who had been on MTX therapy for at least 6 months.
Cohort studies have been able to study effectiveness of these therapies in the real world. The Romanian Registry of Rheumatic Disease (RRBR) collects data on all RA patients treated with biologic therapies [12]. In a cohort of 123 etanercept originator and 119 biosimilar patients there was no difference in outcomes after 6 months: 19% (originator) and 18% (biosimilar) achieved DAS28 remission, and 31% (originator) and 32% (biosimilar) achieved a good EULAR response. However, this study included patients starting etanercept as their first biologic, as well as patients who were biologic-experienced, which may explain why proportions were lower than seen in the current analysis. In a larger cohort study in Portugal (reuma.pt) of patients with RA starting etanercept as their first biologic, disease activity outcomes at 6 months were similar between the 645 originator and 219 biosimilar patients. A good EULAR response was achieved in 46% of patients starting etanercept originator and 38% of patients starting biosimilar (P = 0.07). However, this difference was assessed using a simple chi-squared test, rather than the propensity score adjustment used in the current analysis. The Portuguese study also reported a high amount of missing data; EULAR response data were available for 47% of the originator patients (N = 301) and 48% of biosimilar patients (N = 105). Drug survival was also assessed, with ∼80% remaining on etanercept therapy after 1 year and no difference between treatments. After 36 months, similar proportions of patients remained on therapy (73% originator and 64% biosimilar), although 6% of originator patients had switched to the biosimilar during follow-up. This is much lower than the 35% of originator patients switching to an etanercept biosimilar in the current analysis.
This analysis was completed in a large national prospective cohort study of patients starting biologic therapy for RA; registration into the study is recommended by the British Society for Rheumatology, but is not mandatory. However, there are limitations with real-world observational studies. Due to the approval of the originator in the early 2000s being much earlier than the approval of biosimilars in 2016, there is much less potential follow-up available in the biosimilar cohort (5.0 vs 2.7 years), which may lead to bias [20]. However, the Cox proportional hazard model censors patients at final follow-up and was adjusted for differences in patient characteristics that should counter this bias. RA patients from the early 2000s may also be fundamentally different compared with patients starting biologics more recently. With this in mind, only patients starting etanercept from 2010 onwards were included in the analysis. In addition, outcomes were adjusted for patient characteristics as part of a propensity score.
In conclusion, this is one of the largest analyses of over 1800 patients with RA starting etanercept as their first biologic from 2010 in a national cohort study. It found patients starting etanercept originator achieved similar treatment response with respect to disease activity and drug survival compared with patients starting etanercept biosimilar. More patients on etanercept biosimilar achieved a MCID in functional ability after 6 months of treatment, although this difference did not persist to 1 year. After consideration of originator patients switching to biosimilar therapy, patients starting etanercept originator were no more likely to stop etanercept therapy compared with biosimilar patients, with 71% on originator after 1 year vs 76% on biosimilar. These data are reassuring to both patients and clinical teams when considering starting etanercept originator or biosimilar therapy as their first biologic.
Supplementary Material
Acknowledgements
The authors acknowledge the enthusiastic collaboration of all consultant rheumatologists and their specialist nurses in the UK in providing the data (visit www.bsrbr.org for a full list of contributors). The authors would like to gratefully acknowledge the support of the National Institute for Health Research (NIHR), through the Comprehensive Local Research Networks at participating centres. In addition, the authors acknowledge support from the BSR Executive, the members of the BSRBR Registers Committee and the BSRBR Project Team in London for their active role in enabling the register to undertake its tasks. The authors also acknowledge the seminal role of the BSR Clinical Affairs Committee for establishing national biological guidelines and recommendations for such a register. Finally, the authors would like to acknowledge the Centre for Epidemiology Versus Arthritis (Arthritis Research UK Grant No. 21755), who provided the infrastructure support for the study, and supported by the NIHR Manchester Biomedical Research Centre (NIHR203308). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. Access to the underlying identifiable and potentially re-identifiable pseudonymized electronic health record data is tightly governed by various legislative and regulatory frameworks, and restricted by best practice. The manuscript was presented in part at the 2022 British Society for Rheumatology conference 25–27 April 2022.
Contributor Information
Lianne Kearsley-Fleet, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
Aasiyah Rokad, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
Man-Fung Tsoi, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
Sizheng Steven Zhao, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
Mark Lunt, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
Kath D Watson, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
BSRBR-RA Contributors Group, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.
Kimme L Hyrich, Centre for Epidemiology Versus Arthritis, Division of Musculoskeletal and Dermatological Sciences, School of Biological Sciences, Faculty of Biology Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK; NIHR Manchester Biomedical Research Centre, Manchester University NHS Foundation Trust, Manchester, UK.
Supplementary material
Supplementary material is available at Rheumatology online.
Data availability
The data that support the findings of this study are available from the British Society for Rheumatology. Restrictions apply to the availability of these data (see: https://www.rheumatology.org.uk/practice-quality/registers/requesting-registers-data).
Contribution statement
All the authors have provided substantial contributions to the conception or design of the work, the acquisition of the data and the interpretation of data. L.K-F. and A.R. performed the statistical analysis, and M-F.T., S.S.Z., M.L., K.W. and K.L.H. all contributed to the analysis interpretation. L.K-F. and A.R. wrote the first draft. All the other authors participated in the final drafting of the work or revising it critically for important intellectual content. All authors contributed to the final approval of the version published.
Funding
This work was supported by the British Society for Rheumatology (BSR). The BSR commissioned the BSR Biologics Register in Rheumatoid Arthritis (BSRBR-RA) as a UK-wide national project to investigate the safety of biologic and other targeted therapies in routine medical practice. K.L.H. is the chief investigator. BSR currently receives restricted income from UK pharmaceutical companies, including Abbvie, Amgen, Celltrion Healthcare, Eli Lilly, Galapagos, Pfizer, Samsung Bioepis and Sanofi, and in the past Hospira, MSD, Roche, Sandoz, SOBI and UCB. This income finances a wholly separate contract between the BSR and the University of Manchester. The chief investigator and the BSRBR-RA team at the University of Manchester have full academic freedom and are able to work independently of pharmaceutical industry influence. The funding source did not play any role in study design, collection, analysis and interpretation of data, or writing of the manuscript. Members of the BSRBR-RA University of Manchester team, BSR trustees, committee members and staff complete an annual declaration in relation to conflicts of interest. K.L.H. is supported by the National Institute for Health Research (NIHR) Manchester Biomedical Research Centre (NIHR203308). S.S.Z. is supported by an NIHR Clinical Lectureship.
Disclosure statement: K.L.H. reports honoraria from Abbvie and grants from Pfizer and BMS outside the submitted work, and is supported by the NIHR Manchester Biomedical Research Centre (NIHR203308). S.S.Z. reports travel support and consultancy fees from UCB outside the submitted work. All other authors report no conflicts of interest.
References
- 1. EMA. Summary of Product Characteristics - Enbrel. 2010. https://www.ema.europa.eu/en/documents/product-information/enbrel-epar-product-information_en.pdf (5 January 2023, date last accessed).
- 2. Kearsley-Fleet L, Davies R, De Cock D. et al. ; BSRBR-RA Contributors Group. Biologic refractory disease in rheumatoid arthritis: results from the British Society for Rheumatology Biologics Register for Rheumatoid Arthritis. Ann Rheum Dis 2018;77:1405–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Feist E, Baraliakos X, Behrens F. et al. Effectiveness of etanercept in rheumatoid arthritis: real-world data from the German non-interventional study ADEQUATE with focus on treat-to-target and patient-reported outcomes. Rheumatol Ther 2022;9:621–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. NHS England. Commissioning framework for biological medicines (including biosimilar medicines). 2017. https://www.england.nhs.uk/wp-content/uploads/2017/09/biosimilar-medicines-commissioning-framework.pdf (5 January 2023, date last accessed).
- 5. Azevedo VF, Galli N, Kleinfelder A, D’Ippolito J, Urbano PCM.. Etanercept biosimilars. Rheumatol Int 2015;35:197–209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Emery P, Vencovský J, Sylwestrzak A. et al. A phase III randomised, double-blind, parallel-group study comparing SB4 with etanercept reference product in patients with active rheumatoid arthritis despite methotrexate therapy. Ann Rheum Dis 2017;76:51–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Emery P, Vencovský J, Sylwestrzak A. et al. 52-week results of the phase 3 randomized study comparing SB4 with reference etanercept in patients with active rheumatoid arthritis. Rheumatology (Oxford) 2017;56:2093–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Emery P, Vencovský J, Sylwestrzak A. et al. Long-term efficacy and safety in patients with rheumatoid arthritis continuing on SB4 or switching from reference etanercept to SB4. Ann Rheum Dis 2017;76:1986–991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Girolomoni G, Feldman SR, Emery P. et al. Comparison of injection-site reactions between the etanercept biosimilar SB4 and the reference etanercept in patients with rheumatoid arthritis from a phase III study. Br J Dermatol 2018;178:e215–6. [DOI] [PubMed] [Google Scholar]
- 10. Vulto AG, Jaquez OA.. The process defines the product: what really matters in biosimilar design and production? Rheumatology (Oxford) 2017;56(suppl_4):iv14–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Glintborg B, Loft AG, Omerovic E. et al. To switch or not to switch: results of a nationwide guideline of mandatory switching from originator to biosimilar etanercept. One-year treatment outcomes in 2061 patients with inflammatory arthritis from the DANBIO registry. Ann Rheum Dis 2019;78:192–200. [DOI] [PubMed] [Google Scholar]
- 12. Codreanu C, Popescu CC, Mogoșan C. et al. Efficacy and safety of original and biosimilar etanercept (SB4) in active rheumatoid arthritis – A comparison in a real-world national cohort: original and biosimilar etanercept in RA. Biologicals 2019;62:27–32. [DOI] [PubMed] [Google Scholar]
- 13. Pinto AS, Cunha MM, Pinheiro F. et al. Effectiveness and safety of original and biosimilar etanercept (Enbrel® vs Benepali®) in bDMARD-naïve patients in a real-world cohort of Portugal. ARP Rheumatol 2022;1:109–16. [PubMed] [Google Scholar]
- 14. Hyrich KL, Watson KD, Isenberg DA, Symmons DP; BSRBR-RA. The British Society for Rheumatology Biologics Register: 6 years on. Rheumatology (Oxford) 2008;47:1441–3. [DOI] [PubMed] [Google Scholar]
- 15. Fransen J, Creemers MC, Van Riel PL.. Remission in rheumatoid arthritis: agreement of the disease activity score (DAS28) with the ARA preliminary remission criteria. Rheumatology (Oxford) 2004;43:1252–5. [DOI] [PubMed] [Google Scholar]
- 16. Fransen J, Van Riel PLCM.. The Disease Activity Score and the EULAR response criteria. Clin Exp Rheumatol 2005;23(5 Suppl 39):S93–9. [PubMed] [Google Scholar]
- 17. Goldsmith CH, Boers M, Bombardier C, Tugwell P.. Criteria for clinically important changes in outcomes: development, scoring and evaluation of rheumatoid arthritis patient and trial profiles. OMERACT Committee. J Rheumatol 1993;20:561–5. [PubMed] [Google Scholar]
- 18. Courvoisier DS, Lauper K, Kedra J. et al. EULAR points to consider when analysing and reporting comparative effectiveness research using observational data in rheumatology. Ann Rheum Dis 2022;81:780–5. [DOI] [PubMed] [Google Scholar]
- 19. StataCorp. Stata Statistical Software: release 14. College Station, TX: StataCorp LP, 2015. [Google Scholar]
- 20. Clark TG, Bradburn MJ, Love SB, Altman DG.. Survival Analysis Part I: basic concepts and first analyses. Br J Cancer 2003;89:232–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the British Society for Rheumatology. Restrictions apply to the availability of these data (see: https://www.rheumatology.org.uk/practice-quality/registers/requesting-registers-data).