Radiographic and clinical effects of 10 mg and 25 mg twice-weekly etanercept over 52 weeks in Japanese patients with active rheumatoid arthritis

Abstract

Objectives

To compare the radiographic and clinical effects of 25 versus 10 mg twice-weekly (BIW) etanercept over 52 weeks in Japanese patients with active rheumatoid arthritis (RA).

Methods

This was a post-hoc analysis of a Phase 3 study where Japanese patients with active RA were randomized to receive BIW etanercept 25 mg (n = 182), etanercept 10 mg (n = 192), or methotrexate (n = 176) for 52 weeks (NCT00445770). This analysis included assessments of week-24 and week-52 disease activity, week-52 radiographic progression, and the relationship between baseline characteristics and week 52 clinical outcomes with clinically relevant radiographic progression (CRRP) at week 52.

Results

At week 52, there were no significant differences between 25 and 10 mg etanercept in terms of achieving low disease activity or remission. CRRP was observed in 36% and 32% of patients in the 10 and 25 mg groups, respectively. Predictor analysis suggested that worse background disease status, treatment with methotrexate rather than etanercept, and poorer clinical outcomes at week 52 were associated with CRRP.

Conclusions

The 25 mg BIW etanercept dosage does not appear to be significantly more efficacious than 10 mg in Japanese patients with RA. Further studies evaluating the optimal etanercept dosing regimen in this patient population may be merited.

NCT: NCT00445770

Keywords:

• Etanercept
• rheumatoid arthritis
• Japan
• clinical trial
• methotrexate

Introduction

Etanercept (ETN) is approved in Japan for the treatment of rheumatoid arthritis (RA) [1]. Its efficacy has been demonstrated in several clinical trials and observational studies in Japanese patients [2–7]. The JERA study (Etanercept Early Rheumatoid arthritis study for radiographic analyses in Japan) was a Pharmaceuticals and Medical Devices Agency (PMDA)-requested Phase 3 clinical trial to evaluate the radiographic efficacy of 20 mg/week and 50 mg/week ETN versus methotrexate (MTX) in Japanese patients, and was carried out in individuals with active RA and with inadequate response to ≥1 conventional disease-modifying anti-rheumatic drug (DMARD). Patients were randomized to receive 10 mg twice weekly (BIW) ETN monotherapy, 25 mg BIW ETN monotherapy, or ≤8.0 mg once-weekly MTX monotherapy for 52 weeks [8].

In the JERA trial, statistically significant advantages of the higher versus the lower ETN dose were observed for several secondary and additional outcomes: (1) change in joint narrowing space score from baseline to week 24; (2) the proportion of patients achieving a van der Heijde-modified Total Sharp Score (mTSS) change ≤ the smallest detectable difference (SDD) at week 52; (3) the proportion of patients achieving a good or moderate European League Against Rheumatism (EULAR) response at week 52 and; (4) mean score and percentage change in tender joint count (TJC) and physician global assessment (PGA) from baseline to week 52. This suggests that, at least in part, 25 mg BIW ETN may be more efficacious than 10 mg BIW ETN for the treatment of RA in Japanese patients, and warrants further analysis. The goal of this exploratory post-hoc analysis of the JERA trial was to further investigate potential differences between 10 mg and 25 mg ETN on radiographic and clinical outcomes, as well as to identify baseline characteristics predictive of clinically relevant radiographic progression (CRRP).

Methods

Study design

This was an exploratory post-hoc analysis from the Phase 3, double-dummy JERA study of ETN in Japanese patients with active RA (NCT00445770). Patients in this study were randomized to one of three active treatment groups: subcutaneous ETN 25 mg BIW plus oral MTX placebo, subcutaneous ETN 10 mg BIW plus oral MTX placebo, or oral MTX ≤8.0 mg/week plus subcutaneous ETN placebo. Details of the study design, patient population, and study endpoints and assessments have been published previously [8]. It is important to note that the original study was not powered to detect significant differences between 25 mg and 10 mg BIW etanercept. This study was conducted in accordance with the principles of the Declaration of Helsinki and the International Conference on Harmonisation Good Clinical Practice guidelines, while following all local regulatory requirements.

Efficacy endpoints and assessments

Clinical efficacy was assessed in two ways: by calculating the proportions of patients achieving clinical remission or low disease activity (LDA) based on the disease activity score in 28 joints (DAS28), the Clinical Disease Activity Index (CDAI), and the Simple Disease Activity Index (SDAI) values, as well as American College of Rheumatology (ACR)/EULAR Boolean remission criteria at weeks 24 and 52, and by generating cumulative probability plots of the numeric index of the ACR response (ACR-N) at week 52.

Radiographic efficacy was assessed in several ways. Cumulative probability plots of change in mTSS from baseline to week 52 are presented for each of the three treatments. CRRP, defined as a change in mTSS from baseline of ≥3 units, was assessed based on patients’ achievement of DAS28 remission at weeks 24 and/or 52. The relationship between change in mTSS from baseline and categorized time-averaged DAS28 response (remission, LDA, moderate disease activity [MDA], and high disease activity [HDA]) was assessed by treatment group at week 52. Scatter plots of change in mTSS from baseline and time-averaged DAS28 at week 52 are presented.

Functional efficacy was assessed by generating cumulative probability plots of Health Assessment Questionnaire Disability Index (HAQ-DI) scores at baseline and week 52.

Statistical analysis

Clinical and functional efficacy analyses were based on the modified intent-to-treat (mITT) population, defined as all patients who received at least one dose of ETN or MTX. Missing data were imputed using the last observation carried forward (LOCF) approach. Radiographic efficacy analyses were based on the radiographic intent-to-treat (rITT) population, defined as all patients who received at least one dose of ETN or MTX and had a radiographic assessment at baseline and at least one post-baseline visit. The continuous efficacy endpoints were analyzed using an analysis of covariance (ANCOVA) model with baseline value as covariate, when available, and the study center and treatment received as factors. Clinical remission/LDA rates were analyzed using an unadjusted method, an adjustment based on the Cochran–Mantel–Haenszel estimator, in which patient data were stratified by study center and prior MTX use, and an adjustment based on the logit estimator, in which the value of 0.5 was added to cells with values of zero, in order to include strata with such cells in the calculation. The exception were strata with a sum row of zero or a sum column of zero, which were not included in the calculation.

Two-step predictor analyses were conducted in order to assess which baseline demographic and clinical characteristics, as well as clinical responses at week 52, were significantly associated with CRRP response at week 52. The first step was a univariate analysis, performed for each parameter of interest, in which odds ratios (OR) with 95% confidence intervals (CI) and P-values for comparisons against reference values were calculated. In the second step, baseline parameters and all clinical parameters at week 52 with p ≤ .1 were incorporated in a stepwise regression model that also included ETN dosage.

Results

Baseline characteristics

Baseline demographics and disease characteristics for patients receiving 25 mg BIW ETN, 10 mg BIW ETN, and MTX have been published previously [8], and were comparable among treatment groups, with the exception of mean body mass index (BMI), which was significantly higher in the 25 mg etanercept group versus the MTX group (22.8 ± 3.8 kg/m² versus 21.7 ± 3.3 kg/m², respectively; p = .005).

Clinical efficacy endpoints

For nearly all clinical criteria for remission or LDA activity, a nominally higher proportion of patients in the 25 mg group achieved those outcomes compared with the 10 mg group, but the difference was not statistically significant for any of the outcomes (Figure 1 and Supplementary Table S1).

Figure 1. Odds of achieving LDA or remission (25 mg BIW versus 10 mg BIW ETN) at weeks 24 and 52 (Logit estimator). Logit estimator adjustments were performed by adding 0.5 to cells with values of zero, in order to include strata with such cells in the calculation. The exception were strata with a sum row of zero or a sum column of zero, which were not included in the calculation. Alternative OR calculations, in which ORs were either unadjusted or adjusted using the CMH estimator (with data stratified by study center and prior MTX use) yielded a nominally statistically significant advantage of ETN 25 mg BIW over ETN 10 mg BIW for ACR/EULAR Boolean remission at week 52 (Table S1). ACR: American College of Rheumatology; CDAI: Clinical Disease Activity Index; CMH: Cochran-Mantel-Haenszel; CI: confidence interval; DAS28: 28-joint Disease Activity Score; ETN: etanercept; EULAR: European League Against Rheumatism; LDA: low disease activity; MTX: methotrexate; OR: odds ratio; SDAI: Simple Disease Activity Index.

Cumulative probability plots show a similar distribution of treatment responses in the two ETN groups, with more patients achieving higher response than patients in the MTX group (Figure 2).

Figure 2. Cumulative probability plot of ACR-N at week 52 (mITT population, LOCF). ACR-N: percentage improvement using American College of Rheumatology scale; ETN: etanercept; MTX: methotrexate.

A total of 41/165 (25%) patients receiving MTX experienced disease worsening (negative ACR score at week 52), with final ACR scores ranging from −1 to −182, compared with 25/182 (14%) and 18/172 (10%) patients receiving 10 mg and 25 mg ETN, respectively (score ranges: 10 mg, −5 to −120; 25 mg, −4 to −100). Mean observed values (95% CI) in ACR-N at week 52 for 10 mg ETN (55.8 [49.8, 59.3]) and 25 mg ETN (57.2 [51.4, 61.2]) were significantly higher than that for MTX (47.2 [41.5, 52.2]), with p-values of .027 and .008, respectively. Differences between the 10 mg and 25 mg ETN groups were not significant (p = .590).

Radiographic efficacy endpoints

Change in mTSS and attainment of CRRP

Radiographic progression, assessed as the mean [95% CI] change in annualized mTSS to week 52, was the highest for MTX (9.8 [7.6, 11.6]), which was significantly greater than the changes observed for 10 mg ETN (5.2 [2.7, 6.5], p < .001) and 25 mg ETN (3.3 [1.0, 4.8], p < .001). The difference in annualized mTSS between the 10 and 25 mg ETN groups was not statistically significant (p = .199).

Unlike the patients in the MTX group, the majority of patients in the ETN treatment groups had an annualized mTSS score <3, which corresponds with minimal or no radiographic progression. CRRP was observed in 91 (53%) patients in the MTX group, compared with 68 (36%) and 58 (32%) patients in the 10 mg and 25 mg ETN groups, respectively (Figure 3). In patients with CRRP (i.e. mTSS score change ≥3) treated with ETN 10 mg twice weekly (BIW) and ETN 25 mg BIW, the mTSS score changes ranged from 3 to 87 and from 3 to 52, respectively (Figure 3).

Figure 3. Cumulative probability of annualized change in mTSS from baseline to week 52 (rITT population). CRRP: clinically relevant radiologic progression (mTSS ≥3); ETN: etanercept; mTSS: modified Total Sharp Score; MTX: methotrexate; rITT: radiologic intent-to-treat population.

For patients who did not achieve DAS28 remission at week 24, a higher proportion had CRRP at week 52 in the group treated with 10 mg ETN (45% [61/135]) than with 25 mg ETN (37% [51/137]) (Supplemental Figure S1).

Radiographic progression versus DAS28 response

In each treatment group, the mean ± standard deviation (SD) change from baseline in mTSS was lower in patients who achieved time-averaged DAS28 remission or LDA compared with patients who had MDA or HDA, with MTX-treated patients having the highest change in mTSS score in each DAS28 response category (Figure 4(A)). Patients receiving MTX were more likely to experience radiographic progression from baseline to week 52 than patients receiving ETN, across all DAS28 response categories. Compared with the ETN 10 mg group, the group treated with ETN 25 mg had a nominally lower proportion of patients with MDA or HDA (62% versus 68%) and a higher proportion of those with remission or LDA (47% versus 44%) who experienced radiographic progression (Figure 4(B)).

Figure 4. Radiographic outcomes at week 52 by DAS28 response: (A) change in mTSS and (B) radiographic progression. DAS28: 28-joint disease activity score; ETN: etanercept; HDA: high disease activity; LDA: low disease activity; MDA: moderate disease activity; mTSS: modified Total Sharp Score; MTX: methotrexate; SD: standard deviation.

These findings were corroborated through scatter plots, with Spearman correlations between mTSS change from baseline at week 52 and DAS28 response status of 0.45 (MTX), 0.43 (ETN 10 mg), and 0.28 (ETN 25 mg) (Supplemental Figure S2).

Functional efficacy endpoints

Mean [95% CI] improvements in HAQ-DI from baseline to week 52 were similar between groups treated with 10 mg ETN (–0.56 [–0.60, −0.44]) and 25 mg ETN (–0.58 [–0.66, −0.50]). Both of those changes were significantly larger (ETN 10 mg, p = .009; ETN 25 mg, p < .001) than the improvement observed in the MTX group (–0.36 [–0.46, −0.29]) (Supplemental Figure S3).

Predictors of CRRP

The univariate analysis of baseline characteristics suggests that a positive rheumatoid factor (RF) status, higher C-reactive protein (CRP), higher erythrocyte sedimentation rate (ESR), a higher degree of joint narrowing, and a higher HAQ-DI score were associated with attaining CRRP at week 52, while a higher total joint count (TJC) was a negative predictor (Figure 5(A)). In the follow-up stepwise analysis, these findings were confirmed for RF status, CRP, ESR, and TJC, while both ETN doses were associated with significantly lower odds of attaining CRRP, compared with MTX (Figure 5(B)).

Figure 5. Univariate and stepwise analysis of baseline characteristics as predictors of CRRP at week 52. CRP: C-reactive protein; CRRP: clinically relevant radiographic progression; ESR: erythrocyte sedimentation rate; ETN: etanercept; HAQ-DI: Health Assessment Questionnaire – Disability Index; JSN: joint space narrowing; MTX: methotrexate; RF: rheumatoid factor; TJC: tender joint count.

In addition, both the univariate (Figure 6(A)) and stepwise analysis (Figure 6(B)) suggest that poorer clinical outcomes at week 52 are predictors of CRRP.

Figure 6. Univariate and stepwise analysis of week 52 clinical outcomes as predictors of CRRP at week 52. AUC: area under the curve; CRP: C-reactive protein; CRRP: clinically relevant radiographic progression; DAS28: disease activity score in 28 joints; HAQ-DI: Health Assessment Questionnaire – Disability Index; Q: quartile.

AUC average values were divided into 4 quartiles based on outcome, where 0 = lowest quartile of scores and 3 = highest quartile of scores.

Discussion

This post hoc analysis of data from the JERA study suggests that there may be some differences in clinical and radiographic efficacy between the 25 mg BIW and 10 mg BIW ETN doses in Japanese patients with RA. However, those differences were small and not statistically significant. In another study in Japanese patients with RA that compared 10 mg and 25 mg BIW ETN [9], ACR20 response rates in the two groups were similar (25 mg, 65%; 10 mg, 64%) [10], and consistent with findings from the JERA trial (25 mg, 79%; 10 mg, 76%) [8]. Other studies that compared 25 mg and 10 mg BIW ETN in patients with RA have shown that 25 mg was significantly more effective than 10 mg in respect to ACR20/50/70 responses [11–13] and radiographic progression [12]. However, those studies were conducted in predominately Caucasian populations (>80% of all treatment groups), and there are both genetic and environmental differences between Japanese and Caucasian populations that are relevant for patients with RA. For example, HLA-B27, which is associated with rheumatic diseases, is found in fewer than 1% of Japanese individuals, but its prevalence ranges from 6% to 9% in most Caucasian populations [14]. The prevalence of numerous other genes associated with susceptibility to RA have been shown to differ between European and Asian populations [15]. Relevant environmental differences between the two populations include smoking rates and average body mass index (BMI) [10]. A difference in BMI can result in different odds of developing RA [16] as well as in different results of clinical trials in RA that use a fixed-dose treatment [10].

Our predictor analysis suggests that a worse background disease status, treatment assignment to MTX (as opposed to ETN), and poorer clinical outcomes at week 52 are associated with a clinically relevant radiographic progression of RA. The worse disease status, such as the presence of RF or higher titer of RF, as well as higher values of CRP and ESR as predictors of CRRP were compatible with the previous report, while it is interesting to note that a high baseline TJC is associated with lower odds of developing CRRP at 52 weeks, compared with lower TJC. In general, TJC is an indicator of disease activity, which ultimately leads to joint destruction, which should manifest as CRRP. Some unidentified confounders may account for this unexpected observation.

This analysis had some limitations. The original clinical trial was not powered to detect differences between the two ETN doses, making it less likely that significant differences between 10 mg and 25 mg ETN could be detected in this analysis. In addition, although safety assessments were part of the trial, we did not assess differences in safety between the two ETN doses.

Conclusions

The 25 mg BIW ETN dosage does not appear to be significantly more efficacious than the 10 mg BIW dosage in Japanese patients with RA. Further studies evaluating the optimal ETN dosing regimen in this patient population may be merited.

Acknowledgments

Additional statistical analyses were provided by Annette Szumski of Syneos Health and Kenneth Kwok of Pfizer. Medical writing assistance was provided by Lorna Forse, PhD, and Vojislav Pejović, PhD, of Engage Scientific Solutions, and was sponsored by Pfizer.

Availability of data and materials

Upon request, and subject to certain criteria, conditions, and exceptions (see https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information), Pfizer will provide access to individual de-identified participant data from Pfizer-sponsored global interventional clinical studies conducted for medicines, vaccines, and medical devices (1) for indications that have been approved in the US and/or EU or (2) in programs that have been terminated (i.e. development for all indications has been discontinued). Pfizer will also consider requests for the protocol, data dictionary, and statistical analysis plan. Data may be requested from Pfizer trials 24 months after study completion. The de-identified participant data will be made available to researchers whose proposals meet the research criteria and other conditions, and for which an exception does not apply, via a secure portal. To gain access, data requestors must enter into a data access agreement with Pfizer.

Conflicts of interest

T. T. reports research grants: Astellas, Chugai, Daiichi-Sankyo, Takeda, AbbVie GK, Asahi Kasei, Mitsubishi Tanabe Pharma, Pfizer Japan, Eisai, AYUMI, Nippon Kayaku, and Novartis; Speakers’ fees: AbbVie GK, Bristol-Myers KK, Chugai, Mitsubishi Tanabe Pharma, Pfizer Japan, Astellas, Daiichi-Sankyo, Eisai, Sanofi KK, Teijin Pharma, Takeda, Novartis. N. M. reports none.

R. D. P. reports Pfizer employee at the time the study was conducted and may own company stock. N. S. and T. H. report Pfizer employees and may own company stock.

Additional information

Funding

This study was sponsored by Pfizer.