Introduction

The goals of managing rheumatoid arthritis (RA) are to control or prevent damage to the joints, decrease pain, and prevent loss of function [1]. Remission as defined by the European League against Rheumatism (EULAR; disease activity score in 28 joints [DAS28] <2.6) is currently a realistic and achievable goal of RA treatment [2, 3]. There are various treatment strategies to achieve remission, such as combination therapy with disease-modifying antirheumatic drugs (DMARDs) or therapy with tumor necrosis factor–α (TNF-α) inhibitors (monotherapy or combination therapy) [3, 4]. Combination therapy using TNF-α antagonists plus traditional DMARDs is a common strategy for the treatment of RA and improves the clinical outcomes in patients who previously received monotherapy [5]. In particular, the combination of TNF-α antagonists and methotrexate (MTX) results in better clinical and radiographic responses than when either agent is used alone [6, 7].

A relatively recent concern in RA disease management is the treatment of patients who do not respond to initial anti–TNF-α therapy. Up to approximately one-third of patients treated with TNF-α antagonists fail to respond or develop adverse events (AEs) that lead to treatment discontinuation [8–10]. The use of biologics such as rituximab [11] and abatacept [12] with a different mechanism of action may be an option for such patients; however, several studies have shown that patients who do not tolerate or respond to initial treatment with a TNF-α antagonist can benefit by switching to a different TNF-α antagonist (despite the fact that both agents target TNF-α) [13–21]. Conclusions regarding the benefit of switching TNF-α inhibitors from these studies have, however, been limited by relatively small patient numbers.

The Japanese postmarketing survey for etanercept (ETN; a fully human soluble TNF receptor fusion protein) registered all patients treated with ETN in Japan, including a large number of patients who had switched from infliximab (IFX) to ETN therapy. The purpose of this post hoc analysis of interim data (for the first 7,099 patients; analysis for all 13,894 patients is still ongoing) from the postmarketing survey is to evaluate the safety and effectiveness of ETN among patients who discontinued IFX treatment because of a lack of effectiveness or AEs.

Patients and methods

Patients

This study enrolled 13,894 Japanese patients with RA (aged ≥ 18 years) with and without prior IFX experience participating in a postmarketing surveillance study of ETN use between March 2005 and April 2007. This Post-Marketing Surveillance Program was conducted under Pharmaceutical Affairs Law, which collects data and information from usage of products within product labeling under daily clinical practice. Patient eligibility for ETN treatment was evaluated according to the Japan College of Rheumatology guidelines [22], and the protocol was reviewed and approved by the Japanese Ministry of Health, Labor, and Welfare. ETN was indicated for patients with active RA despite DMARD treatment for >3 months, ≥6 tender joints, ≥6 swollen joints, and erythrocyte sedimentation rate (ESR) ≥28 mm/h or C-reactive protein (CRP) ≥2.0 mg/dL. Patients also had to have a low risk for opportunistic infections (white blood cell count ≥4,000/mm3, peripheral blood lymphocyte count ≥1,000/mm3, and negative serum β-d-glucan). Contraindications for ETN treatment included ongoing infection, history of serious infection in the previous 6 months, history of tuberculosis infection or Pneumocystis carinii pneumonia, congestive heart failure, or malignancy or demyelinating disease. Mandatory chest radiographs and tuberculin tests were conducted before initiation of treatment.

Study design/assessments

Patients receiving ETN were divided into 2 groups based on their history of IFX use and were categorized as either having previous IFX treatment (pre-IFX) or no prior IFX treatment (non-IFX). Etanercept 10–25 mg was administered subcutaneously twice weekly. Patients could self-inject after receiving appropriate training from healthcare professionals. Safety and effectiveness data were collected every 4 weeks for 24 weeks. The primary effectiveness endpoint was EULAR response.

Statistical analysis

In this post hoc analysis, missing effectiveness data were accounted for using last-observation-carried-forward (LOCF) methods, except for baseline values, which were not carried forward. Fisher’s exact test and t-test were used to assess differences between pre-IFX patients and non-IFX patients in baseline characteristics, treatment effectiveness (DAS28 response according to EULAR criteria), and serious AEs. Continuation rates of etanercept between the pre-IFX and non-IFX groups were compiled to generate a Kaplan–Meier plot and were tested using log-rank test. Furthermore, Cox proportional hazard models were applied to estimate relative risks and 95% CIs of serious AEs in relation to pre-IFX treatment, after adjustment for major confounders including age, sex, Steinbrocker functional class 4, duration of RA, history of infectious disease, history of tuberculosis, presence of any comorbidities, and concomitant DMARD use. Also, multiple logistic regression models were used to estimate the effect of pre-IFX treatment on the likelihood of achieving a moderate to good EULAR response after adjustment for major confounders. All statistical analyses were performed using SAS, version 8.2, software (SAS Institute, Inc., Cary, NC). Two-sided P values of less than 0.05 were considered to indicate statistical significance.

Results

Patient characteristics

This interim analysis evaluated the safety and effectiveness of ETN among the first 7,099 patients (908 pre-IFX and 6,191 non-IFX patients) out of 13,894 patients enrolled. Most baseline characteristics differed significantly between pre-IFX and non-IFX patients (Table 1). Patients in the pre-IFX group tended to be younger than patients in the non-IFX group (mean age, 54.2 and 58.9 years, respectively, P < 0.001) and had a shorter duration of RA (9.0 and 9.9 years, respectively, P < 0.001) and fewer comorbidities (52.0 and 59.1%, respectively, P < 0.001). In the pre-IFX patients, concomitant DMARD use at the enrollment to the study was higher compared with non-IFX patients (87.1 and 68.7%, respectively, P < 0.001), as was the concomitant use of MTX (80.9 and 48.6%, respectively, P < 0.001).

Table 1 Patient characteristics
Full size table

Treatment rationale and compliance

Within the cohort of pre-IFX patients (n = 908), most patients (94%, n = 856) switched from IFX to ETN treatment because of a lack of IFX effectiveness. In cases in which initial IFX treatment was effective (6%, n = 52), the majority of patients (73%, n = 38) switched to ETN because of treatment-related AEs.

Etanercept compliance was monitored in pre- and non-IFX cohorts for 24 weeks. Pre-IFX group showed significantly higher continuation rate than non-IFX group (87.9% vs. 82.3%, P < 0.001) (Fig. 1). Of the entire study population, 17.0% (n = 1,208) of patients treated with ETN discontinued use, and most discontinued because of AEs (8.6%, n = 608) rather than a loss of effectiveness (2.6% n = 185). Among the patients who discontinued treatment because of AEs, pre-IFX patients had a significantly lower rate of ETN discontinuation compared with non-IFX patients (5.0% vs. 9.1%; P < 0.001). There was no statistical difference in discontinuations due to lack of effectiveness between pre- and non-IFX patients.

Fig. 1
figure 1

Kaplan–Meier analysis for continuation rates of etanercept between the pre-IFX and non-IFX groups. IFX = infliximab. P < 0.001 for pre-IFX vs. non-IFX (log-rank test)

Full size image

Safety

Approximately 34% (n = 2,424) of the patients in this study reported an AE during the observation period. For both groups, the most common AE was infection. Overall, 6.4% (n = 452) of patients reported a serious AE during the study. The incidence of AEs and serious AEs (SAEs) was significantly lower in the pre-IFX group than in the non-IFX group (P < 0.001 and P = 0.004, respectively; Table 2). In patients who received MTX concomitant with ETN, there was no increase in the risk of SAEs relative to ETN monotherapy (Fig. 2). The incidence of SAEs in non-IFX patients treated with doses of MTX ≥ 10 mg/wk was significantly lower than in patients treated with ETN alone (P = 0.019). Compared with the non-IFX group, pre-IFX patients had a non-significantly lower risk for SAE occurrence (multivariate hazard ratio, 0.75; 95% CI, 0.54–1.06; P = 0.098).

Table 2 Incidence of adverse events
Full size table
Fig. 2
figure 2

Incidence rate of serious adverse events by dose of MTX. IFX infliximab; MTX methotrexate. *P < 0.05 for all doses versus 0 mg MTX (Fisher’s exact test). Values in parentheses indicate number of subjects per group

Full size image

Effectiveness

Etanercept was effective, as measured by EULAR response through the treatment period, in both pre-IFX and non-IFX patients. The majority of pre-IFX patients (>80%) responded to ETN treatment (Fig. 3). According to the EULAR response criteria of no response, moderate response, and good response, the number of good responses increased significantly (P < 0.05) at week 8 and beyond in the non-IFX group and at week 12 and beyond in the pre-IFX group (P < 0.05). Remission was achieved in 14.4% of the patients in the pre-IFX group after switching to ETN (vs. 16.6% of the patients in the non-IFX group). There was no significant difference in achieving a good or moderate EULAR response between the non-IFX and pre-IFX groups (multivariate odds ratio, 0.93; 95% CI, 0.70–1.25; P = 0.64).

Fig. 3
figure 3

European League Against Rheumatism responses. IFX infliximab. *P < 0.05 versus week 4 for good responders (Fisher’s exact test)

Full size image

Discussion

In clinical practice, it is important to consider the benefit of prescribing a second TNF-α antagonist if an initial anti–TNF-α treatment has not been effective. Indeed, multiple TNF-α inhibitors are currently in clinical use, and although all TNF-α inhibitors have the same target [23], each drug is unique in its mechanism of action and pharmacologic properties (Table 3) [24–26], which can result in variations in clinical outcomes. In this study of 7,099 Japanese patients with RA, the largest study to date to evaluate the benefit of TNF-α switching, ETN was effective in patients with a history of prior IFX use as well as in those who were naive to IFX therapy. In the case of patients for whom prior IFX therapy had failed and who were subsequently treated with ETN, >80% were able to attain a moderate to good EULAR response by the end of the study period, similar to the effectiveness observed in IFX-naive patients treated with ETN. Moreover, ETN was able to induce remission in a subset of patients for whom prior IFX therapy had failed. Importantly, the majority (810 of 860; 94%) of pre-IFX patients in this study had discontinued initial IFX treatment because of a lack of effectiveness; therefore, our data are consistent with smaller scale studies that showed that patients who discontinue IFX can respond effectively to subsequent ETN treatment [15–21, 27]. In contrast to a study demonstrating that patients switching from ETN to IFX required higher doses of IFX when compared with ETN-naive patients [13], IFX-treated patients in the present study did not require higher doses of ETN to achieve responses compared to IFX-naive patients.

Table 3 Differences between ETN and IFX
Full size table

Overall, treatment continuation was good in both groups, regardless of previous IFX experience, and was at least as good as continuation rates observed in other IFX-to-ETN switching studies [20, 27, 28]. Regardless of prior treatment history, most patients continued to receive ETN for the duration of the study. Pre-IFX patients had a slightly higher rate of continuation throughout the study and significantly higher participation at the end of the treatment period compared with non-IFX patients. Significantly, fewer pre-IFX patients experienced AEs that led to discontinuation compared with IFX-naive patients. Generally, pre-IFX patients were younger and healthier (fewer comorbidities) compared with IFX-naive patients, which may contribute to the lower rate of ETN discontinuation observed in the pre-IFX group.

The most common AE reported among all patients during the study was non-serious infection. There were no significant differences in the types of AEs reported by pre- and non-IFX patients, suggesting that the safety profile of ETN is not changed by previous IFX experience. Overall, the incidence of AEs and SAEs was lower among patients with previous IFX exposure, which is encouraging given that a proportion of these patients discontinued previous IFX treatment because of AEs related to IFX.

Many RA treatment regimens that use biologic DMARDs also incorporate MTX, and the tolerability of multidrug treatment regimens is a common concern. Although MTX is not consistently used in ETN regimens, MTX is usually administered with IFX because MTX inhibits the production of human antichimeric antibodies, which can interfere with treatment effectiveness and induce autoimmune sequelae [29, 30]. MTX is currently approved as a second-line agent in Japan, and the recommended dose is lower (upper limit is 8 mg/wk) than that in the European Union or the United States, owing to the higher incidence of AEs observed among Japanese patients in MTX clinical trials [31]. In this study, the concomitant use of MTX and ETN did not cause an increase in SAEs in pre- or non-IFX patients compared with patients receiving ETN monotherapy, suggesting that MTX use is not a major factor in predicting SAEs in these patients. It should be noted that more pre-IFX patients received concomitant MTX therapy compared with non-IFX patients. Although the reasons for increased tolerability to the combination of ETN and MTX are not clear, pre-IFX patients tended to be younger, had a shorter duration of RA, and had fewer comorbidities when compared with non-IFX patients. These data imply that patients healthy enough to be treated with MTX may have a lower incidence of SAEs resulting from combination therapy.

The present study is limited, in part, by its observational nature. Patients were followed for only up to 6 months, and radiographic analysis was not performed to confirm effectiveness. In addition, the period of IFX treatment and the period between the final infusion of IFX and the first injection of ETN were confirmed only for a subset of patients. This study does not definitively demonstrate whether ETN treatment is effective for all patients with RA who are non-responsive to IFX, nor does it address the issue of recurrence of AEs in patients who switched from IFX treatment because of AEs. In addition, differences in baseline patient demographics with regard to age, disease onset and duration, background DMARD therapy, and comorbidities may have affected the results.

The interim analysis of this large observational registry study demonstrated, in a real-world setting, the safety and effectiveness of ETN treatment in patients with active RA who were switched from previous IFX treatment as a result of lack of effectiveness or AEs. ETN treatment was effective and well tolerated in both pre-IFX and non-IFX patients. Overall, patients with prior IFX experience had safety and effectiveness outcomes that were as good as those of patients who were naive to IFX treatment.