Abstract
Introduction
The persistence of golimumab (GLM) treatment in Japanese patients with rheumatoid arthritis (RA) has been evaluated previously, but evidence of long-term real-world use is lacking. This study assessed the long-term persistence of GLM use, its influencing factors, and impact of prior medications in patients with RA in actual clinical practice in Japan.
Methods
This is a retrospective cohort study of patients with RA using data from a hospital insurance claims database in Japan. The identified patients were stratified as only GLM treatment (naïve), had one biological disease-modifying anti-rheumatic drug (bDMARD)/Janus kinase (JAK) inhibitor treatment prior to GLM [switch (1)] and had at least two bDMARDs/JAK prior to GLM treatment [switch (≥ 2)]. Patient characteristics were evaluated using descriptive statistics. Kaplan–Meier survival and Cox regression methods were used to analyze GLM persistence at 1, 3, 5, and 7 years and the associated factors. Treatment differences were compared using a log-rank test.
Results
GLM persistence rate in the naïve group was 58.8%, 32.1%, 21.4%, and 11.4% at 1, 3, 5, and 7 years, respectively. Overall persistence rates in the naïve group were higher than in switch groups. Higher GLM persistence was observed among patients aged 61–75 years and those concomitantly using methotrexate (MTX). Also, women were less likely to discontinue treatment compared to men. Higher Charlson Comorbidity Index score, initial GLM dose of 100 mg, and switch from bDMARDs/JAK inhibitor were related to a lower persistence rate. As a prior medication, infliximab showed the longest persistence for subsequent GLM, and using this as a reference, tocilizumab, sarilumab, and tofacitinib subgroups had significantly shorter persistence, respectively (p = 0.001, 0.025, 0.041).
Conclusion
This study presents the long-term real-world results for persistence of GLM and its potential determinants. These most recent and long-term observations demonstrated that GLM and other bDMARDs continue to benefit patients with RA in Japan.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Why carry out this study? |
Long-term treatment with golimumab (GLM) and other biological disease-modifying anti-rheumatic drugs (bDMARDs) over the past 7 years has not yet been reviewed in day-to-day clinical practice. |
There is insufficient evidence on patient characteristics associated with long-term use of GLM after switching from bDMARDs/Janus kinase (JAK) inhibitor treatments in Japan. |
There is an ambiguity about which bDMARDs/JAK inhibitor prior to GLM treatment could have a potential impact on the persistence of GLM. |
What was learned from the study? |
The persistence of GLM in patients with RA was relatively higher than other bDMARDs and JAK inhibitor treatments throughout the 7-year observation period. |
Patients who received multiple bDMARDs prior to GLM treatment had a lower persistence rate. |
Patients being female, naïve to GLM treatment, and receiving concomitant methotrexate were associated with longer persistence of GLM. These findings will be helpful to healthcare providers in developing effective treatment strategies in the future. |
Introduction
Rheumatoid arthritis (RA) is a systemic autoimmune disease that primarily starts with symmetric synovitis and may lead to progressive joint-related disabilities [1]. An inadequately diagnosed or treated joint destruction progresses and causes irreversible physical dysfunction, resulting in permanent disabilities that affect patients’ quality of life and subsequently contribute to an increased socioeconomic burden [2]. The estimated prevalence of RA in Japan is 825,000, which equates to nearly 0.65% of the Japanese population [3]. Current treatments for RA include traditionally used non-steroidal anti-inflammatory drugs (NSAIDs) and corticosteroids as well as conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) including methotrexate, sulfasalazine, bucillamine, tacrolimus, and iguratimod. Furthermore, Janus kinase (JAK) inhibitors, as well as biological disease-modifying anti-rheumatic drugs (bDMARDs) such as tumor necrosis factor (TNF) inhibitors, interleukin-6 (IL-6) inhibitors, IL-1 inhibitors, B cell and T cell co-stimulation inhibitors, have recently been approved (Supplementary Table S1) [3]. The introduction of bDMARDs has transformed RA treatment; apart from preventing disease progression, they have been shown to improve quality of life and reduce mortality [4,5,6]. Despite these benefits, complete long-term disease remission is not achieved in many patients. Clinical trial data indicates that approximately 30% of patients with RA do not respond to treatment with the first bDMARDs, and 40% of such patients fail on second bDMARD treatment, resulting in a significant number of patients receiving treatment with three or more bDMARDs [7, 8]. However, there is little evidence to support the bDMARD switching strategy [9], and a clear consensus has not been reached.
According to published reports from observational studies [10, 11], meta-analysis [12], and systematic reviews [13], patients who switched to the second or third bDMARDs after an inadequate response (IR) to the first bDMARD used had a lower probability of achieving a clinical response. The clinical efficacy of TNF-targeting monoclonal antibody golimumab (GLM) has been demonstrated in patients with RA who had previously discontinued anti-TNF therapy because of IR, intolerance, or other reasons [8]. The post hoc analysis of post-marketing surveillance (PMS) data for GLM reports that switching to GLM caused a clinical response regardless of the number of previous bDMARDs and mechanism of action (MoA) of pretreatment bDMARDs; however, the evaluation period considered was as short as 24 weeks [14]. Furthermore, the retention rate (i.e., persistence) of bDMARDs varies considerably depending on the country and drug [15]. A systematic review of 52 global studies indicated that the retention rates of bDMARDs at 1 year ranged from 32.0% to 90.9% [16]. These large variations can be attributed to institutional and cultural factors in the region or during the evaluation period, making direct comparison of survey results for each country difficult.
This retrospective analysis of records from the Japanese Medical Data Vision (MDV) health claims database investigated the background of patients with RA and the long-term retention rate (persistence) of GLM in association with prior experience of bDMARDs/JAK inhibitor treatments in a real-world setting in Japan.
Methods
Data Source
This retrospective cohort study was conducted using data sets extracted from the MDV health claims database for a period of 13 years, from 2008 to 2020. MDV is a hospital-based database that is recognized as one of the largest and most credible commercially available medical databases in Japan. Briefly, the MDV database contains health insurance claims, Diagnosis Procedure Combination (DPC) data, and administrative data from over 400 hospitals covering over 30 million patients. This database was chosen over other databases, as it provides data on elderly patients, making it more suitable for research of RA. It provides data on medical treatment information which includes age, gender, disease name, disease diagnoses, in- or out-patient status, drug prescriptions, dosage, blood test, medical procedures, and reimbursement costs. These data were analyzed to investigate the background of patients with RA and the long-term persistence of GLM in Japan. The extracted data are unlinkable (to any personal patient information) and were anonymized before receipt. Diagnoses of patients were coded according to the International Classification of Diseases, 10th Revision (ICD-10) coding scheme and indicated as either “confirmed” or “suspected” diagnosis. Data of only patients with a “confirmed” diagnosis of RA were used. As per the ethical guidelines for medical and health research involving humans issued by the Japanese Ministry of Health, Labor and Welfare and Ministry of Education, Culture, Sports, Science, and Technology the process of informed consent was not applicable for this study.
Study Design and Population
This study investigated the background of patients with RA and the long-term persistence of GLM in Japan after 1, 3, 5, and 7 years of the treatment. Patients were stratified into two groups: the “GLM naïve group”, in which GLM was prescribed as the first bDMARD; and the “treatment switch group”, in which GLM was prescribed as subsequent bDMARD. The treatment switch group was further divided into the switch (1) group and the switch (≥ 2) group based on the number of times patient had switched treatment, and then further subdivided into specific subgroups (i.e., TNF inhibitors, IL-6 inhibitors, CTLA4-Ig, and JAK inhibitors) depending on the bDMARD the patient received before being administered GLM (Supplementary Table S1). Patients in switch groups were not further stratified into separate treatment subgroups and were evaluated as a whole.
For the GLM naïve group, data from patients with a confirmed RA diagnosis (excluding suspected cases) during the study period defined by ICD-10 codes M05 and M06 (Supplementary Table S2), who were aged at least 18 years on the index date and received a prescription for GLM at least once after the index date, were analyzed. The index date was defined as the first date in the database when GLM was prescribed post first RA diagnosis date. Patients were considered as discontinued if they switched to another target treatment or if the actual refill prescription date was greater than 60 days post prescription period. The next prescription date was regarded as the discontinuation date (Fig. 1). Patients who could not be traced 3 months before and 1 year after the index date were excluded, as information from 3 months before the index date was used to collect baseline characteristics and long-term endpoint assessments required information from 1 year after the index date. Patients with RA who also had Crohn’s disease, ankylosing spondylitis, juvenile arthritis, psoriasis, psoriatic arthritis, ulcerative colitis, or Behçet’s disease were excluded (Supplementary Table S3). The persistence rate was defined as the time (days) from the index date (treatment initiation) till the medication discontinuation; this definition is in agreement with earlier database studies in RA or other disease areas [17,18,19].
For the treatment switch group, patients prescribed a JAK inhibitor or a bDMARD other than GLM during the study period who met the same inclusion and exclusion criteria as the primary endpoints were analyzed. For this group, the index date was defined as the first date in the database when bDMARDs or a JAK inhibitor was prescribed after the first RA diagnosis. Patients who received prescriptions for bDMARDs or a JAK inhibitor at least once after the index date were included.
Endpoints
This study investigated the GLM persistence rate after 1, 3, 5, and 7 years of GLM treatment in the GLM-naïve and treatment switch groups. The treatment duration (i.e., time between the index date and the GLM discontinuation date) and influence of patient demographics or treatment patterns on the persistence rate were estimated. The persistence rate in both treatment groups was further examined on the basis of demographic group categories of age, gender, concomitant use of methotrexate (MTX), Charlson Comorbidity Index (CCI) score, initial dose of GLM, and number of previous biologic therapies to assess the effect of these factors on the persistence rate.
Statistical Analysis
Descriptive statistics were used to estimate the number of patients, mean, and standard deviation for continuous variables and the frequency distribution for categorical variables. The persistence rate in each treatment group was estimated using the Kaplan–Meier method. The numbers at risk, the cumulative persistence rate and its two-sided 95% confidence interval (CI) were provided at each timepoint. The median duration and its two-sided 95% CI were calculated for each treatment group. An event was defined as the discontinuation of treatment, and patients were censored at the end of their follow-up. A log-rank test was performed to compare treatment differences. Multiple testing was adjusted using the Bonferroni correction. The Cox regression model was used to perform univariate analyses on the factors potentially affecting the persistence rate (i.e., age, gender, CCI score, concomitant use and dose of MTX, initial dose of GLM, and number of previous biologic therapies). For each factor, an estimate of hazard ratio (HR) and its associated 95% CI were obtained. Multivariate analysis was performed with all factor variables (without performing model selection). In all analyses, a p value of less than 0.05 was considered statistically significant. All statistical analyses were performed using SAS version 9.4 (TS1M6) (SAS/STAT 15.1).
Results
Patient Characteristics
A total of 384,983 records of patients with RA were identified from the MDV database, of which 7524 were prescribed GLM. A total of 39,545 patients were prescribed JAK inhibitors or any other bDMARDs but not GLM and 11,995 of them who met the inclusion/exclusion criteria were considered as the reference cohort for analysis (Fig. 2). Of the 7524 GLM-treated patients, 2283 were never prescribed bDMARDs or JAK inhibitors prior to GLM (naïve), 942 were prescribed one bDMARD or JAK inhibitor prior to GLM [switch (1)], and 322 were prescribed at least two other bDMARDs or JAK inhibitors prior to GLM [switch (≥ 2)]. Figure 2 depicts an overview of the stratified patient demographics of naïve, switch (1), and switch (≥ 2) patients.
The mean age of the patients in the naïve group (67.3 years) was slightly but significantly higher than the mean age of the patients in the switch (overall) group (66.5 years), p = 0.047. There was no significant difference in the mean age of patients in the switch (1) group (p = 0.106) versus the switch (≥ 2) group (p = 0.717). The majority of patients were women (naïve group, 78.84%; switch group, 81.41%) and switched patients had slightly higher CCI scores (CCI score of 3–5, naïve group [23.04%] vs switch group [30.30%]) and had higher morbidity than naïve patients at baseline (Table 1). Patients in the switch (1) group had a significantly higher rate of congestive heart failure (p = 0.010), peripheral vascular diseases (p = 0.009), chronic lung diseases (p < 0.001), ulcers (p = 0.003), diabetes without chronic complications (p = 0.005), renal disease (p = 0.030), and acquired immune deficiency syndrome/human immunodeficiency virus (AIDS/HIV) (p = 0.028) than the naïve group. However, none of the differences were significant when compared to patients in the switch (≥ 2) group (p > 0.05). The prevalence of hemiplegia or paraplegia (0.57%) was found only in naïve patients. A higher percentage of naïve patients (naïve group 37.14% vs overall switch group 22.47%) were receiving concomitant high dose MTX (> 8 mg/week), while more switched patients received no MTX (MTX free, 41.46%) or low dose (≤ 8 mg/week) of concomitant MTX (36.08%). A high proportion of naïve patients (80.29% vs overall switch group 70.02%) used 50 mg of GLM as their first dose, while a high proportion of switched patients (overall switch group 29.35% vs naïve group 19.23%) used 100 mg of GLM as their first dose.
GLM Persistence in Naïve vs Switch Groups
As per the Kaplan–Meier analyses, the persistence rate of GLM naïve patients was 58.78% (95% CI 56.73–60.77) at 1 year, 32.08% (95% CI 29.93–34.25) at 3 years, 21.42% (95% CI 18.96–23.99) at 5 years, and 11.36% (95% CI 7.54–16.02) at 7 years (Fig. 3). The persistence rate of GLM decreased with the number of drug switches compared to naïve patients. The persistence rate for switch (1) patients was 49.04% (95% CI 45.81–52.19) at 1 year, 30.29% (95% CI 27.17–33.45) at 3 years, 18.76% (95% CI 15.49–22.27) at 5 years, and 11.62% (95% CI 7.70–16.41) at 7 years. In switch (≥ 2) patients, the rates were 45.34% (95% CI 39.83–50.68%) at 1 year, 19.12% (95% CI 14.56–24.15) at 3 years, 13.98% (95% CI 9.49–19.31) at 5 years, and 5.24% (95% CI 1.40–13.05%) at 7 years (Fig. 3). The Kaplan–Meier analysis showed longer persistence in the GLM naïve group compared with the switch (1) group (p < 0.001). Similar significant difference was also observed when the switch (1) group was compared with the switch (≥ 2) group (p < 0.001). The median time to discontinuation of GLM was shorter in the switch groups compared to the naïve group [naïve, 529 days; switch (1), 358 days; switch (≥ 2), 319 days]. The time to discontinuation of GLM, especially in the naïve group (529 days), was comparable or longer than the reference cohort that includes other bDMARDs and JAK inhibitors (329 days) (Fig. 3).
Effect of Baseline Factors on GLM Persistence
The Cox HR analysis was used to examine baseline demographic and clinical factors affecting GLM persistence at each GLM administration timing (Table 2). The GLM persistence rate was significantly increased by age. In the GLM-naïve group patients aged between 61 and 75 years had an increased GLM persistence rate (HR 0.84, p = 0.008). However, age had no significant effect on persistence rate in patients aged > 75 years or in the switch groups. Gender was associated with an increase in the GLM persistence rate in the naïve group (HR 0.86, p = 0.019), but there was no significant difference in switch groups. Both high-dose (> 8 mg/week) and low-dose (≤ 8 mg/week) concomitant use of MTX were significant factors in increasing the GLM persistence rate in the naïve and switch (1) groups. The HR for concomitant MTX high dose was 0.71 (p < 0.001) for the naïve group and 0.78 (p < 0.01) for switch (1) group. The HR for concomitant MTX low dose was 0.66 (p < 0.001) for the naïve group and 0.74 (p < 0.001) for the switch (1) group. CCI scores of 3–5 had higher hazards of medication discontinuation and thus were less persistent compared to the reference group in the naïve group (HR 1.19, p = 0.005) and switch group (1) (HR 1.18, p = 0.0048). Regarding the initial dose of GLM, a dose of 100 mg significantly reduced the GLM persistence rate in the switch (1) group compared to patients receiving a dose of 50 mg (HR 1.23, p = 0.022). There was no significant difference in the naïve group amongst the patients taking a dose of 100 mg or 50 mg. Overall, in the switch (≥ 2) group, there was no statistically significant influence of age, gender, CCI score (Supplementary Table S4), concomitant MTX, or initial dose of GLM on persistence rate.
Baseline Factors Predictive of > 5-Year Persistence with GLM
Logistic regression analysis was used to identify baseline factors associated with achieving GLM treatment persistence for more than 5 years (Table 3). Age (61–75 years), gender (HR 0.86, p = 0.002), concomitant MTX high dose (HR 0.73, p < 0.001), concomitant MTX low dose (HR 0.70, p < 0.001) were positively associated with GLM persistence for over 5 years. Conversely, CCI score of 3–5 (HR 1.17, p < 0.001) and score > 5 (HR 1.29, p = 0.010), an initial GLM dose of 100 mg (HR 1.14, p = 0.009), one prior biologic use (HR 1.14, p = 0.006) or two or more prior biologics use (HR 1.31, p < 0.001) were identified as factors negatively associated with GLM persistence for over 5 years.
Effect of Prior Biologics on GLM Persistence
To analyze whether the persistence of GLM was altered by the use of bDMARDs or JAK inhibitors prior to GLM, Kaplan–Meier analysis of each bDMARD or JAK inhibitor subgroup was performed for up to 7 years (Fig. 4). Although there were no significant differences in patient background between the subgroups (Table 4), the duration of GLM for each subgroup was as follows in descending order (Fig. 4): infliximab (IFX), 602 days (95% CI 398–882); adalimumab (ADA), 427 days (95% CI 259–693); abatacept (ABT), 355 days (95% CI 280–469); tofacitinib (TOF), 341 days (95% CI 28–not reached); etanercept (ETN), 266 days (95% CI 176–410); certolizumab pegol (CEL), 258 days (95% CI 154–772); tocilizumab (TCZ), 214 days (95% CI 154–302); and sarilumab (SAR), 196 days (95% CI 86–225).
GLM persistence rates by subgroup over time at 1, 3, 5, and 7 years were IFX (59%, 38%, 16%, 11%), ETN (45%, 28%, 20, 13%), ADA (54%, 54%, 24%, 13%), CEL (49%, 49%. 29%, 29%), TCZ (38%, 38%, 14%, 11%), SAR (0%, 0%, 0%, 0%), ABT (48%, 48%, 18%, 9%), TOF (47%, 47%, 31%, 31%), and BAR (80%, 80%, 53%, 53%), respectively.
Furthermore, using the IFX subgroup with the longest persistence rate as a reference, the log-rank test with Bonferroni correction was used to analyze the persistence rate of each subgroup up to 7 years. The results showed significant differences in TCZ (p = 0.001), SAR (p = 0.025), and TOF (p = 0.041), indicating that switching from these drugs to GLM significantly reduced the persistence rate compared to switching from IFX to GLM (Fig. 4). After switching from other drugs to GLM, the persistence rate was lower in groups with high MTX-free rates (Table 4). In particular, in the naïve and switch (1) groups, the persistence of the GLM with MTX was improved (Fig. 5). Amongst the GLM naïve population, the persistence rate in the MTX high group was 63.92%, 34.85%, 21.58%, and 10.66%; in the MTX low group, it was 63.89%, 38.41%, 29.22%, and 21.65% when compared to MTX-free group (47.3%, 21.89%, 12.47%, and 3.57%) at 1, 3, 5, and 7 years, respectively. Similarly, in the switch (1) population, the persistence rate in the MTX high group was 54.02%, 37.01%, 21.8%, and 9.91%; in the MTX low group, it was 53.52%, 35.36%, 27.13%, and 18.41% when compared to the MTX-free group (41.6%, 21.07%, 9.21%, and 9.21%) at 1, 3, 5, and 7 years, respectively (Supplementary Table S4).
Discussion
The current analysis of the GLM persistence in patients with RA provides an update of 7 years of long-term hospital claims cohort data that has not previously been reported in Japan. In a previous study in naïve Japanese patients, MDV data was analyzed, and high RA persistence rates for GLM (100%) were reported at 1 year; however, this study included a shorter prescription history of 3 years (2012–2014), only 1.5 years of GLM follow-up period, and small sample size (GLM naïve, n = 27; switch, n = 87) [15]. In contrast, a large population of patients with RA from long-term prescription records for the period of 13 years (2008–2020) was analyzed in this study. This study demonstrated that the persistence rates for the GLM naïve or switch (1) patients were approximately 60% in 1 year, 30% in 3 years, 20% in 5 years, and 10% in 7 years, and with these GLM subgroups the persistence rates were found to be significantly improved compared to all bDMARDs or JAK inhibitors subgroup (Fig. 3). These persistence rates were comparable to or lower than those of reported in Japanese and global studies [7, 8, 20,21,22,23]. However, in this study, the persistence rate for GLM was comparable to the reference cohort, which included all other bDMARDs and JAK inhibitors that have recently been approved in Japan as oral agents (Fig. 3). Furthermore, a 6-year retrospective observational study of GLM in Japan reported the persistence rate of GLM as 66.3%, 48.3%, and 24.5% at 12, 36, and 72 months, respectively [24], which are in agreement with the persistence rates observed in this study. Therefore, the results presented here may be considered as a real-world indicator of the GLM persistence in Japan.
The possible reasons for the variation in persistence rates in the Japanese studies using the same database includes changes in medical technology over time, changes in guidelines, and other medical information such as remission criteria and timing of judgment, as well as an increase in treatment choices [25]. Previous reports using MDV limited the choice of biologics to five drugs (GLM, ETN, ADA, IFX, TCZ, and ABT). In this analysis, we have extended the drug list with the addition of data on SAR, CPZ, and JAK inhibitors (TOF and BAR). Additionally, advances in methods for assessing disease activity in RA and setting treatment goals that aim for higher treatment satisfaction and deeper remission have resulted in earlier judgments of drug effectiveness and shorter duration of each drug.
Various randomized clinical trials assessed the GLM persistence rates in the Japanese population. The 3-year persistence rate of GLM in the study of MTX-IR patients (GO-FORTH) was 73.2% [23], and the 120-week persistence rate in the study of csDMARD-IR patients (GO-MONO) was 73.1% [22]. Also, a 24-week PMS study has reported the GLM persistence after switching from other biologics with persistence rates of 84.4%, 78.2%, 72.9%, 72.7%, and 72.3% for IFX, ADA, ABT, ETN, and TCZ, respectively [26].
In this study, the factors affecting the long-term use of GLM were age (61–75 years), gender (female), and the concomitant use of MTX as extension factors. On the contrary, CCI score of ≥ 3, initial GLM dose of 100 mg, and the number of previous biologics switches were shortening factors (Tables 2 and 3). It has been shown that younger patients generally tend to have a higher persistence rate of TNF agents than elderly patients [27, 28], and in a previous MDV analysis, both gender and CCI showed a similar trend in comparison with bDMARDs overall [15]. Although other studies have shown that CCI has no effect on the persistence of biologic agents [17, 29], it is possible that many clinicians were attempting to make the best possible use of one agent because of the limited drug options available at the time. In the current analysis, the presence or absence of concomitant MTX and the initial dose of GLM were newly identified as factors affecting the persistence of GLM. In the Japanese PMS study, GLM combined with MTX clearly improved the disease activity index (DAS28-ESR [erythrocyte sedimentation rate]) and the persistence rate of GLM in patients with RA [30], and the current study supported these results. Although the same PMS study showed that the dose of GLM (50 or 100 mg) has no effect on the persistence rate [30], the findings of the current study were based on the initial GLM dose, and the persistence rate may have been affected by the patient’s condition at the start of GLM administration. In general, patients who started high-dose administration were expected to have a more severe condition. Furthermore, since GLM has already been approved for use at 100 mg without MTX in Japan, it is possible that the proportion of patients with serious complications was higher than in the group starting with 50 mg. However, we were unable to identify the precise cause of the difference.
It is commonly known that the initial choice of bDMARD influences the clinical response to the next drug when switching bDMARDs, considering the principles of the treat-to-target strategy [9, 31]. With reference to the previous subanalysis of PMS [26], we further provided real-world data on GLM persistence rates in patients with RA who had received primary treatment with another bDMARDs or JAK inhibitors prior to GLM prescription. In a subanalysis of GLM PMS, improvements in clinical signs and symptoms (DAS28-CRP, DAS28-ESR, SDAI [Simplified Disease Activity Index], and CDAI [Clinical Disease Activity Index] scores) were observed in the entire patient population [26]. Several other studies have reported that the reason for discontinuation of the first bDMARD (inefficiency and intolerance) may affect the effectiveness of the second bDMARD [10, 11, 32]. In addition, patients who discontinued the first bDMARD because of adverse events are more likely to respond to the second bDMARD than those who discontinued because of effectiveness-related reasons [13], posing an inherent challenge in interpreting effectiveness and persistence. Several observational cohort studies have demonstrated the effectiveness of switching from one TNF inhibitor to another in actual clinical practice [33,34,35]. Although evidence is lacking on which bDMARD treatment pattern is most effective, a subanalysis of the PMS that stratified primary bDMARDs prior to switching to GLM was conducted to assess clinical effectiveness, remission rates, and persistence. The results showed that GLM was effective regardless of the type of previous bDMARDs. It was also shown that the persistence rate was the highest when switching from IFX to GLM and was lower when switching from ETN to GLM [26]. Similarly, in the current MDV cohort study, the highest persistence rate was observed when switching from IFX to GLM, and this was confirmed to be reproducible (Fig. 4). Although our analysis does not identify compelling reasons for these results, further research investigating the presence of antibody levels and anti-drug antibodies after IFX treatment and their impact on outcomes would be beneficial to better understand the mechanisms underlying improved drug response and persistence to subsequent bDMARDs. Additionally, it was noteworthy that in previous subanalyses of PMS, patients switched from ETN to GLM exhibited lower persistence rates than those who switched from IFX, despite having similar baseline characteristics. On the other hand, the results from our study showed that the subgroup of patients who switched to GLM from drugs with different MoA, such as IL-6 inhibitors (TCZ [p = 0.001], SAR [p = 0.025]) or JAK inhibitor (TOF [p = 0.041]), had a significantly lower retention rate than the group switching from IFX. Furthermore, persistence rate after switching from other drugs to GLM in each group was relatively low in the subgroup with a high MTX-free rate (Table 4). A previous PMS study of GLM reported improved persistence in the patient using concomitant MTX irrespective of GLM (50 mg or 100 mg) dose [28]. Unlike clinical trials, in real-world data, GLM may not be used concomitantly with MTX depending on the patient’s clinical conditions, which could be the reason for the relatively low persistence in real practice.
Considering the increased number of drug options, switching to bDMARDs with a different MoA is now proposed as a therapeutic option. Also, several lines of evidence from randomized controlled trials [36], retrospective [37,38,39], and prospective cohort studies [39] suggest that switching to a non-TNF inhibitor may be more effective than looping TNF inhibitors in patients with RA who have not responded adequately to TNF inhibitors.
Conversely, there is limited evidence that TNF inhibitors can adequately control disease activity in patients with RA who have had an inadequate response to non-TNF bDMARDs [40,41,42]. Although the effectiveness of switching from non-TNF bDMARDs to GLM has been reported in patients with RA who had inadequate response to TCZ [40], we believe that further analysis of the effectiveness of switching from other non-TNF bDMARDs to GLM in actual clinical practice is needed in the future.
Limitations
Several limitations of this study merit comment. The first limitation was the heterogeneous patient population across GLM subgroups. Particularly, among the non-TNF subgroups, the subgroups of anti-IL-6 agents (SAR) and JAK inhibitors (TOF and BAR), which are relatively new therapies, have a much smaller population than the TNF subgroup, which may compromise the statistical power of this study. Furthermore, limitations exist because of the nature of the database which did not allow for the tracing of typical measures of disease activity such as DAS28, the inability to determine the reason for discontinuation of each drug, and the inability to follow up if a patient is transferred. The MDV database includes both inpatient and outpatient data unless the patient has stopped visiting the hospital. Thus, once the patients leave one hospital, data is not available for evaluation, and it remains unclear if the patient has been lost to follow-up or discontinued the treatment. However, the short evaluation period of 24 weeks, which was a limitation in the GLM PMS subanalysis study, has been sufficiently improved in this study. While real-world data provide valuable insights into the context of more drug options and various medical advances, as well as historical background over time, the data presented in this report needs to be understood and interpreted with caution.
Conclusions
This study demonstrated a high GLM persistence rate with naïve treatment and after switching from other bDMARDs or JAK inhibitor in a real-world clinical setting with a long follow-up period and a large sample size. The long-term persistence rate of GLM was comparable to that reported in real-world clinical practice, but lower persistence rates were observed in patients who switched drugs more frequently than in naïve patients. Age (61–75 years), gender (female), and concomitant use of MTX were identified as factors associated with long-term persistence of GLM, whereas high CCI score, initial GLM dose (100 mg), and history of use of other biologic agents were shown to be factors associated with shorter persistence rates. Overall, this real-world evidence can serve as a foundation to develop future treatment strategies; however, further research is needed to optimize specific drug and patient selection.
References
Sparks JA. Rheumatoid Arthritis. Ann Intern Med. 2019;170(1):ITC1–ITC16.
Kojima M, Nakayama T, Tsutani K, et al. Epidemiological characteristics of rheumatoid arthritis in Japan: prevalence estimates using a nationwide population-based questionnaire survey. Mod Rheumatol. 2020;30(6):941–7.
Nakajima A, Sakai R, Inoue E, Harigai M. Prevalence of patients with rheumatoid arthritis and age-stratified trends in clinical characteristics and treatment, based on the National Database of Health Insurance Claims and Specific Health Checkups of Japan. Int J Rheum Dis. 2020;23(12):1676–84.
Smolen JS, Landewé R, Breedveld FC, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs. Ann Rheum Dis. 2010;69(6):964–75.
Curtis JR, Singh JA. Use of biologics in rheumatoid arthritis: current and emerging paradigms of care. Clin Ther. 2011;33(6):679–707.
Listing J, Kekow J, Manger B, et al. Mortality in rheumatoid arthritis: the impact of disease activity, treatment with glucocorticoids, TNFα inhibitors and rituximab. Ann Rheum Dis. 2015;74(2):415–21.
Keystone EC, Genovese MC, Klareskog L, et al. Golimumab, a human antibody to tumour necrosis factor alpha given by monthly subcutaneous injections, in active rheumatoid arthritis despite methotrexate therapy: the GO-FORWARD Study. Ann Rheum Dis. 2009;68(6):789–96.
Smolen JS, Kay J, Doyle MK, et al. Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor alpha inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial. Lancet. 2009;374(9685):210–21.
Favalli EG, Raimondo MG, Becciolini A, Crotti C, Biggioggero M, Caporali R. The management of first-line biologic therapy failures in rheumatoid arthritis: current practice and future perspectives. Autoimmun Rev. 2017;16(12):1185–95.
Karlsson JA, Kristensen LE, Kapetanovic MC, Gülfe A, Saxne T, Geborek P. Treatment response to a second or third TNF-inhibitor in RA: results from the South Swedish Arthritis Treatment Group Register. Rheumatology (Oxford). 2008;47(4):507–13.
Bombardieri S, Ruiz AA, Fardellone P, et al. Effectiveness of adalimumab for rheumatoid arthritis in patients with a history of TNF-antagonist therapy in clinical practice. Rheumatology (Oxford). 2007;46(7):1191–9.
Lloyd S, Bujkiewicz S, Wailoo AJ, Sutton AJ, Scott D. The effectiveness of anti-TNF-alpha therapies when used sequentially in rheumatoid arthritis patients: a systematic review and meta-analysis. Rheumatology (Oxford). 2010;49(12):2313–21.
Rendas-Baum R, Wallenstein GV, Koncz T, et al. Evaluating the efficacy of sequential biologic therapies for rheumatoid arthritis patients with an inadequate response to tumor necrosis factor-α inhibitors. Arthritis Res Ther. 2011;13(1):R25.
Shimizu H, Kobayashi H, Kanbori M, Ishii Y. Clinical response among golimumab-treated Japanese patients with rheumatoid arthritis by number of previous biologic therapies: real-world evidence from post-hoc analysis of post-marketing surveillance data. Mod Rheumatol. 2021;31(3):566–74.
Mahlich J, Sruamsiri R. Persistence with biologic agents for the treatment of rheumatoid arthritis in Japan. Patient Prefer Adherence. 2016;10:1509–19.
Blum MA, Koo D, Doshi JA. Measurement and rates of persistence with and adherence to biologics for rheumatoid arthritis: a systematic review. Clin Ther. 2011;33(7):901–13.
Neubauer S, Cifaldi M, Mittendorf T, Ganguli A, Wolff M, Zeidler J. Biologic TNF inhibiting agents for treatment of rheumatoid arthritis: persistence and dosing patterns in Germany. Heal Econ Rev. 2014;4(1):32.
Wu E, Chen L, Birnbaum H, Yang E, Cifaldi M. Retrospective claims data analysis of dosage adjustment patterns of TNF antagonists among patients with rheumatoid arthritis. Curr Med Res Opin. 2008;24(8):2229–40.
Yeaw J, Benner JS, Walt JG, Sian S, Smith DB. Comparing adherence and persistence across 6 chronic medication classes. J Manag Care Pharm. 2009;15(9):728–40.
Emery P, Keystone E, Tony HP, et al. IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial. Ann Rheum Dis. 2008;67(11):1516–23.
Okazaki M, Kobayashi H, Ishii Y, Kanbori M, Yajima T. Real-world treatment patterns for golimumab and concomitant medications in Japanese rheumatoid arthritis patients. Rheumatol Ther. 2018;5(1):185–201.
Takeuchi T, Harigai M, Tanaka Y, et al. Clinical efficacy, radiographic, and safety results of golimumab monotherapy in Japanese patients with active rheumatoid arthritis despite prior therapy with disease-modifying antirheumatic drugs: final results of the GO-MONO trial through week 120. Mod Rheumatol. 2018;28(5):770–9.
Tanaka Y, Harigai M, Takeuchi T, et al. Clinical efficacy, radiographic progression, and safety through 156 weeks of therapy with subcutaneous golimumab in combination with methotrexate in Japanese patients with active rheumatoid arthritis despite prior methotrexate therapy: final results of the randomized GO-FORTH trial. Mod Rheumatol. 2016;26(4):481–90.
Mutoh T, Nagai T, Shirai T, Okazaki S, Sato H, Fujii H. Predictive factors for retention of golimumab over a median 4-year duration in Japanese patients with rheumatoid arthritis in a real-world setting: a retrospective study and literature review. Int J Rheumatic Dis. 2022;25(3):335–43.
Fraenkel L, Bathon JM, England BR, et al. 2021 American College of Rheumatology guideline for the treatment of rheumatoid arthritis. Arthritis Rheumatol. 2021;73(7):1108–23.
Shimizu H, Kobayashi H, Kanbori M, Ishii Y. Effectiveness of golimumab in rheumatoid arthritis patients with inadequate response to first-line biologic therapy: results from a Japanese post-marketing surveillance study. Mod Rheumatol. 2021;31(3):556–65.
Cho SK, Sung YK, Choi CB, Bae SC. Impact of comorbidities on TNF inhibitor persistence in rheumatoid arthritis patients: an analysis of Korean National Health Insurance claims data. Rheumatol Int. 2012;32(12):3851–6.
Matsubara H, Kojima T, Kaneko A, et al. Longterm retention rate and risk factor for discontinuation due to insufficient efficacy and adverse events in Japanese patients with rheumatoid arthritis receiving etanercept therapy. J Rheumatol. 2014;41(8):1583–9.
Kang JH, Park DJ, Lee JW, et al. Drug survival rates of tumor necrosis factor inhibitors in patients with rheumatoid arthritis and ankylosing spondylitis. J Korean Med Sci. 2014;29(9):1205–11.
Okazaki M, Kobayashi H, Shimizu H, Ishii Y, Yajima T, Kanbori M. Safety, effectiveness, and treatment persistence of golimumab in elderly patients with rheumatoid arthritis in real-world clinical practice in Japan. Rheumatol Ther. 2018;5(1):135–48.
Smolen JS, Aletaha D, Barton A, et al. Rheumatoid arthritis. Nat Rev Dis Primers. 2018;4:18001.
Smolen JS, Kay J, Matteson EL, et al. Insights into the efficacy of golimumab plus methotrexate in patients with active rheumatoid arthritis who discontinued prior anti-tumour necrosis factor therapy: post-hoc analyses from the GO-AFTER study. Ann Rheum Dis. 2014;73(10):1811–8.
Furst DE, Gaylis N, Bray V, et al. Open-label, pilot protocol of patients with rheumatoid arthritis who switch to infliximab after an incomplete response to etanercept: the opposite study. Ann Rheum Dis. 2007;66(7):893–9.
Schiff MH, von Kempis J, Goldblum R, Tesser JR, Mueller RB. Rheumatoid arthritis secondary non-responders to TNF can attain an efficacious and safe response by switching to certolizumab pegol: a phase IV, randomised, multicentre, double-blind, 12-week study, followed by a 12-week open-label phase. Ann Rheum Dis. 2014;73(12):2174–7.
Weinblatt ME, Fleischmann R, van Vollenhoven RF, et al. Twenty-eight-week results from the REALISTIC phase IIIb randomized trial: efficacy, safety and predictability of response to certolizumab pegol in a diverse rheumatoid arthritis population. Arthritis Res Ther. 2015;17:325.
Gottenberg JE, Brocq O, Perdriger A, et al. Non-TNF-targeted biologic vs a second anti-TNF drug to treat rheumatoid arthritis in patients with insufficient response to a first anti-TNF drug: a randomized clinical trial. JAMA. 2016;316(11):1172–80.
Favalli EG, Biggioggero M, Marchesoni A, Meroni PL. Survival on treatment with second-line biologic therapy: a cohort study comparing cycling and swap strategies. Rheumatology (Oxford). 2014;53(9):1664–8.
Bergman MJ, Elkin EP, Ogale S, Kamath T, Hamburger MI. Response to biologic disease-modifying anti-rheumatic drugs after discontinuation of anti-tumor necrosis factor alpha agents for rheumatoid arthritis. Rheumatol Ther. 2014;1(1):21–30.
Harrold LR, Reed GW, Kremer JM, et al. The comparative effectiveness of abatacept versus anti-tumour necrosis factor switching for rheumatoid arthritis patients previously treated with an anti-tumour necrosis factor. Ann Rheum Dis. 2015;74(2):430–6.
Matsuno H, Katayama K. Effectiveness of golimumab for rheumatoid arthritis in patients with an inadequate response to tocilizumab. Mod Rheumatol. 2017;27(2):246–51.
Akiyama M, Kaneko Y, Kondo H, Takeuchi T. Comparison of the clinical effectiveness of tumour necrosis factor inhibitors and abatacept after insufficient response to tocilizumab in patients with rheumatoid arthritis. Clin Rheumatol. 2016;35(11):2829–34.
Umeda M, Koga T, Ichinose K, et al. Efficacy of infliximab as a switched biologic in rheumatoid arthritis patients in daily clinical practice. Immunol Med. 2018;41(4):181–6.
Acknowledgements
Funding
Sponsorship for this study and its publication, including the journal’s Rapid Service Fee were provided by Janssen Pharmaceutical K.K. (Tokyo, Japan).
Medical Writing, Editorial, and Other Assistance
Medical writing support, under the guidance of the authors, was provided by Kalpana Tilekar (PhD) from SIRO Clinpharm Pvt. Ltd, India and was funded by Janssen Pharmaceutical K.K. The authors thank Seiji Yokoyama for supporting the first draft of manuscript and Shigefumi Tsukamoto (EPS Corp.) for performing the statistical analysis. Statistical analysis support was funded by Janssen Pharmaceutical K.K.
Authorship
All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.
Author Contributions
All authors contributed to the study conception and design. The first draft of the manuscript was written by Masahiko Miyashiro. Yutaka Ishii, Celine Miyazaki, Hirohito Shimizu and Junya Masuda reviewed the manuscript. All authors read and approved the final manuscript.
Disclosures
Masahiko Miyashiro and Hirohito Shimizu are employees of Janssen Pharma K.K., a wholly owned subsidiary of Johnson & Johnson. Celine Miyazaki, Yutaka Ishii and Junya Masuda are employees of Janssen Pharmaceutical K.K., a wholly owned subsidiary of Johnson & Johnson, and may hold stock and/or stock options in the company.
Compliance with Ethics Guidelines
All the necessary permissions were obtained to use the database. The data in this study were de-identified by the database provider; therefore, no informed consent was necessary. The study was in line with the guidelines of Johnson & Johnson and approved by the Janssen approval committee and conducted according to the Declaration of Helsinki.
Data Availability
We used commercially available hospital claims data from Medical Data Vision Co., Ltd., the data set supporting the conclusions is not publicly available.
Author information
Authors and Affiliations
Corresponding author
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc/4.0/.
About this article
Cite this article
Miyashiro, M., Ishii, Y., Miyazaki, C. et al. A Real-World Claims Database Study Assessing Long-Term Persistence with Golimumab Treatment in Patients with Rheumatoid Arthritis in Japan. Rheumatol Ther 10, 615–634 (2023). https://doi.org/10.1007/s40744-023-00539-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40744-023-00539-z