PDF
Citation
Print
INTRODUCTION
Rheumatoid arthritis (RA) is related to autoantibodies and several autoantibodies are used for diagnosis or evaluation of disease activity of RA [1]. Rheumatoid factor (RF) is positive in 70% of patients with RA, and RF is concerned with immune complexes that affect the disease pathogenesis [2]. Anti-citrullinated peptide antibodies (ACPA) are highly specific for RA [3]. The serum titer of anti-agalactosyl immunoglobulin (Ig) G antibody (anti-Gal (0) IgG) is also useful biomarker for RA diagnosis [4]. Abatacept (ABT) is a fusion protein that prevent the CD80/CD86 antigen on the antigen-presenting cell and CD28 antigen on the T-cell [5]. Human leukocyte antigen (HLA) strongly influence T-cell recognition [6]. The shared epitope containing HLA-DRB1 alleles contribute to the development of ACPA [7]. Previous study reported that one year treatment of ABT decreased serum levels of ACPA and RF in patients with early RA [8]. Japanese study reported that significantly larger decreases in ACPA and RF titer were observed in ABT compared to anti-tumor necrosis factor (anti-TNF) therapy at 3 months [9]. However, other studies showed that autoantibody titers decreased after 24 weeks treatment with biological disease-modifying anti-rheumatic drugs (bDMARDs) other than ABT such as interleukin-6 (IL-6) receptor inhibitor [10]. Anti-TNF inhibitor also reduced autoantibody titers after 46 weeks treatment [11].
It is possible that bDMARDs not only reduce disease activity but also affect autoantibody change in RA patients. While various bDMARDs had the potential to decrease the autoantibody levels, it was unclear what factors really effected on autoantibody change in RA patients. In addition, most studies reporting the effect of ABT on lowering autoimmune antibody levels have been conducted on patients with early RA [8,9]. We hypothesized that different results might be obtained in patients with established RA. The objective of this study was to evaluate the impact of ABT and other bDMARDs on autoantibody levels in patients with established RA in the AIRTIGHT (Can abatacept induce immunological remission under tight control in rheumatoid arthritis patient?) study.
MATERIALS AND METHODS
Study protocol
The AIRTIGHT study was a prospective, comparative non-randomized study to assess the clinical and immunological impacts of ABT and other bDMARDs on patients with established RA. Patients were treated at Osaka City University Hospital, Hamayu Hospital, or Kitade Hospital in Japan between May 2011 and May 2014. The primary endpoints were changes in autoantibody levels (ACPA, IgM-RF, IgG-RF, and anti-Gal (0) IgG) after 48 weeks treatment. This study was an investigator-initiated trial and treating physicians and patients were not blinded to the medication prescribed, but clinical parameters and outcomes were assessed during treatment by trained evaluators blinded to treatment groups and the clinical response of each patient.
Patients
Eligible patients were ≥20 years old, fulfilled the American College of Rheumatology/European League Against Rheumatism classification criteria [12], had been treated for ≥3 months with any conventional synthetic DMARDs, and required treatment with bDMARDs including switching. We continued to register patients in the AIRTIGHT study to the point of 50 patients started on ABT. A total of 165 RA patients were enrolled to the AIRTIGHT study. At baseline, 50 patients were treated with ABT and 115 patients were treated with other bDMARDs. ABT was administered intravenously with the dose determined by body weight every 4 weeks for a total of 48 weeks. Other bDMARDs were administered in compliance with approved dosing methods and dosages. This study was registered with the University Hospital Medical Information Network (UMIN) Clinical Trials Registry (UMIN000005570). This study was conducted after written informed consent was obtained from all study subjects in accordance with the Declaration of Helsinki, and the research protocol was approved by the Ethics Committee of Osaka City University Hospital (approval number: 2720, the name of the university is now changed to Osaka Metropolitan University).
Assessment
Serum samples were collected and stored at –80°C until measurements from all patients before administration of bDMARDs at baseline and after 48 weeks of treatment. All the serum samples of patients were analyzed in a single session, and were measured at the same time for avoiding the error of measurement. We used the chemiluminescence enzyme immunoassay for ACPA (STACIA MEBlux test for CCP; Medical & Biological Laboratories Co., Ltd., Nagano, Japan), latex agglutination turbidimetry for IgM-RF (IATRO RF II; LSI Medience Corp., Tokyo, Japan), enzyme immunoassay for IgG-RF (Tokiwa measuring kit; Tokiwa Chemical Industries Co., Ltd., Tokyo, Japan), and electrochemiluminescence immunoassay for Anti-Gal (0) IgG (PICOLUMI CA•RF; Eisai Co., Ltd., Tokyo, Japan). Cut-off value was 4.5 U/mL in ACPA, 15 IU/mL in IgM-RF, 2.0 IU/mL in IgG-RF and 6.0 AU/mL in Anti-Gal (0) IgG. All of measurements were performed at LSI Medience Corporation.
Propensity score matching
In this study, changes in autoantibody levels were compared between ABT group and non-ABT group. However, patients background such as sex, age, disease activity of RA and medication status were not matched. To overcome this problem, we used propensity score matching method. The propensity score matching method aims to balance the two groups with regard to their covariates. Sex, age, disease duration, disease activity score in 28 joints using C-reactive protein (DAS28-CRP), methotrexate dose, glucocorticoid dose, autoantibody levels (ACPA, IgM-RF, IgG-RF, and anti-Gal (0) IgG) at baseline, and previous bDMARDs use were chosen as matched factors.
Statistical analysis
Data were expressed as median or percentage. Comparisons of baseline characteristics between groups were performed using the Mann-Whitney U-test or chi-square test. The Wilcoxson signed rank test was used to compare changes in ACPA, IgM-RF, IgG-RF, and anti-Gal (0) IgG levels, DAS28-CRP, the Simplified Disease Activity Index (SDAI), and the Clinical Disease Activity Index (CDAI) between baseline and 48 weeks after treatment. Comparisons of average dose of methotrexate, and glucocorticoid after 48 weeks treatment between groups were performed using the Mann-Whitney U-test. Comparisons of achieving DAS28-CRP remission rate after 48 weeks treatment between groups were performed using the chi-square test. Statistical significance was determined for values of p<0.05. All analyses were performed using Excel Statistics 2021 (SSRI, Social Survey Research Information Co., Tokyo, Japan).
RESULTS
Patient characteristics
One hundred four patients with RA (ABT, n=32; infliximab, n=10; etanercept, n=14; adalimumab, n=20; tocilizumab, n=19; golimumab, n=9) completed the study. Percentages continuing treatment were 64.0% in the ABT group and 75.7% in the non-ABT group, showing no significant difference (p=0.14). Adverse events were reported in 3 patients (severe anemia, sinusitis, and hives in 1 patient each) treated with ABT, and in 5 patients (pneumonia in 2 patients, and cancer of the tongue, arrhythmia, and myelodysplastic syndrome in 1 patient each) treated with non-ABT. Nine patients treated with ABT and 14 patients treated with non-ABT were switched bDMARDs for inadequate response. After propensity score matching, 25 patients with ABT and 25 patients with non-ABT (infliximab, n=1; etanercept, n=6; adalimumab, n=7; tocilizumab, n=7; golimumab, n=4) were finally analyzed (Figure 1). Baseline demographics and clinical characteristics were comparable between the ABT and non-ABT groups (Table 1). Before propensity score matching, methotrexate dose was lower (p=0.02) and glucocorticoid dose was higher (p=0.04) in ABT group compared to non-ABT group, and half of the patients in ABT group had a history of previous use of bDMARDs.
Changes in autoantibody levels and disease activity
Median levels at baseline and 48 weeks were 62.7 and 57.8 U/mL for ACPA, 35.0 and 39.0 IU/mL for IgM-RF, 0.5 and 0.5 IU/mL for IgG-RF, and 50.4 and 53.5 AU/mL for anti-Gal (0) IgG, respectively, in ABT group (Figure 2). The changes of level in ACPA (p=0.22), IgM-RF (p=0.21), IgG-RF (p=0.19) and anti-Gal (0) IgG (p=0.22) were not significant after 48 weeks of ABT exposure. In non-ABT group, median levels at baseline and 48 weeks were 143.0 and 57.8 U/mL for ACPA (p=0.03), 50.0 and 37.0 IU/mL for IgM-RF (p<0.01), 0.6 and 0.5 IU/mL for IgG-RF (p=0.22), and 93.2 and 61.8 AU/mL for anti-Gal (0) IgG (p<0.01), respectively (Figure 2). In patients using anti-TNF inhibitors in non-ABT group (n=18), median levels at baseline and 48 weeks were 143.5 and 59.9 U/mL for ACPA (p=0.16), 87.0 and 43.5 IU/mL for IgM-RF (p<0.01), 0.6 and 0.6 IU/mL for IgG-RF (p=0.90), and 124.1 and 75.7 AU/mL for anti-Gal (0) IgG (p<0.01), respectively. In patients with tocilizumab in non-ABT group (n=7), median levels at baseline and 48 weeks were 143.0 and 46.2 U/mL for ACPA (p<0.05), 50.0 and 18.0 IU/mL for IgM-RF (p=0.35), 0.7 and 0.3 IU/mL for IgG-RF (p=0.09), and 61.6 and 19.0 AU/mL for anti-Gal (0) IgG (p=0.13), respectively. After 48 weeks of ABT treatment, no patient (0%) with ACPA, 1 patient (4.0%) with IgM-RF, 1 patient (4.0%) with IgG-RF, and 1 patient (4.0%) with anti-Gal (0) IgG showed seroconversion. In case of non-ABT treatment, no patient (0%) with ACPA, 3 patients (12.0%) with IgM-RF, 1 patient (4.0%) with IgG-RF, and no patient (0%) with anti-Gal (0) IgG showed seroconversion.
As shown in Figure 3, all composite measures of DAS28-CRP (Figure 3A), SDAI (Figure 3B), and CDAI (Figure 3C) significantly decreased after 48 weeks of treatment in both the ABT and non-ABT groups. Four patients in ABT group (16%) and 13 patients in non-ABT group (52%) achieved DAS28-CRP remission. The rate of achieving DAS28-CRP remission was significantly higher in non-ABT group than in ABT group (p=0.02). Even in patients who achieved DAS28-CRP remission with ABT treatment, decreases in ACPA (p=0.50), IgM-RF (p=0.50), and anti-Gal (0) IgG (p=0.46) were not significant, except IgG-RF (0.3 at baseline to 0.1 IU/mL at week 48, p=0.04). Similar results were obtained in the non-ABT group, where only IgM-RF (49.0 at baseline to 37.0 IU/mL at week 48, p<0.05) and anti-Gal (0) IgG (88.8 to 56.6 AU/mL, p=0.02) among the four autoantibodies showed a significant decrease even in patients who achieved remission.
Average dosage of methotrexate and glucocorticoid during 48 weeks treatment were 8.5±3.2 mg/weeks and 2.9±1.6 mg/day in ABT group, and 8.6±2.9 mg/weeks and 2.0±0.9 mg/day in non-ABT group. There were no differences in average dose of methotrexate (p=0.58) and glucocorticoid (p=0.37) between groups.
DISCUSSION
Although previous studies such as ADJUST (Abatacept study to Determine the effectiveness in preventing the development of rheumatoid arthritis in patients with Undifferentiated inflammatory arthritis and to evaluate Safety and Tolerability) showed that treatment with ABT decreased serum levels of ACPA and RF in very early RA patients [8,9], our study revealed that autoantibody levels did not decrease during ABT treatment in patients with RA. Possible explanation for this inconsistency could be that subjects in our study were patients with established RA, whereas the ADJUST trial targeted patients with very early RA [8]. The autoimmune nature of RA has been clearly established, with involvement of both innate and adaptive immune systems. Inflammation of RA was involved multiple lymphoid cells in innate and adaptive responses. It was possible that these autoimmune systems change depending on time elapsed since onset. Sustained RF expression likely activates B-cells and monocytes, and cytokine profiles switched from T-cell to B-cell cytokines over 5 years [13]. Other study showed that T-cell cytokines are elevated in the synovial fluid of patients with RA within 3 months after symptom onset, but this profile was no longer present in established RA [14]. Previous longitudinal study that examined the fluctuations of ACPA levels for five years revealed that significant drop in ACPA levels was observed during the first year [15]. Those results suggest that T-cells play an important role particularly at early stage in RA. The ability of ABT to modulate the activation of T-cells, including naive T-cells, and the role of T-cells in initiating disease [6]. The effect of ABT on autoantibody may be more effective in the early stage of RA when T-cell is significantly activated.
Our study showed that autoantibody levels decreased during treatment with bDMARDs other than ABT. Cytokines were associated with disease duration of RA [13]. In ACPA positive RA patients, IL-6 and TNF begin to increase at pre-clinical phase and remain high even in chronic phase [16]. A pilot study showed that tocilizumab reduced ACPA levels for 24 weeks treatment in RA patients with not early stage (median disease duration: 3.6 years) [10]. Previous study demonstrated that tocilizumab treatment affected the frequency of pre-switch and post-switch memory B-cells in patients with RA [17]. The change in post-switch memory B-cells was negatively correlated with the change in naïve B-cells during treatment with tocilizumab, suggesting that tocilizumab has a potential to reduce ACPA production by affecting the balance of B-cell subpopulation in circulation and in lymphoid tissues [10]. IL-6 inhibitor has a potential to reduce autoantibody levels by affecting the distribution of B-cell subsets in circulation.
ACPA and IgM-RF decreased significantly in established RA patients (median disease duration: 10 years) with infliximab treatment for 1 year [11]. They reported that there was no relationship between change of autoantibody levels and disease activity or disease duration. The effects of anti-TNF inhibitor in RA have been attributed to the induction of apoptosis of inflammatory cells [18]. Citrullination is a post-translational modification of proteins in the apoptotic process and therefore anti-TNF could influence the formation and presence of antibodies against citrullinated proteins by directly interfering with the regulation of the apoptotic process [11]. In addition, previous study reported that anti-TNF inhibitor affected B-cell phenotype and IgD-CD27- memory B-cells in circulation [19]. The frequency of IgD-CD27- memory B-cells in circulation as well as IgM expression by IgD-CD27- memory B-cells were significantly decreased after anti-TNF inhibitor treatment in RA patients [19]. Considering these mechanisms, anti-TNF inhibitor has a potential to reduce autoantibody levels in established RA.
Some limitations must be considered when interpreting our findings. First, eligible patients were not allocated to groups randomly. It was possible that baseline imbalance between groups influenced the results even if adjustments were made by propensity score matching. Second, patient numbers were relatively small and small numbers of patients remained in the study over time. Therefore, we should work on further studies such as randomized controlled trials and large-scale ones. Finally, our analysis did not take into account the changes in concomitant medications or dosage changes during the observation period. Prospective studies considering time-dependent covariates should be planned in the future.
XML Download