This multicenter study was conducted at Kindai University Hospital, Osaka Medical and Pharmaceutical University, Zenjinkai Miyazaki Hospital, Tenri Hospital, and Niigata Rheumatic Center, all of which are in Japan’s Niigata prefecture. We collected the clinical and laboratory data of the 76 RA patients for whom the REMIQ kit was applied at 14 weeks after their initial infliximab induction during the period 2015–2021.

To be eligible for the study, the patient’s diagnosis of RA had to be based on the American College of Rheumatology (ACR) criteria [7] before 2009 and the 2010 ACR and EULAR classification criteria [8] after 2010. The patients were also screened for latent and active tuberculosis. Patients with other connective tissue disease with joint symptoms or had a dose escalation within 14 weeks of infliximab induction were excluded. The infliximab dose escalation protocol is to increase the dose from 3 mg/kg to 6 mg/kg and 10 mg/kg, with dosing intervals of 8 weeks after week 14.

All of the patients were treated with infliximab + MTX at a maximum dose of 4–12 mg/week. The dose assignments were determined by the patients’ treating physicians. A patient’s dose of prednisolone (PSL) could be decreased when treatment efficacy was observed. The follow-up period was 54 weeks, starting at the patient’s initial dose of infliximab 3 mg/kg.

Table 1 summarizes the patients’ background. The group clinical data recorded at baseline and at weeks 14, 24, and 54 included the following: the 28 swollen joint count and tender joint counts (SJC28 and TJC28); a patient visual analogue scale for pain (Pt-VAS) in which 0 = best and 100 = worst; the Steinbrocker stage and class, and the Health Assessment Questionnaire Disability Index (HAQDI). We obtained the following laboratory data: the rheumatoid factor (RF) and anti-citrullinated protein/peptide antibody (ACPA) titers and positivity, the C-reactive protein (CRP) value, the erythrocyte sedimentation rate (ESR), and the metalloproteinase-3 (MMP-3) value. The DAS28 using the ESR (DAS28-ESR) and Clinical Disease Activity Index (CDAI) were obtained from the patient’s clinical examination and calculated with/without laboratory data. With respect to the DAS28-ESR disease activity categories, we divided the patients into those at remission (DAS28-ESR 2.6) and those in non-remission (DAS28-ESR low: $\ge$2.6 to 3.2, moderate: 3.2–5.1, high: $>$5.1). In the the REMIQ kit, an infliximab blood concentration of $\ge$1.0 $\mu$g/mL was defined as positive, and a concentration 1.0 $\mu$g/mL was defined as negative.

The endpoint for clinical response was the DAS28-ESR in the REMIQ-positive (n = 46) and REMIQ-negative (n = 30) groups at baseline and 14, 24, and 54 weeks after the start of infliximab treatment. The EULAR response [9] was also evaluated at 14, 24, and 54 weeks. The treatment efficacy parameters included the proportion of patients with a complete set of data of the DAS28-ESR, CDAI, and TJC28 and SJC28 (both of which were determined by the patient’s treating physician), the Pt-VAS, and the laboratory parameters of CRP (mg/dL), ESR (mm/hr), RF (U/mL), and MMP-3 (ng/mL) assessed at every visit during the follow-up period. The patient’s physical function at baseline was evaluated based on his or her HAQDI score [9].

The treating physicians recorded adverse events (AEs) if any occurred and made treatment adjustments at his/her discretion. AEs were defined as any adverse reaction associated with the infliximab treatment.

Summary statistics of the mean $\pm$ standard deviation (SD) or the median and interquartile range (IQR) are presented for continuous variables as appropriate. Categorical variables are presented as percentages. Comparisons between independent means were analyzed using the Mann-Whitney test and paired t-test. The retention rates of the REMIQ-positive and -negative groups were assessed by the Kaplan-Meier method, and the difference in retention curves was examined by log-rank test. Multivariate logistic regression was applied in a Cox regression analysis of the achievement of low disease activity and remission in order to calculate the hazard ratios (HRs) and 95% confidence intervals (95% CIs) including the REMIQ test result at 14 weeks, the CRP and ESR values as inflammatory biomarkers, and RF and ACPA as immunological tests at baseline [10].

A receiver-operator characteristic (ROC) curve was constructed using CRP and ESR at baseline and 14 weeks in both the positive- and negative-REMIQ patients. Sensitivity, specificity, and predictive values were calculated using ROC curves. The last observation carried forward (LOCF) method was applied to assess the patients who discontinued the infliximab therapy [11]. p-values 0.05 were considered significant. Statistical analysis software (GraphPad Prism, GraphPad Software, San Diego, CA, USA) and JMP Statistical Software (SAS Institute, Cary, NC, USA) were used for the statistical analyses.

Table 1 summarizes the baseline characteristics, clinical assessments, laboratory results, and doses and/or the percentage of MTX and PSL for all patients (n = 76) and the REMIQ- positive (n = 46) and -negative groups (n = 30) at 14 weeks after their initial infliximab induction. The disease duration in the REMIQ-negative group (54.0 months) tended to be longer than that in the REMIQ-positive group (12.0 months). At baseline, the REMIQ-positive group had significantly lower CRP levels than the REMIQ-negative group (**p 0.01). The MTX doses also tended to be higher in the REMIQ-positive group (10.0 mg/week) compared to the -negative group (8.0 mg/week). The Steinbrocker class proportions were significantly different between the two groups (*p 0.05).

The RA patients’ retention rates at 54 weeks are illustrated in Fig. 1. The overall retention rate at 54 weeks based on withdrawal from infliximab treatment for any reason (e.g., lack of efficacy, remission, at the patient’s request due to economic problems or moving away from the area) or sustained remission is illustrated in Fig. 1A. The infliximab retention rates in the REMIQ- positive and -negative groups were 100.0% and 100.0% at 14 weeks, 100.0% and 90.0% at 24 weeks, 93.5% and 73.3% at 36 weeks, and 93.5% and 70.0% at 54 weeks, respectively. Thus, the overall retention rate at 54 weeks in the two groups was significantly higher for the REMIQ-positive group. Fig. 1B depicts the retention rate for infliximab 3 mg/kg at 54 weeks. The patients who received infliximab 3 mg/kg but whose dose was escalated by the treating physician’s decision due to an inadequate response were considered withdrawals. The REMIQ-positive and -negative groups’ retention rates for infliximab 3 mg/kg were 91.3% and 80.0% at 14 weeks, 87.0% and 26.7% at 24 weeks, 84.7% and 26.7% at 36 weeks, and 82.6% and 6.7% at 54 weeks (p 0.0005), respectively. The details of the infliximab dose escalation procedure in the REMIQ-positive and -negative groups are provided in Supplementary Figs. 1,2.

Fig. 1.

Retention rates of the 76 patients with rheumatoid arthritis (RA). Kaplan-Meier curves for the REMIQ-positive (n = 46) and REMIQ-negative (n = 30) groups treated with infliximab 3, 6, and 12 mg/kg regarding the lenghts of time to continue or withdraw for any reason, e.g., due to lack of efficacy, remission, or the patient’s request (A) and the length of time to continue at infliximab 3 mg/kg (B) from week 0 to week 54. **p 0.01 and ****p 0.0001, REMIQ-positive group vs. negative.

We analyzed the changes from the two groups’ baseline values in the clinical assessment. Fig. 2 illustrates the disease activity and clinical response at 14, 24, and 54 weeks: the DAS28-ESR score, the CDAI, the TJC28 and SJC28 values, and the Pt-VAS ratings. The DAS28-ESR score, CDAI, TJC28, and SJC28 of both the REMIQ-positive and -negative patients were significantly reduced compared to the baseline values. At 14 weeks, the DAS28-ESR, TJC, and Pt-VAS in the REMIQ-positive group were significantly reduced compared to those in the REMIQ-negative group. The REMIQ-positive patients also showed a decrease in Pt-VAS ratings at all observation points compared to the REMIQ-negative group.

Fig. 2.

Serial changes in the DAS28-ESR, CDAI, TJC28 (0–28 joints), SJC28 (0–28 joints), and Pt-VAS (cm) in the REMIQ-positive and -negative groups at baseline, 14, 24, 36, and 54 weeks. The LOCF method was applied to assess the group who discontinued the treatment. Comparisons between the two groups’ mean values were analyzed using the Mann-Whitney test: *p 0.05, REMIQ-positive vs. -negative. CDAI, Clinical Disease Activity Index; DAS28-ESR, Disease Activity Score assessing 28 joints with erythrocyte sedimentation rate; Pt-VAS, patient Visual Analogue Scale; SJC, swollen joint count; TJC, tender joint count.

The ESR and MMP-3 values at 14 and 54 weeks in the REMIQ-positive group were significantly reduced compared to those in the -negative group (Fig. 3). At 14, 24, and 54 weeks, the CRP, ESR, and MMP-3 of both the REMIQ-positive and -negative groups were significantly reduced compared to the baseline values. The RF titers at each observation period were reduced from the baseline values in both groups but not significantly different between the groups.

Fig. 3.

Serial changes in the C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and rheumatoid factor (RF) values in the REMIQ-positive and -negative groups at baseline, 14, 24, 36, and 54 weeks. The LOCF method was applied to assess the group who discontinued treatment. Comparisons between the two groups’ mean values were analyzed using the Mann-Whitney test: *p 0.05, REMIQ-positive vs. -negative.

When we evaluated the disease activity based on the DAS28 criteria at 14, 24, and 54 weeks (Fig. 4A), we observed that the proportions of patients achieving low disease activity including remission at 14 weeks was significantly increased for the DAS28 in the REMIQ-positive patients compared to the -negative patients (63.0% vs. 28.6%, respectively). At 24 and 54 weeks, the percentage of patients achieving low disease activity tended to increase with the escalating infliximab dose in the REMIQ-negative group at 24 and 54 weeks. The rate of responders (good response) at 14 weeks was 58.7% in the REMIQ-positive group but only 32.1% in the REMIQ-negative group (Fig. 4B). However, the percentage of responders (good response) at week 54 was 70.0% in the REMIQ-positive group and 52.9% in the REMIQ-negative group, with a smaller difference between the two groups when infliximab was escalated in the REMIQ-negative group based on physicians’ judgments of disease activity. The percentages of non-responders at 12, 24, and 54 weeks were not significantly different between the REMIQ-positive and -negative groups.

Fig. 4.

Disease activity in DAS28-ESR and the EULAR (European League Against Rheumatism) response. (A) Time course of the disease activity in DAS28-ESR through 54 weeks following the REMIQ-positive and -negative groups. (B) By comparing the groups’ DAS28-ESR scores at 14, 24, and 54 weeks, it is possible to define an improved response. The disease activity and DAS28-ESR criteria were categorized as follows. The DAS28-ESR criteria: High: $>$5.1, Moderate: 3.2–5.1, Low: 2.6–3.2. Remission: 2.6. The EULAR responses were categorized as follows: No response: DAS28-ESR improvement $\le$0.6 and DAS28-ESR improvement $>$0.6 and $\le$1.2 in present DAS28-ESR $>$5.1. Moderate response: DAS28-ESR improvement $>$0.6 and $\le$1.2 in present DAS28-ESR $\le$3.2, and $>$3.2 and $\le$5.1; DAS28-ESR improvement $>$1.2 in present DAS28-ESR $>$5.1, and $>$3.2 and $\le$5.1. Good response: DAS28-ESR improvement $>$1.2 in present DAS28-ESR $\le$3.2. DAS28-ESR: Disease Activity Score assessing 28 joints with erythrocyte sedimentation rate.

When the percentage of patients achieving low disease activity was examined in the REMIQ-positive and -negative groups, the REMIQ-positive group showed a significant increase compared to the -negative group after 14 weeks (Fig. 5). Conversely, there was no significant between-group difference in the length of time to the achievement of remission.

Fig. 5.

Achievement of low disease activity and the patients’ infliximab blood concentrations. Kaplan-Meier curves for the REMIQ positive- (n = 46) and -negative (n = 30) groups treated with infliximab 3, 6, and 12 mg/kg regarding the length of time to achieve low disease activity and remission in the 76 RA patients. **p 0.01, REMIQ-positive vs. -negative group.

We also obtained the Cox proportional HRs for the number of days to the achievement of low disease activity and the number of days to remission. The HR to achieve low disease activity in the REMIQ-positive group was significantly higher than that in the REMIQ-negative group (HR 2.10, 95% CI: 1.55–5.71). Conversely, the number of days to remission was not significantly different between the REMIQ-positive and -negative groups; nor was the HR (HR 1.13, 95% CI: 0.53–2.55).

The REMIQ kit was used to determine the infliximab blood concentrations and RA disease activity at 14 weeks (Fig. 6A). The percentage of patients with positive REMIQ at 14 weeks was 0.0% in the high disease activity group and 39.1% in the high/moderate disease activity group. We examined the ROC curves for CRP and ESR at 14 weeks and observed that at CRP 0.45 mg/dL and ESR 20.0 mm/hr, the ROC curves showed 31.0% and 26.9% specificity and 93.1% and 67.9% sensitivity (area under the curve [AUC] 0.78 and 0.70) (Fig. 6B). The Cox regression analysis with multiple variables showed that REMIQ positivity at baseline (HR 2.10, 95% CI: 1.55–5.71) and ACPA negativity at baseline (HR 0.53, 95% CI: 0.11–0.87) were each associated with the length of time to achieve low disease activity (Fig. 5, Table 2). In addition, RF positivity and ACPA positivity at baseline were each associated with the length of time to achieve remission (HR 0.44, 95% CI: 0.09–0.82) and (HR 0.35, 95% CI: 0.04–0.48), respectively. Conversely, the patients’ CRP and ESR values at baseline were not significantly associated with the lengths of time to achieve low disease activity or remission.

Fig. 6.

Predictors for the achievement to the low disease activity. (A) The patients in the REMIQ-positive group were divided based on their disease activity at 14 weeks of infliximab treatment. (B) The ROC curves for the examination of the CRP and ESR cut-off values for predicting REMIQ positivity were assessed.

Table 3 lists the adverse events over the 54 weeks in the REMIQ positive- and -negative groups. At 54 weeks, an AE had occurred in 13.3% of the REMIQ-negative patients and in 6.5% of REMIQ-positive patients. None of the patients in either group experienced an AE that was serious enough to result in treatment discontinuation or hospitalization. Infusion reactions occurred in three of the REMIQ-negative patients and one REMIQ-positive patient. No significant between-group differences were detected in the occurrence of other AEs such as rash, gastrointestinal disorders, and liver disorders.