Real-World Comparative Effectiveness of Sarilumab Versus Janus Kinase Inhibitors as Monotherapy in Rheumatoid Arthritis

doi:10.21203/rs.3.rs-7787959/v1

Download PDF

Research Article

Real-World Comparative Effectiveness of Sarilumab Versus Janus Kinase Inhibitors as Monotherapy in Rheumatoid Arthritis

https://doi.org/10.21203/rs.3.rs-7787959/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 02 Jan, 2026

Read the published version in Arthritis Research & Therapy →

You are reading this latest preprint version

Background

Sarilumab (SAR), an interleukin-6 receptor inhibitor, and Janus kinase inhibitors (JAKi) are approved options for rheumatoid arthritis (RA) when methotrexate (MTX) cannot be used. Real world evidence for MTX-free monotherapy remains limited.

Methods

We conducted a multicenter retrospective cohort study of RA patients receiving SAR or JAKi as MTX-free monotherapy. To reduce confounding, 1 to 1 propensity score matching was performed in the overall cohort (n = 252, 126 per group) and separately within treatment-line strata: Phase 2 first-line biologic/targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs: 45 per group), Phase 3 second-line b/tsDMARDs (53 per group), and Phase 3 ≥ third-line b/tsDMARDs (47 per group). Outcomes over 12 months included drug retention, change in Clinical Disease Activity Index (CDAI), glucocorticoid (GC) tapering and discontinuation, low disease activity (LDA, CDAI ≤ 10), and safety profiles. Predictors of LDA were evaluated with logistic regression.

Results

Across matched strata by prior b/tsDMARDs, retention and CDAI change were similar between SAR and JAKi through 12 months. When classified by cause, adverse events (AEs) related discontinuation was higher with JAKi, yielding lower AEs specific retention. Both groups demonstrated GC sparing over time, with a greater increase in GC discontinuation for SAR than for JAKi in Phase 2. Baseline predictors of achieving LDA at 12 months included higher C reactive protein (CRP) and platelet count (Plt) in both groups, with additional associations of younger age and lower hemoglobin in the SAR. In safety analyses, overall AEs were less frequent with SAR than with JAKi, driven by lower risks of infection including herpes zoster, while other categories were similarly infrequent.

Conclusion

In MTX-free monotherapy, SAR and JAKi achieved comparable 12 month retention and disease control. Higher CRP and Plt with lower Hb, particularly in younger patients, identified better response to SAR and support biomarker guided selection between IL-6Ri and JAKi. In Phase 2, GC discontinuation with SAR suggests a practical strategy to reduce AEs while maintaining efficacy. Prospective studies should validate these findings and define actionable thresholds.

Rheumatoid arthritis

methotrexate

biological DMARDs

Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory disease characterized by persistent synovitis, systemic inflammation, and progressive joint destruction. The therapeutic landscape of RA has advanced substantially with the introduction of biologic or targeted synthetic DMARDs (b/tsDMARDs) [1, 2], enabling more individualized treatment strategies based on disease activity, prognostic factors, and patient preferences.

Interleukin-6 (IL-6) plays a pivotal role in RA pathogenesis [3]. IL-6 receptor inhibitors (IL-6Ri), such as sarilumab (SAR), have demonstrated clinical efficacy both in combination with conventional synthetic DMARDs (csDMARDs) and monotherapy [4, 5]. Similarly, Janus kinase inhibitors (JAKi) show robust efficacy and are frequently prescribed for patients who are refractory to or intolerant of csDMARDs [6].

Although both IL-6Ri and JAKi are recommended as monotherapy options for patients unable to use methotrexate (MTX), real-world, MTX-free comparative data remain limited. Randomized controlled trials have demonstrated the efficacy of both classes as monotherapy in MTX-intolerant or MTX-naïve patients [7–9], but observational studies that evaluate these agents in MTX-free settings—incorporating detailed patient backgrounds, treatment histories, and longitudinal outcomes—are scarce. Moreover, to our knowledge, no published studies have directly compared predictors of treatment response between SAR and JAKi.

The objective of this study was to evaluate treatment retention and clinical disease activity—assessed using the Clinical Disease Activity Index (CDAI) [10]—and to examine glucocorticoid (GC)-sparing effects in patients with RA receiving MTX-free SAR or JAKi in a multicenter observational cohort. To minimize baseline imbalances, 1:1 propensity score matching (PSM) was performed in the overall cohort and separately within treatment-line strata—Phase 2 first-line, Phase 3 second-line, and Phase 3 third- line or later—to align patient characteristics. In addition, patients were stratified by prior exposure to b/tsDMARDs and by baseline biomarkers—including CDAI, C-reactive protein (CRP), white blood cell count (WBC), hemoglobin (Hb), platelet count (Plt), anti-citrullinated peptide antibody (ACPA), and rheumatoid factor (RF)—to identify predictors of treatment response to SAR and JAKi.

Patients

This study used data from 13 facilities, including university-affiliated hospitals and specialized rheumatology centers: Kindai University Hospital; Osaka Medical and Pharmaceutical University Hospital; Miyazaki Zenjinkai Hospital; Izumi City General Medical Center; Miyamoto Internal Medicine and Rheumatology Clinic; Seirei Hamamatsu General Hospital; Tenri Hospital; Osaka Metropolitan University Hospital; the Center for Senile Degenerative Disorders at Osaka Metropolitan University; Kobe University Hospital; Kobe University Graduate School of Medicine; Okayama University Hospital; and the Musculoskeletal Pain Center at Okayama University Hospital.

RA was classified according to either the 1987 American College of Rheumatology (ACR) criteria or the 2010 ACR/European League Against Rheumatism (EULAR) criteria [11, 12]. During the study period, both frameworks were in routine use in Japan, with the 1987 criteria commonly applied in patients with long-standing disease established before or around 2010. Eligibility did not depend on the specific classification set as long as at least one of the two definitions was fulfilled prior to the index date. We included patients who initiated treatment with either SAR (IL-6Ri monoclonal antibody) or a JAKi. In total, 260 SAR-treated and 212 JAKi-treated patients were analyzed. To minimize temporal bias and ensure both agents were available in the same treatment era, we included patients who initiated treatment between September 2017 (the date of sarilumab approval in Japan) and May 2024.

Safety outcomes

Adverse events (AEs) were recorded and categorized as infections, rash, hematologic, malignancy, RA-ILD (RA associated interstitial lung disease) exacerbation, renal injury, and other. We specifically identified AE discontinuations, defined as investigator attributed AEs that resulted in permanent cessation of SAR or JAKi therapy. AE counts and risks(%)were summarized overall and by category. Analyses were performed in the overall cohorts after PSM (JAKi and SAR: n = 126).

Statistical Analysis

To reduce baseline imbalances between groups, 1:1 PSM was performed using nearest-neighbor matching without replacement and a caliper of 0.2 standard deviations of the logit of the propensity score, based on established methods [13, 14]. Variables included in the matching model were age, sex, disease duration, baseline CDAI, and concurrent use of GCs. Covariate balance before and after matching was assessed using standardized mean differences (SMDs), with SMD < 0.2 indicating sufficient balance. Patients with missing key covariates were excluded. Treatment continuation was evaluated using Kaplan–Meier survival curves and compared with log-rank tests and Cox proportional hazards models. The proportional hazards assumption was assessed using Schoenfeld residuals. Changes in GC dosage were assessed at each time point using the Wilcoxon signed-rank test, while discontinuation rates were compared using chi-square tests. All analyses were conducted using JMP Pro 18.0 (SAS Institute, Cary, NC) and GraphPad Prism 10 (GraphPad Software, San Diego, CA).

Patient characteristics

Table 1 summarizes baseline characteristics before and after PSM and values are reported as mean ± SD or median [IQR] as appropriate. We prespecified |SMD| < 0.10 as the balance target and did not perform hypothesis testing for baseline differences. Before matching SAR n=260 vs. JAKi n=212 showed marked imbalance in treatment line distribution with 1st vs. 2nd vs. ≥3rd line 47.7% vs. 26.5% vs. 25.8% in SAR vs. 24.5% vs. 32.1% vs. 43.4% in JAKi and higher inflammatory burden with SAR with CRP 1.8 [0.5–4.9] vs. 0.6 [0.1–2.1] mg/dL and ESR 56.4 ± 34.5 vs. 39.5 ± 31.4 mm/h. Hematologic indices were also higher with SAR with platelets 29.1 ± 11.4 vs. 25.8 ± 8.7 ×10⁴/μL, neutrophils 5652.5 ± 2680.4 vs. 4891.4 ± 2298.8/μL, and WBC 7995 ± 2826 vs. 7246 ± 2689/μL, while age and sex were comparable and disease activity was modestly higher with SAR with CDAI 22.0 ± 12.0 vs. 19.1 ± 10.4. After one to one matching the treatment line distribution in SAR vs. JAKi was 25.4% vs. 26.2% for first-line, 35.7% vs. 34.1% for second-line, and 38.9% vs. 39.7% for ≥third-line, and baseline characteristics were comparable with age 68.6 ± 13.3 vs. 69.4 ± 12.1 years, female proportion 77.0% vs. 77.8%, and CDAI 20.7 ± 11.6 vs. 19.6 ± 9.0. Residual differences persisted mainly in inflammatory and patient reported measures with higher CRP and ESR and differences in patient VAS, swollen joint count, and HAQ-DI, while most other covariates were similar. Tables 2 and 3 show the same pattern within strata of prior bDMARD exposure with PSM equalizing demographics and most clinical measures across first line, second line, and ≥third line groups, while modest residual imbalances in inflammatory markers and selected patient reported outcomes remained without compromising overall comparability of the matched cohorts. The composition of JAKi by agent is summarized in Table 1 for tofacitinib, baricitinib, peficitinib, upadacitinib, and filgotinib with percentages before and after matching. All five agents used during the study period were represented, and given small numbers in some post-match strata, drug level findings are presented descriptively without hypothesis testing.

Overall Clinical Outcomes

Following PSM, clinical outcomes were compared between the SAR and JAKi groups over a 12-month period. Figure 1A provides the groups' overall retention rates, including all reasons for treatment discontinuation. At 6 months, the retention rates were 82.5% for the JAKi group and 78.3% for the SAR group. At 12 months, the rates declined to 56.3% and 60.3% for the JAKi and SAR groups, respectively. Kaplan–Meier survival analysis revealed no significant difference in treatment continuation between the groups (log-rank test, p=0.60). Cox proportional hazards modeling demonstrated a statistically no significant difference favoring SAR, with a hazard ratio (HR) of 1.1 (95% confidence interval [CI]: 0.75–1.66, p=0.6). In Figure 1B, time to discontinuation for ineffectiveness did not differ between SAR and JAKi (log-rank p=0.67), with survival curves closely overlapping throughout follow-up, and the retention rate for AEs likewise showed no statistically significant difference (log-rank p=0.29), although the JAKi curve trended toward slightly lower retention, suggesting a modestly higher rate of AE-related discontinuation that did not reach significance. As illustrated in Figure 1D, both groups demonstrated comparable CDAI change from baseline through 3, 6, and 12 months. No statistically significant differences in CDAI were observed between groups at any point, indicating similar levels of disease control. Rates of CDAI-low disease activity (LDA) (CDAI-LDA ≤10.0) and remission (CDAI ≤2.8) were also comparable between groups at 3, 6, and 12 months (Figure 1E). At 12 months, the proportion of patients achieving LDA (72.3% vs. 65.9%) and remission (38.3% vs. 33.0%) was not significantly different between SAR and JAKi groups.

Phase-based clinical outcomes in MTX-free RA with sarilumab versus JAK inhibitors by prior b/tsDMARD use
We evaluated 6- and 12-month drug retention and corresponding changes in CDAI by prior treatment line as displayed in Figure 2 and no between-group differences reached statistical significance with log-rank p=0.30 in each stratum. In Phase 2 corresponding to first-line therapy panels A and D the 6-month retention was SAR 88.8% vs. JAKi 85.8% and the 12-month retention was SAR 72.0% vs. JAKi 60.0% and the CDAI changes were SAR −14.3 ± 11.7 vs. JAKi −13.9 ± 8.5 at 6 months and SAR −16.5 ± 14.2 vs. JAKi −14.5 ± 8.6 at 12 months. In Phase 3 corresponding to second-line therapy panels B and E the 6-month retention was SAR 75.4% vs. JAKi 82.0% and the 12-month retention was SAR 50.9% vs. JAKi 58.5% and the CDAI changes were SAR −10.1 ± 8.7 vs. JAKi −15.6 ± 12.2 at 6 months and SAR −9.6 ± 10.3 vs. JAKi −13.4 ± 12.1 at 12 months. In Phase 3 corresponding to third-line or later panels C and F the 6-month retention was SAR 80.0% vs. JAKi 80.5% and the 12-month retention was SAR 65.6% vs. JAKi 52.1% and the CDAI changes were SAR −7.9 ± 9.6 vs. JAKi −14.1 ± 15.6 at 6 months and SAR −8.1 ± 8.1 vs. JAKi −12.4 ± 17.6 at 12 months. Collectively these phase-based analyses indicate broadly comparable retention and disease control between SAR and JAKi across prior-line strata within MTX-free regimens.

Glucocorticoid outcomes over 12 months in sarilumab and JAK inhibitors by phase and treatment-line

Figure 3 summarizes oral GC outcomes from baseline to 12 months where panels A through D display the change in daily GC dose at 3, 6, 9, and 12 months and panels E through H display cumulative GC discontinuation at baseline, 3, 6, and 12 months. In the overall cohort panels A and E, the median change in daily GC dose was SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [0.0–0.0] mg per day at 3 months and SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [−1.0–0.0] at 12 months and the proportion discontinuing GC increased from 3.3% to 6.7% with SAR and from 1.6% to 4.0% with JAKi. In Phase 2 first-line panels B and F, median dose changes were SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [0.0–0.0] at 3 months and SAR 0.0 [−2.0–0.0] vs. JAKi 0.0 [0.0–0.0］at 12 months and GC discontinuation rose from 2.2% to 11.3% with SAR and from 2.3% to 4.6% with JAKi. In Phase 3 second line panels C and G, the median change was SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [0.0–0.0] at 3 months and SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [−1.0–0.0] at 12 months and the discontinuation proportion increased from 1.9% to 7.5% with SAR and from 3.9% to 5.8% with JAKi. In Phase 3 third-line or later panels D and H, median dose changes were SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [0.0–0.0] at 3 months and SAR 0.0 [0.0–0.0] vs. JAKi 0.0 [−1.0–0.0] at 12 months and GC discontinuation increased from 0.0% to 0.0% with SAR and from 0.0% to 2.1% with JAKi. Overall these data show stable to modestly decreasing daily doses over time together with gradual increases in GC discontinuation and no material divergence between SAR and JAKi within each prior line stratum.

CDAI Improvement Stratified by Baseline Prognostic Factors

ΔCDAI at 12 months was evaluated after stratifying baseline CDAI, CRP, RF, ACPA, WBC, Hb, and Plt into quartiles (Suppl. Table 1 and Figure 4). For baseline CDAI, both treatment groups showed the largest improvement in Q4. In the SAR, ΔCDAI increased stepwise from Q1 to Q4 (−2.7, −7.2, −12.8, −28.0), with Q2–Q4 each greater than Q1 (all p < 0.001). A similar pattern was observed with JAKi (p < 0.001). For CRP, improvement rose across quartiles in both groups. In SAR, changes were −5.2, −11.2, −11.6, and −20.3 from Q1 to Q4; in JAKi, −7.2, −9.8, −11.4, and −16.5, with Q4 greater than Q1 (p < 0.05) in JAKi. For WBC, only JAKi showed a quartile effect (Q4 −17.5 vs. Q1 −5.4; p < 0.05). For Hb, SAR showed greater improvement in Q1 (7.3–<10.5 g/dL) than in Q3 and Q4 (−20.9 vs. −4.3 and −7.0; p < 0.001); no Hb-related differences were seen with JAKi. For Plt, both groups improved more in Q4 than Q1 (SAR −24.0 vs. −6.7; JAKi −18.4 vs. −8.2; both p < 0.05). By serological status (Figure 5), RF-negative patients on SAR improved more than RF-positive patients (−18.5 vs. −11.5; p < 0.001). ACPA status did not affect ΔCDAI in either group. Quartile analyses of RF and ACPA titers showed no significant differences, although numerically larger improvements were seen in the lowest quartiles.

Factors Associated with Achieving Low Disease Activity at 12 Months

We identified baseline predictors of achieving CDAI-LDA (CDAI ≤10) at 12 months using logistic regression conducted separately for SAR and JAKi (Table 4). In the SAR cohort, multivariable analysis showed that younger age (adjusted OR 0.98; 95% CI: 0.96–0.99; p = 0.01), b/tsDMARD-naïve status (adjusted OR 2.59; 95% CI: 2.21–3.70; p = 0.01), lower Hb (adjusted OR 0.54; 95% CI: 0.31–0.95; p = 0.05), higher Plt (adjusted OR 1.24; 95% CI: 1.12–1.37; p = 0.001), and higher CRP (adjusted OR 1.83; 95% CI: 1.32–2.45; p = 0.001) independently predicted LDA. In the JAKi cohort, higher Plt (adjusted OR 1.16; 95% CI: 1.07–1.28; p = 0.001) and higher CRP (adjusted OR 1.23; 95% CI: 1.16–1.78; p = 0.03) remained independent predictors. In univariate analyses, additional factors were associated with LDA but did not persist after adjustment: for SAR, RF positivity, lower Hb, higher Plt, and higher CRP (along with age and b/tsDMARD-naïve status) were significant; for JAKi, b/tsDMARD-naïve status, ACPA positivity, lower WBC, lower Hb, higher Plt, and higher CRP were significant.

Safety outcomes and adverse events leading to treatment discontinuation.

Table 5 presented safety outcomes in the matched cohorts (n=126 per group). Any AEs occurred in 22 of 126 patients (17.5%) with SAR and in 42 of 126 (33.3%) with JAKi, an absolute difference of 15.8 percentage points corresponding to approximately 1.9 fold higher risk with JAKi. By category, infection occurred in 3.2% versus 10.3%, herpes zoster in 0.0% versus 2.4%, cancer in 0.8% versus 1.6%, rash in 2.4% versus 0.8%, exacerbation of RA associated ILD in 0.8% versus 1.6%, renal injury in 0.0% versus 1.6%, bone marrow suppression in 3.2% versus 2.4%, and others in 4.8% versus 0.0% for SAR and JAKi respectively.

This multicenter real-world study compared sarilumab an IL-6Ri and JAKi in MTX-free rheumatoid arthritis and showed overall similarity in clinical effectiveness with variability in response according to patient phenotype. Across the cohort CDAI improvement was broadly comparable between SAR and JAKi while baseline hematologic and serologic features modified the magnitude of response. These observations support a precision medicine approach in MTX-free care in which routinely available laboratory measures inform initial treatment selection.

Patients with low baseline hemoglobin experienced greater improvement with SAR, which is consistent with the biology of IL-6–driven anemia. IL-6 upregulates hepatic hepcidin, disrupts iron handling, and contributes to anemia of chronic disease, and IL-6 receptor inhibition can mitigate this process [15]. The larger CDAI improvement observed in the lowest hemoglobin quartile aligns with preferential effectiveness of IL-6Ri in an IL-6–dominant hematologic phenotype and is concordant with prior reports that hemoglobin recovery during IL-6Ri treatment tracks with disease control [16].

Platelet count was associated with greater clinical improvement across both SAR and JAKi. In univariate and multivariable models’ higher baseline Plt predicted better response within each class. Any apparent advantage for SAR was modest with overlapping confidence intervals and should be considered exploratory. This pattern is biologically plausible because IL-6 promotes thrombopoiesis and thrombocytosis mirrors inflammatory activity in RA [17], and Plt often decline after IL-6 receptor inhibition in parallel with clinical improvement [18]. Platelet-derived mediators such as sCD40L can amplify synovial inflammation through fibroblast-like synoviocyte activation and induction of IL-6, establishing a feed-forward loop [19, 20]. Reductions in circulating sCD40L after IL-6Ri have been reported and correlate with improvement in disease activity [21]. Taken together, elevated Plt should be viewed not as a selective indicator for SAR but as a hematologic marker of active inflammation that may forecast response to either class while potentially enriching for IL-6 biology in a subset.

CRP emerged as an independent predictor of achieving CDAI low disease activity in both treatment groups, consistent with CRP as an integrative measure of systemic inflammation responsive to either pathway [22–24]. Based on this finding, patients with high CRP and clinically high disease activity by CDAI should be considered priority treatment targets in MTX-free settings for both SAR and JAKi. When this inflammatory state coexists with low Hb or high Plt that suggest an IL-6–dominant hematologic phenotype SAR may provide greater benefit. In highly inflammatory cases without these hematologic features JAKi remain a reasonable option. For either class treatment intensity should be optimized on the basis of baseline CRP and CDAI, and structured tapering and discontinuation of oral glucocorticoids is recommended.

Phase-based analyses also demonstrated clinical context. In first-line therapy retention, SAR while CDAI improvement was similar. Differences were small in second-line and in third-line or later. These patterns suggest that prior treatment exposure and baseline inflammatory biology may be more informative than drug class alone when MTX is not used. Glucocorticoid outcomes were concordant with these findings. Daily oral doses were stable to modestly decrease over twelve months and the proportion discontinuing glucocorticoids rose gradually in both groups. In first-line use SAR was associated with greater glucocorticoid discontinuation which may have contributed to better persistence. These observations are hypothesis generating and require confirmation.

Safety findings were directionally different yet statistically underpowered. After matching, discontinuations due to infection occurred in 3.2% with SAR versus 10.3% with JAKi and herpes zoster in 0.0% versus 2.4% respectively. Event counts were small and confidence intervals were wide, so the study was not powered for definitive safety comparisons. Even so, the data support routine infection risk assessment at initiation and during follow up, review of vaccination status for herpes zoster where available, careful monitoring in older adults, and attention to background glucocorticoid exposure.

This study has limitations. The design was retrospective and observational and residual confounding and treatment selection bias may persist despite one to one propensity score matching and multivariable adjustment. Some covariates remained imbalanced after matching and drug specific strata within the JAKi class were small, which limits subgroup inference. The IL-6Ri group comprised SAR, so generalizability to other agents in the class may be limited. Results derive from participating centers, which may constrain external validity. Safety analyses were limited by few events and wide intervals, so the study was not powered for group safety comparisons.

These results have practical implications. Baseline WBC, Hb, Plt, CRP, and serologies including RF and ACPA can be incorporated into routine triage for MTX-free regimens. In patients with inflammatory anemia or thrombocytosis both SAR and JAKi are reasonable options. When thrombocytosis coexists with low Hb or other features suggestive of IL-6-dominant biology SAR may be prioritized, whereas in thrombocytosis without such features either class can be selected through shared decision making that considers comorbidity profile and treatment access. Regardless of class, structured efforts to taper and discontinue glucocorticoids appear feasible and may support treatment persistence.

Future research should include prospective biomarker-stratified trials that compare SAR and JAKi in MTX-free RA with a focus on patients with anemia, thrombocytosis, or high CRP. Studies should incorporate IL-6–related biomarkers including hepcidin, ferritin, and transferrin saturation and should evaluate radiographic and functional outcomes along with safety signals with adequate power. Validation of platelet-derived mediators as predictive or pharmacodynamic biomarkers and formal testing of treatment by biomarker interactions are also warranted.

In summary SAR and JAKi achieved broadly similar clinical effectiveness in MTX-free RA. Signals suggest that IL-6 receptor inhibition may confer greater benefit in patients with low Hb and possibly in subsets with IL-6-enriched biology while both classes appear effective in patients with thrombocytosis. First-line use of SAR was associated with glucocorticoid sparing. These findings support a pragmatic precision approach that uses routine laboratory data including CRP and CDAI to guide class selection while highlighting the need for prospective confirmation.

RA: rheumatoid arthritis

MTX: methotrexate

SAR: sarilumab

JAKi: Janus kinase inhibitor

DMARDs: disease modifying antirheumatic drugs

csDMARDs: conventional synthetic DMARDs

bDMARDs: biologic DMARDs

tsDMARDs: targeted synthetic DMARDs

b/tsDMARDs: biologic or targeted synthetic DMARDs

CDAI: Clinical Disease Activity Index

LDA: low disease activity (CDAI ≤10)

GC: glucocorticoid

AE(s): adverse event(s)

RA-ILD: rheumatoid arthritis associated interstitial lung disease

IL-6: interleukin 6

IL-6R: interleukin 6 receptor

IL-6Ri: interleukin 6 receptor inhibitor

CRP: C reactive protein

ESR: erythrocyte sedimentation rate

Hb: hemoglobin

WBC: white blood cell

Plt: platelet

VAS: visual analogue scale

HAQ-DI: Health Assessment Questionnaire Disability Index

PSM: propensity score matching

SMD: standardized mean difference

KM: Kaplan–Meier

HR: hazard ratio

OR: odds ratio

CI: confidence interval

SD: standard deviation

IQR: interquartile range

Q1–Q4: quartiles 1 through 4

TOF: tofacitinib

BAR: baricitinib

PEF: peficitinib

UPA: upadacitinib

FIL: filgotinib

Supplementary data

Supplementary data are available online at Arthritis Research & Therapy.

Authors' contributors: YN conceptualized and designed the study, analyzed the data, and drafted the manuscript. KK, TI, DT, YW, TK, TT, TH, SH, TM, HM, KH, KM, YY, MO, RO, JS, MH, KN, KK, and SR contributed to converting electronic medical records into datasets for analysis. YN is the guarantor and is responsible for the study. All of the authors actively participated in the data acquisition and read and approved the final manuscript.

Acknowledgements: We are deeply grateful to all the medical staff at each participating institution for their cooperation in this study. We also extend our sincere thanks to all hospital personnel involved in this clinical research and to Ms. Kanako Kanzaki of Kindai University Hospital.

Funding: No funding

Data availability statement:

The datasets used and analyzed in this study are available from the corresponding author upon reasonable request.

Ethics approval

This multicenter observational study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee of Kindai University School of Medicine (Approval No. 31-020), the coordinating institution. Ethics approval was also obtained from the institutional review boards of Kobe University Graduate School of Medicine, Osaka Metropolitan University, Osaka Medical and Pharmaceutical University, Miyazaki Zenjinkai Hospital, Izumi City General Medical Center, Seirei Hamamatsu General Hospital, Tenri Hospital, and Okayama University Hospital. The requirement for written informed consent was waived at Kindai University via an opt-out process disclosed on the hospital website. At all other institutions, written informed consent was obtained.

Consent for publication: All authors approved the final manuscript and the submission to this journal.

Conflict of interest statement: YN has received speaker fees from AbbVie, Astellas, Asahi Kasei, Chugai, Eisai, Eli Lilly, GlaxoSmithKline, and Mitsubishi Tanabe. TO has received speaker fees and/or research grants from AbbVie, Asahi Kasei, Astellas, Daiichi Sankyo, Eisai, Eli Lilly, Janssen, Novartis Pharma, Mitsubishi Tanabe, and UCB. TH has received speaker fees from AbbVie, Eli Lilly, Pfizer, Asahi Kasei, Bristol-Myers Squibb, Chugai, Janssen, Taisho, and Eisai. KN has received speaker fees from Eisai, Mitsubishi Tanabe, Asahi Kasei, Taisho, Ayumi, Chugai, and Daiichi Sankyo. HM has received speaker fees from AbbVie, Asahi Kasei, AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Chugai, Eisai, Eli Lilly, GlaxoSmithKline, Sanofi, and Mitsubishi Tanabe. JS has received speaker fees from Eli Lilly, Asahi Kasei, and GlaxoSmithKline. TT has received speaker fees and/or research grants from AstraZeneca, Boehringer Ingelheim, Chugai, GlaxoSmithKline, Eli Lilly, Otsuka, Taisho, and Mitsubishi Tanabe. TK has received speaker fees from AbbVie, Bristol-Myers Squibb, Chugai, Eisai, Eli Lilly, Pfizer, and Boehringer Ingelheim. KM has received speaker fees from AbbVie, Mitsubishi Tanabe, Asahi Kasei, Eisai, Eli Lilly, and Taisho.

Aletaha D, Smolen JS. Diagnosis and management of rheumatoid arthritis: a review. JAMA. 2018;320:1360–72.
Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis. Lancet. 2016;388:2023–38.
Taylor PC, Feist E, Pope JE, et al. What have we learnt from the inhibition of IL-6 in RA and what are the clinical opportunities for patient outcomes? Ther Adv Musculoskelet Dis. 2024;16:1–19.
Genovese MC, Fleischmann R, Kivitz A, et al. Efficacy and safety of sarilumab in combination with csDMARDs or as monotherapy in subpopulations of patients with moderately to severely active rheumatoid arthritis in three phase III randomized, controlled studies. Arthritis Res Ther. 2020;22:139.
Burmester GR, Lin Y, Patel R, et al. Efficacy and safety of sarilumab monotherapy versus adalimumab monotherapy for the treatment of patients with active rheumatoid arthritis (MONARCH): a randomised, double-blind, parallel-group phase III trial. Ann Rheum Dis. 2017;76:840–7.
Correction: Baricitinib in patients with inadequate response or intolerance to conventional synthetic DMARDs: results from the RA-BUILD study. Ann Rheum Dis. 2017;76:1634.
Fleischmann R, Kremer J, Cush J, et al. Placebo-controlled trial of tofacitinib monotherapy in rheumatoid arthritis. N Engl J Med. 2012;367:495–507.
Keystone EC, Taylor PC, Tanaka Y, et al. Patient-reported outcomes from a phase 3 study of baricitinib versus placebo or adalimumab in rheumatoid arthritis: secondary analyses from the RA-BEAM study. Ann Rheum Dis. 2017;76:1853–61.
van Vollenhoven R, Takeuchi T, Pangan AL, et al. Efficacy and safety of upadacitinib monotherapy in methotrexate-naive patients with moderately-to-severely active rheumatoid arthritis (SELECT-EARLY): a multicenter, multi-country, randomized, double-blind, active comparator-controlled trial. Arthritis Rheumatol. 2020;72:1607–20.
Aletaha D, Nell VP, Stamm T, et al. Acute phase reactants add little to composite disease activity indices for rheumatoid arthritis: validation of a clinical activity score. Arthritis Res Ther. 2005;7:R796–806.
Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum. 1988;31:315–24.
Aletaha D, Neogi T, Silman AJ, et al. 2010 Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 2010;62:2569–81.
Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70:41–55.
D’Agostino RB Jr. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med. 1998;17:2265–81.
Nemeth E, Rivera S, Gabayan V, et al. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest. 2004;113:1271–6.
Nakayama Y, Watanabe R, Yamamoto W, et al. IL-6 inhibitors and JAK inhibitors as favourable treatment options for patients with anaemia and rheumatoid arthritis: ANSWER cohort study. Rheumatology (Oxford). 2024;63:349–57.
Jiang SZ, To JL, Hughes MR, et al. Platelet signaling at the nexus of innate immunity and rheumatoid arthritis. Front Immunol. 2022;13:977828.
Matsuno H. Remarkable efficacy of tocilizumab for treating rheumatoid arthritis in patients with high platelet counts. Mod Rheumatol. 2015;25:38–42.
Habets KL, Trouw LA, Levarht EW, et al. Anti-citrullinated protein antibodies contribute to platelet activation in rheumatoid arthritis. Arthritis Res Ther. 2015;17:209.
Zhang F, Wei K, Slowikowski K, et al. Defining inflammatory cell states in rheumatoid arthritis joint synovial tissues by integrating single-cell transcriptomics and mass cytometry. Nat Immunol. 2019;20:928–42.
Zamora C, Diaz-Torne C, et al. Platelet-derived soluble CD40L and its impact on immune modulation and anti-IL6R antibody treatment outcome in rheumatoid arthritis. Cells. 2025;14:625.
Shafran IH, Alasti F, Smolen JS, et al. Implication of baseline levels and early changes of C-reactive protein for subsequent clinical outcomes of patients with rheumatoid arthritis treated with tocilizumab. Ann Rheum Dis. 2020;79:874–82.
Nakayama Y, Hashimoto M, Watanabe R, et al. Favorable clinical response and drug retention of anti-IL-6 receptor inhibitor in rheumatoid arthritis with high CRP levels: the ANSWER cohort study. Scand J Rheumatol. 2022;51:431–40.
Wang J, Devenport J, Low JM, et al. Relationship between baseline and early changes in C-reactive protein and interleukin-6 levels and clinical response to tocilizumab in rheumatoid arthritis. Arthritis Care Res (Hoboken). 2016;68:882–5.

Tables 1 to 5 are available in the Supplementary Files section

Competing interest reported. YN has received speaker fees from AbbVie, Astellas, Asahi Kasei, Chugai, Eisai, Eli Lilly, GlaxoSmithKline, and Mitsubishi Tanabe.

Download PDF

Journal Publication

published 02 Jan, 2026

Read the published version in Arthritis Research & Therapy →

Editorial decision: Revision requested
24 Nov, 2025
Reviews received at journal
18 Nov, 2025
Reviewers agreed at journal
09 Nov, 2025
Reviews received at journal
02 Nov, 2025
Reviewers agreed at journal
24 Oct, 2025
Reviewers invited by journal
22 Oct, 2025
Editor assigned by journal
12 Oct, 2025
Submission checks completed at journal
12 Oct, 2025
First submitted to journal
06 Oct, 2025

You are reading this latest preprint version

Real-World Comparative Effectiveness of Sarilumab Versus Janus Kinase Inhibitors as Monotherapy in Rheumatoid Arthritis

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

PATIENTS AND METHODS

Patients

Safety outcomes

Statistical Analysis

RESULTS

Discussion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1