Abstract
Objective:
To assess the effectiveness and continuation rate of half-dose Janus kinase (JAK) inhibitors in patients with rheumatoid arthritis (RA).
Methods:
We conducted a retrospective cohort study of 212 patients who initiated JAK inhibitors at our hospital (standard-dose group: 142; half-dose group: 70). Among 141 patients who continued the same dose for more than 6 months (standard-dose: 99; half-dose: 42), propensity score matching was performed based on age, disease duration, sex, number of prior biologics/JAK inhibitors, baseline CRP, CDAI, glucocorticoid dose, methotrexate dose, and mHAQ score. Matched groups (37 per group) were compared for CRP, CDAI, and mHAQ at 6 months and for changes from baseline. Reasons for discontinuation were analysed in 71 patients who discontinued treatment within 6 months.
Results:
At 6 months, the half-dose group showed significantly higher CRP levels and less improvement in mHAQ scores than the standard-dose group. No significant differences were observed in CDAI or other parameters. Discontinuation due to insufficient effectiveness was significantly more frequent in the half-dose group, whereas discontinuation due to adverse events did not differ between groups.
Conclusions:
Compared with standard-dose therapy, half-dose JAK inhibitors are associated with reduced effectiveness and higher discontinuation rates owing to insufficient clinical response.
Introduction
Janus kinase (JAK) inhibitors are oral disease-modifying antirheumatic drugs (DMARDs) that block the JAK-STAT pathway, a central mediator of inflammation in rheumatoid arthritis (RA). They provide rapid symptom relief—often within 2 weeks—and are effective even in patients who do not respond to conventional DMARDs or biologics.1–3 Their oral route also offers a convenient alternative to injectable therapies. However, risks such as infections, thrombosis, and lipid abnormalities necessitate careful monitoring.4–7
Despite their clinical benefits, the high cost of JAK inhibitors can place a substantial financial burden on patients. During the study period, generic JAK inhibitors were not available in our healthcare setting, and the cost of JAK inhibitors remained substantially higher than that of conventional synthetic DMARDs. Consequently, half-dose regimens are sometimes used in real-world practice to improve affordability. Although early clinical trials compared multiple dosing strategies,8–11 there is limited evidence specifically evaluating the effectiveness and safety of half-dose therapy in routine care. Moreover, the impact of half-dose treatment on treatment persistence—including adherence and long-term therapeutic success—remains unclear.
To address this gap, the present study examined the clinical effectiveness and drug retention of JAK inhibitors administered at half doses compared with standard doses. These findings aim to support optimised treatment strategies, particularly for patients who struggle to access JAK inhibitors at full recommended doses.
Materials and Methods
This study was based on an observational, single-centre registry of patients with RA at our institution (Kobe University Hospital Institutional Review Board, approval No 1738, 20th March 2013). Written general consent for the use of clinical data for research purposes was obtained in advance from all patients. The registry was established in October 2013 to improve RA clinical outcomes, and clinical data from each visit between October 2013 and December 2023 were recorded. RA was diagnosed according to the 2010 ACR/EULAR criteria 12 or the 1987 ACR criteria. 13 During the study period, 223 patients were treated with tofacitinib (TOF), baricitinib (BAR), peficitinib (PEF), upadacitinib (UPA), or filgotinib (FIL). Eleven patients were excluded because of missing data regarding the Clinical Disease Activity Index (CDAI), modified Health Assessment Questionnaire (mHAQ), or C-reactive protein (CRP), resulting in 212 eligible patients.
Of these 212 patients, 142 initiated therapies at the standard dose (Group S: TOF 39, BAR 68, PEF 4, UPA 25, FIL 6), and 70 started at half doses (Group H: TOF 21, BAR 33, PEF 10, UPA 5, FIL 1). Standard doses were TOF 10 mg, BAR 4 mg, PEF 150 mg, UPA 15 mg, and FIL 200 mg; half doses were TOF 5 mg, BAR 2 mg, PEF 100 mg, UPA 7.5 mg, and FIL 100 mg. In Group H, TOF was administered as 5 mg once daily instead of the standard regimen of 5 mg twice daily. For the other JAK inhibitors, the reduced dose was administered once daily using the corresponding half-strength dose, maintaining the same dosing frequency as the standard regimen. Patients were categorised according to treatment continuity during the first 6 months. A total of 141 patients continued the initial dosage for 6 months (Group S: 99; Group H: 42), whereas 71 patients either discontinued treatment or changed dosage within 6 months (Group S: 43; Group H: 28). The baseline characteristics of this cohort reflect the clinical profile of patients with RA commonly encountered in Japan, where the patient population tends to be older and elderly-onset RA is relatively common. In addition, methotrexate (MTX) doses used in Japanese clinical practice are often lower than those reported in Western countries, particularly in elderly patients, due to safety and tolerability considerations. Therefore, the treatment background observed in this study is considered representative of real-world RA management in Japan.
For patients who maintained a consistent dose for 6 months, propensity score matching (PSM) was conducted to minimise baseline imbalances. Matching variables included age, disease duration, sex, number of previously used biologics/JAK inhibitors, baseline CRP, CDAI, mHAQ, glucocorticoid dose, and MTX dose. Matching was performed using a caliper width of 20% of the standard deviation of the propensity score, with random selection among patients with identical scores. This process yielded 37 matched pairs (n = 74). CRP, CDAI, and mHAQ values at 6 months were compared between the matched groups, and changes from baseline to 6 months were analysed.
For the 71 patients who discontinued or adjusted therapy within 6 months, the duration from treatment initiation to discontinuation was recorded. Reasons for discontinuation or dose modification were classified as (a) insufficient effectiveness, (b) complications or adverse events, or (c) other reasons such as financial constraints or relocation. The distribution of these categories was compared between Groups S and H.
Statistical Analyses
All data are expressed as the mean ± standard deviation unless otherwise indicated. Patients’ background characteristics between the two groups were compared using the Mann-Whitney U test and paired t-test before and after propensity matching, respectively (Table 1). Paired t-test was used to compare CRP, CDAI, and mHAQ between the two groups at 6 months after the start of administration, as well as to analyse the changes in these parameters before and after administration. For the group of cases in which administration was discontinued, or dosage was adjusted within 6 months after the start of administration, cumulative probabilities of the drug retention and discontinuation rates were estimated using the Kaplan-Meier method. Survivorship curves for the various subgroups were compared using the log-rank test. All analyses were performed using BellCurve for Excel (Social Survey Research Information Co., Ltd., Tokyo, Japan).
Patients’ Background Characteristics Before (A) and After (B) Propensity Score Matching.
Results
The results of the comparisons of CRP, CDAI, and mHAQ values at 6 months after treatment initiation between patients who continued treatment with the same dose of JAK inhibitors for more than 6 months are shown in Table 2. Comparisons of changes in these parameters from baseline to 6 months are presented in Table 3. In Group H, CRP levels at 6 months were significantly higher, and improvements in mHAQ scores were significantly smaller than those observed in Group S.
Values at 6 Months After the Initiation of Treatment.
Reduction from Baseline to 6 Months After the Initiation of Treatment.
Discontinuation or adjustment due to insufficient effectiveness within the first 6 months was significantly more frequent in Group H than in Group S, occurring in 30.0% (N = 21) in Group H compared with 13.4% (N = 19) in Group S (P = .0036). In contrast, complications or adverse effects led to treatment changes in 11.3% (N = 16) in Group S and 5.7% (N = 4) in Group H, with no significant difference between groups (P = .193, the 95% confidence interval for the difference in proportions ranged from −0.139 to 0.028).
Other reasons were reported in 5.6% (N = 8) in Group S and 4.3% (N = 3) in Group H, also without a significant difference (P = .68). In the log-rank test, no significant difference was found between the two groups regarding the overall continuation rate within 6 months (Figure 1A). However, when the analysis was restricted to discontinuation or dosage adjustment due to insufficient effectiveness, Group H showed a significantly higher rate than Group S (Figure 1B), indicating that differences between the groups became apparent when focusing specifically on treatment effectiveness.
(A) Retention Rates of JAK Inhibitor Treatments: No Significant Difference Was Found Between Group S and Group H Regarding the Overall Continuation Rate Within 6 Months of Drug Initiation. Group S: Dashed Line; Group H: Solid Line. (B) Discontinuation Rates of JAK Inhibitor Treatments Due to Insufficient Effectiveness: Group H Had a Significantly Higher Discontinuation or Dosage Adjustment Rate Within 6 Months of Drug Initiation Than Did Group S. Group S: Dashed Line; Group H: Solid Line.
Discussion
In this retrospective cohort study using real-world clinical data, we demonstrated that initiating JAK inhibitors at a half-dose was associated with inferior clinical outcomes compared with standard-dose therapy in patients with RA. Specifically, patients receiving half-dose JAK inhibitors showed less effective suppression of inflammation, as reflected by higher CRP levels at 6 months, and smaller improvements in physical function assessed by mHAQ. In addition, treatment discontinuation or dose modification due to insufficient effectiveness occurred more frequently in the half-dose group, indicating compromised treatment persistence attributable to inadequate disease control. Importantly, no reduction in discontinuation related to adverse events was observed with half-dose therapy.
These findings suggest that although half-dose JAK inhibitor therapy is sometimes adopted in clinical practice—often driven by financial considerations or concerns regarding safety—it may not provide sufficient therapeutic benefit for maintaining optimal disease control. The observed difference in inflammatory markers and functional improvement implies that a certain exposure threshold may be required to achieve sustained suppression of synovial inflammation and to translate this effect into meaningful functional gains. From a clinical perspective, suboptimal disease control during the early treatment phase may predispose patients to persistent symptoms, functional impairment, and eventual treatment discontinuation.
Previous randomised controlled trials have evaluated different dosing regimens of JAK inhibitors, primarily during drug development. For example, Genovese et al. compared baricitinib 2 mg and 4 mg and reported broadly comparable rates of major adverse events, although infection rates tended to be higher at the higher dose. 14 Similarly, van Vollenhoven et al. demonstrated dose-dependent efficacy of upadacitinib, with higher doses achieving more rapid and robust disease control at the expense of increased adverse events. 15 These studies highlight the trade-off between efficacy and safety across dosing ranges. Although these studies were performed under controlled conditions with strict inclusion criteria, the present study provides real-world evidence from routine clinical practice, including a more heterogeneous patient population and physician-driven treatment decisions. Therefore, our findings may provide complementary information regarding the effectiveness of reduced-dose JAK inhibitor therapy in real-world clinical settings.
In contrast, the present study focused on half-dose initiation in a real-world setting, where patient characteristics, prior treatment histories, and comorbidities are more heterogeneous. Our results indicate that, in this context, half-dose therapy did not confer a measurable safety advantage, as discontinuation due to adverse events was comparable between groups. Although no statistically significant difference was observed, the incidence of adverse events was numerically higher in the half-dose group than in the standard-dose group (11.3% vs. 5.7%), and this lack of statistical significance may be attributable to the limited sample size. This finding is clinically relevant, as it challenges the assumption that dose reduction necessarily improves tolerability. Instead, it suggests that adverse events associated with JAK inhibitors may not be sufficiently dose-dependent within this reduced range to justify routine half-dose initiation, while efficacy appears more sensitive to dose reduction.
The higher rate of discontinuation due to insufficient effectiveness in the half-dose group further underscores the potential drawbacks of this strategy. Early discontinuation or dose escalation driven by poor response may ultimately negate any initial benefits of dose reduction, including cost savings, and may delay achievement of treatment targets such as low disease activity or remission. In treat-to-target strategies, inadequate initial dosing could therefore compromise long-term outcomes.
Several limitations of this study warrant consideration. First, despite the use of propensity score matching, the matched sample size was modest, which may limit statistical power and generalisability. Second, the reasons for selecting half-dose therapy were not systematically captured; factors such as frailty, comorbidities, patient preference, or socioeconomic constraints may have influenced dosing decisions and outcomes. Third, laboratory parameters related to organ function, which could affect both dosing decisions and safety outcomes, were not included in the matching process. Fourth, the severity and nature of insufficient effectiveness leading to discontinuation were not analysed in detail, nor were adverse events stratified by type or severity. In addition, although documentation of individual adverse events would have provided further insight, this was not feasible due to the retrospective nature of the study and limitations in the available data. Fifth, this study was conducted exclusively in a Japanese population, and ethnic differences in drug metabolism and treatment practices may limit the generalisability of the findings to other populations. Therefore, caution is warranted when extrapolating these results to non-Japanese populations. Finally, analyses were not performed separately for individual JAK inhibitors, precluding conclusions regarding agent-specific dose-response relationships.
Despite these limitations, this study provides clinically relevant evidence regarding the effectiveness and treatment persistence of half-dose JAK inhibitor therapy in routine practice. The findings suggest that half-dose initiation may compromise efficacy without offering a clear safety advantage, thereby limiting its overall clinical value. While individualised dosing strategies remain important, particularly in vulnerable patient populations, standard-dose therapy appears more appropriate for achieving and maintaining adequate disease control in most patients with RA.
Conclusion
In patients with RA, half-dose JAK inhibitors were associated with reduced effectiveness and higher discontinuation rates due to insufficient clinical response compared with standard doses. Further studies are needed to identify patient populations in whom individualised dose reduction may be feasible without compromising treatment outcomes.
Footnotes
Authors’ Contribution
Conception and design of study: NN, SH, JS, RK. Acquisition of data: NN, SH, MT, TM. Analysis and/or interpretation of data: NN, SH. Drafting the manuscript: NN. Revising the manuscript critically for important intellectual content: NN, SH, MT, TM, JS, RK.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Ethical Approval
All procedures were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The full name of the approving ethics committee: Kobe University Hospital Institutional Review Board, date of approval: March 20th, 2013, document submission number: No. 1738.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
Patient Consent
Written general consent was obtained from all patients.