Weekly tocilizumab dosing linked to lower rheumatoid arthritis activity in real-world study

Patients with rheumatoid arthritis whose disease remained inadequately controlled on subcutaneous tocilizumab given every two weeks showed a meaningful decline in disease activity after their dose was escalated to once weekly, according to a 52-week retrospective observational study conducted at a single tertiary rheumatology center in Japan between January 2017 and March 2024. The study, published in BMC Rheumatology, followed 58 patients who met the 2010 American College of Rheumatology/European League Against Rheumatism classification criteria for rheumatoid arthritis and who were escalated by their treating rheumatologists from 162 mg of subcutaneous tocilizumab every two weeks to the same 162 mg dose once weekly because of persistent clinical disease activity. All patients received the higher-frequency regimen after escalation, and clinical assessments were performed at baseline, defined as the date of escalation, and again at weeks 12, 24, 36, and 52.

The primary outcome was the overall longitudinal change in the Clinical Disease Activity Index, or CDAI, a composite score that the researchers prioritized because, unlike indices that incorporate inflammatory markers, it does not include acute-phase reactants such as C-reactive protein or the erythrocyte sedimentation rate. This distinction matters because tocilizumab blocks the interleukin-6 receptor, a signaling pathway central to inflammation, and thereby suppresses the acute-phase reactants that feed into measures like the Disease Activity Score in 28 joints using the erythrocyte sedimentation rate, or DAS28-ESR. When such markers are artificially lowered by the drug itself, composite indices that rely on them can underestimate how much clinical disease activity actually remains. The CDAI, which relies only on tender joint counts, swollen joint counts, and patient and evaluator global assessments, therefore offers a complementary window on residual disease during interleukin-6 receptor blockade.

In the complete-case analysis restricted to the 42 patients with CDAI data at baseline and all four follow-up visits, the Friedman test indicated a significant overall longitudinal change over the 52 weeks, with a P value below 0.001. The median CDAI fell from 16.00 at the time of escalation to 9.75 at week 12, 9.06 at week 24, 8.80 at week 36, and 8.20 at week 52. At baseline, only 2 of the 58 patients, or 3.4 percent, were in CDAI remission, while 15 had low disease activity, 26 had moderate disease activity, and 15 had high disease activity. By week 52, among the 43 patients with available data, 7 patients, or 16.3 percent, were in remission, 20 had low disease activity, 14 had moderate disease activity, and only 2, or 4.7 percent, still had high disease activity.

Sensitivity analyses using linear mixed-effects models fitted to all available observations, without imputation of missing values, reinforced the direction of these findings. The model-estimated change from baseline to week 52 was a decrease of 5.89 points in CDAI, with a 95 percent confidence interval spanning 8.15 to 3.63 points. The DAS28-ESR likewise declined, with a model-estimated week-52 change of 1.29 points, and the median DAS28-ESR fell from 3.36 at baseline to 1.74 at week 52 in the descriptive complete-case summaries. Matrix metalloproteinase-3, a biomarker associated with synovial inflammation, dropped from a median of 146.6 ng/mL at baseline to 79.8 ng/mL at week 52, with a model-estimated percentage change of 45.8 percent. The authors cautioned, however, that the MMP-3 decline should be read as a change in an inflammatory biomarker rather than as evidence of reduced structural joint damage, because radiographic progression was not assessed in this study.

The researchers also took care to characterize who these patients actually were, and the picture that emerged complicates any simple label. Under the European Alliance of Associations for Rheumatology framework, difficult-to-treat rheumatoid arthritis requires persistent or problematic signs or symptoms despite treatment failure of at least two biological or targeted synthetic disease-modifying antirheumatic drugs with different mechanisms of action, after conventional synthetic DMARD therapy has failed or is contraindicated. When the researchers re-adjudicated each patient’s treatment history at the individual level, they found that 40 patients, or 69 percent, had at least two distinct failed mechanisms of action if the ongoing every-two-week tocilizumab regimen was counted as a current interleukin-6 receptor inhibitor failure. But under a stricter classification that excluded the ongoing tocilizumab regimen, since dose escalation within the same drug should not be counted as failure of a distinct mechanism, only 5 patients, or 8.6 percent, fulfilled the difficult-to-treat definition. The cohort, the authors concluded, should not be interpreted as predominantly difficult-to-treat disease.

Baseline characteristics at escalation underscored how refractory these patients’ disease was in other respects. The median age was 58.2 years, 44 patients, or 75.9 percent, were female, and the median disease duration was 8.7 years. Methotrexate intolerance was widespread: 33 patients, or 56.9 percent, had discontinued methotrexate because of adverse events, and 45 patients, or 77.6 percent, had a recorded history of methotrexate adverse events or intolerance, leaving only 24 patients, or 41.4 percent, actually receiving methotrexate at escalation. Notably, C-reactive protein and erythrocyte sedimentation rate were already low at escalation, with medians of 0.3 mg/L and 6.0 mm/h respectively, despite a median CDAI of 16.00 and elevated matrix metalloproteinase-3, a pattern that directly illustrates why acute-phase-reactant-free measures are valuable during interleukin-6 blockade.

Follow-up completeness emerged as an important caveat. Fifteen patients, or 25.9 percent, lacked week-52 CDAI data: 13 discontinued or interrupted once-weekly treatment before complete follow-up, for reasons including primary inefficacy in five, secondary inefficacy or inadequate control in five, patient preference in one, a switch to intravenous tocilizumab in one, and a surgery-related interruption in one, while two patients transferred care. Critically, the incomplete-follow-up group had significantly higher baseline disease activity than the complete-case group, with a median baseline CDAI of 22.85 versus 12.50, a higher baseline DAS28-ESR, and a higher baseline prednisolone dose. This pattern raises concerns about attrition bias, since the complete-case estimates may preferentially reflect patients with lower baseline disease activity. The mixed-effects sensitivity analyses, which used all available observations, yielded directionally consistent estimates but could not fully eliminate bias if missingness was related to the clinical course.

To guard against attributing improvement to concomitant treatment changes, the researchers tracked initiation or dose escalation of regular systemic medications at prespecified visits. Seven of 58 patients, or 12.1 percent, had documented regular systemic treatment intensification during the first 12 weeks, involving prednisolone in four patients, methotrexate in two, and another conventional synthetic DMARD in one. Even after excluding these seven patients, the median 12-week CDAI change among the remaining 49 patients with paired data was a decrease of 3.00 points, with a P value below 0.001, suggesting the early improvement was not simply an artifact of added therapies. Over the full 52 weeks, 15 patients, or 25.9 percent, had documented treatment intensification at some prespecified visit, while at the group level prednisolone use, prednisolone dose, and methotrexate use all decreased during follow-up.

The study’s most sobering result came from its contextual comparison. The researchers identified 242 patients who continued every-two-week tocilizumab during the same period without escalation, and among the 215 with week-12 CDAI data, 66 had residual disease activity, defined as a CDAI above 10, and 54 had paired data at weeks 12 and 24. Using week 12 after tocilizumab initiation as a pseudo-index for this comparator group, the median 12-week change in CDAI was a decrease of 2.85 points in the escalation cohort and a decrease of 3.40 points in the residual-activity subgroup continuing every-two-week dosing, with no significant between-group difference at P equal to 0.754. Because this comparison was nonrandomized and unmatched, the absence of a difference cannot be read as evidence of equivalence, but it does suggest that natural fluctuation, delayed response to ongoing therapy, and regression to the mean may have contributed substantially to the improvement seen after escalation.

Safety and laboratory findings offered reassurance within the study’s limits. No serious infections or treatment discontinuations due to adverse events were documented during the 52-week follow-up, although the authors noted that mild or transient events may have been underreported in a retrospective design. Estimated glomerular filtration rate declined modestly, from a median of 80.2 to 75.3 mL/min/1.73 m², with 18 of 43 patients showing a decline of at least 10 percent; exploratory subgroup analyses hinted at larger declines among patients using tacrolimus but were underpowered and cannot establish renal safety. The authors emphasized that escalation decisions were made at rheumatologists’ discretion without protocol-defined thresholds, so selection bias and confounding by indication cannot be excluded. Their conclusion was deliberately restrained: disease activity decreased after once-weekly escalation, the descriptive findings are hypothesis-generating, and prospective controlled studies are needed to determine whether weekly dosing offers genuine incremental benefit and, if so, which patients are most likely to gain from it.

Subject of Research: Real-world longitudinal outcomes of escalating subcutaneous tocilizumab dosing in rheumatoid arthritis patients with inadequate disease control

Article Title: Longitudinal clinical outcomes after the escalation of subcutaneous tocilizumab from every 2 weeks to once weekly in rheumatoid arthritis with inadequate disease control: A 52-week real-world observational study

Article References: Sudo, M., Takamura, S., Abe, A., Ishikawa, H., Kobayashi, D., Nakazono, K., Murasawa, A., Yamamoto, S., & Ito, S. (2026). Longitudinal clinical outcomes after the escalation of subcutaneous tocilizumab from every 2 weeks to once weekly in rheumatoid arthritis with inadequate disease control: A 52-week real-world observational study. BMC Rheumatology, 10(1), Article 74. https://doi.org/10.1186/s41927-026-00696-y

Image Credits: AI Generated

DOI: 10.1186/s41927-026-00696-y

Keywords: rheumatoid arthritis, tocilizumab, dose escalation, CDAI, DAS28-ESR, difficult-to-treat RA, interleukin-6 receptor inhibitor, observational study, disease activity, eGFR, methotrexate intolerance, real-world evidence