POS0592 EFFECTIVENESS OF TOCILIZUMAB VS PREDNISONE IN RHEUMATOID ARTHRITIS PATIENTS WITH INSUFFICIENT RESPONSE TO DISEASE-MODIFYING ANTIRHEUMATIC DRUGS (TOPIRA): A PRAGMATIC RANDOMISED CLINICAL TRIAL

Background: Early induction and maintenance of remission through pharmacological intervention is the cornerstone of treatment of rheumatoid arthritis (RA). After failure of a conventional (cs) DMARD, add-on therapy with another class of drugs is recommended, typically biological DMARDs (bDMARDS) such as an anti-TNF or IL-6 agent. Low to medium doses of glucocorticoids (GCs) have also been shown to have disease-modifying properties when given for an extended period of time. However, the data on adverse events for prolonged low-dose usage of GCs are equivocal. While both bDMARDS and GCs are effective treatments for managing RA, GCs incur far less costs. A direct comparison of the effectiveness and safety of these treatments has not yet been conducted. Objectives: The primary objective was to compare the clinical effectiveness of tocilizumab (TCZ), an IL-6 receptor blocker, with daily 10 mg prednisone in a treat-to-target strategy for RA patients inadequately controlled on csDMARD therapy. The secondary objective was to assess safety through glucocorticoid-related adverse events. Methods: This study was a multicentre, pragmatic, randomised, open-label trial. Patients were randomised 1:1 to receive either TCZ (162 mg weekly, subcutaneously) or prednisone 10 mg daily in a treat-to-target strategy. To be eligible, patients had to meet the 2010 ACR/EULAR classification criteria and have active RA, defined by a Clinical Disease Activity Index (CDAI)>10 with at least one swollen joint. After randomisation, patients had monthly visits until month 3, after which visits occurred 3 monthly until month 12. After month 3, patients could switch treatments in case of inefficacy or adverse events. Inefficacy was defined as CDAI >10 at two consecutive visits, CDAI>22 at any visit or due to unacceptable side effects determined by the treating rheumatologist. The primary endpoint was the CDAI in the intention-to-treat (ITT) population between 6 and 12 months. Key secondary endpoints included adverse events (AE's), including GC-associated events, using the Glucocorticoid Toxicity Index (GTI). A per-protocol (PP) analysis and an on-treatment analysis were performed as sensitivity analyses to account for medication switching during the study. All outcomes were analysed using linear mixed models except for the analysis of total AE's, which was analysed using a Kruskal–Wallis test. Results: In total, 65 patients were randomised. Overall, baseline demographics were balanced between the groups. In the ITT analysis, TCZ showed better results than prednisone during the period from 6 up to 12 months with on average a 3 points lower CDAI (p = 0.07; 95% CI -3.02 to 0.27; Figure 1). The per-protocol analysis was in line with the ITT analysis. At the end of follow-up, 17 patients from the prednisone group had switched treatment, while 9 patients from the TCZ group changed their treatment. In the prednisone group, 13 switches were due to inefficacy, and 4 switches were due to adverse events. In the TCZ group, 4 patients switched because the treatment was ineffective, and 5 due to adverse events. The on-treatment analysis also favoured TCZ with on average a 7 points lower CDAI (p < 0.001; 95% CI -10.06 to -3.83). The total number of AEs did not differ significantly between treatment groups. In the ITT population, Patients had a median of 5 AEs in the prednisone group and 4 AEs in the TCZ group, respectively (p = 0.348). There were a total of seven serious adverse events: five in the prednisone group and two in the TCZ group, mostly resulting from hospitalizations due to infections. The total glucocorticoid toxicity did not differ between groups over 12 months in ITT analysis (2.76 points in favour of TCZ, 95% CI -13.63 to 8.11). In the on-treatment analysis, patients in the TCZ group had significantly less glucocorticoid toxicity (10.33 points in favour of TCZ; p=0.049; 95% CI -20.13 to -0,54). Toxicity was mainly driven by a large difference in sleep quality between the groups (see Figure 2). The per-protocol analysis was in line with ITT analysis. Conclusion: In this pragmatic clinical trial, treatment with TCZ compared to prednisone in patients with active established RA showed a greater reduction in disease activity, as measured by the CDAI. This result was not statistically significant in the ITT analysis, but the on-treatment analysis strongly favoured TCZ in terms of CDAI reduction. While the ITT analysis showed no statistically significant difference in overall AE's and glucocorticoid toxicity between the groups, the on-treatment analysis suggested lower glucocorticoid toxicity associated with TCZ as expected, largely driven by better sleep quality compared to the TCZ group. These findings suggest that TCZ is more effective than prednisone in the treatment of RA patients with active disease on csDMARD therapy. Further research assessing cost-effectiveness and long-term safety could be of additional interest. REFERENCES: NIL . Figure 1Average CDAI measured over a 12-month period. Figure 2Mean GTI by component in the on-treatment analysis. Acknowledgements: NIL . Disclosure of Interests: Sina Fadaei: None declared, Matthijs van der Leeuw: None declared, Paco M.J. Welsing: None declared, Suzanne Linn-Rasker: None declared, Bo de Klerk: None declared, Judith Peeters: None declared, Debby den Uyl: None declared, Paul Baudoin: None declared, Reinhard Bos: None declared, Jacob M. van Laar I have received research grants from Boehringer Ingelheim and Alfasigma and honoraria from Abbvie, Astra Zeneca, Alfasigma, Boehringer Ingelheim, GSK, Novartis. No other disclosures. © The Authors 2025. This abstract is an open access article published in Annals of Rheumatic Diseases under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Neither EULAR nor the publisher make any representation as to the accuracy of the content. The authors are solely responsible for the content in their abstract including accuracy of the facts, statements, results, conclusion, citing resources etc.