Rene Westhovens’s research while affiliated with Leuven University College and other places

Objectives Long-term safety is fundamental for treatment decision-making. This integrated analysis of filgotinib clinical trials in rheumatoid arthritis (RA) and ulcerative colitis (UC) assessed adverse events of interest (AEI): major adverse cardiovascular events (MACE), venous thromboembolism (VTE) and malignancies. Methods Data were integrated from all phase II and III trials that have investigated filgotinib 100 mg or 200 mg once daily in RA and UC to date. Results Analyses represent >12 500 (RA) and >2800 (UC) patient-years of exposure (PYE) to filgotinib. Incidences of AEI in the integrated analysis population were low. Modest numerical increases in incidence rates occurred in patients aged ≥65 years, including MACE (patients with RA), and malignancies (excluding non-melanoma skin cancer (NMSC)) and NMSC (patients with RA or UC). VTE was rare; in patients with RA aged ≥65 years receiving filgotinib 200 mg, exposure-adjusted incidence rate (95% CI) for VTE was 0.3 (0.1, 0.8)/100 PYE; no VTE events occurred in patients with UC aged ≥65 years. In patients with RA aged ≥65 years, MACE incidence rates were identical between filgotinib 100 mg and 200 mg; rates of malignancies and NMSC were numerically higher with 200 mg compared with 100 mg. Conclusions Data are consistent with previous overall safety analyses demonstrating low rates of AEI in the overall study population. Numerically increased rates of AEI occurred in patients aged ≥65 years; further data are needed to assess the effect of CV risk factors. Overall, in this analysis, there was no consistent filgotinib dose effect on AEI.

Objectives DARWIN 3 (ClinicalTrials.gov: NCT02065700 ) assessed the safety and efficacy of filgotinib in a long-term extension (LTE) of two phase II randomised controlled rheumatoid arthritis (RA) trials. Methods Eligible patients completing the 24-week DARWIN 1 (filgotinib plus methotrexate) and DARWIN 2 (filgotinib monotherapy) trials could enrol. Patients received filgotinib 200 mg/day, except 15 men who received filgotinib 100 mg/day. The primary endpoints were safety and tolerability, which were assessed by the incidence of treatment-emergent adverse events (TEAEs). Safety and efficacy analyses included all enrolled patients who received ≥1 dose of filgotinib in DARWIN 3. Results 739 patients entered the LTE. The total patient-years of exposure (PYE) to filgotinib was 3706.3 years; the mean exposure duration was 259.8 weeks. 497 patients (67.3%) discontinued prematurely (including 266 TEAEs and 172 withdrawals due to the patient’s decision or ‘sponsor request’). Overall exposure-adjusted incidence rate (EAIR) was 67 (95% CI 62 to 72.2)/100 PYE for TEAEs and 3.8 (95% CI 3.2 to 4.5)/100 PYE for serious TEAEs. EAIR of infections was 23.3 (95% CI 21.2 to 25.6)/100 PYE, 1.3 (95% CI 0.9 to 1.7)/100 PYE for serious infections and 1.3 (95% CI 0.9 to 1.7)/100 PYE for herpes zoster. EAIRs of major adverse cardiovascular events (0.19 (95% CI 0.8 to 0.39)/100 PYE) and malignancies (0.6 (95% CI 0.4 to 0.9)/100 PYE) were low. Disease response assessed using non-responder imputation plateaued at LTE week 12 before slowly declining over time, with overall American College of Rheumatology (ACR)20/50/70 response rates of 26.9%/20.2%/14.7% at week 396. Conclusion Filgotinib was well tolerated in patients with RA for up to 8 years. Safety and efficacy profiles were maintained in patients previously receiving either filgotinib plus methotrexate or filgotinib monotherapy.

Objectives We aimed to study if smartphone applications could support the self-management of RA, while investigating engagement and potential negative psychological effects with app-use. Methods App-based Education and GOal-setting in RA (AEGORA) was a multicentre randomised controlled trial with 2:1:1-allocation to usual care or two versions of an app-based self-management intervention for RA. The 16-week programme involved patient education, goal-setting, and remote monitoring of the Rheumatoid Arthritis Impact of Disease (RAID) instrument, either weekly or monthly depending on randomisation. The primary end point was improvement in the Arthritis Self-Efficacy Scale (ASES) after 16 weeks. Secondary endpoints included non-inferiority regarding the Pain Catastrophizing Scale (PCS) and superiority regarding patient-reported physical activity, sleep quality and RAID. App engagement and RAID-scores were analysed descriptively. Results Overall, 122 patients were included: mean (SD) disease duration 12 (9) years, mean (SD) age 58(11), 68% female, mean (SD) DAS28-CRP 2.4(0.9). The intervention did not improve the ASES-score over usual care (β: 0.44, p= 0.87). Non-inferiority was established for the PCS (β -0.95 [95% CI -3.30 to + 1.40] favouring the intervention). Other predefined outcomes did not differ. App retention steadily declined to 43% by 16 weeks. Although the RAID remained stable over time overall, 35% of app users reported ≥1 episode of clinically relevant worsening over 16 weeks. Conclusion This app-based self-management intervention was not superior to usual care regarding self-efficacy improvement. However, remote symptom monitoring provided valuable insight and did not increase pain catastrophising, alleviating concerns regarding the psychological impact of remote monitoring with apps. Trial registration number clinicaltrials.gov, NCT05888181.

Background The optimal retreatment strategy with rituximab for rheumatoid arthritis (RA) remains a point of discussion. Depending on local guidelines, rituximab can either be administered at fixed intervals or when losing disease control, balancing therapeutic effectiveness with drug overexposure. However, treatment based on loss of disease control may significantly affect patients’ lives, provoking uncertainty and potentially leading to progressive joint damage. Moreover, as low-dose rituximab proved to be effective in treating RA while decreasing toxicity, drug exposure may be limited by tapering down rituximab doses guided by disease activity. Methods RITUXERA is a 104-week open-label multicentre randomised controlled superiority trial. In total, 134 patients with RA treated with rituximab will be 1:1 randomised when in need of retreatment (DAS28-CRP ≥ 3.2 with previous rituximab administration at least 24 weeks earlier) to either a treat-to-target-driven fixed dose retreatment strategy (usual care group) or fixed interval disease-activity guided dose optimisation strategy (experimental group). The usual care group will be retreated with fixed rituximab doses (1 × 1000 mg IV) in case of loss of disease control (DAS28-CRP ≥ 3.2). The experimental group will receive a 24-weekly rituximab treatment while tapering down the dose in a decreasing sequence if DAS28-CRP ≤ 3.2: 1 × 1000 mg IV (maximal dose), 1 × 500 mg IV, and 1 × 200 mg IV (minimal dose). If DAS28-CRP exceeds 3.2 at the six-monthly retreatment, patients will receive and remain on the previous effective dose. Study visits are planned every 12 weeks. Primary outcome is the comparison of longitudinal patient-reported disease impact over 104 weeks, measured with the Rheumatoid Arthritis Impact of Disease (RAID) instrument, analysed using a linear mixed model. Main secondary outcome is the comparison of longitudinal disease activity (DAS28-CRP) over 104 weeks. Discussion The RITUXERA trial aims to explore the optimal retreatment strategy with rituximab for RA in terms of long-term patient-reported disease impact, by proposing a fixed interval disease activity-guided dose optimisation strategy as compared to a treat-to-target fixed dose strategy. Trial registration CTIS 2023–506638-59–01 (registration date: 07 September 2023), ClinicalTrials.gov NCT06003283 (registration date: 17 August 2023).

Objectives To investigate if patients with early rheumatoid arthritis responding insufficiently to initial methotrexate (MTX) and bridging glucocorticoids (GCs) could benefit from early but temporary etanercept introduction as a second remission-induction attempt. Methods CareRA2020 ( NCT03649061 ) was a 2-year, open-label, multicentre, pragmatic randomised controlled trial. Treatment-naïve patients started MTX and GC bridging (COBRA-Slim: CS). Within a time window from week (W) 8 until W32, early insufficient responders (28-joint Disease Activity Score – C-reactive Protein (DAS28-CRP) >3.2 between W8 and W32 or ≥2.6 at W32) were randomised to a Standard-CS strategy (adding leflunomide first) or Bio-induction-CS strategy (adding etanercept for 24 weeks). Additional treatment adaptations followed the treat-to-target principle. Longitudinal disease activity (DAS28-CRP) over 104 weeks (primary outcome), achievement of DAS28-CRP <2.6 28 weeks after randomisation, and biologic or targeted synthetic disease-modifying antirheumatic drug (b/tsDMARD) use at W104 were compared between randomisation groups. Results Following CS treatment, 142 patients were early responders; 55 early insufficient responders received Standard-CS and 55 Bio-induction-CS. Superiority of Bio-induction-CS over Standard-CS could not be demonstrated (ß=−0.204, (95% CI –0.486 to 0.078), p=0.157) for the primary outcome. More patients on Bio-induction-CS achieved DAS28-CRP <2.6 at 28 weeks after randomisation (59% (95% CI 44% to 72%) vs 44% (95% CI 31% to 59%) in Standard-CS) and they were treated less frequently with b/tsDMARDs at W104 (19/55, 35%) compared with Standard-CS (29/55, 53%). Conclusion Half of the patients responded well to initial COBRA-Slim induction therapy. In early insufficient responders, adding etanercept for 6 months did not improve disease control over 104 weeks versus adding leflunomide first. However, temporary introduction of etanercept resulted in improved disease control early after randomisation and less patients on b/tsDMARDs at W104. Trial registration number NCT03649061 . CTR pilot approval Belgium S59474, EudraCT number: 2017-004054-41.

Background In the CareRA2020 trial, early access to 24 weeks etanercept (Bio-induction COBRA-Slim, CS) was not superior in terms of disease control over 2 years compared to addition of leflunomide (Standard-CS) in patients with early Rheumatoid Arthritis (RA) insufficiently responding to an initial remission-induction regimen of methotrexate (MTX) and step-down glucocorticoids. However, b/tsDMARD and oral glucocorticoid use at Week (W) 104 was higher in Standard-CS compared to Bio-induction-CS. Objectives To determine the speed and long-term stability of disease control in patients with early RA after initial COBRA-Slim remission-induction followed by a treat-to-target (T2T) approach. Methods DMARD-naïve patients with early RA (diagnosis ≤1y) were included in the 2-year, open-label, multicentre randomised controlled CareRA2020 trial. All patients started a remission-induction regimen with MTX and a step-down prednisone scheme starting from 30 mg/d. In case of an early insufficient response between W8 – W32, defined as DAS28-CRP >3.2 from W8 onwards, or DAS28-CRP ≥2.6 at W32, patients were 1:1 randomised to a “second-hit” remission-induction regimen: addition of leflunomide (Standard-CS) or etanercept for 24 weeks (Bio-induction-CS). Further following a T2T-approach, a DAS28-CRP threshold of ≤3.2 was aimed for. To investigate the speed of response to the second-hit remission-induction regimen in early insufficient responders, the time until first achievement of DAS28-CRP <2.6 after randomisation was determined. The proportion of patients with sustained disease control defined as DAS28-CRP <2.6 for at least 24 weeks, as well as sustained remission based on SDAI (≤3.3) was determined for the early responder as well as both early insufficient responder groups. Data were multiple imputed (m=100), and were analysed descriptively. Results In CareRA2020, 142 patients were early responders (67% female, mean ± SD age 54.9 ± 13.9y) to the initial remission-induction regimen, 55 early insufficient responders were randomised to Standard-CS (75% female, age 53.4 ± 13.1y), and 55 to Bio-induction-CS (69% female, age 52.5 ± 12.9y). Between BL-W104, 94% [95% CI 88 - 97] early responders achieved sustained DAS28-CRP disease control for at least 24 weeks (sustained SDAI disease control: 64% [55 - 72]). This was 86% [79 - 91] between BL-W64 (SDAI: 50% [41 - 58]), and 76% [68 - 83] between W64-W104 (SDAI: 41% [32 - 50]). Moreover, 38% [30 - 47] of early responders had a DAS28-CRP <2.6 at every study visit between BL-W104 (SDAI ≤3.3 at every study visit: 9% [5 - 15]). After randomisation, Bio-induction-CS achieved DAS28-CRP <2.6 more rapidly compared to Standard-CS (Figure 1). However, a comparable proportion of patients in both randomisation groups maintained DAS28-CRP <2.6 for at least 24 weeks after randomisation until W104 (Standard-CS: 51% [37 - 65], Bio-induction-CS: 45% [32 - 60]) (SDAI: 14% [7 - 26] and 13% [6 - 26], respectively). When considering all visits after randomisation, the proportion of patients maintaining DAS28-CRP <2.6 until the study end was also comparable between the two groups (Standard-CS: 9% [4 - 22], Bio-induction-CS: 8% [3 - 20]) (SDAI: 0% for both randomisation groups). Conclusion CareRA2020 confirmed that a remission-induction regimen with MTX and bridging prednisone is a highly effective treatment for patients with early RA, as 94% of early responders achieved sustained disease control for at least 24 weeks between BL and W104. In early insufficient responders to the initial regimen, a second-hit remission-induction regimen with 24 weeks of etanercept yielded a more rapid achievement of DAS28-CRP <2.6 compared to addition of leflunomide. However, after randomisation both groups had comparable sustained disease control for at least 24 weeks during the trial. • Download figure • Open in new tab • Download powerpoint Figure 1. First achievement of DAS28-CRP <2.6 after randomisation in the Standard-CS (addition of leflunomide), and Bio-induction-CS group (addition of 24 weeks of etanercept). The figure shows the first 28 weeks after randomisation. REFERENCES NIL Acknowledgements NIL Disclosure of Interests None declared

Background The Janus Kinase inhibitor Upadacitinib (UPA) has shown efficacy in pivotal trials in rheumatoid arthritis (RA), yet real-world effectiveness data (RWD) are scarce. Objectives To investigate the RWD effectiveness of UPA in RA using Tumour Necrosis Factor inhibitors (TNFi) as benchmark. Methods Participants were included from TARDIS, that collects each reimbursement request for drug initiation or prolongation from all Belgian RA patients on advanced therapy. Patients starting on UPA or TNFi between 01/11/2020-31/12/2021, ≥18 years and DAS28>3.7 (required for reimbursement) and with follow-up were included. The proportion of patients achieving remission (DAS-28 <2.6), low disease activity (LDA, DAS-28 ≤3.2), a clinical meaningful HAQ-DI-decrease (MCID_HAQ) since baseline=0.22, HAQ-DI≤0.25 (HAQ25), and HAQ-DI≤0.50 (HAQ50) after 2 registrations in TARDIS (after 15 months for UPA patients and 18 months for TNFi) were analysed. If DAS28CRP was unavailable, DAS28ESR was used. Patients who discontinued/switched treatment prior to assessment were non-responders. Outcomes within 1st, 2nd and 3th+ treatment line were also investigated. To control for confounding, logistic regression was used. First, models included standardized average treatment effect (sATE) propensity-score based weights using age, disease duration, baseline joint counts, DAS28, HAQ-DI, PGA, and number of prior advanced treatment lines. Secondly, multivariable logistic regression models included the UPA vs TNFi variable, along with the other potential confounders. To facilitate interpretation, adjusted odds ratios for UPA vs TNFi were transformed into adjusted risk ratios (RR). Group imbalance was examined using standardized mean differences. Results In total, 617/996 (62%) UPA and 822/1274 (65%) TNFi patients were included. Table 1 details the two populations. After applying sATE weighting, baseline characteristics were balanced between groups (standardized mean differences all <0.1). Remission was reached in 48% (297/617) of UPA and 35% (288/822) of TNFi patients. Proportion of patients reaching LDA (59% vs 44%) and MCID_HAQ (53% vs 37%) was also higher for UPA, while differences in HAQ25 (13% vs 10%) and HAQ50 (22% vs 21%) seemed similar. Figure 1 shows subgroup results by treatment line and indicates decreasing outcomes with increasing treatment line. Logistic regression models with sATE-propensity-weighting yielded adjusted RR (95%CI) for UPA versus TNFi of 1.5 (1.4-1.7), 1.5 (1.4-1.6), 1.5 (1.4-1.7), 1.7 (1.3-2.3) and 1.3 (1.1-1.6) for remission (p<0.001), LDA (p<0.001), MCID_HAQ (p<0.001), HAQ25 (p<0.001) and HAQ50 (p=0.005) respectively. The multivariable logistic regression approach yielded similar results. To note, UPA was continued in 72.4% (447/617) while TNFi was continued in 57.1% (504/882) of patients. Table 1 Baseline population characteristics • Download figure • Open in new tab • Download powerpoint • Download figure • Open in new tab • Download powerpoint Figure 1. Effectiveness outcomes at the second TARDIS visit (15 months for UPA; 18 months for TNFi). Remission=DAS28<2.6; LDA=Low disease activity (DAS28≤3.2); HAQ25=HAQ-DI score≤0.25; HAQ50=HAQ-DI score≤0.5; MCID_HAQ=clinical meaningful HAQ-DI-decrease since baseline of 0.22. Conclusion In a real-world setting, UPA outperformed TNFi in improving disease activity and functionality outcomes for moderately to severely active RA, after accounting for baseline differences. Reliance on mostly CRP for DAS-remission may have biased the results. REFERENCES NIL Acknowledgements O behalf of the Royal Belgian Society for Rheumatology AbbVie sponsored the study, contributed to the design and participated in reviewing and approving the final version. No honoraria/payments were made for authorship. Disclosure of Interests None declared

Background Advanced therapy responses in patients with rheumatoid arthritis (RA) exhibit remarkable similarity across trials. However, Janus-kinase-inhibitors (JAKi) stand out with a consistent 10% improvement in trials, referred to as the JAKi-bonus. Real-world data on this bonus are lacking. Objectives To explore the existence and potential explanations for the JAKi-bonus in a RA registry. Methods Patients were included from the electronic platform “Tool for Administrative Reimbursement Drug Information Sharing” (TARDIS). Data from all Belgian RA patients on biologic and targeted therapy are collected in this platform during the submission of a request for initiation or prolongation of reimbursement of these drugs.Adult patients ≥18 years- and a DAS28 score>3.7 starting for a first time a JAKi or a comparator population treated with non-JAKi were selected between 1 December 2017 and 1 January 2024. Patients who attended a first follow-up moment, structurally planned in TARDIS after 3 months (JAKi) or 6 months (non-JAKi) were included. The JAKi bonus was determined by the patient proportion achieving remission (DAS28<2.6). DAS28 was calculated via CRP and if missing via the ESR. A propensity-adjusted logistic regression model was constructed to adjust for confounding. The standardized differences between treatment groups from baseline to first follow-up for each of the DAS components were analysed to find putative mechanisms of a JAK bonus. A standardized difference greater than 0.1 after propensity adjustment was considered meaningful. Results In total, 2867 patients on JAKi and 5986 patients on non-JAKi were included (Table 1). Remission was reached in 41.4% (1187/2867) JAKi versus 32.8% (1964/5986) non-JAKi patients (p<0.001). DAS28-ESR (39.9% vs 33.3%) and DAS28-CRP (41.2% vs 33.0%) remission rates confirmed this result. A logistic regression model with a propensity score using baseline age, disease duration, joint counts, PGA, DAS and HAQ, and treatment line showed an odds ratio (CI95%) for remission of 1.6 (1.4-1.7, p<0.001). Propensity-adjusted standardized differences were 0.12, 0,13, 0.21, 0.05 and 0,02 for TJC28, SJC28, PGA, CRP and ESR respectively(Figure 1).View this table: • View inline • View popup Table 1. Baseline characteristics • Download figure • Open in new tab • Download powerpoint Figure 1. Standardised differences of DAS28 components between JAKi and non-JAKi TJC= tender joint count; SJC= Swollen joint Count; PGA= Patient Global assessment; CRP= C-reactive protein; ESR= erythrocyte sedimentation rate. A positive effect implies an improvement of JAKi over non-JAKi. Conclusion A JAKi bonus is present in real-world data in patients with RA. The drivers might be the subjective components of the DAS28 score. This cohort was enriched with advanced treatment-naive patients, and the JAKi bonus per treatment line will be investigated in the near future. REFERENCES [1] Landewé et al. Why most (but perhaps not all) DMARDs work equally well. Sar. 2023. Acknowledgements NIL Disclosure of Interests None declared

Background Filgotinib (FIL), a preferential Janus Kinase (JAK)1 inhibitor has demonstrated efficacy in rheumatoid arthritis (RA) in pivotal clinical trials. Limited data exist on the real-world effectiveness of FIL in RA. Objectives To examine the short-term effectiveness of FIL compared to other modes of action in a heterogeneous Belgian RA population. Methods Patients were included from the electronic platform “Tool for Administrative Reimbursement Drug Information Sharing” (TARDIS). Data from all Belgian RA patients on biologic and targeted therapy are collected in this platform during the submission of a request for initiation or prolongation of reimbursement of these drugs.Adult patients ≥18 years- and a Disease Activity Score 28 (DAS28 score) >3.7 starting FIL or a comparator population treated with bDMARDs were selected between August 2021 and December 2023. Patients who attended a first follow-up moment, structurally planned in TARDIS after 3 months (FIL) or 6 months (bDMARDs) were included. Effectiveness was determined by the patient proportion achieving remission (DAS28CRP <2.6, if DAS28-CRP was unavailable DAS28-ESR was used), Low Disease Activity (LDA) (DAS28 ≤3.2), Health Assessment Questionnaire Disability Index (HAQ-DI) ≤0.25 (HAQ25), HAQ-DI ≤0.5 (HAQ50) and minimum HAQ-DI-decrease (MCID_HAQ) since baseline of 0.22. Logistic regression was adjusted for confounding variables. Results In total, 714 patients on FIL and 3741 patients on bDMARDs were included (Table 1). Disease duration was longer in FIL population (9.9±10.0 years) versus the bDMARD population (6.5±8.1 years). Remission and LDA was reached respectively in 45.1% (322/714) and 58.3% (416/714) FIL versus 29.7% (1110/3741) and 38.1% (1425/3741) bDMARD patients (p<0.001). MCID_HAQ was reached in 53.5% (382/714) FIL versus 34.0% (1272/3741) bDMARD patients (p<0.001). Logistic regression models adjusted for baseline age, disease duration, joint counts, Patient Global Assessment, DAS28, HAQ-DI, and treatment line confirmed these results. Odds Ratio (CI95%) for remission was 2.4 (2.0-2.8, p<0.001), for LDA 2.6 (2.2-3.1, p<0.001), for MCID_HAQ 2.6 (2.2-3.1, p<0.001), for HAQ25 2.5 (1.9-3.2, p<0.001), and finally for HAQ50 2.3 (1.9-2.9, p<0.001). FIL efficacy slightly decreased with increasing line of therapy (Figure 1).View this table: • View inline • View popup Table 1. Baseline population characteristics (Table 1 linked to abstract AB0589) • Download figure • Open in new tab • Download powerpoint Figure 1. Effectiveness outcomes of FIL per treatment line remission (DAS28<2.6), LDA - low disease activity (DAS28≤3.2), HAQ-DI≤0.25 or ≤0. 5 (HAQ25/50) and minimum HAQ-DI-decrease (MCID_HAQ) since baseline=0.22. Conclusion FIL treatment showed enhanced clinical and functional outcomes versus a contemporary bDMARD cohort on the short-term in moderately to severely active RA in real-world data. Although adjusted analyses were used, selection bias still could influence the results. Moreover, remission based on CRP was mostly used, and this could bias remission rates towards FIL. REFERENCES NIL Acknowledgements This work was supported by Galapagos. Disclosure of Interests None declared