Background Many new medicines in SLE have not met endpoints in placebo-controlled RCTs even when other evidence indicates that they are effective. Reasons suggested are inclusion of patients with inactive disease, inaccurate or subjective outcome measures that seek to address diverse manifestations, and comparator arms that include excessive glucocorticoid dose.

Objectives (i) To evaluate the feasibility of a novel trial design focusing on musculoskeletal SLE with objective eligibility criteria and endpoints and a low-dose glucocorticoid standard of care.

(ii) To provide preliminary evidence for efficacy of rituximab.

(iii) To provide additional validation of outcome measures.

Methods Adults with SLE were enrolled if they had clinical synovitis and/or ultrasound (US) tenosynovitis and/or positive power Doppler (PD) in ≥1 joint despite stable background therapies including a maximum 10mg prednisolone. Patients were randomized to 1000mg rituximab (Rixathon, RTX) or placebo, on days 1 and 15. Blinded infusions were preceded by 100mg methylprednisolone. Outcome measures, including BILAG-2004, SLEDAI-2K, The Lupus Arthritis and Musculoskeletal Disease Activity Score (LAMDA), tender and swollen joint counts, physician global, patient MSK pain and global VAS, patient reported outcome measures, BICLA, SRI-4 were evaluated monthly. US of both hands and wrists was performed at 0 and 16 weeks. The primary endpoint was overall feasibility. The key efficacy timepoint was 16 weeks. After 16 weeks placebo patients with active disease were eligible for rescue rituximab with repeat follow up timepoints. US and LAMDA were validated against BILAG-MSK improvement at 16 weeks using regression models adjusted for baseline.

Results Of 35 patients invited, 35 attended screened and consented, 6 were ineligible and 2 withdrew prior to baseline. Of 27 patients randomised, 12 were randomised to placebo, of whom 9 received rescue therapy. All completed ≥16 weeks follow up and 25/27 completed all follow up visits. 24/27 (89%) patients were female, 17/27 (63%) were white, 7/27 (26%) were South Asian. Mean (SD) age was 49.7 (12.7) and disease duration 6.7 (9.0). BILAG MSK domain at baseline was scored A in 7/27 (26%), B in 16/27 (59%) and C in 2/27 (7.4%). B scores were also present in mucocutaneous in 2/27 patients and constitutional in 1/27. At 16 weeks, BILAG-MSK response was significantly associated with improvement in LAMDA (OR 0.48, 95% CI 0.18, 0.84); US joints grey-scale (OR 0.56, 95% CI 0.28, 0.85) and US tendons PD (OR 0.33, 95% CI 0.04, 1.01). No substantive difference between arms in efficacy variables was found at week 16. Unexpectedly, results suggested greater improvement in some outcomes in patients who received methylprednisolone and placebo compared to methylprednisolone and rituximab. These measures then converged by 16 weeks (Figure 1). Pooling all rituximab cycles administered (as initial therapy or as rescue) showed improvement in PGA, Physician MSK VAS, joints GS and joints PD scores and LAMDA, but not Global BILAG, SLEDAI-2K, tender or swollen joint counts or and LupusQoL domain or patient VAS at week 16.

Conclusion design dedicated to MSK SLE was feasible. The new outcome measures of LAMDA and MSK US were validated against BILAG-MSK; these outcomes should be included in a definitive trial. While not powered to measure efficacy, exploratory analysis suggests unexpectedly greater improvement in patients who received methylprednisolone with placebo compared to methylprednisolone with rituximab at early time points after infusion, later converging by week 16. Meanwhile there was an overall improvement in LAMDA and US at 16 weeks across all RTX cycles administered. In a definitive trial, early assessment of efficacy should be retrained, but longer observation without retreatment may be required to assess the efficacy of RTX in this population.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

REFERENCES NIL.

Acknowledgements NIL.

Disclosure of Interests Khaled Mahmoud: None declared, Michelle Wilson: None declared, Md Yuzaiful Md Yusof Consultant of: Aurinia, UCB, Sarah Brown: None declared, Elizabeth Hensor: None declared, Edward Vital Consultant of: AstraZeneca, Novartis, Roche, UCB, Aurinia, Lilly, Otsuka, Alumis, Capella, Pfizer, Speakers bureau: AstraZeneca, Novartis, Grant/research support from: AstraZeneca, Sandoz.