POS0193 EVALUATION OF CRESS IN THE PHASE 2 RANDOMISED PLACEBO-CONTROLLED STUDY OF SEQUENTIAL BELIMUMAB/RITUXIMAB ADMINISTRATION IN PATIENTS WITH PRIMARY SJÖGREN’S SYNDROME

Background EULAR Sjögren’s syndrome disease activity index (ESSDAI) assesses systemic disease activity in patients (pts) with primary Sjögren’s syndrome (pSS); however, weaknesses include exclusion of patient-reported symptoms, tear and salivary gland function, and a marked placebo (PBO) response. Composite of Relevant Endpoints for Sjögren’s Syndrome (CRESS) is a recently developed composite outcome measure validated using data from three Phase 3 randomised controlled trials of pts with pSS. 1 Concise CRESS (cCRESS) is used when ocular staining score and salivary gland ultrasonography are unavailable. ESSDAI was an endpoint in a Phase 2, PBO-controlled study, evaluating the safety and efficacy of belimumab (BEL) and rituximab (RTX) sequential administration (BEL/RTX), and BEL and RTX monotherapies in pts with pSS. Although the results numerically favoured BEL/RTX over PBO, this was not statistically significant. Objectives To evaluate the efficacy of BEL/RTX and monotherapies using cCRESS overall responses at Weeks (Wks) 24, 52, and 68, and individual item responses at Wk 24 in pts with pSS who completed the Phase 2 study. Methods In the Phase 2, double-blind, 68-Wk study ( NCT02631538 ) adults were randomised (2:2:2:1) into 4 treatment arms: BEL/RTX (n=24; weekly BEL 200 mg subcutaneous [SC] to Wk 24 followed by weekly PBO SC to Wk 52 + RTX 1000 mg intravenous [IV], Wk 8 + 10), BEL monotherapy (n=24; weekly BEL 200 mg SC to Wk 52), RTX monotherapy (n=25; RTX 1000 mg IV, Wk 8 + 10), or PBO (n=13). Pts were classified post hoc as cCRESS responders when ≥3 of the following 5 items were met: 1) Clinical (Clin)ESSDAI score <5 (low disease state); 2) decrease of ≥1 point or ≥15% from baseline (BL) in EULAR Sjögren’s Syndrome Patient Reported Index (ESSPRI); 3) increase of ≥5 mm from BL in abnormal Schirmer’s test, or no change to abnormal if normal at BL; 4) unstimulated whole saliva (UWS) increase of ≥25% from BL, or any increase from BL if score was 0 at BL; 5) decrease of ≥25% in the rheumatoid factor (RF) titre from BL, or decrease of ≥10% in IgG from BL. 1 Results Of 86 randomised pts, 60 completed follow-up to Wk 68 (completer population) and were included in the analysis. Most pts were female (95%, n=57); mean (SD) age was 49.6 (13.0) years. BL disease characteristics are presented in the Table 1. Table 1. Clinical, functional, and laboratory parameters at BL and cCRESS responders at Wks 24, 52, and 68 (completer population) PBO (n=8) BEL/RTX (n=17) BEL (n=19) RTX (n=16) Pt parameters at BL, mean (SD ) ClinESSDAI 11.1 (3.76) 11.7 (5.47) 9.2 (3.77) 11.7 (4.76) ESSPRI 6.4 (2.05) 6.0 (1.97) 6.5 (1.68) 5.9 (2.20) Schirmer, mm/5 min 2.7 (3.25) 5.3 (6.44) 3.3 (3.16) 2.8 (3.15) UWS, ml/min 0.1 (0.11) 0.1 (0.12) 0.1 (0.09) 0.1 (0.14) RF, KU/l 60.8 (42.24) 30.9 (38.20) 37.0 (34.98) 105.0 (200.97) IgG, g/l 20.4 (6.65) 16.7 (5.00) 18.1 (7.19) 16.5 (6.09) cCRESS responders, n (% ) Wk 24 4 (50.0) 9 (52.9) 7 (36.8) 5 (31.3) Wk 52 4 (50.0) 10 (58.8) 8 (42.1) 4 (25.0) Wk 68 1 (12.5) 6 (35.3) 7 (36.8) 3 (18.8) At Wks 24 and 52, the proportion of cCRESS responders was numerically higher with BEL/RTX than with either BEL, RTX, or PBO, but the difference was not significant (Table 1). At Wk 68, the proportion of cCRESS responders was numerically higher with BEL/RTX than with RTX or PBO (Table 1). The 5 cCRESS items contributed relatively equally to total cCRESS response, with the highest response observed in the RF/IgG item and the lowest in the tear gland item (Schirmer’s test; Figure 1). Figure 1. cCRESS and individual item responders at Wk 24 (completer population ) Conclusion At Wks 24, 52, and 68, BEL/RTX was generally associated with a numerically higher cCRESS response rate compared with the monotherapies or PBO. The PBO response for cCRESS was notable and similar to the PBO response for ClinESSDAI. The PBO response in the tear and salivary gland items was greater than in the other treatment arms, perhaps due to the use of cCRESS instead of CRESS. Due to the small sample size, the results should be interpreted with caution. References [1]Arends S, et al. Lancet Rheumatol 2021;3:553–62 Acknowledgements This post hoc analysis of the GSK Study 201842 was funded by GlaxoSmithKline (GSK). Medical writing support was provided by Casmira Brazaitis, PhD, Fishawack Indicia Ltd, UK, part of Fishawack Health, and was funded by GSK. Disclosure of Interests Hendrika Bootsma Consultant of: BSM, Roche, Novartis, Medimmune and Union Chimique Belge, Grant/research support from: BSM and Roche, Suzanne Arends: None declared, Liseth de Wolff: None declared, Kenneth L Clark Shareholder of: GSK, Employee of: GSK, Andre van Maurik Shareholder of: GSK, Employee of: GSK, Prafull Mistry Shareholder of: GSK, Employee of: GSK, Pragya Shukla Shareholder of: GSK, Employee of: GSK, Svetlana Nihtyanova Shareholder of: GSK, Consultant of: Roche, Employee of: GSK, Norma Lynn Fox Shareholder of: GSK, Employee of: GSK, David Roth Shareholder of: GSK, Employee of: GSK