AB0632 Essdai domain evaluation of primary sjÖgren’s syndrome (PSS) patients enrolled in two independent poc studies indicates differential utility of domains for trial inclusion and composite endpoints in pss trials

Background According to international consensus, disease activity in primary Sjogren’s syndrome (pSS) patients shall be scored across 12 different domains according to the European Sjogren Syndrome Disease Activity Index (ESSDAI 1 ). Two separate phase 2 proof-of-concept (PoC) studies using either BAFF receptor inhibitor (VAY736; ianalumab) or CD40 inhibitor (CFZ533) have recently been completed in comparable pSS cohorts with ESSDAI as the primary endpoint. Objectives To evaluate contribution of individual domains to composite ESSDAI scores at baseline and after interventional treatment with either VAY736 or CFZ533, or placebo. Aggregate efficacy and safety results were presented at previous meetings. 2 3 Methods Key inclusion criteria for both studies were fulfilling revised European US consensus criteria for pSS, 4 autoantibody positivity, exclusion of secondary SS and ESSDAI ≥6. Permitted background medications in both trials included stable doses of hydroxychloroquine, methotrexate and prednisone ≤10 mg/d and also azathioprine in the CFZ533 trial. The primary endpoint was change in ESSDAI at week 12. Descriptive and principal component analysis were done with the goal to identify distinct subgroups of patients based on ESSDAI involvement at baseline (BL), and the relative importance of single domain contribution to overall ESSDAI responses were explored. Results 27 patients received single i.v. dose VAY736 10 mg/kg (n=12) or 3 mg/kg (n=6) or placebo, and 44 patients received multiple doses of CFZ533 10 mg/kg i.v. (n=21) or 3 mg/kg s.c. (n=8) or placebo. ESSDAI breakdown at BL showed a predominance of the articular, glandular, biological, constitutional and lymphadenopathy domains in both trials. Activity in more than 3 domains was recorded for 12/27 (44%) and 12/44 (27%) of patients in the VAY736 and CFZ533 studies, respectively. Principal component analysis identified the articular domain as the key component describing the difference between patients in their ESSDAI domains at BL. Two other domains that explained the variability between patients were the biological and glandular domains. Treatment effects in domains with low BL scores were more difficult to assess using ESSDAI. The majority of ESSDAI domains were not amenable to quantitative analysis, due to absence or low incidence at baseline. Conclusions The most frequently observed ESSDAI domain was articular involvement. Our results provide insights into ESSDAI domain frequency and distribution in the randomised controlled trial setting that may have implication for future trial design in pSS. Reference [1] Seror R, et al. (2016)Ann Rheum Dis75:382–389. [2] Dorner Th, et al. (2016)Arthritis Rheumatol68 Suppl 10; [3] Fisher B, et al. (2017) Arthritis Rheumatol69; Suppl 10; [4] Vitali C, et al. (2002) Ann Rheum Dis61:554–558. Disclosure of Interest T. Dorner Grant/research support from: Novartis, Charite CRO, Consultant for: Novartis, B. Fisher Consultant for: Novartis, Roche, BMS and AstraZeneca/MedImmune., X. Ren Employee of: Novartis Pharmaceuticals, P. Faerber Employee of: Novartis Pharmaceuticals, P. Gergely Shareholder of: Novartis Pharmaceuticals, Employee of: Novartis Pharmaceuticals, L. Mooney Employee of: Novartis Pharmaceuticals, Y. Li Shareholder of: Novartis Pharmaceuticals, Employee of: Novartis Pharmaceuticals, S. Oliver Shareholder of: Novartis Pharmaceuticals, Employee of: Novartis Pharmaceuticals, W. Hueber Shareholder of: Novartis Pharmaceuticals, Employee of: Novartis Pharmaceuticals, A. Wright Shareholder of: Novartis Pharmaceuticals, Employee of: Novartis Pharmaceuticals