Background/Purpose: Sjögren’s is a chronic autoimmune disease symptomatically characterized by dryness, pain and fatigue, which are typically assessed in clinical trials using EULAR Sjögren’s Syndrome Patient Reported Index (ESSPRI). However high placebo response rates are observed when responder definition guidelines (ESSPRI score ≤5 and improvement of ≥1-point or ≥15%) are implemented. The aim of this study was to explore possible options for alternative responder definitions that indicate the optimal balance of sensitivity (the ability to correctly identify improved patients) and specificity (the ability to correctly identify stable patients) to reduce possible misclassification of patients in Sjögren’s clinical trial analyses.

Methods: Thresholds for alternative responder definitions were derived using blinded data from a Phase 2b randomized, double-blind, placebo-controlled, multicenter, parallel-group trial (NCT02962895) in primary Sjögren’s (n=190 at Week 24) and were validated using blinded pooled data from three Phase 2 randomized, double-blind, placebo-controlled, parallel studies (NCT02775916; NCT02291029; NCT02149420) in primary Sjögren’s (n=116 at Week 12; n=68 at Week 24).

Anchor- and distribution-based methods were conducted and triangulated to derive thresholds for an alternative responder definition based on a minimal symptom severity threshold and a more conservative minimal meaningful improvement threshold. A range of anchors were included in the anchor-based analyses based on suitable correlations with ESSPRI ( >0.30), including Patient Global Assessment (PaGA), Physician Global Assessment (PhGA), Short Form-36 (SF-36) physical component score (PCS) and mental component score (MCS; validation analyses only), Functional Assessment of Chronic Illness Therapy – Fatigue (FACIT-F; derivation analyses only), and Multidimensional Fatigue Inventory (MFI; validation analyses only).

Results: Both analysis populations were predominantly female ( >90%), white (90%), and with a mean age of 50 years (range: 20-75). A realistic meaningful change threshold range between 1- and 2-points was identified through derivation analysis; ROC estimates and correlation weighted minimal important mean change supported a minimal threshold of 1.5-1.6 for ESSPRI total score. Correlation weighted mean change estimates and upper 95% confidence intervals supported a more conservative value of 2 or more. The minimal symptom severity threshold range was between 3.0 and 5.3, however 3 or less showed the greatest sensitivity and specificity. Validation analyses generally supported the threshold ranges suggested by the derivation analyses, with a meaningful change threshold range of 1.46-2.11 and a minimal symptom severity range of 2.7 to 4.0 based on the most suitable anchors (PaGA and PhGA).

Conclusion: The results of these analyses suggest that an alternative, more conservative composite ESSPRI responder definition could be used in future trials: an ESSPRI score of ≤3 and an improvement in ESSPRI score of 1.5-points or more. Further analysis in unblinded data is warranted to further test the utility of this new endpoint.

« Back to ACR Convergence 2021

ACR Meeting Abstracts - https://acrabstracts.org/abstract/exploring-alternative-responder-definitions-for-eular-sjogrens-syndrome-patient-reported-index-esspri-for-use-in-sjogrens-clinical-trials/