Development of the Sjogren's Syndrome Responder Index, a data-driven composite endpoint for assessing treatment efficacy

Objectives. To determine which outcome measures detected rituximab efficacy in the Tolerance and Efficacy of Rituximab in Sjogren's Disease (TEARS) trial and to create a composite endpoint for future trials in primary SS (pSS). Methods. Post hoc analysis of the multicentre randomized placebo-controlled double-blind TEARS trial. The results were validated using data from two other randomized controlled trials in pSS, assessing rituximab (single-centre trial in the Netherlands) and infliximab, respectively. Results. Five outcome measures were improved by rituximab in the TEARS trial: patient-assessed visual analogue scale scores for fatigue, oral dryness and ocular dryness, unstimulated whole salivary flow and ESR. We combined these measures into a composite endpoint, the SS Responder Index (SSRI), and we defined an SSRI-30 response as a 530% improvement in at least two of five outcome measures. In TEARS, the proportions of patients with an SSRI-30 response in the rituximab and placebo groups at 6, 16 and 24 weeks were 47% vs 21%, 50% vs 7% and 55% vs 20%, respectively (P<0.01 for all comparisons). SSRI-30 response rates after 12 and 24 weeks in the single-centre rituximab trial were 68% (13/19) vs 40% (4/10) and 74% (14/19) vs 40% (4/10), respectively. No significant differences in SSRI-30 response rates were found between infliximab and placebo at any of the time points in the infliximab trial. Conclusion. A core set of outcome measures used in combination suggests that rituximab could be effective and infliximab ineffective in pSS. The SSRI might prove useful as the primary outcome measure for future therapeutic trials in pSS.