Development and performance of the Clinical Trials ESSDAI (ClinTrialsESSDAI), consisting of frequently active clinical domains, in two randomised controlled trials in primary Sjögren's syndrome.

ABSTRACT

OBJECTIVES: To develop and evaluate the Clinical Trials EULAR Sjögren's Syndrome Disease Activity Index (ClinTrialsESSDAI), consisting of frequently active clinical domains of the ESSDAI, using two randomised controlled trials in primary Sjögren's syndrome (pSS). METHODS: The ASAP-III trial in abatacept (80 pSS patients) and TRACTISS trial in rituximab (133 pSS patients) were analysed. The most frequently active clinical domains were selected, and ClinTrialsESSDAI total score was calculated using existing weightings of the ClinESSDAI (which also excludes the biological domain). Performance of the ClinTrialsESSDAI was compared to ClinESSDAI and ESSDAI. Responsiveness was assessed using standardised response mean (SRM), and discrimination was assessed using adjusted mean difference. RESULTS: Besides the biological domain, the most frequently active domains were glandular, articular, haematological, constitutional, lymphadenopathy and cutaneous. These domains were selected for the ClinTrialsESSDAI. At primary endpoint visits, SRM values of ClinTrialsESSDAI, ClinESSDAI and ESSDAI were respectively -0.65/-0.59, -0.63/-0.59 and -0.64/-0.61 for abatacept/placebo and -0.33/-0.13, -0.34/-0.12 and -0.41/-0.16 for rituximab/placebo. Adjusted mean differences between active treatment and placebo groups were respectively -1.7, -1.4 and -1.1 for ASAP-III and -1.1, -1.1 and -1.2 for TRACTISS. CONCLUSIONS: The ClinTrialsESSDAI, consisting of six frequently active clinical domains of the ESSDAI, shows closely similar responsiveness and discrimination between treatment groups compared to the ClinESSDAI and ESSDAI. Therefore, this ClinTrialsESSDAI is not preferable to ClinESSDAI and ESSDAI for use as primary endpoint. A composite endpoint combining response at multiple clinically relevant items seems more suitable as primary study endpoint in pSS.