External Validation of the Lupus Multivariable Outcome Score for Systemic Lupus Erythematosus Trials

Objective Development of new systemic lupus erythematosus (SLE) treatments requires an effective responder index. Toward this end, we have recently developed a new Lupus Multivariable Outcome Score (LuMOS) to optimize discrimination between actively treated patients and those on placebo. We now report on external validation of LuMOS in two independent clinical trials. Methods Validation was performed with the Illuminate data sets that evaluated tabalumab (TB) in SLE. To accommodate laboratory results assessed on different platforms, we developed a standardized LuMOS 2.0 model that uses z score transformations of biomarker values. For validation, we calculated LuMOS 2.0 scores at week 52 for all participants. Effect size (ES), with 95% confidence intervals (CIs), compared the ability of LuMOS and the SLE Responder Index‐5 (SRI‐5) to discriminate between outcomes in patients randomized to TB dosage and outcomes in those randomized to a placebo. Results Mean LuMOS 2.0 scores were significantly higher ( P  < 0.0001) for the TB groups than the placebo group, including the Illuminate‐1 trial, in which the SRI‐5 did not identify significant treatment effects. For both TB groups in both trials, LuMOS 2.0–based ES indicated moderately strong treatment effects (>0.4) in contrast to weak SRI‐5 effects (<0.25). For monthly TB, LuMOS 2.0–based ES were 0.44 (95% CI: 0.30‐0.59) and 0.54 (95% CI: 0.39‐0.68) for the Illuminate‐1 and Illuminate‐2 trials versus corresponding SRI‐5‐based ES of 0.13 (95% CI: −0.02 to +0.27) and 0.15 (95% CI: 0.01‐0.30). Conclusion LuMOS 2.0 detected significantly greater treatment effects compared with the SRI‐5 in the Illuminate trials. Additional validation of LuMOS 2.0 in trials of non–B cell–directed therapies will be necessary to document its universality as an outcome measure.