Clinical trials of therapies for lupus nephritis have used many different primary outcome measures, ranging from complete response to time to end-stage renal disease. The objective of this study was to compare several possible outcome measures, using data from a large, multicenter trial of abatacept in lupus nephritis, to gain insight into which outcome measure, if any, was best able to discern differences among treatment groups.
Study patients received either abatacept or placebo, on a background of mycophenolate mofetil and glucocorticoids. Using data from this trial, the following primary outcome measures at 24 and 52 weeks were compared: complete response rate, major clinical response rate, total response rate (complete plus partial response), improvement in proteinuria, improvement in estimated glomerular filtration rate, and frequency of treatment failure. Time to complete response was also evaluated.
Complete response rate, major clinical response rate, and time to complete response were the measures that best discriminated between the abatacept groups and placebo, and the sensitivities of these 3 measures were comparable. For these measures, sample sizes of 50 patients would have been sufficient to demonstrate a statistically significant difference between treatment and control at 52 weeks. Each of the other measures also discriminated between treatment and control, but much larger group sizes would have been required to determine statistical significance.
The choice of primary outcome measure can substantially influence the ability to detect therapeutic benefit in lupus nephritis trials. This study suggests that complete response rate, major clinical response rate at 52 weeks, and time to complete response may be the most sensitive outcome measures for detecting differences among therapeutic regimens.