Abstract
Time to sustained worsening in the expanded disability status scale as the standard for evaluating the accumulation of disability has been used as a measure of clinical efficacy in many relapsing-remitting multiple sclerosis (RRMS) clinical trials. However, this measurement usually requires a large sample and long-term study to demonstrate the treatment effect. Annualized relapse rate or time to first relapse is also widely used as alternative measurements of clinical efficacy. A formal statistical validation of short-term relapse activity as a surrogate endpoint for long-term sustained progression of disability could potentially permit smaller, shorter, and less expensive clinical trials in RRMS. Four statistical validation/evaluation approaches consistently showed that relapse activity through one year of treatment serves as statistically valid surrogate endpoint for time to sustained progression of disability. The analysis demonstrates that long-term sustained progression of disability can be predicted by short-term relapse measures with 4 consistent validations of statistical approaches, including a formal statistical hypothesis test. This was demonstrated in a large phase III trial of natalizumab and showed that the beneficial clinical effect of natalizumab on sustained progression of disability at 2 years in patients with RRMS can be predicted by the total number of relapses at 1 year.