Design and Statistical Issues Related to Testing Experimental Therapy in Multiple Sclerosis

Summary

The authors believe that neurologists planning a clinical trial of an experimental therapy for multiple sclerosis can select certain design elements that will enhance the sensitivity of the trial.

The typical trial will be multi-centered, double-blinded and randomized, with experimental therapy and placebo-treated patient groups.

The authors recommend that:

• 1. Admission to the trial should be restricted to those patients whose Extended Disability Status Scale (EDSS) scores do not exceed 3.5 at entry

• 2. Admission should be restricted to patients with relapsing forms of multiple sclerosis

• 3. One entry criterion should be evidence of disease activity: relapses during a pretrial period

• 4. There should be one primary response variable, based on the EDSS, and a restricted number of secondary response variables

• 5. An event should be defined as an increase in the EDSS score from baseline of at least one unit; it should be confirmed at a subsequent 3- or 6-month examination

• 6. The primary response variable should be defined as the “time to a confirmed event”

• 7. The clinical trial should include 3 years of treatment and follow-up for each patient

• 8. It should be estimated that 50% of the placebo patients will progress at least one confirmed EDSS unit during the trial

• 9. The minimum clinical effect of the experimental therapy to be demonstrated with high probability should fall in the range 30%–50% reduction in progression as compared to that in the placebo group

• 10. Interim analyses may be performed at 2 and 2½ years to permit termination of the trial if a statistically significant difference between the therapy and placebo groups is observed