Improving power to detect disease progression in multiple sclerosis through alternative analysis strategies

Area of Science:

• Neurology
• Clinical Trial Methodology
• Biostatistics

Background:

• The Expanded Disability Status Scale (EDSS) is commonly used in multiple sclerosis (MS) clinical trials to assess disease progression.
• The efficacy of sustained progression as an outcome measure for detecting treatment effects in MS is not fully established.

Purpose of the Study:

• To compare the statistical power of different modeling strategies for the EDSS outcome measure in detecting treatment effects on MS disease progression.
• To determine the most effective modeling approach for analyzing EDSS data in the context of potential therapeutic interventions.

Main Methods:

• Simulated EDSS measurements at 6-month intervals over 24 months for patients with initial EDSS scores from 0 to 3.
• Applied nine different modeling strategies, including those based on sustained progression and direct EDSS score modeling.
• Evaluated modeling performance under three distinct treatment effect scenarios: reducing higher EDSS, increasing lower EDSS, and both.

Main Results:

• Modeling approaches that directly analyzed EDSS scores demonstrated greater statistical power than those based on sustained progression.
• The advantage of direct EDSS modeling was particularly evident when treatments increased the probability of patient improvement.
• The choice of modeling strategy significantly impacts the ability to detect treatment effects on MS clinical progression.

Conclusions:

• Direct modeling of EDSS scores is a more powerful outcome measure for assessing treatment effects in MS clinical progression compared to sustained progression.
• Optimizing EDSS data analysis through appropriate modeling strategies can enhance the efficiency and sensitivity of clinical trials for MS therapies.