A predictive model for corticosteroid response in individual patients with MS relapses

Area of Science:

• Neurology
• Clinical Trials
• Biostatistics

Background:

• Acute relapses in relapsing-remitting multiple sclerosis (MS) necessitate effective treatment strategies.
• Corticosteroids are commonly used, but predicting individual response remains challenging.

Purpose of the Study:

• To develop a predictive model for corticosteroid response in MS relapses.
• To identify clinical variables guiding corticosteroid use.

Main Methods:

• Analysis of randomized controlled trial data (n=98) using binary logistic regression.
• Inclusion of age, gender, baseline disability (EDSS, MSIS-29), and treatment timing.
• Evaluation of response based on EDSS improvement cut-offs (≥0.5 and ≥1.0).

Main Results:

• Younger age and lower MSIS-29 physical scores predicted better corticosteroid response at 6 weeks.
• The predictive model achieved 71.2% - 73.1% fit.
• Identified key predictors for treatment efficacy.

Conclusions:

• Pilot study suggests age and MSIS-29 physical score as predictors of corticosteroid response in MS relapses.
• Highlights potential for a simple predictive model.
• Replication in larger prospective studies is recommended to distinguish treatment response from spontaneous recovery.