An algorithm using clinical data to predict the optimal individual glucocorticoid dosage to treat multiple sclerosis...

Area of Science:

• Neurology
• Immunology
• Pharmacology

Background:

• Glucocorticoid (GC) pulse therapy is a standard treatment for multiple sclerosis (MS) relapses.
• GC resistance is a significant clinical challenge, necessitating personalized dosing strategies.
• A predictive model to determine the maximum effective GC dose is crucial for optimizing MS relapse management.

Purpose of the Study:

• To develop and validate a predictive model for glucocorticoid (GC) dosing in multiple sclerosis (MS) relapse treatment.
• To identify key factors influencing GC response and dose requirements in MS patients.
• To support clinicians in estimating the optimal GC dose, avoiding sub-therapeutic or excessive administration.

Main Methods:

• Two independent retrospective cohorts of MS patients were established for model generation and validation.
• Multivariate regression analysis was employed, with GC dose as the dependent variable.
• Independent variables included serum vitamin D (25D) concentration, sex, age, EDSS, contrast enhancement on MRI, immunotherapy, and optic nerve involvement.

Main Results:

• Serum vitamin D (25D) concentration and the presence of optic neuritis were identified as independent predictors of the required GC dose in the explorative cohort (n=113).
• The developed multivariate linear regression model was validated in a second cohort (n=30), showing no significant difference between predicted and administered GC doses (p=0.173).
• The model demonstrated the ability to predict GC doses used in routine MS relapse care where further benefit is unlikely.

Conclusions:

• A validated model can predict the effective glucocorticoid (GC) dose for multiple sclerosis (MS) relapses, assisting clinicians in optimizing treatment.
• Serum vitamin D levels and optic neuritis presence are key predictors for GC dosing in MS relapse management.
• Further research is warranted to refine and implement this predictive algorithm into clinical practice for personalized GC therapy.