Establishment of a new classification system for chronic inflammatory demyelinating polyneuropathy based on...

Area of Science:

• Neurology
• Immunology

Background:

• Predicting treatment response in chronic inflammatory demyelinating polyneuropathy (CIDP) is challenging due to varied clinical presentations.
• A robust classification model for CIDP patients is needed to guide therapeutic strategies.

Purpose of the Study:

• To develop and validate a novel classifier for CIDP patients.
• To identify key clinical, laboratory, and electrophysiological features influencing disease severity and treatment responsiveness.

Main Methods:

• Unsupervised hierarchical clustering was applied to data from 172 treatment-naïve CIDP patients.
• A tree-based model was constructed using pivotal features identified through clustering.
• Patient clusters were analyzed for differences in disability, demyelination, comorbidities, and response to pulse steroid therapy.

Main Results:

• Three distinct patient clusters emerged, differing in baseline disability, fulfillment of demyelinating criteria, and prevalence of chronic kidney disease and hypoalbuminemia.
• Cluster 2 showed significantly higher responsiveness to pulse steroid therapy compared to Clusters 1 and 3.
• The developed tree-based model accurately classified CIDP patients into these clusters with 89.5% accuracy.

Conclusions:

• The established classification system effectively identifies key features associated with CIDP heterogeneity.
• This model can aid clinicians in selecting appropriate treatments for individual CIDP patients.
• The classification facilitates patient stratification for clinical trials, improving research efficiency.