Subphenotyping aneurysmal subarachnoid Hemorrhage using clinical and biological data clustering

Area of Science:

• Neurosurgery
• Intensive Care Medicine
• Biomarker Research

Background:

• Aneurysmal subarachnoid hemorrhage (aSAH) presents heterogeneous outcomes despite similar initial severity.
• Improved patient stratification is crucial for optimizing aSAH management.
• Current methods may not fully capture the variability in aSAH patient trajectories.

Purpose of the Study:

• To identify distinct clinical subphenotypes within the aSAH patient population.
• To utilize clustering analysis on comprehensive patient data for subphenotype discovery.
• To explore the association of identified subphenotypes with clinical outcomes and biomarkers.

Main Methods:

• Retrospective cohort study of 511 aSAH patients admitted to the ICU (2010-2021).
• K-means clustering applied to standardized demographic, clinical, and laboratory admission data.
• Principal component analysis for dimensionality reduction; analysis of serum biomarkers (S100B, HMGB1, TLR4).

Main Results:

• Two distinct aSAH subphenotypes identified: High-Risk (58.9%) and Low-Risk (41.1%).
• High-Risk cluster exhibited severe systemic complications and higher mortality.
• Elevated serum S100B levels in the High-Risk cluster (p=0.008) with moderate discriminatory power (AUC=0.72).

Conclusions:

• Clustering analysis successfully delineated two aSAH subphenotypes with differential associations with outcomes.
• Subphenotypes correlate with delayed cerebral ischemia (DCI), mortality, and functional status.
• Integrating early clinical data and biomarkers can significantly enhance aSAH patient stratification.