Subphenotyping aneurysmal subarachnoid Hemorrhage using clinical and biological data clustering.

BACKGROUND

Aneurysmal subarachnoid hemorrhage (aSAH) is an heterogeneous disease with variable outcomes, even among patients with similar clinical and radiological severity. Additional research is needed to better stratify aSAH patients.

OBJECTIVES

To identify distinct clinical subphenotypes of aSAH, we applied clustering analysis using clinical, radiological, and laboratory data.

METHODS

We conducted a retrospective cohort study of adult patients with aSAH admitted to the ICU between 2010 and 2021. K-means clustering was applied to standardized demographic, clinical, and laboratory variables collected at admission. Principal component analysis was used for dimensionality reduction and visualization. Additionally, we analyzed whether these clusters were associated with serum biomarkers (S100B, HMGB1, and TLR4) in a subset of patients.

RESULTS

The study included 511 patients with aSAH. Two distinct subphenotypes were identified: a High-Risk Cluster (n = 301, 58.9 %) characterized by severe systemic complications, and higher mortality, and a Low-Risk Cluster (n = 210, 41.1 %) with less severe symptoms and better outcomes. Serum S100B levels were significantly elevated in the High-Risk Cluster (0.077 [0.056-0.179] vs. 0.055 [0.040-0.079] μg/L, p = 0.008) and showed moderate discriminatory power (AUC = 0.72).

CONCLUSIONS

Clustering analysis revealed two aSAH subphenotypes associated with DCI, mortality and functional. Integrating early clinical and biomarker data could enhance patient stratification.