Unsupervised Clustering Identifies High-Risk Phenotypic Subgroup in Crescentic Glomerulonephritis Patients

Area of Science:

• Nephrology
• Internal Medicine
• Clinical Epidemiology

Background:

• Current crescentic glomerulonephritis (cGN) classification focuses on renal outcomes, with controversial ability to predict overall patient survival.
• Need for improved risk stratification for both overall survival and end-stage renal disease (ESRD) in cGN patients.

Purpose of the Study:

• To develop a risk stratification system for overall survival in cGN.
• To validate a renal risk score for predicting ESRD in cGN patients.

Main Methods:

• Retrospective analysis of 224 cGN patients.
• Utilized K-means clustering, principal component analysis, and decision tree analysis.
• Primary outcome: all-cause mortality; Secondary outcome: ESRD.

Main Results:

• K-means clustering identified high-risk and low-risk groups with significantly different 10-year survival rates (63.1% vs. 89.6%, p=0.004).
• High-risk status independently predicted mortality (HR=3.28).
• A cGN kidney risk score (serum creatinine, normal glomeruli percentage, tubular atrophy/interstitial fibrosis) independently predicted ESRD (p<0.001).

Conclusions:

• Identified two distinct survival risk groups in cGN.
• Developed a novel renal risk score for ESRD prediction.
• The developed risk stratification system provides comprehensive information for clinical management of cGN.