Clinical subphenotypes and molecular endotypes in sepsis: toward an integrated and dynamic framework

Area of Science:

• Critical Care Medicine
• Genomics and Computational Biology
• Translational Medicine

Background:

• Sepsis is a life-threatening condition with high mortality, necessitating better patient stratification.
• Previous clinical trials have failed, highlighting the need for precise patient subgroups.
• Machine learning and multi-omics have identified clinical subphenotypes and molecular endotypes.

Purpose of the Study:

• To review advances in static and dynamic subphenotyping of sepsis.
• To discuss omics-derived endotypes for understanding sepsis pathophysiology.
• To outline strategies for integrating these into precision medicine frameworks.

Main Methods:

• Review of recent literature on sepsis subphenotyping and endotyping.
• Analysis of machine learning and multi-omics applications in sepsis research.
• Discussion of dynamic modeling integrating longitudinal data.

Main Results:

• Clinical subphenotypes offer aggregated views but may lack specific mechanisms.
• Molecular endotyping reveals pathophysiological drivers but often lacks bedside actionability.
• Dynamic subtyping integrating longitudinal data shows promise for optimizing patient stratification.

Conclusions:

• Sepsis is a dynamic process requiring approaches that capture temporal changes.
• Integrating static and dynamic subphenotyping with molecular endotyping is crucial.
• Development of clinically actionable frameworks for precision sepsis medicine is essential.