Biomarker-Based Predictive Enrichment in Critically Ill Sepsis Patients: Systematic Review

Overview

This systematic review identifies randomized controlled trials (RCTs) employing biomarker-based predictive enrichment in critically ill sepsis patients. It characterizes biomarkers used to select patient subgroups more likely to respond to interventions and evaluates the impact of such strategies on trial outcomes.

Background

Sepsis and septic shock are leading causes of ICU admission with high mortality rates ranging from 30% to 50%. Despite numerous RCTs, few have demonstrated effective targeted therapies, partly due to heterogeneity of treatment effect (HTE) within sepsis populations. Predictive enrichment using biomarkers to identify responsive subphenotypes has transformed treatment in other fields but remains underutilized in sepsis research. This review aims to assess the use and impact of biomarker-based predictive enrichment in sepsis RCTs.

Data Highlights

The review included RCTs enrolling adult patients meeting Sepsis II or III criteria, using blood-based genomic, transcriptomic, proteomic, or metabolomic biomarkers measured at enrollment for predictive enrichment. Studies were identified through systematic searches in Medline, Embase, and Cochrane Central up to January 15, 2025. Data extracted included study characteristics, biomarker types and thresholds, and outcomes. Risk of bias was assessed using the Cochrane Risk of Bias 2.0 tool.

Key Findings

• Sepsis mortality remains high despite advances in supportive care and standardized protocols.
• Heterogeneity of treatment effect dilutes therapeutic signals in unselected sepsis populations, contributing to negative trial results.
• Predictive enrichment strategies use biomarkers to identify biologically distinct subgroups more likely to benefit from targeted interventions.
• Biomarker-based enrichment has revolutionized treatment in oncology, rheumatology, and neurology but is rarely applied in sepsis RCTs.
• The review identified RCTs that used blood-based biomarkers at enrollment for patient selection, focusing on genomic, proteomic, transcriptomic, or metabolomic profiles.
• Barriers to wider adoption include complexity of sepsis pathobiology, biomarker validation challenges, and trial design limitations.

Clinical Implications

Incorporating biomarker-based predictive enrichment in sepsis trials may improve identification of patient subgroups that respond to specific therapies, potentially enhancing trial success and clinical outcomes. Clinicians should be aware of emerging biomarker tools that could guide precision treatment approaches in sepsis. Future research should focus on validating biomarkers and integrating them into trial designs to overcome heterogeneity challenges.

Conclusion

Biomarker-based predictive enrichment holds promise to address heterogeneity in sepsis treatment response, but its application in critical care trials remains limited. Expanding this approach could facilitate precision medicine strategies and improve therapeutic efficacy in sepsis.

References

1. Surviving Sepsis Campaign -- Standardized best practices in sepsis care
2. Cochrane Handbook/PRISMA -- Systematic review methodology
3. PROSPERO ID: CRD420250653215 -- Registered systematic review protocol