Pentraxin-3, MyD88, GLP-1, and PD-L1: Performance assessment and composite algorithmic analysis for sepsis...

Area of Science:

• Biomarker discovery
• Clinical diagnostics
• Infectious disease research

Background:

• Accurate sepsis diagnosis is critical for initiating life-saving treatments.
• Novel host-response biomarkers are needed to aid clinicians in identifying acute infection and sepsis.

Purpose of the Study:

• To evaluate nine emerging host-response biomarkers for their diagnostic accuracy in patients with suspected acute infections.
• To assess the performance of a machine learning model combining these biomarkers for sepsis prediction.

Main Methods:

• Recruited 388 patients with suspected acute infections in the ED.
• Quantified nine biomarkers (MyD88, MMP-8, leptin, ENA-78, fractalkine, PD-L1, pentraxin-3, TRAIL, GLP-1) using a multiparameter assay.
• Performed AUROC analysis and machine learning (XGBoost) for bacterial infection, sepsis, and 30-day mortality endpoints.

Main Results:

• MyD88, PD-L1, and pentraxin-3 demonstrated high AUROCs for bacterial infection (≥0.87), sepsis (≥0.81), and 30-day mortality (≥0.71).
• Seven of nine biomarkers significantly discriminated across all three endpoints.
• An XGBoost model combining pentraxin-3, MyD88, and GLP-1 achieved an AUROC of 0.89 for sepsis prediction, outperforming NEWS-2 (0.83) and procalcitonin (0.81).

Conclusions:

• Pentraxin-3, MyD88, GLP-1, and PD-L1 are valuable biomarkers for risk assessment and stratification in sepsis.
• A multiparameter classifier using these combined biomarkers offers a valid tool for sepsis diagnosis.