Complementary role of transcriptomic endotyping and protein-based biomarkers for risk stratification in sepsis-associated acute kidney injury.

Summary

In 167 septic ICU patients, transcriptomic endotypes (IE, AE, CE) exhibited distinct biology and risks. Non-functional biomarkers mapped to endotypes (NGAL/suPAR high in IE; bio-ADM strongest in CE). Combining endotyping with bio-ADM or suPAR improved prediction of KRT/death (AUC 0.80) and 7-day mortality (AUC 0.85).

Key Findings

• Endotype distribution: IE 33%, AE 42%, CE 24%.
• Primary endpoint (KRT or death): IE 30% vs AE 17% vs CE 10%.
• NGAL and suPAR were disproportionately elevated in IE, independent of AKI severity.
• Bio-ADM was the strongest predictor of outcomes in CE.
• Endotyping + bio-ADM achieved AUC 0.80 (KRT/death); endotyping + suPAR achieved AUC 0.85 (7-day mortality).

Clinical Implications

Consider integrating endotyping and biomarkers (bio-ADM, suPAR, NGAL) to stratify SA-AKI risk: prioritize endothelial-targeted strategies for CE and immune-modulating strategies for IE; allocate monitoring and resources accordingly.

Limitations

• Secondary analysis with modest sample size (n=167) and no external validation
• Observational; clinical utility and treatment guidance not prospectively tested

Future Directions

External validation across centers; endotype-adaptive interventional trials targeting endothelial dysfunction or innate immune dysregulation.