Skip to main content

Weekly Sepsis Research Analysis

3 papers

This week’s sepsis literature emphasized rapid diagnostics and early prediction, mechanistic insights into organ-specific injury, and precision interventions. High-impact reports validated time-series deep learning to predict bloodstream infections before culture results and prospectively validated a plasma ddPCR panel that dramatically increases Gram-negative detection. Mechanistic studies (tissue-resident bladder macrophages; S100A8/A9–RAGE–Drp1 mitochondrial pathway) nominate new prevention a

Summary

This week’s sepsis literature emphasized rapid diagnostics and early prediction, mechanistic insights into organ-specific injury, and precision interventions. High-impact reports validated time-series deep learning to predict bloodstream infections before culture results and prospectively validated a plasma ddPCR panel that dramatically increases Gram-negative detection. Mechanistic studies (tissue-resident bladder macrophages; S100A8/A9–RAGE–Drp1 mitochondrial pathway) nominate new prevention and therapeutic targets, while trials and guidelines push toward microcirculation-guided resuscitation and standardized outcome measures for neonates.

Selected Articles

1. A bladder-blood immune barrier constituted by suburothelial perivascular macrophages restrains uropathogen dissemination.

88.5Immunity · 2025PMID: 40015270

This mechanistic study identifies suburothelial perivascular macrophages (suPVMs) in the bladder that capture uropathogenic E. coli, maintain inflamed vessel integrity, and deploy METosis with MMP-13 to sequester bacteria and recruit neutrophils. Monocyte-derived replenishment of suPVMs after infection confers protection against recurrence, defining a tissue-resident bladder–blood immune barrier that limits systemic dissemination and suggests prevention strategies for urosepsis.

Impact: Reveals a previously unrecognized tissue-resident immune barrier that mechanistically limits UTI-to-urosepsis transition, opening new preventive and immunomodulatory strategies (macrophage training, METosis/MMP-13 modulation).

Clinical Implications: Translational work should validate suPVM signatures in human bladder tissue and test interventions to boost barrier function (vaccination, monocyte priming, targeted modulation of METosis/MMP-13) to reduce urosepsis risk.

Key Findings

  • Identified suburothelial perivascular macrophages (suPVMs) that capture UPEC and preserve inflamed vessel integrity.
  • suPVMs undergo METosis releasing extracellular DNA traps and MMP-13 to sequester bacteria and promote neutrophil transuroepithelial migration.
  • Monocyte-derived replenishment of suPVMs after prior infection confers protection against recurrent UTIs.

2. Elevated levels of S100A8 and S100A9 exacerbate muscle mitochondrial fragmentation in sepsis-induced muscle atrophy.

83Communications biology · 2025PMID: 40021770

This translational study links sepsis-associated skeletal muscle atrophy to an S100A8/A9–RAGE–Drp1 pathway that drives Drp1 phosphorylation, mitochondrial fission, and myocyte atrophy. The work integrates retrospective clinical association (ΔSMI linked to 60-day mortality) with mouse CLP models where inhibition of S100A8/A9, RAGE ablation, or Drp1 inhibition restored mitochondrial function and reduced atrophy, nominating a druggable axis to prevent ICU-acquired weakness.

Impact: Identifies a mechanistic, pharmacologically tractable pathway underlying septic myopathy, bridging clinical risk signal to validated in vivo and in vitro interventions.

Clinical Implications: Supports development of biomarker strategies (S100A8/A9) and early-phase trials of RAGE or Drp1 modulators to prevent or mitigate ICU-acquired muscle wasting and weakness in sepsis survivors.

Key Findings

  • ΔSMI (change in skeletal muscle index) was independently associated with 60-day mortality in septic patients.
  • Sepsis increased S100a8/a9 expression and mitochondrial dysfunction in mouse skeletal muscle; blocking S100a8/a9 or inhibiting Drp1 improved mitochondrial morphology and reduced atrophy.
  • S100a8/a9 binds RAGE to induce Drp1 phosphorylation and mitochondrial fragmentation; RAGE ablation mitigates these effects.

3. Utilising routinely collected clinical data through time series deep learning to improve identification of bacterial bloodstream infections: a retrospective cohort study.

81.5The Lancet. Digital health · 2025PMID: 40015765

In a large single-system retrospective cohort (n=20,850), an LSTM time-series model using up to 14 days of prior laboratory data predicted pathogenic bloodstream infections with AUROC 0.97 on a temporal hold-out set, substantially outperforming static models. CRP, eosinophil, and platelet trajectories were important predictors. The approach offers a path to earlier, individualized diagnostic decision support to guide expedited testing and stewardship.

Impact: Demonstrates that routinely collected longitudinal data can power high-performance, clinically actionable models to predict bloodstream infection ahead of culture, enabling earlier targeted diagnostics and antimicrobial stewardship interventions.

Clinical Implications: Integrating time-series LSTM models into EHR workflows could triage patients for expedited cultures/therapy and reduce unnecessary empiric antibiotics for low-risk patients; prospective impact and external multi-center validation are needed prior to deployment.

Key Findings

  • LSTM using up to 14 days of prior labs achieved AUROC 0.97 and AUPRC 0.65 in temporal hold-out testing, outperforming static logistic models (AUROC 0.74).
  • Time-series dynamics (CRP, eosinophils, platelets) were key predictors, and removing temporal information degraded performance, especially for hospital-acquired BSIs.
  • Large-scale retrospective dataset (n=20,850) with temporal hold-out validation demonstrates clinical feasibility.