Sepsis, September 30 edition

The authors developed single-cell host-bacterial 16S co-sequencing to directly associate microbial signals with host transcriptional states. In Klebsiella pneumoniae sepsis, the approach identified discrete myeloid populations and implicated a CD38–NAD immunometabolic axis in impaired lysosomal acidification, providing a mechanistic link to defective bacterial clearance.

Impact: This study provides a technically innovative framework for resolving host-cell states defined by their associated microbial signals, overcoming a major limitation of conventional single-cell transcriptomics. The CD38–NAD axis represents a potentially actionable mechanism for restoring innate immune function.

Clinical Implications: The findings could support biomarker-guided identification of immune-cell states associated with impaired bacterial clearance and motivate development of therapies targeting CD38, NAD metabolism, or lysosomal function. Translation requires validation in human sepsis cohorts and confirmation that modulation improves outcomes without increasing inflammation or infection risk.

Future Directions: Future work should reproduce scH16S-seq findings across pathogens and patient subgroups, define whether CD38 or NAD metabolism can be safely modulated, and test whether the identified cell states predict treatment response or mortality.

The investigators engineered neutrophil-membrane-camouflaged nanoparticles containing a prodrug derived from 4-octyl-itaconate and retinol. In vitro and murine septic acute lung injury models, M@ORNPs targeted inflamed lungs, suppressed reactive oxygen species and NF-κB, STAT1, and NLRP3 signaling, reduced lung injury, and improved survival under lethal challenge.

Impact: This work addresses two major therapeutic barriers in septic lung injury—poor delivery to inflamed tissue and simultaneous inflammatory and oxidative amplification—using one biomimetic platform. The survival benefit in a lethal animal model strengthens its translational rationale, although clinical applicability remains unproven.

Clinical Implications: The platform could eventually provide targeted treatment for sepsis-associated acute lung injury and other inflammatory lung disorders, potentially reducing systemic exposure. Before clinical translation, pharmacokinetics, manufacturing consistency, immunogenicity, toxicity, dose optimization, and efficacy in clinically relevant polymicrobial sepsis models must be evaluated.

Future Directions: Further studies should assess biodistribution and safety after repeated dosing, compare efficacy with current supportive and anti-inflammatory strategies, test polymicrobial and clinically relevant sepsis models, and establish scalable production and regulatory specifications.

Using 98,233 adult ICU patients from three independent critical-care databases, the study compared nine algorithms to predict 28-day mortality using information available within 24 hours of ICU admission. XGBoost performed best, with AUROCs of 0.884 internally and 0.877 and 0.875 in two external cohorts; mechanical ventilation, comorbidity burden, lactate dehydrogenase, age, respiratory rate, and blood pressure were important predictors.

Impact: The large sample and independent external validation address important weaknesses of many single-center sepsis prediction studies. The use of routinely collected variables and explainable predictors improves the model’s potential for clinical integration, while prospective impact assessment is still required.

Clinical Implications: The model could support early mortality risk stratification, escalation of monitoring, and ICU resource allocation at the 24-hour landmark. It should not yet replace clinical assessment or determine treatment independently because prospective validation, workflow evaluation, calibration monitoring, and assessment of effects on patient outcomes are incomplete.

Future Directions: Future research should prospectively evaluate whether model-guided risk stratification changes treatment timing, monitoring, resource use, and mortality, while testing fairness across demographic and clinical subgroups and implementing site-specific recalibration.