Weekly ReportJun 23–29, 2026
Sepsis, week 26 edition
We read 236 papers and selected 3.
Summary
This week’s sepsis literature emphasized mechanistic advances linking cellular stress and programmed cell death to organ injury (notably ER stress/IκBζ–XBP1s and DAPK2–HSPA5–IRE1α axes), alongside pragmatic translational and clinical work that could change practice (an LLM-enabled quality-improvement RCT improving SEP-1 compliance and randomized evidence on oxygen targets). Diagnostic and prognostic innovation continued with head-to-head presepsin data and non-invasive microbial cfDNA targeted sequencing supporting faster, more specific pathogen detection. Computational microvascular modeling and large-scale model meta-analysis offered frameworks to re-evaluate hemodynamic management and prognostic tool deployment.
Selected Articles
1. ER stress amplifies inflammation via a dual mechanism involving IκBζ-XBP1s synergism and Regnase-1 degradation.
This mechanistic study shows ER stress amplifies IL-6 and selected secondary-response genes by (1) stabilizing Nfkbiz mRNA through IKK-dependent Regnase-1 degradation and (2) transcriptional synergy between IκBζ and XBP1s; both layers were required for excessive IL-6 in septic mice, nominating IκBζ accumulation and Regnase-1 preservation as therapeutic angles.
Impact: Delineates a dual transcriptional/post-transcriptional ER stress mechanism driving hyperinflammation in sepsis with in vivo validation, revealing actionable molecular nodes (IκBζ, XBP1s, Regnase-1) for therapeutic development.
Clinical Implications: Targeting IκBζ accumulation, preserving Regnase-1 function, or disrupting IκBζ–XBP1s cooperativity could attenuate IL-6–driven immunopathology in sepsis; translation requires selective modulators and biomarker-driven patient selection.
Key Findings
- ER stress synergizes with TLR signaling to strongly upregulate IκBζ in macrophages.
- Ca2+-dependent IKK activity degrades Regnase-1, stabilizing Nfkbiz mRNA and promoting IκBζ accumulation.
- IκBζ cooperates with XBP1s to drive selective secondary-response genes (e.g., Il6, Nos2) and this synergy is required for excessive IL-6 in septic mice.
2. Death-associated protein kinase 2 (DAPK2) propagates endoplasmic reticulum stress in macrophages to worsen sepsis through HSPA5-IRE1α axis.
This translational preclinical study identifies macrophage DAPK2 as a TLR4–MyD88–NF-κB–induced kinase that phosphorylates HSPA5 at Ser588, promoting its proteasomal degradation and consequent IRE1α activation; macrophage-specific DAPK2 deletion reduced ER stress and sepsis severity, positioning the DAPK2–HSPA5–IRE1α axis as a druggable pathway.
Impact: Uncovers a kinase–chaperone mechanism directly linking innate immune activation to macrophage ER stress and organ injury, backed by genetic perturbation and proteomic interactome data—high translational potential.
Clinical Implications: Pharmacologic inhibition of DAPK2 or strategies to stabilize HSPA5/attenuate IRE1α signaling could mitigate macrophage ER stress in sepsis; DAPK2 expression may serve as a stratification biomarker for targeted trials.
Key Findings
- DAPK2 is transcriptionally upregulated in sepsis macrophages via TLR4–MyD88–NF-κB.
- DAPK2 phosphorylates HSPA5 at Ser588, promoting proteasomal degradation and activating IRE1α.
- Macrophage-specific DAPK2 deletion attenuates ER stress and sepsis severity; pathway perturbation studies confirm causality.
3. Medical Record Abstraction for Quality Improvement in Sepsis Care Using Artificial Intelligence: A Cluster Randomized Trial.
In a single-blind cluster RCT across two EDs, near-real-time LLM-enabled abstraction and targeted feedback improved SEP-1 compliance from 70.1% to 82.9% (absolute +13%; OR 2.10, P=0.02) with 92% agreement between LLM determinations and expert review, though no difference in ICU admissions or 30-day mortality was observed.
Impact: Provides randomized evidence that scalable LLM-driven record abstraction with timely feedback can materially improve sepsis process adherence (SEP-1), establishing feasibility, effect size, and high LLM–expert concordance for health-system implementation.
Clinical Implications: Health systems can pilot LLM-based near-real-time SEP-1 feedback to improve bundle adherence; implementation should monitor for documentation-driven biases and be paired with trials assessing patient-centered outcomes.
Key Findings
- LLM-enabled feedback increased SEP-1 compliance from 70.1% to 82.9% (absolute +13%; OR 2.10; P=0.02).
- Largest improvement was in completion of the 30 mL/kg fluid bolus component (documentation-sensitive).
- LLM determinations agreed with expert review at 92%; no change in ICU admissions or 30-day mortality detected.