Explore the vast range of titles available.

Inflammation induced by lung infection is a double-edged sword, moderating both anti-viral and immune pathogenesis effects; the mechanism of the latter is not fully understood. Previous studies suggest the vasculature is involved in tissue injury. Here, we report that expression of Sparcl1, a secreted matricellular protein, is upregulated in pulmonary capillary endothelial cells (EC) during influenza-induced lung injury. Endothelial overexpression of SPARCL1 promotes detrimental lung inflammation, with SPARCL1 inducing ‘M1-like’ macrophages and related pro-inflammatory cytokines, while SPARCL1 deletion alleviates these effects. Mechanistically, SPARCL1 functions through TLR4 on macrophages in vitro, while TLR4 inhibition in vivo ameliorates excessive inflammation caused by endothelial Sparcl1 overexpression. Finally, SPARCL1 expression is increased in lung ECs from COVID-19 patients when compared with healthy donors, while fatal COVID-19 correlates with higher circulating SPARCL1 protein levels in the plasma. Our results thus implicate SPARCL1 as a potential prognosis biomarker for deadly COVID-19 pneumonia and as a therapeutic target for taming hyperinflammation in pneumonia. The molecular basis underlying infection infection-mediated lung pathology is not fully revealed. Here the authors report that SPARCL1 expressed in pulmonary capillary endothelial cells contributes to immune pathology in mouse model via pro-inflammatory macrophage induction, while circulating SPARCL1 levels corelate with COVID-19 lethality.

Red blood cells (RBCs), traditionally recognized for their role in transporting oxygen, play a pivotal role in the body’s immune response by expressing TLR9 and scavenging excess host cell-free DNA. DNA capture by RBCs leads to accelerated RBC clearance and triggers inflammation. Whether RBCs can also acquire microbial DNA during infections is unknown. Murine RBCs acquire microbial DNA in vitro, and bacterial DNA–induced (bDNA-induced) macrophage activation was augmented by WT, but not Tlr9-deleted, RBCs. In a mouse model of polymicrobial sepsis, RBC-bound bDNA was elevated in WT mice but not in erythroid Tlr9–deleted mice. Plasma cytokine analysis in these mice revealed distinct sepsis clusters characterized by persistent hypothermia and hyperinflammation in the most severely affected mice. RBC Tlr9 deletion attenuated plasma and tissue IL-6 production in the most severely affected group. Parallel findings in humans confirmed that RBCs from patients with sepsis harbored more bDNA than did RBCs from healthy individuals. Further analysis through 16S sequencing of RBC-bound DNA illustrated distinct microbial communities, with RBC-bound DNA composition correlating with plasma IL-6 in patients with sepsis. Collectively, these findings unveil RBCs as overlooked reservoirs and couriers of microbial DNA, capable of influencing host inflammatory responses in sepsis.

PurposeAlthough dozens of studies have associated vancomycin + piperacillin–tazobactam with increased acute kidney injury (AKI) risk, it is unclear whether the association represents true injury or a pseudotoxicity characterized by isolated effects on creatinine secretion. We tested this hypothesis by contrasting changes in creatinine concentration after antibiotic initiation with changes in cystatin C concentration, a kidney biomarker unaffected by tubular secretion.MethodsWe included patients enrolled in the Molecular Epidemiology of SepsiS in the ICU (MESSI) prospective cohort who were treated for ≥ 48 h with vancomycin + piperacillin–tazobactam or vancomycin + cefepime. Kidney function biomarkers [creatinine, cystatin C, and blood urea nitrogen (BUN)] were measured before antibiotic treatment and at day two after initiation. Creatinine-defined AKI and dialysis were examined through day-14, and mortality through day-30. Inverse probability of treatment weighting was used to adjust for confounding. Multiple imputation was used to impute missing baseline covariates.ResultsThe study included 739 patients (vancomycin + piperacillin–tazobactam n = 297, vancomycin + cefepime n = 442), of whom 192 had cystatin C measurements. Vancomycin + piperacillin–tazobactam was associated with a higher percentage increase of creatinine at day-two 8.04% (95% CI 1.21, 15.34) and higher incidence of creatinine-defined AKI: rate ratio (RR) 1.34 (95% CI 1.01, 1.78). In contrast, vancomycin + piperacillin–tazobactam was not associated with change in alternative biomarkers: cystatin C: − 5.63% (95% CI − 18.19, 8.86); BUN: − 4.51% (95% CI − 12.83, 4.59); or clinical outcomes: dialysis: RR 0.63 (95% CI 0.31, 1.29); mortality: RR 1.05 (95%CI 0.79, 1.41).ConclusionsVancomycin + piperacillin–tazobactam was associated with creatinine-defined AKI, but not changes in alternative kidney biomarkers, dialysis, or mortality, supporting the hypothesis that vancomycin + piperacillin–tazobactam effects on creatinine represent pseudotoxicity.

As big data, machine learning, and artificial intelligence continue to penetrate into and transform many facets of our lives, we are witnessing the emergence of these powerful technologies within health care. The use and growth of these technologies has been contingent on the availability of reliable and usable data, a particularly robust resource in critical care medicine where continuous monitoring forms a key component of the infrastructure of care. The response to this opportunity has included the development of open databases for research and other purposes; the development of a collaborative form of clinical data science intended to fully leverage these data resources, and the creation of data-driven applications for purposes such as clinical decision support. Most recently, data levels have reached the thresholds required for the development of robust artificial intelligence features for clinical purposes. The systematic capture and analysis of clinical data in both individuals and populations allows us to begin to move toward precision medicine in the intensive care unit (ICU). In this perspective review, we examine the fundamental role of data as we present the current progress that has been made toward an artificial intelligence (AI)-supported, data-driven precision critical care medicine.

Illness severity scores are regularly employed for quality improvement and benchmarking in the intensive care unit, but poor generalization performance, particularly with respect to probability calibration, has limited their use for decision support. These models tend to perform worse in patients at a high risk for mortality. We hypothesized that a sequential modeling approach wherein an initial regression model assigns risk and all patients deemed high risk then have their risk quantified by a second, high-risk-specific, regression model would result in a model with superior calibration across the risk spectrum. We compared this approach to a logistic regression model and a sophisticated machine learning approach, the gradient boosting machine. The sequential approach did not have an effect on the receiver operating characteristic curve or the precision-recall curve but resulted in improved reliability curves. The gradient boosting machine achieved a small improvement in discrimination performance and was similarly calibrated to the sequential models.

Red blood cells (RBCs), traditionally recognized for their role in transporting oxygen, play a pivotal role in the body's immune response by expressing TLR9 and scavenging excess host cell-free DNA. DNA capture by RBCs leads to accelerated RBC clearance and triggers inflammation. Whether RBCs can also acquire microbial DNA during infections is unknown. Murine RBCs acquire microbial DNA in vitro, and bacterial DNA-induced (bDNA-induced) macrophage activation was augmented by WT, but not TIr9-deleted, RBCs. In a mouse model of polymicrobial sepsis, RBC-bound bDNA was elevated in WT mice but not in erythroid TIr9-deleted mice. Plasma cytokine analysis in these mice revealed distinct sepsis clusters characterized by persistent hypothermia and hyperinflammation in the most severely affected mice. RBC TIr9 deletion attenuated plasma and tissue IL-6 production in the most severely affected group. Parallel findings in humans confirmed that RBCs from patients with sepsis harbored more bDNA than did RBCs from healthy individuals. Further analysis through 16S sequencing of RBCbound DNA illustrated distinct microbial communities, with RBC-bound DNA composition correlating with plasma IL-6 in patients with sepsis. Collectively, these findings unveil RBCs as overlooked reservoirs and couriers of microbial DNA, capable of influencing host inflammatory responses in sepsis.

Large language models (LLMs) are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm Assessment for Risk in Medicine), a benchmark using 100 real primary care-to-specialist consultation cases to measure frequency and severity of harm from LLM-generated medical recommendations. NOHARM covers 10 specialties, with 12,747 expert annotations for 4,249 clinical management options. Across 31 LLMs, potential for severe harm from LLM recommendations occurs in up to 22.2% (95% CI 21.6-22.8%) of cases, with harm of omission accounting for 76.6% (95% CI 76.4-76.8%) of errors. Safety performance is only moderately correlated (r = 0.61-0.64) with existing AI and medical knowledge benchmarks. The best models outperform generalist physicians on safety (mean difference 9.7%, 95% CI 7.0-12.5%), and a diverse multi-agent approach improves safety compared to solo models (mean difference 8.0%, 95% CI 4.0-12.1%). Therefore, despite strong performance on existing evaluations, widely used AI models can produce severely harmful medical advice at nontrivial rates, underscoring clinical safety as a distinct performance dimension necessitating explicit measurement.

Inflammation upon infectious lung injury is a double-edged sword: while tissue-infiltrating immune cells and cytokines are necessary to control infection, these same factors often aggravate injury. Full appreciation of both the sources and targets of inflammatory mediators is required to facilitate strategies to maintain antimicrobial effects while minimizing off-target epithelial and endothelial damage. Recognizing that the vasculature is centrally involved in tissue responses to injury and infection, we observed that pulmonary capillary endothelial cells (ECs) exhibit dramatic transcriptomic changes upon influenza injury punctuated by profound upregulation of Sparcl1 . Endothelial deletion and overexpression of SPARCL1 implicated this secreted matricellular protein in driving key pathophysiologic symptoms of pneumonia, which we demonstrate result from its effects on macrophage polarization. SPARCL1 induces a shift to a pro-inflammatory \"M1-like\" phenotype (CD86 + CD206 - ), thereby increasing associated cytokine levels. Mechanistically, SPARCL1 acts directly on macrophages in vitro to induce the pro-inflammatory phenotype via activation of TLR4, and TLR4 inhibition in vivo ameliorates inflammatory exacerbations caused by endothelial Sparcl1 overexpression. Finally, we confirmed significant elevation of SPARCL1 in COVID-19 lung ECs in comparison with those from healthy donors. Survival analysis demonstrated that patients with fatal COVID-19 had higher levels of circulating SPARCL1 protein compared to those who recovered, indicating the potential of SPARCL1 as a biomarker for prognosis of pneumonia and suggesting that personalized medicine approaches might be harnessed to block SPARCL1 and improve outcomes in high-expressing patients.Inflammation upon infectious lung injury is a double-edged sword: while tissue-infiltrating immune cells and cytokines are necessary to control infection, these same factors often aggravate injury. Full appreciation of both the sources and targets of inflammatory mediators is required to facilitate strategies to maintain antimicrobial effects while minimizing off-target epithelial and endothelial damage. Recognizing that the vasculature is centrally involved in tissue responses to injury and infection, we observed that pulmonary capillary endothelial cells (ECs) exhibit dramatic transcriptomic changes upon influenza injury punctuated by profound upregulation of Sparcl1 . Endothelial deletion and overexpression of SPARCL1 implicated this secreted matricellular protein in driving key pathophysiologic symptoms of pneumonia, which we demonstrate result from its effects on macrophage polarization. SPARCL1 induces a shift to a pro-inflammatory \"M1-like\" phenotype (CD86 + CD206 - ), thereby increasing associated cytokine levels. Mechanistically, SPARCL1 acts directly on macrophages in vitro to induce the pro-inflammatory phenotype via activation of TLR4, and TLR4 inhibition in vivo ameliorates inflammatory exacerbations caused by endothelial Sparcl1 overexpression. Finally, we confirmed significant elevation of SPARCL1 in COVID-19 lung ECs in comparison with those from healthy donors. Survival analysis demonstrated that patients with fatal COVID-19 had higher levels of circulating SPARCL1 protein compared to those who recovered, indicating the potential of SPARCL1 as a biomarker for prognosis of pneumonia and suggesting that personalized medicine approaches might be harnessed to block SPARCL1 and improve outcomes in high-expressing patients.

We introduce the first generative foundation model for chest radiograph synthesis trained from scratch at the billion-parameter scale. Existing radiographic AI models often suffer from poor generalisation across patient subpopulations, institutions, and acquisition settings, resulting in limited real-world clinical utility. Controlled, high-fidelity synthesis of chest radiographs is a promising path toward diversifying clinical datasets and evaluating the robustness of diagnostic models. Therefore, we present the largest specialist generative foundation model for chest radiographs to date, with over 1.3B parameters, trained for 1.6T tokens on a curated, heterogeneous dataset comprising 1.2M radiographs and clinical expert-guided metadata. Our model supports controllable radiograph generation and editing across multiple demographic subgroups, acquisition views, and a dozen pathologies. Moreover, we significantly advance the state of the art in radiograph synthesis fidelity, producing images that are indistinguishable from real radiographs to clinical experts.