Intensive Care Medicine Experimental最新文献

Background: Several approaches for setting PEEP in patients with (COVID-19-related) ARDS have been proposed. It is unclear whether a best approach exist, and how the recommended PEEP and resulting transpulmonary pressure, overdistension and collapse relate.

Objectives: To compare approaches based on electrical impedance tomography (EIT) (including targeting the crossing point of overdistension/collapse curves, EIT_CP) with targeting positive end-expiratory transpulmonary pressure (P_L,EE) and targeting highest respiratory system compliance (C_RS).

Methods: Post-hoc analysis of 29 patients with COVID-19-related ARDS from cohorts of two Dutch hospitals. Patients underwent a decremental PEEP trial, while EIT data and esophageal pressure data were recorded. We compared the recommended PEEP, as well as resulting P_L,EE and amounts of overdistension and collapse at the suggested PEEP.

Results: Targeting EIT_CP resulted in higher recommended PEEP (14 [12-16] cmH₂O) compared to a positive P_L,EE (12 [8-14] cmH₂O), while highest C_RS resulted in intermediate PEEP levels. Individually, the difference between the highest and lowest recommended PEEP level were 6 [4-8] cmH₂O. P_L,EE at the recommended PEEP was generally higher when targeting EIT_CP compared to and positive P_L,EE (1.4 [0.6-2.1] cmH₂O). The amount of collapse was lowest with EIT_CP (3.0 [2.0-4.0]%) and highest when targeting P_L,EE (5.4 [2.0-12.0]%). No significant differences in the amount of overdistension were found. Targeting positive P_L,EE resulted in 51% patients with high (> 10%) values for either overdistension or collapse, more than any other method.

Conclusions: Targeting EIT_CP results in slightly higher recommended PEEP and P_L,EE levels compared to positive P_L,EE, leading to less collapse, but not more overdistension. EIT-based methods protect better against high values of either overdistension or collapse.

Background: The ventilatory ratio (VR) is frequently used as a surrogate marker of ventilatory efficiency in patients with ARDS. However, its ability to reflect changes in alveolar ventilation (V̇Talv/VT) when respiratory mechanics are modified remains unknown. This study aimed to evaluate the relationship between VR and V̇Talv/VT during sequential changes in respiratory mechanics=, tidal volume (VT), and minute ventilation (V̇E) in patients with ARDS.

Methods: This was a secondary analysis of a quasi-experimental, repeated-measures study conducted in a single-center adult ICU. Twenty-two patients with ARDS were evaluated across three sequential 60 min controlled periods, during which trunk inclination was adjusted to induce changes in VT. At the end of each period, VR was calculated, and V̇Talv/VT was measured using volumetric capnography. A total of 66 paired measurements were analyzed in this study.

Results: By design, VT increased from Time 1 to Time 2 by + 62 mL and decreased from Time 2 to Time 3 by - 68 mL. These changes in VT were associated with the following: VR was not significantly different between Time 1 and Time 2 [- 0.23 (95% CI: - 0.44 to - 0.02; p = 0.071)] or between Time 2 and Time 3 [+ 0.17 (95% CI: - 0.04 to + 0.38; p = 0.086)]. The alveolar ventilation ratio (V̇Talv/VT) increased significantly from Time 1 to Time 2 by + 0.080 (95% CI: + 0.039 to + 0.121; p < 0.001), and decreased from Time 2 to Time 3 by - 0.060 (95% CI: - 0.101 to - 0.019; p < 0.001). Association between VR and V̇Talv/VT: no significant relationship was found (β =  - 0.056, marginal R² = 0.052, conditional R² = 0.205, p = 0.111).

Conclusions: In this cohort of patients with ARDS, VR did not correlate with V̇Talv/VT following controlled modifications of respiratory mechanics. These findings suggest that VR may not reliably represent ventilatory efficiency under changing ventilatory conditions, and its use as a surrogate variable should be approached with caution.

Background: Sepsis is a major cause of morbidity and mortality worldwide, with its heterogeneous and dynamically evolving clinical presentation complicating diagnosis, treatment, and prognosis. The identification of clinically meaningful sub-phenotypes within the sepsis population could help tailor interventions and improve outcomes. However, existing phenotyping studies have yielded inconsistent results with limited clinical utility. In this study, we propose a novel, guided machine-learning approach to identify clinically relevant sub-phenotypes within the sepsis condition by integrating deep representation learning with prediction-guided clustering to capture temporal disease trajectories.

Methods: We trained a recurrent neural network-based encoder to generate compact, predictive representations of sepsis patients over time. During training, the encoder is guided by four auxiliary prediction objectives (i.e., 90-day mortality, remaining length of stay, need for mechanical ventilation, and need for renal replacement therapy), which encourage the model to create representations that are relevant with respect to patient-centred outcomes. After training, patient representations were clustered using the K-means algorithm. The identified sub-phenotypes were compared across two large ICU data sets (AmsterdamUMCdb and MIMIC-IV) and interpreted using Integrated Gradients-based attribution maps. Practical and clinical utility of the phenotypes was evaluated using a reinforcement learning framework to evaluate optimal treatment strategies within each sepsis sub-phenotype.

Results: Through our approach, we identified six clinically distinct sub-phenotypes with varying risk profiles and presentations. The learned representations demonstrated robust generalisability across the different data sets, and the reinforcement learning results indicated that the different sub-phenotypes were associated with different optimal treatment strategies, highlighting the potential for phenotype-informed decision-making.

Conclusions: This study introduces a flexible and effective framework for the identification of robust and clinically meaningful sub-phenotypes within the population of sepsis patients. Moreover, the identified sub-phenotypes are clinically interpretable, and the proposed trajectory-aware phenotyping approach may support the future development of personalised and precision medicine strategies.

Background: Acute Respiratory Distress Syndrome (ARDS) is characterized by severe hypoxemia from heterogeneous impairments in regional ventilation and perfusion. Metabolic syndrome, a combination of central obesity, insulin resistance, hypertension, and dyslipidemia, affects over one-third of adults worldwide and is associated with systemic inflammation, endothelial dysfunction, and increased ARDS risk. Whether metabolic syndrome alters regional ventilation and perfusion distributions in ARDS remains unknown.

Methods: We performed a retrospective cohort study of 25 mechanically ventilated patients with ARDS evaluated by the Massachusetts General Hospital Lung Rescue Team from 2020 to 2025. After a recruitment maneuver and decremental PEEP titration in the supine position, we quantified regional ventilation and perfusion distributions using electrical impedance tomography across a 32 × 32 pixel impedance matrix. We used Bayesian regression modeling adjusted for age, severity of illness, and ARDS etiology to define associations between metabolic syndrome and regional distributions.

Results: 25 patients were included, of whom 44% (n = 11) had metabolic syndrome. The mean age was 52 (16) years, PaO₂/FiO₂ was 151 (61), and 48% (n = 12) had pulmonary ARDS. Patients with metabolic syndrome were younger and had higher respiratory system compliance. In dorsal lung regions, metabolic syndrome was associated with reduced ventilation distribution (adjusted mean difference: - 5.84%, 95% credible interval: - 12.29 to 1.17, posterior probability of decrease = 95.2%) and modest, uncertain reductions in perfusion distribution (- 2.01%, 95% credible interval: - 9.25 to 5.41). The proportion of pixels with a low ventilation-to-perfusion impedance ratio showed a trend toward increase in dorsal lung regions in patients with metabolic syndrome (4.25%, 95% credible interval: - 4.49 to 13.1, posterior probability = 80.3%). Body mass index accounted for 40% of the difference in dorsal ventilation. The association between regional ventilation and perfusion distributions was similar between groups (interaction coefficient: 0.01, 95% credible interval: - 0.34 to 0.36).

Conclusions: In ARDS, metabolic syndrome may be associated with reduced ventilation distribution to dorsal lung regions with smaller changes in perfusion distribution. The association between regional ventilation and perfusion distributions did not differ by metabolic syndrome. These exploratory findings suggest metabolic syndrome may primarily affect ventilation distribution rather than perfusion distribution in ARDS.

Background: Sepsis and septic shock remain major causes of morbidity and mortality worldwide, and management is largely based on source control, antimicrobial therapy, and supportive care. Despite limited high-quality evidence, adjunctive therapies targeting the dysregulated host response involving immune dysfunction, coagulopathy, and endothelial injury, steroids, vasopressors, and adjunctive therapies are frequently used in clinical practice. This survey aimed to describe real-world patterns of sepsis therapy across Europe.

Methods: We conducted an open, web-based, multinational survey endorsed by the European Society of Intensive Care Medicine (ESICM) and the Italian Society of Anesthesia, Analgesia, Resuscitation and Intensive Care (SIAARTI). The survey was distributed through professional mailing lists, newsletters, and national society networks. A structured 30-item questionnaire collected information on respondent demographics, ICU characteristics, availability of adjunctive therapies, clinical indications, triggers for initiation, and duration of treatment. Participation was voluntary and anonymous. A total of 442 physicians completed the survey. Data were analyzed descriptively and are presented as proportions and frequencies.

Results: More than 80% of respondents reported use of at least one adjunctive therapy for septic shock within the previous year. Corticosteroids were used by over 90% of clinicians, predominantly hydrocortisone for septic shock. Considerable variability was observed regarding indications, timing of initiation, and duration of therapy. Extracorporeal blood purification techniques were used by approximately 75% of respondents, most frequently hemoadsorption in patients with refractory shock; high cost and limited availability were the main barriers to broader implementation. Intravenous immunoglobulins were used by approximately one-third of clinicians, often guided by measured immunoglobulin levels or perceived immune dysfunction. Additional vasoactive and inotropic agents, including levosimendan, methylene blue, and beta-blockers, were employed in selected cases. In contrast, specific immunomodulatory therapies such as interleukin (IL)-1receptor antibodies were rarely used. Across all adjunctive strategies, marked heterogeneity in practice patterns was evident.

Conclusion: Adjunctive therapies are widely used in European ICUs, particularly in patients with severe or refractory sepsis, despite limited supporting evidence. The substantial variability in practice highlights ongoing clinical uncertainty and underscores the need for well-designed randomized trials to inform evidence-based and individualized treatment strategies.

Background: Acute community-acquired pneumonia (CAP) is a leading cause of infection-related mortality worldwide. Endotoxemia, characterized by elevated plasma lipopolysaccharide (LPS), is a key driver of inflammation and thrombosis in Gram-negative sepsis and has been suggested to occur in severe pneumonia, irrespective of etiology. However, current immunoassays for LPS quantification lack sensitivity and specificity. We aimed to quantify plasma LPS in severe CAP patients, including COVID-19, using a validated mass spectrometry method, and to explore associations with immune activation, coagulation, gut translocation, and clinical outcomes.

Methods: In this prospective ancillary study of the LYMPHONIE cohort, we included 34 non-COVID-19 severe CAP (sCAP), 34 severe COVID-19 (sCOVID-19) and 34 matched healthy volunteers. Plasma LPS was measured by LC-MS/MS detecting 3-hydroxy fatty acids of lipid A. Clinical data, immune biomarkers, coagulation biomarkers, and gut injury markers were measured.

Results: Unexpectedly, median plasma LPS concentrations were significantly lower in sCAP patients (724 pmol/ml in sCAP; 750 pmol/ml in sCOVID-19) compared to healthy volunteers (1009 pmol/ml, p < 0.001). LPS levels did not correlate with severity scores or mortality. Low positive correlations were observed between LPS and markers of endothelial activation (sVCAM-1) and coagulation (D-dimer). However, patients with high LPS showed no increased risk of thrombotic or cardiovascular events.

Conclusions: Using a highly specific LC-MS/MS method, we found no evidence of increased circulating LPS in severe pneumonia patients, challenging the hypothesis of gut-derived endotoxemia as a major contributor to systemic inflammation in severe CAP, including COVID-19. Take-home message Using a highly specific mass-spectrometry assay, we found no evidence of elevated circulating lipopolysaccharide in severe community-acquired pneumonia, including COVID-19. These findings challenge the concept that gut-derived endotoxemia is a major driver of systemic inflammation in severe pneumonia. Tweet Mass spectrometry reveals no rise in plasma LPS in severe pneumonia or COVID-19, questioning gut endotoxemia's role in inflammation.

Background: The Aggregate Index of Systemic Inflammation (AISI) is a novel index based on platelets, neutrophils, and lymphocytes associated with the prognosis of patients with cancer and infectious diseases. However, its application in acute ischemic stroke (AIS) has rarely been reported. This study evaluated stroke prognosis using AISI by examining the relationship between levels of systemic immunoinflammatory indices at admission and patient outcomes at different times after onset.

Methods: This was a retrospective cohort study. The data from 1222 stroke patients were obtained from multiparameter intelligent monitoring in the Intensive Care III database (MIMIC-III). Cox proportional risk model was conducted to estimate the relation between AISI, all-cause mortality, and ischemic. The findings were further validated with restricted cubic spline (RCS) and subgroup analyses.

Results: A total of 1222 patients with AIS were classified into tertiles based on AISI levels, tertile 1: low AISI, AISI levels less than 3881 (n = 408), tertile 2: medium AISI, AISI levels 3881 to 8354 (n = 408), and tertile 3: high AISI, AISI levels greater than 8354 (n = 407). After adjusting for multiple covariates, it was found that AISI was related to all-cause mortality in stroke patients. Patients with high AISI had a 31% increased risk of death after 90 days (HR = 1.31, 95% CI 1.03-1.67, P = 0.03) compared to patients with low AISI. Patients with high AISI had a 27% increased risk of death after 365 days (HR = 1.27, 95% CI 1.03-1.58, P = 0.029) than low AISI patients. Furthermore, compared with patients with low AISI, patients with high AISI had a 30% increased risk of death after two years (HR = 1.30, 95% CI 1.05-1.60, P = 0.014). During the 2-year follow-up period, the use of RCS showed that the mortality rate gradually increased with the increase of AISI value after 5841.5.

Conclusion: Systemic immunoinflammatory indices are related to long-term adverse outcomes in patients with AIS. Therefore, AISI is a promising inflammatory index for predicting the long-term prognosis of stroke.

Background: Microcirculatory dysfunction is a key pathophysiological feature of sepsis and contributes to organ failure and mortality. In the gastrointestinal tract, impaired barrier function due to microcirculatory injury promotes translocation of inflammatory mediators and bacteria, worsening systemic inflammation and multiorgan dysfunction. Inotropic and vasoactive agents may improve microvascular perfusion through vasodilation in addition to their inotropic effects. Milrinone, a phosphodiesterase-3 inhibitor, and levosimendan, a calcium sensitizer, have shown promising but inconsistent effects in sepsis, while data on their direct microcirculatory and mitochondrial effects in abdominal organs remain limited. Sub-therapeutic vasopressin has demonstrated beneficial effects on gut microcirculation in experimental models, but its combination with inotropes has not been investigated. We hypothesized that (1) milrinone and levosimendan increase colonic and hepatic microvascular blood flow and oxygenation, (2) adjunctive low-dose vasopressin further enhances gastrointestinal microcirculation, and (3) mitochondrial respiration does not differ between treatment groups.

Methods: Male Wistar rats (n = 105) underwent colon ascendens stent peritonitis (CASP) or sham surgery to induce moderate sepsis. Twenty-four hours later, animals received intravenous infusions of vehicle, milrinone, levosimendan, or the respective inotrope with low-dose vasopressin. Colonic and hepatic microvascular oxygenation and blood flow were assessed using tissue-reflectance spectrophotometry and laser Doppler flowmetry. Mitochondrial respiration in colonic and hepatic tissue homogenates from septic animals was analyzed by respirometry. Statistical analyses included mixed-effects models with Tukey or Dunnett post-hoc tests and Kruskal-Wallis tests with Dunn's correction, using a two-sided significance level of α = 0.05.

Results: In septic animals, milrinone and levosimendan increased colonic and hepatic microvascular blood flow. With adjunctive vasopressin, colonic perfusion remained increased, whereas hepatic blood flow did not increase. Microvascular oxygenation remained unchanged in both organs. In sham-operated animals, microvascular blood flow and oxygenation did not differ between treatment groups. Mitochondrial respiration in colon and liver was unchanged across treatments, as indicated by respiratory control index and ADP/O ratio.

Conclusions: In experimental abdominal sepsis, milrinone and levosimendan increase colonic and hepatic microvascular blood flow without affecting mitochondrial respiration. Adjunctive vasopressin alters hepatic but not colonic microvascular responses during combined inotropic therapy.

Animal models of critical illness span diverse species and experimental approaches, reflecting the biological complexity of severe disease states while being constrained by animal welfare requirements and country-specific regulatory, infrastructural, and workforce factors. Persistent challenges remain, including limited reproducibility, fragmented standards, and the need for ethical alignment across borders. This review examines these shared structural challenges in critical illness animal research across the Asia-Pacific region. While alternative and complementary methodologies are increasingly incorporated into preclinical research, their adoption remains uneven. We argue that alignment with globally recognized preclinical frameworks, including the 3Rs and disease-specific standards, such as MQTiPSS, is essential. This review discusses actionable strategies-centered on harmonized standards, shared resources, and international collaboration-to strengthen research rigor, support early career researchers, and enhance the translational relevance of critical illness animal research.