Development and validation of parsimonious algorithms to classify acute respiratory distress syndrome phenotypes: a secondary analysis of randomised controlled trials

Precision Medicine in Acute Respiratory Distress Syndrome: Progress, Challenges, and the Road ahead

Several novel high-dimensional biologic measurements are increasingly being applied to biomedical sciences. Acute respiratory distress syndrome (ARDS) is a theoretically fertile ground for such approaches. Not only are these biologic and analytic tools available to better understand ARDS but also arguably, simpler approaches such as respiratory physiology has been vastly underutilized as a means of delivering precision-based care in the field. Here we review the progress made in ARDS toward discovering biologically homogeneous phenotypes, treatment responsive subgroups, the challenges to implement these discoveries at the bedside, and the road ahead that will enable precision medicine in ARDS.

Acute respiratory distress syndrome

The understanding of acute respiratory distress syndrome (ARDS) has evolved greatly since it was first described in a 1967 case series, with several subsequent updates to the definition of the syndrome. Basic science advances and clinical trials have provided insight into the mechanisms of lung injury in ARDS and led to reduced mortality through comprehensive critical care interventions. This review summarizes the current understanding of the epidemiology, pathophysiology, and management of ARDS. Key highlights include a recommended new global definition of ARDS and updated guidelines for managing ARDS on a backbone of established interventions such as low tidal volume ventilation, prone positioning, and a conservative fluid strategy. Future priorities for investigation of ARDS are also highlighted.

Glucocorticoid therapy for acute respiratory distress syndrome: Current concepts

Acute respiratory distress syndrome (ARDS), a fatal critical disease, is induced by various insults. ARDS represents a major global public health burden, and the management of ARDS continues to challenge healthcare systems globally, especially during the pandemic of the coronavirus disease 2019 (COVID-19). There remains no confirmed specific pharmacotherapy for ARDS, despite advances in understanding its pathophysiology. Debate continues about the potential role of glucocorticoids (GCs) as a promising ARDS clinical therapy. Questions regarding GC agent, dose, and duration in patients with ARDS need to be answered, because of substantial variations in GC administration regimens across studies. ARDS heterogeneity likely affects the therapeutic actions of exogenous GCs. This review includes progress in determining the GC mechanisms of action and clinical applications in ARDS, especially during the COVID-19 pandemic.

The Role of Inflammation in the Pathogenesis of Cardiogenic Shock Secondary to Acute Myocardial Infarction: A Narrative Review

Cardiogenic shock (CS) is one of the most serious complications of myocardial infarction (MI) with a high mortality rate. The timely and effective prevention and early suppression of this adverse event may influence the prognosis and outcome in patients with MI complicated by CS (MI CS). Despite the use of existing pharmaco-invasive options for maintaining an optimal pumping function of the heart in patients with MI CS, its mortality remains high, prompting the search for new approaches to pathogenetic therapy. This review considers the role of the systemic inflammatory response in the pathogenesis of MI CS. The primary processes involved in its initiation are described, including the progression from the onset of MI to the generalization of the inflammatory response and the development of multiple organ dysfunction. The approaches to anti-inflammatory therapy in patients with CS are discussed, and further promising research directions are outlined. In this review, we updated and summarized information on the inflammatory component of MI CS pathogenesis with a particular focus on its foundational aspects. This will facilitate the identification of specific inflammatory phenotypes and endotypes in MI CS and the development of targeted therapeutic strategies for this MI complication.

Early inflammatory profiles predict maximal disease severity in COVID-19: An unsupervised cluster analysis

Background

The inflammatory changes that underlie the heterogeneous presentations of COVID-19 remain incompletely understood. In this study we aimed to identify inflammatory profiles that precede the development of severe COVID-19, that could serve as targets for optimised delivery of immunomodulatory therapies and provide insights for the development of new therapies.

Methods

We included individuals sampled <10 days from COVID-19 symptom onset, recruited from both inpatient and outpatient settings. We measured 61 biomarkers in plasma, including markers of innate immune and T cell activation, coagulation, tissue repair and lung injury. We used principal component analysis and hierarchical clustering to derive biomarker clusters, and ordinal logistic regression to explore associations between cluster membership and maximal disease severity, adjusting for known risk factors for severe COVID-19.

Results

In 312 individuals, median (IQR) 7 (4-9) days from symptom onset, we found four clusters. Cluster 1 was characterised by low overall inflammation, cluster 2 was characterised by higher levels of growth factors and markers of endothelial activation (EGF, VEGF, PDGF, TGFα, PAI-1 and p-selectin). Cluster 3 and 4 both had higher overall inflammation. Cluster 4 had the highest levels of most markers including markers of innate immune activation (IL6, procalcitonin, CRP, TNFα), and coagulation (D-dimer, TPO), in contrast cluster 3 had the highest levels of alveolar epithelial injury markers (RAGE, ST2), but relative downregulation of growth factors and endothelial activation markers, suggesting a dysfunctional inflammatory pattern. In unadjusted and adjusted analysis, compared to cluster 1, cluster 3 had the highest odds of progressing to more severe disease (unadjusted OR (95%CI) 9.02 (4.53-17.96), adjusted OR (95%CI) 6.02 (2.70-13.39)).

Conclusion

Early inflammatory profiles predicted subsequent maximal disease severity independent of risk factors for severe COVID-19. A cluster with downregulation of growth factors and endothelial activation markers, and early evidence of alveolar epithelial injury, had the highest risk of severe COVID-19.

Phenotyping COVID-19 respiratory failure in spontaneously breathing patients with AI on lung CT-scan

BACKGROUND

Automated analysis of lung computed tomography (CT) scans may help characterize subphenotypes of acute respiratory illness. We integrated lung CT features measured via deep learning with clinical and laboratory data in spontaneously breathing subjects to enhance the identification of COVID-19 subphenotypes.

METHODS

This is a multicenter observational cohort study in spontaneously breathing patients with COVID-19 respiratory failure exposed to early lung CT within 7 days of admission. We explored lung CT images using deep learning approaches to quantitative and qualitative analyses; latent class analysis (LCA) by using clinical, laboratory and lung CT variables; regional differences between subphenotypes following 3D spatial trajectories.

RESULTS

Complete datasets were available in 559 patients. LCA identified two subphenotypes (subphenotype 1 and 2). As compared with subphenotype 2 (n = 403), subphenotype 1 patients (n = 156) were older, had higher inflammatory biomarkers, and were more hypoxemic. Lungs in subphenotype 1 had a higher density gravitational gradient with a greater proportion of consolidated lungs as compared with subphenotype 2. In contrast, subphenotype 2 had a higher density submantellar-hilar gradient with a greater proportion of ground glass opacities as compared with subphenotype 1. Subphenotype 1 showed higher prevalence of comorbidities associated with endothelial dysfunction and higher 90-day mortality than subphenotype 2, even after adjustment for clinically meaningful variables.

CONCLUSIONS

Integrating lung-CT data in a LCA allowed us to identify two subphenotypes of COVID-19, with different clinical trajectories. These exploratory findings suggest a role of automated imaging characterization guided by machine learning in subphenotyping patients with respiratory failure.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT04395482. Registration date: 19/05/2020.

Early predictive values of clinical assessments for ARDS mortality: a machine-learning approach

Acute respiratory distress syndrome (ARDS) is a devastating critical care syndrome with significant morbidity and mortality. The objective of this study was to evaluate the predictive values of dynamic clinical indices by developing machine-learning (ML) models for early and accurate clinical assessment of the disease prognosis of ARDS. We conducted a retrospective observational study by applying dynamic clinical data collected in the ARDSNet FACTT Trial (n = 1000) to ML-based algorithms for predicting mortality. In order to compare the significance of clinical features dynamically, we further applied the random forest (RF) model to nine selected clinical parameters acquired at baseline and day 3 independently. An RF model trained using clinical data collected at day 3 showed improved performance and prognostication efficacy (area under the curve [AUC]: 0.84, 95% CI: 0.78-0.89) compared to baseline with an AUC value of 0.72 (95% CI: 0.65-0.78). Mean airway pressure (MAP), bicarbonate, age, platelet count, albumin, heart rate, and glucose were the most significant clinical indicators associated with mortality at day 3. Thus, clinical features collected early (day 3) improved performance of integrative ML models with better prognostication for mortality. Among these, MAP represented the most important feature for ARDS patients' early risk stratification.

Longitudinal analysis of the lung proteome reveals persistent repair months after mild to moderate COVID-19

In order to assess homeostatic mechanisms in the lung after COVID-19, changes in the protein signature of bronchoalveolar lavage from 45 patients with mild to moderate disease at three phases (acute, recovery, and convalescent) are evaluated over a year. During the acute phase, inflamed and uninflamed phenotypes are characterized by the expression of tissue repair and host defense response molecules. With recovery, inflammatory and fibrogenic mediators decline and clinical symptoms abate. However, at 9 months, quantified radiographic abnormalities resolve in the majority of patients, and yet compared to healthy persons, all showed ongoing activation of cellular repair processes and depression of the renin-kallikrein-kinin, coagulation, and complement systems. This dissociation of prolonged reparative processes from symptom and radiographic resolution suggests that occult ongoing disruption of the lung proteome is underrecognized and may be relevant to recovery from other serious viral pneumonias.

From ICU Syndromes to ICU Subphenotypes: Consensus Report and Recommendations for Developing Precision Medicine in the ICU

Critical care uses syndromic definitions to describe patient groups for clinical practice and research. There is growing recognition that a "precision medicine" approach is required and that integrated biologic and physiologic data identify reproducible subpopulations that may respond differently to treatment. This article reviews the current state of the field and considers how to successfully transition to a precision medicine approach. To impact clinical care, identification of subpopulations must do more than differentiate prognosis. It must differentiate response to treatment, ideally by defining subgroups with distinct functional or pathobiological mechanisms (endotypes). There are now multiple examples of reproducible subpopulations of sepsis, acute respiratory distress syndrome, and acute kidney or brain injury described using clinical, physiological, and/or biological data. Many of these subpopulations have demonstrated the potential to define differential treatment response, largely in retrospective studies, and that the same treatment-responsive subpopulations may cross multiple clinical syndromes (treatable traits). To bring about a change in clinical practice, a precision medicine approach must be evaluated in prospective clinical studies requiring novel adaptive trial designs. Several such studies are underway, but there are multiple challenges to be tackled. Such subpopulations must be readily identifiable and be applicable to all critically ill populations around the world. Subdividing clinical syndromes into subpopulations will require large patient numbers. Global collaboration of investigators, clinicians, industry, and patients over many years will therefore be required to transition to a precision medicine approach and ultimately realize treatment advances seen in other medical fields.

Early serum biomarkers to characterise different phenotypes of primary graft dysfunction after lung transplantation: a systematic scoping review

Background

Lung transplantation (LUTX) is often complicated by primary graft dysfunction (PGD). Plasma biomarkers hold potential for PGD phenotyping and targeted therapy. This scoping review aims to collect the available literature in search of serum biomarkers for PGD phenotyping.

Methods

Following JBI and PRISMA guidelines, we conducted a systematic review searching MEDLINE, Web of Science, EMBASE and The Cochrane Library for papers reporting the association between serum biomarkers measured within 72 h of reperfusion and PGD, following International Society for Heart and Lung Transplantation (ISHLT) guidelines. We extracted study details, patient demographics, PGD definition and timing, biomarker concentration, and their performance in identifying PGD cases.

Results

Among the 1050 papers screened, 25 prospective observational studies were included, with only nine conducted in the last decade. These papers included 1793 unique adult patients (1195 double LUTX, median study size 100 (IQR 44-119)). Most (n=21) compared PGD grade 3 to less severe PGD, but only four adhered to 2016 PGD definitions. Enzyme-linked immunosorbent assays and the multiplex bead array technique were utilised in 23 and two papers, respectively. In total, 26 candidate biomarkers were identified, comprising 13 inflammatory, three endothelial activation, three epithelial injury, three cellular damage and two coagulation dysregulation markers. Only five biomarkers (sRAGE, ICAM-1, PAI-1, SP-D, FSTL-1) underwent area under the receiver operating characteristic curve analysis, yielding a median value of 0.58 (0.51-0.78) in 406 patients (276 double LUTX).

Conclusions

Several biomarkers exhibit promise for future studies aimed at PGD phenotyping after LUTX. To uncover the significant existing knowledge gaps, further international prospective studies incorporating updated diagnostic criteria, modern platforms and advanced statistical approaches are essential.

A systematic review of machine learning models for management, prediction and classification of ARDS

AIM

Acute respiratory distress syndrome or ARDS is an acute, severe form of respiratory failure characterised by poor oxygenation and bilateral pulmonary infiltrates. Advancements in signal processing and machine learning have led to promising solutions for classification, event detection and predictive models in the management of ARDS.

METHOD

In this review, we provide systematic description of different studies in the application of Machine Learning (ML) and artificial intelligence for management, prediction, and classification of ARDS. We searched the following databases: Google Scholar, PubMed, and EBSCO from 2009 to 2023. A total of 243 studies was screened, in which, 52 studies were included for review and analysis. We integrated knowledge of previous work providing the state of art and overview of explainable decision models in machine learning and have identified areas for future research.

RESULTS

Gradient boosting is the most common and successful method utilised in 12 (23.1%) of the studies. Due to limitation of data size available, neural network and its variation is used by only 8 (15.4%) studies. Whilst all studies used cross validating technique or separated database for validation, only 1 study validated the model with clinician input. Explainability methods were presented in 15 (28.8%) of studies with the most common method is feature importance which used 14 times.

CONCLUSION

For databases of 5000 or fewer samples, extreme gradient boosting has the highest probability of success. A large, multi-region, multi centre database is required to reduce bias and take advantage of neural network method. A framework for validating with and explaining ML model to clinicians involved in the management of ARDS would be very helpful for development and deployment of the ML model.

Identifying novel clinical phenotypes of acute respiratory distress syndrome using trajectories of daily fluid balance: a secondary analysis of randomized controlled trials

BACKGROUND

Previously identified phenotypes of acute respiratory distress syndrome (ARDS) could not reveal the dynamic change of phenotypes over time. We aimed to identify novel clinical phenotypes in ARDS using trajectories of fluid balance, to test whether phenotypes respond differently to different treatment, and to develop a simplified model for phenotype identification.

METHODS

FACTT (conservative vs liberal fluid management) trial was classified as a development cohort, joint latent class mixed models (JLCMMs) were employed to identify trajectories of fluid balance. Heterogeneity of treatment effect (HTE) for fluid management strategy across phenotypes was investigated. We also constructed a parsimonious probabilistic model using baseline data to predict the fluid trajectories in the development cohort. The trajectory groups and the probabilistic model were externally validated in EDEN (initial trophic vs full enteral feeding) trial.

RESULTS

Using JLCMM, we identified two trajectory groups in the development cohort: Class 1 (n = 758, 76.4% of the cohort) had an early positive fluid balance, but achieved negative fluid balance rapidly, and Class 2 (n = 234, 24.6% of the cohort) was characterized by persistent positive fluid balance. Compared to Class 1 patients, patients in Class 2 had significantly higher 60-day mortality (53.5% vs. 17.8%, p < 0.001), and fewer ventilator-free days (0 vs. 20, p < 0.001). A significant HTE between phenotypes and fluid management strategies was observed in the FACTT. An 8-variables model was derived for phenotype assignment.

CONCLUSIONS

We identified and validated two novel clinical trajectories for ARDS patients, with both prognostic and predictive enrichment. The trajectories of ARDS can be identified with simple classifier models.

Patient stratification using plasma cytokines and their regulators in sepsis: relationship to outcomes, treatment effect and leucocyte transcriptomic subphenotypes

RATIONALE

Heterogeneity of the host response within sepsis, acute respiratory distress syndrome (ARDS) and more widely critical illness, limits discovery and targeting of immunomodulatory therapies. Clustering approaches using clinical and circulating biomarkers have defined hyper-inflammatory and hypo-inflammatory subphenotypes in ARDS associated with differential treatment response. It is unknown if similar subphenotypes exist in sepsis populations where leucocyte transcriptomic-defined subphenotypes have been reported.

OBJECTIVES

We investigated whether inflammatory clusters based on cytokine protein abundance were seen in sepsis, and the relationships with previously described transcriptomic subphenotypes.

METHODS

Hierarchical cluster and latent class analysis were applied to an observational study (UK Genomic Advances in Sepsis (GAinS)) (n=124 patients) and two clinical trial datasets (VANISH, n=155 and LeoPARDS, n=484) in which the plasma protein abundance of 65, 21, 11 circulating cytokines, cytokine receptors and regulators were quantified. Clinical features, outcomes, response to trial treatments and assignment to transcriptomic subphenotypes were compared between inflammatory clusters.

MEASUREMENTS AND MAIN RESULTS

We identified two (UK GAinS, VANISH) or three (LeoPARDS) inflammatory clusters. A group with high levels of pro-inflammatory and anti-inflammatory cytokines was seen that was associated with worse organ dysfunction and survival. No interaction between inflammatory clusters and trial treatment response was found. We found variable overlap of inflammatory clusters and leucocyte transcriptomic subphenotypes.

CONCLUSIONS

These findings demonstrate that differences in response at the level of cytokine biology show clustering related to severity, but not treatment response, and may provide complementary information to transcriptomic sepsis subphenotypes.

TRIAL REGISTRATION NUMBER

ISRCTN20769191, ISRCTN12776039.

Time to split: biomarker trajectories in pediatric acute respiratory distress syndrome hint at underlying disease

Pediatric acute respiratory distress syndrome (ARDS) is severe, noncardiac hypoxemic respiratory failure that carries a substantial risk of death. Given the complexity of this clinically defined syndrome and the repeated failure of therapeutic trials, there has been an effort to identify subphenotypes of ARDS that may share targetable mechanisms of disease. In this issue of the JCI, Yehya and colleagues measured 19 plasma biomarkers in 279 children over the first seven days of ARDS. Increases in select tissue injury makers and inflammatory cytokines in peripheral blood were associated with multiple organ dysfunction syndrome and death, but not persistent ARDS. These findings argue that splitting patients by clinical and molecular phenotype may be more informative than lumping them under the umbrella diagnosis of ARDS. However, future studies are needed to determine whether these plasma factors represent targetable pathways in lung injury or are a consequence of systemic organ dysfunction.

Endotyping in ARDS: one step forward in precision medicine

BACKGROUND

The Berlin definition of acute respiratory distress syndrome (ARDS) includes only clinical characteristics. Understanding unique patient pathobiology may allow personalized treatment. We aimed to define and describe ARDS phenotypes/endotypes combining clinical and pathophysiologic parameters from a Canadian ARDS cohort.

METHODS

A cohort of adult ARDS patients from multiple sites in Calgary, Canada, had plasma cytokine levels and clinical parameters measured in the first 24 h of ICU admission. We used a latent class model (LCM) to group the patients into several ARDS subgroups and identified the features differentiating those subgroups. We then discuss the subgroup effect on 30 day mortality.

RESULTS

The LCM suggested three subgroups (n1 = 64, n2 = 86, and n3 = 30), and 23 out of 69 features made these subgroups distinct. The top five discriminating features were IL-8, IL-6, IL-10, TNF-a, and serum lactate. Mortality distinctively varied between subgroups. Individual clinical characteristics within the subgroup associated with mortality included mean PaO2/FiO2 ratio, pneumonia, platelet count, and bicarbonate negatively associated with mortality, while lactate, creatinine, shock, chronic kidney disease, vasopressor/ionotropic use, low GCS at admission, and sepsis were positively associated. IL-8 and Apache II were individual markers strongly associated with mortality (Area Under the Curve = 0.84).

PERSPECTIVE

ARDS subgrouping using biomarkers and clinical characteristics is useful for categorizing a heterogeneous condition into several homogenous patient groups. This study found three ARDS subgroups using LCM; each subgroup has a different level of mortality. This model may also apply to developing further trial design, prognostication, and treatment selection.

Inflammatory subphenotypes previously identified in ARDS are associated with mortality at intensive care unit discharge: a secondary analysis of a prospective observational study

BACKGROUND

Intensive care unit (ICU)-survivors have an increased risk of mortality after discharge compared to the general population. On ICU admission subphenotypes based on the plasma biomarker levels of interleukin-8, protein C and bicarbonate have been identified in patients admitted with acute respiratory distress syndrome (ARDS) that are prognostic of outcome and predictive of treatment response. We hypothesized that if these inflammatory subphenotypes previously identified among ARDS patients are assigned at ICU discharge in a more general critically ill population, they are associated with short- and long-term outcome.

METHODS

A secondary analysis of a prospective observational cohort study conducted in two Dutch ICUs between 2011 and 2014 was performed. All patients discharged alive from the ICU were at ICU discharge adjudicated to the previously identified inflammatory subphenotypes applying a validated parsimonious model using variables measured median 10.6 h [IQR, 8.0-31.4] prior to ICU discharge. Subphenotype distribution at ICU discharge, clinical characteristics and outcomes were analyzed. As a sensitivity analysis, a latent class analysis (LCA) was executed for subphenotype identification based on plasma protein biomarkers at ICU discharge reflective of coagulation activation, endothelial cell activation and inflammation. Concordance between the subphenotyping strategies was studied.

RESULTS

Of the 8332 patients included in the original cohort, 1483 ICU-survivors had plasma biomarkers available and could be assigned to the inflammatory subphenotypes. At ICU discharge 6% (n = 86) was assigned to the hyperinflammatory and 94% (n = 1397) to the hypoinflammatory subphenotype. Patients assigned to the hyperinflammatory subphenotype were discharged with signs of more severe organ dysfunction (SOFA scores 7 [IQR 5-9] vs. 4 [IQR 2-6], p < 0.001). Mortality was higher in patients assigned to the hyperinflammatory subphenotype (30-day mortality 21% vs. 11%, p = 0.005; one-year mortality 48% vs. 28%, p < 0.001). LCA deemed 2 subphenotypes most suitable. ICU-survivors from class 1 had significantly higher mortality compared to class 2. Patients belonging to the hyperinflammatory subphenotype were mainly in class 1.

CONCLUSIONS

Patients assigned to the hyperinflammatory subphenotype at ICU discharge showed significantly stronger anomalies in coagulation activation, endothelial cell activation and inflammation pathways implicated in the pathogenesis of critical disease and increased mortality until one-year follow up.

Post-pandemic acute respiratory distress syndrome: A New Global Definition with extension to lower-resource regions

Acute respiratory distress syndrome (ARDS), first described in 1967, is characterized by acute respiratory failure causing profound hypoxemia, decreased pulmonary compliance, and bilateral CXR infiltrates. After several descriptions, the Berlin definition was adopted in 2012, which established three categories of severity according to hypoxemia (mild, moderate and severe), specified temporal aspects for diagnosis, and incorporated the use of non-invasive ventilation. The COVID-19 pandemic led to changes in ARDS management, focusing on continuous monitoring of oxygenation and on utilization of high-flow oxygen therapy and lung ultrasound. In 2021, a New Global Definition based on the Berlin definition of ARDS was proposed, which included a category for non-intubated patients, considered the use of SpO2, and established no particular requirement for oxygenation support in regions with limited resources. Although debates persist, the continuous evolution seeks to adapt to clinical and epidemiological needs, and to the search of personalized treatments.

Artificial Intelligence in the Intensive Care Unit: Current Evidence on an Inevitable Future Tool

Artificial intelligence (AI) is a technique that attempts to replicate human intelligence, analytical behavior, and decision-making ability. This includes machine learning, which involves the use of algorithms and statistical techniques to enhance the computer's ability to make decisions more accurately. Due to AI's ability to analyze, comprehend, and interpret considerable volumes of data, it has been increasingly used in the field of healthcare. In critical care medicine, where most of the patient load requires timely interventions due to the perilous nature of the condition, AI's ability to monitor, analyze, and predict unfavorable outcomes is an invaluable asset. It can significantly improve timely interventions and prevent unfavorable outcomes, which, otherwise, is not always achievable owing to the constrained human ability to multitask with optimum efficiency. AI has been implicated in intensive care units over the past many years. In addition to its advantageous applications, this article discusses its disadvantages, prospects, and the changes needed to train future critical care professionals. A comprehensive search of electronic databases was performed using relevant keywords. Data from articles pertinent to the topic was assimilated into this review article.

Development and External Validation of Models to Predict Persistent Hypoxemic Respiratory Failure for Clinical Trial Enrichment

OBJECTIVES

Improving the efficiency of clinical trials in acute hypoxemic respiratory failure (HRF) depends on enrichment strategies that minimize enrollment of patients who quickly resolve with existing care and focus on patients at high risk for persistent HRF. We aimed to develop parsimonious models predicting risk of persistent HRF using routine data from ICU admission and select research immune biomarkers.

DESIGN

Prospective cohorts for derivation ( n = 630) and external validation ( n = 511).

SETTING

Medical and surgical ICUs at two U.S. medical centers.

PATIENTS

Adults with acute HRF defined as new invasive mechanical ventilation (IMV) and hypoxemia on the first calendar day after ICU admission.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We evaluated discrimination, calibration, and practical utility of models predicting persistent HRF risk (defined as ongoing IMV and hypoxemia on the third calendar day after admission): 1) a clinical model with least absolute shrinkage and selection operator (LASSO) selecting Pa o2 /F io2 , vasopressors, mean arterial pressure, bicarbonate, and acute respiratory distress syndrome as predictors; 2) a model adding interleukin-6 (IL-6) to clinical predictors; and 3) a comparator model with Pa o2 /F io2 alone, representing an existing strategy for enrichment. Forty-nine percent and 69% of patients had persistent HRF in derivation and validation sets, respectively. In validation, both LASSO (area under the receiver operating characteristic curve, 0.68; 95% CI, 0.64-0.73) and LASSO + IL-6 (0.71; 95% CI, 0.66-0.76) models had better discrimination than Pa o2 /F io2 (0.64; 95% CI, 0.59-0.69). Both models underestimated risk in lower risk deciles, but exhibited better calibration at relevant risk thresholds. Evaluating practical utility, both LASSO and LASSO + IL-6 models exhibited greater net benefit in decision curve analysis, and greater sample size savings in enrichment analysis, compared with Pa o2 /F io2 . The added utility of LASSO + IL-6 model over LASSO was modest.

CONCLUSIONS

Parsimonious, interpretable models that predict persistent HRF may improve enrichment of trials testing HRF-targeted therapies and warrant future validation.

Heterogeneity of immune cells and their communications unveiled by transcriptome profiling in acute inflammatory lung injury

Background

Acute Respiratory Distress Syndrome (ARDS) or its earlier stage Acute lung injury (ALI), is a worldwide health concern that jeopardizes human well-being. Currently, the treatment strategies to mitigate the incidence and mortality of ARDS are severely restricted. This limitation can be attributed, at least in part, to the substantial variations in immunity observed in individuals with this syndrome.

Methods

Bulk and single cell RNA sequencing from ALI mice and single cell RNA sequencing from ARDS patients were analyzed. We utilized the Seurat program package in R and cellmarker 2.0 to cluster and annotate the data. The differential, enrichment, protein interaction, and cell-cell communication analysis were conducted.

Results

The mice with ALI caused by pulmonary and extrapulmonary factors demonstrated differential expression including Clec4e, Retnlg, S100a9, Coro1a, and Lars2. We have determined that inflammatory factors have a greater significance in extrapulmonary ALI, while multiple pathways collaborate in the development of pulmonary ALI. Clustering analysis revealed significant heterogeneity in the relative abundance of immune cells in different ALI models. The autocrine action of neutrophils plays a crucial role in pulmonary ALI. Additionally, there was a significant increase in signaling intensity between B cells and M1 macrophages, NKT cells and M1 macrophages in extrapulmonary ALI. The CXCL, CSF3 and MIF, TGFβ signaling pathways play a vital role in pulmonary and extrapulmonary ALI, respectively. Moreover, the analysis of human single-cell revealed DCs signaling to monocytes and neutrophils in COVID-19-associated ARDS is stronger compared to sepsis-related ARDS. In sepsis-related ARDS, CD8+ T and Th cells exhibit more prominent signaling to B-cell nucleated DCs. Meanwhile, both MIF and CXCL signaling pathways are specific to sepsis-related ARDS.

Conclusion

This study has identified specific gene signatures and signaling pathways in animal models and human samples that facilitate the interaction between immune cells, which could be targeted therapeutically in ARDS patients of various etiologies.

Consensus statements on the utility of defining ARDS and the utility of past and current definitions of ARDS-protocol for a Delphi study

INTRODUCTION

Acute respiratory distress syndrome (ARDS), marked by acute hypoxemia and bilateral pulmonary infiltrates, has been defined in multiple ways since its first description. This Delphi study aims to collect global opinions on the conceptual framework of ARDS, assess the usefulness of components within current and past definitions and investigate the role of subphenotyping. The varied expertise of the panel will provide valuable insights for refining future ARDS definitions and improving clinical management.

METHODS

A diverse panel of 35-40 experts will be selected based on predefined criteria. Multiple choice questions (MCQs) or 7-point Likert-scale statements will be used in the iterative Delphi rounds to achieve consensus on key aspects related to the utility of definitions and subphenotyping. The Delphi rounds will be continued until a stable agreement or disagreement is achieved for all statements.

ANALYSIS

Consensus will be considered as reached when a choice in MCQs or Likert-scale statement achieved ≥80% of votes for agreement or disagreement. The stability will be checked by non-parametric χ2 tests or Kruskal Wallis test starting from the second round of Delphi process. A p-value ≥0.05 will be used to define stability.

ETHICS AND DISSEMINATION

The study will be conducted in full concordance with the principles of the Declaration of Helsinki and will be reported according to CREDES guidance. This study has been granted an ethical approval waiver by the NMC Healthcare Regional Research Ethics Committee, Dubai (NMCHC/CR/DXB/REC/APP/002), owing to the nature of the research. Informed consent will be obtained from all panellists before the start of the Delphi process. The study will be published in a peer-review journal with the authorship agreed as per ICMJE requirements.

TRIAL REGISTRATION NUMBER

NCT06159465.

Inflammatory and tissue injury marker dynamics in pediatric acute respiratory distress syndrome

BACKGROUNDThe molecular signature of pediatric acute respiratory distress syndrome (ARDS) is poorly described, and the degree to which hyperinflammation or specific tissue injury contributes to outcomes is unknown. Therefore, we profiled inflammation and tissue injury dynamics over the first 7 days of ARDS, and associated specific biomarkers with mortality, persistent ARDS, and persistent multiple organ dysfunction syndrome (MODS).METHODSIn a single-center prospective cohort of intubated pediatric patients with ARDS, we collected plasma on days 0, 3, and 7. Nineteen biomarkers reflecting inflammation, tissue injury, and damage-associated molecular patterns (DAMPs) were measured. We assessed the relationship between biomarkers and trajectories with mortality, persistent ARDS, or persistent MODS using multivariable mixed effect models.RESULTSIn 279 patients (64 [23%] nonsurvivors), hyperinflammatory cytokines, tissue injury markers, and DAMPs were higher in nonsurvivors. Survivors and nonsurvivors showed different biomarker trajectories. IL-1α, soluble tumor necrosis factor receptor 1, angiopoietin 2 (ANG2), and surfactant protein D increased in nonsurvivors, while DAMPs remained persistently elevated. ANG2 and procollagen type III N-terminal peptide were associated with persistent ARDS, whereas multiple cytokines, tissue injury markers, and DAMPs were associated with persistent MODS. Corticosteroid use did not impact the association of biomarker levels or trajectory with mortality.CONCLUSIONSPediatric ARDS survivors and nonsurvivors had distinct biomarker trajectories, with cytokines, endothelial and alveolar epithelial injury, and DAMPs elevated in nonsurvivors. Mortality markers overlapped with markers associated with persistent MODS, rather than persistent ARDS.FUNDINGNIH (K23HL-136688, R01-HL148054).

Molecular Phenotypes of Acute Respiratory Distress Syndrome in the ROSE Trial Have Differential Outcomes and Gene Expression Patterns That Differ at Baseline and Longitudinally over Time

Rationale: Two molecular phenotypes have been identified in acute respiratory distress syndrome (ARDS). In the ROSE (Reevaluation of Systemic Early Neuromuscular Blockade) trial of cisatracurium in moderate to severe ARDS, we addressed three unanswered questions: 1) Do the same phenotypes emerge in a more severe ARDS cohort with earlier recruitment; 2) Do phenotypes respond differently to neuromuscular blockade? and 3) What biological pathways most differentiate inflammatory phenotypes?Methods: We performed latent class analysis in ROSE using preenrollment clinical and protein biomarkers. In a subset of patients (n = 134), we sequenced whole-blood RNA using enrollment and Day 2 samples and performed differential gene expression and pathway analyses. Informed by the differential gene expression analysis, we measured additional plasma proteins and evaluated their abundance relative to gene expression amounts.Measurements and Main Results: In ROSE, we identified the hypoinflammatory (60.4%) and hyperinflammatory (39.6%) phenotypes with similar biological and clinical characteristics as prior studies, including higher mortality at Day 90 for the hyperinflammatory phenotype (30.3% vs. 61.6%; P < 0.0001). We observed no treatment interaction between the phenotypes and randomized groups for mortality. The hyperinflammatory phenotype was enriched for genes associated with innate immune response, tissue remodeling, and zinc metabolism at Day 0 and collagen synthesis and neutrophil degranulation at Day 2. Longitudinal changes in gene expression patterns differed dependent on survivorship. For most highly expressed genes, we observed correlations with their corresponding plasma proteins' abundance. However, for the class-defining plasma proteins in the latent class analysis, no correlation was observed with their corresponding genes' expression.Conclusions: The hyperinflammatory and hypoinflammatory phenotypes have different clinical, protein, and dynamic transcriptional characteristics. These findings support the clinical and biological potential of molecular phenotypes to advance precision care in ARDS.

Exploring the translational landscape of the long noncoding RNA transcriptome in acute respiratory distress syndrome: it is a long way to the top

Acute respiratory distress syndrome (ARDS) poses a significant and widespread public health challenge. Extensive research conducted in recent decades has considerably improved our understanding of the disease pathophysiology. Nevertheless, ARDS continues to rank among the leading causes of mortality in intensive care units and its management remains a formidable task, primarily due to its remarkable heterogeneity. As a consequence, the syndrome is underdiagnosed, prognostication has important gaps and selection of the appropriate therapeutic approach is laborious. In recent years, the noncoding transcriptome has emerged as a new area of attention for researchers interested in biomarker development. Numerous studies have confirmed the potential of long noncoding RNAs (lncRNAs), transcripts with little or no coding information, as noninvasive tools for diagnosis, prognosis and prediction of the therapeutic response across a broad spectrum of ailments, including respiratory conditions. This article aims to provide a comprehensive overview of lncRNAs with specific emphasis on their role as biomarkers. We review current knowledge on the circulating lncRNAs as potential markers that can be used to enhance decision making in ARDS management. Additionally, we address the primary limitations and outline the steps that will be essential for integration of the use of lncRNAs in clinical laboratories. Our ultimate objective is to provide a framework for the implementation of lncRNAs in the management of ARDS.

Immunocompromised-Associated Pediatric Acute Respiratory Distress Syndrome: Experience From the 2016/2017 Pediatric Acute Respiratory Distress Syndrome Incidence and Epidemiology Prospective Cohort Study

OBJECTIVES

To characterize immunocompromised-associated pediatric acute respiratory distress syndrome (I-PARDS) and contrast it to PARDS.

DESIGN

This is a secondary analysis of the 2016-2017 PARDS incidence and epidemiology (PARDIE) study, a prospective observational, cross-sectional study of children with PARDS.

SETTING

Dataset of 145 PICUs across 27 countries.

PATIENTS

During 10 nonconsecutive weeks (from May 2016 to June 2017), data about immunocompromising conditions (ICCs, defined as malignancy, congenital/acquired immunodeficiency, posttransplantation, or diseases requiring immunosuppression) were collected.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Of 708 subjects, 105 (14.8%) had ICC. Before the development of I-PARDS, those with ICC were more likely to be hospitalized (70% vs. 35%, p < 0.001), have more at-risk for PARDS ( p = 0.046), and spent more hours at-risk (20 [interquartile range, IQR: 8-46] vs. 11 [IQR: 4-33], [ p = 0.002]). Noninvasive ventilation (NIV) use was more common in those with ICC ( p < 0.001). Of those diagnosed with PARDS on NIV ( n = 161), children with ICC were more likely to be subsequently intubated ( n = 28/40 [70%] vs n = 53/121 [44%], p = 0.004). Severe PARDS was more common (32% vs 23%, p < 0.001) in I-PARDS. Oxygenation indices were higher at diagnosis and had less improvement over the first 3 days of PARDS ( p < 0.001). Children with I-PARDS had greater nonpulmonary organ dysfunction. Adjusting for Pediatric Risk of Mortality IV and oxygenation index, children with I-PARDS had a higher severity of illness-adjusted PICU mortality (adjusted hazard ratio: 3.0 [95% CI, 1.9-4.7] p < 0.001) and were less likely to be extubated alive within 28 days (subdistribution hazard ratio: 0.47 [95% CI, 0.31-0.71] p < 0.001).

CONCLUSIONS

I-PARDS is a unique subtype of PARDS associated with hospitalization before diagnosis and increased: time at-risk for PARDS, NIV use, hypoxia, nonpulmonary organ dysfunction, and mortality. The opportunity for early detection and intervention seems to exist. Dedicated study in these patients is imperative to determine if targeted interventions will benefit these unique patients with the ultimate goal of improving outcomes.

Machine learning for the early prediction of acute respiratory distress syndrome (ARDS) in patients with sepsis in the ICU based on clinical data

Background

Acute respiratory distress syndrome (ARDS) is a fatal outcome of severe sepsis. Machine learning models are helpful for accurately predicting ARDS in patients with sepsis at an early stage.

Objective

We aim to develop a machine-learning model for predicting ARDS in patients with sepsis in the intensive care unit (ICU).

Methods

The initial clinical data of patients with sepsis admitted to the hospital (including population characteristics, clinical diagnosis, complications, and laboratory tests) were used to predict ARDS, and screen out the crucial variables. After comparing eight different algorithms, namely, XG boost, logistic regression, light GBM, random forest, GaussianNB, complement NB, support vector machine (SVM), and K nearest neighbors (KNN), rebuilding a prediction model with the best one. When remodeling with the best algorithm, 10% was randomly selected to test, and the remaining was trained for cross-validation. Using the area under the curve (AUC), sensitivity, accuracy, specificity, positive and negative predictive value, F1 score, kappa value, and clinical decision curve to evaluate the model's performance. Eventually, the application in the model illustrated by the SHAP package.

Results

Ten critical features were screened utilizing the lasso method, namely, PaO2/PAO2, A-aDO2, PO2(T), CRP, gender, PO2, RDW, MCH, SG, and chlorine. The prior ranking of variables demonstrated that PaO2/PAO2 was the most significant variable. Among the eight algorithms, the performance of the Gaussian NB algorithm was significantly better than that of the others. After remodeling with the best algorithm, the AUC in the training and validation sets were 0.777 and 0.770, respectively, and the algorithm performed well in the test set (AUC = 0.781, accuracy = 78.6%, sensitivity = 82.4%, F1 score = 0.824). A comparison of the overlap factors with those of previous models revealed that the model we developed performs better.

Conclusion

Sepsis-associated ARDS can be accurately predicted early via a machine learning model based on existing clinical data. These findings are helpful for accurate identification and improvement of the prognosis in patients with sepsis-associated ARDS.

Monocyte state 1 (MS1) cells in critically ill patients with sepsis or non-infectious conditions: association with disease course and host response

BACKGROUND

Sepsis is a life-threatening condition arising from an aberrant host response to infection. Recent single-cell RNA sequencing investigations identified an immature bone-marrow-derived CD14+ monocyte phenotype with immune suppressive properties termed "monocyte state 1" (MS1) in patients with sepsis. Our objective was to determine the association of MS1 cell profiles with disease presentation, outcomes, and host response characteristics.

METHODS

We used the transcriptome deconvolution method (CIBERSORTx) to estimate the percentage of MS1 cells from blood RNA profiles of patients with sepsis admitted to the intensive care unit (ICU). We compared these profiles to ICU patients without infection and to healthy controls. Host response dysregulation was further studied by gene co-expression network and gene set enrichment analyses of blood leukocytes, and measurement of 15 plasma biomarkers indicative of pathways implicated in sepsis pathogenesis.

RESULTS

Sepsis patients (n = 332) were divided into three equally-sized groups based on their MS1 cell levels (low, intermediate, and high). MS1 groups did not differ in demographics or comorbidities. The intermediate and high MS1 groups presented with higher disease severity and more often had shock. MS1 cell abundance did not differ between survivors and non-survivors, or between patients who did or did not acquire a secondary infection. Higher MS1 cell percentages were associated with downregulation of lymphocyte-related and interferon response genes in blood leukocytes, with concurrent upregulation of inflammatory response pathways, including tumor necrosis factor signaling via nuclear factor-κB. Previously described sepsis host response transcriptomic subtypes showed different MS1 cell abundances, and MS1 cell percentages positively correlated with the "quantitative sepsis response signature" and "molecular degree of perturbation" scores. Plasma biomarker levels, indicative of inflammation, endothelial cell activation, and coagulation activation, were largely similar between MS1 groups. In ICU patients without infection (n = 215), MS1 cell percentages and their relation with disease severity, shock, and host response dysregulation were highly similar to those in sepsis patients.

CONCLUSIONS

High MS1 cell percentages are associated with increased disease severity and shock in critically ill patients with sepsis or a non-infectious condition. High MS1 cell abundance likely indicates broad immune dysregulation, entailing not only immunosuppression but also anomalies reflecting exaggerated inflammatory responses.

The Rise of Adaptive Platform Trials in Critical Care

As durable learning research systems, adaptive platform trials represent a transformative new approach to accelerating clinical evaluation and discovery in critical care. This Perspective provides a brief introduction to the concept of adaptive platform trials, describes several established and emerging platforms in critical care, and surveys some opportunities and challenges for their implementation and impact.

The Relationship Between Metabolic Syndrome and Mortality Among Patients With Acute Respiratory Distress Syndrome in Acute Respiratory Distress Syndrome Network and Prevention and Early Treatment of Acute Lung Injury Network Trials

OBJECTIVES

Metabolic syndrome is known to predict outcomes in COVID-19 acute respiratory distress syndrome (ARDS) but has never been studied in non-COVID-19 ARDS. We therefore aimed to determine the association of metabolic syndrome with mortality among ARDS trial subjects.

DESIGN

Retrospective cohort study of ARDS trials' data.

SETTING

An ancillary analysis was conducted using data from seven ARDS Network and Prevention and Early Treatment of Acute Lung Injury Network randomized trials within the Biologic Specimen and Data Repository Information Coordinating Center database.

PATIENTS

Hospitalized patients with ARDS and metabolic syndrome (defined by obesity, diabetes, and hypertension) were compared with similar patients without metabolic syndrome (those with less than three criteria).

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

The primary outcome was 28-day mortality. Among 4288 ARDS trial participants, 454 (10.6%) with metabolic syndrome were compared with 3834 controls (89.4%). In adjusted analyses, the metabolic syndrome group was associated with lower 28-day and 90-day mortality when compared with control (adjusted odds ratio [aOR], 0.70 [95% CI, 0.55-0.89] and 0.75 [95% CI, 0.60-0.95], respectively). With each additional metabolic criterion from 0 to 3, adjusted 28-day mortality was reduced by 18%, 22%, and 40%, respectively. In subgroup analyses stratifying by ARDS etiology, mortality was lower for metabolic syndrome vs. control in ARDS caused by sepsis or pneumonia (at 28 d, aOR 0.64 [95% CI, 0.48-0.84] and 90 d, aOR 0.69 [95% CI, 0.53-0.89]), but not in ARDS from noninfectious causes (at 28 d, aOR 1.18 [95% CI, 0.70-1.99] and 90 d, aOR 1.26 [95% CI, 0.77-2.06]). Interaction p = 0.04 and p = 0.02 for 28- and 90-day comparisons, respectively.

CONCLUSIONS

Metabolic syndrome in ARDS was associated with a lower risk of mortality in non-COVID-19 ARDS. The relationship between metabolic inflammation and ARDS may provide a novel biological pathway to be explored in precision medicine-based trials.

ERS International Congress 2023: highlights from the Respiratory Intensive Care Assembly

Early career members of Assembly 2 (Respiratory Intensive Care) attended the 2023 European Respiratory Society International Congress in Milan, Italy. The conference covered acute and chronic respiratory failure. Sessions of interest to our assembly members and to those interested in respiratory critical care are summarised in this article and include the latest updates in respiratory intensive care, in particular acute respiratory distress syndrome and mechanical ventilation.

Prediction of 30-day, 90-day, and 1-year mortality after colorectal cancer surgery using a data-driven approach

PURPOSE

To develop prediction models for short-term mortality risk assessment following colorectal cancer surgery.

METHODS

Data was harmonized from four Danish observational health databases into the Observational Medical Outcomes Partnership Common Data Model. With a data-driven approach using the Least Absolute Shrinkage and Selection Operator logistic regression on preoperative data, we developed 30-day, 90-day, and 1-year mortality prediction models. We assessed discriminative performance using the area under the receiver operating characteristic and precision-recall curve and calibration using calibration slope, intercept, and calibration-in-the-large. We additionally assessed model performance in subgroups of curative, palliative, elective, and emergency surgery.

RESULTS

A total of 57,521 patients were included in the study population, 51.1% male and with a median age of 72 years. The model showed good discrimination with an area under the receiver operating characteristic curve of 0.88, 0.878, and 0.861 for 30-day, 90-day, and 1-year mortality, respectively, and a calibration-in-the-large of 1.01, 0.99, and 0.99. The overall incidence of mortality were 4.48% for 30-day mortality, 6.64% for 90-day mortality, and 12.8% for 1-year mortality, respectively. Subgroup analysis showed no improvement of discrimination or calibration when separating the cohort into cohorts of elective surgery, emergency surgery, curative surgery, and palliative surgery.

CONCLUSION

We were able to train prediction models for the risk of short-term mortality on a data set of four combined national health databases with good discrimination and calibration. We found that one cohort including all operated patients resulted in better performing models than cohorts based on several subgroups.

Using a targeted metabolomics approach to explore differences in ARDS associated with COVID-19 compared to ARDS caused by H1N1 influenza and bacterial pneumonia

RATIONALE

Acute respiratory distress syndrome (ARDS) is a life-threatening critical care syndrome commonly associated with infections such as COVID-19, influenza, and bacterial pneumonia. Ongoing research aims to improve our understanding of ARDS, including its molecular mechanisms, individualized treatment options, and potential interventions to reduce inflammation and promote lung repair.

OBJECTIVE

To map and compare metabolic phenotypes of different infectious causes of ARDS to better understand the metabolic pathways involved in the underlying pathogenesis.

METHODS

We analyzed metabolic phenotypes of 3 ARDS cohorts caused by COVID-19, H1N1 influenza, and bacterial pneumonia compared to non-ARDS COVID-19-infected patients and ICU-ventilated controls. Targeted metabolomics was performed on plasma samples from a total of 150 patients using quantitative LC-MS/MS and DI-MS/MS analytical platforms.

RESULTS

Distinct metabolic phenotypes were detected between different infectious causes of ARDS. There were metabolomics differences between ARDSs associated with COVID-19 and H1N1, which include metabolic pathways involving taurine and hypotaurine, pyruvate, TCA cycle metabolites, lysine, and glycerophospholipids. ARDSs associated with bacterial pneumonia and COVID-19 differed in the metabolism of D-glutamine and D-glutamate, arginine, proline, histidine, and pyruvate. The metabolic profile of COVID-19 ARDS (C19/A) patients admitted to the ICU differed from COVID-19 pneumonia (C19/P) patients who were not admitted to the ICU in metabolisms of phenylalanine, tryptophan, lysine, and tyrosine. Metabolomics analysis revealed significant differences between C19/A, H1N1/A, and PNA/A vs ICU-ventilated controls, reflecting potentially different disease mechanisms.

CONCLUSION

Different metabolic phenotypes characterize ARDS associated with different viral and bacterial infections.

Novel subtypes of severe COVID-19 respiratory failure based on biological heterogeneity: a secondary analysis of a randomized controlled trial

BACKGROUND

Despite evidence associating inflammatory biomarkers with worse outcomes in hospitalized adults with COVID-19, trials of immunomodulatory therapies have met with mixed results, likely due in part to biological heterogeneity of participants. Latent class analysis (LCA) of clinical and protein biomarker data has identified two subtypes of non-COVID acute respiratory distress syndrome (ARDS) with different clinical outcomes and treatment responses. We studied biological heterogeneity and clinical outcomes in a multi-institutional platform randomized controlled trial of adults with severe COVID-19 hypoxemic respiratory failure (I-SPY COVID).

METHODS

Clinical and plasma protein biomarker data were analyzed from 400 trial participants enrolled from September 2020 until October 2021 with severe COVID-19 requiring ≥ 6 L/min supplemental oxygen. Seventeen hypothesis-directed protein biomarkers were measured at enrollment using multiplex Luminex panels or single analyte enzyme linked immunoassay methods (ELISA). Biomarkers and clinical variables were used to test for latent subtypes and longitudinal biomarker changes by subtype were explored. A validated parsimonious model using interleukin-8, bicarbonate, and protein C was used for comparison with non-COVID hyper- and hypo-inflammatory ARDS subtypes.

RESULTS

Average participant age was 60 ± 14 years; 67% were male, and 28-day mortality was 25%. At trial enrollment, 85% of participants required high flow oxygen or non-invasive ventilation, and 97% were receiving dexamethasone. Several biomarkers of inflammation (IL-6, IL-8, IL-10, sTNFR-1, TREM-1), epithelial injury (sRAGE), and endothelial injury (Ang-1, thrombomodulin) were associated with 28- and 60-day mortality. Two latent subtypes were identified. Subtype 2 (27% of participants) was characterized by persistent derangements in biomarkers of inflammation, endothelial and epithelial injury, and disordered coagulation and had twice the mortality rate compared with Subtype 1. Only one person was classified as hyper-inflammatory using the previously validated non-COVID ARDS model.

CONCLUSIONS

We discovered evidence of two novel biological subtypes of severe COVID-19 with significantly different clinical outcomes. These subtypes differed from previously established hyper- and hypo-inflammatory non-COVID subtypes of ARDS. Biological heterogeneity may explain inconsistent findings from trials of hospitalized patients with COVID-19 and guide treatment approaches.

Identification of molecular subphenotypes in two cohorts of paediatric ARDS

BACKGROUND

Two subphenotypes of acute respiratory distress syndrome (ARDS), hypoinflammatory and hyperinflammatory, have been reported in adults and in a single paediatric cohort. The relevance of these subphenotypes in paediatrics requires further investigation. We aimed to identify subphenotypes in two large observational cohorts of paediatric ARDS and assess their congruence with prior descriptions.

METHODS

We performed latent class analysis (LCA) separately on two cohorts using biomarkers as inputs. Subphenotypes were compared on clinical characteristics and outcomes. Finally, we assessed overlap with adult cohorts using parsimonious classifiers.

FINDINGS

In two cohorts from the Children's Hospital of Philadelphia (n=333) and from a multicentre study based at the University of California San Francisco (n=293), LCA identified two subphenotypes defined by differential elevation of biomarkers reflecting inflammation and endotheliopathy. In both cohorts, hyperinflammatory subjects had greater illness severity, more sepsis and higher mortality (41% and 28% in hyperinflammatory vs 11% and 7% in hypoinflammatory). Both cohorts demonstrated overlap with adult subphenotypes when assessed using parsimonious classifiers.

INTERPRETATION

We identified hypoinflammatory and hyperinflammatory subphenotypes of paediatric ARDS from two separate cohorts with utility for prognostic and potentially predictive, enrichment. Future paediatric ARDS trials should identify and leverage biomarker-defined subphenotypes in their analysis.

Subphenotypes of Acute Respiratory Distress Syndrome: Advancing Towards Precision Medicine

Acute respiratory distress syndrome (ARDS) is a common cause of severe hypoxemia defined by the acute onset of bilateral non-cardiogenic pulmonary edema. The diagnosis is made by defined consensus criteria. Supportive care, including prevention of further injury to the lungs, is the only treatment that conclusively improves outcomes. The inability to find more advanced therapies is due, in part, to the highly sensitive but relatively non-specific current syndromic consensus criteria, combining a heterogenous population of patients under the umbrella of ARDS. With few effective therapies, the morality rate remains 30% to 40%. Many subphenotypes of ARDS have been proposed to cluster patients with shared combinations of observable or measurable traits. Subphenotyping patients is a strategy to overcome heterogeneity to advance clinical research and eventually identify treatable traits. Subphenotypes of ARDS have been proposed based on radiographic patterns, protein biomarkers, transcriptomics, and/or machine-based clustering of clinical and biological variables. Some of these strategies have been reproducible across patient cohorts, but at present all have practical limitations to their implementation. Furthermore, there is no agreement on which strategy is the most appropriate. This review will discuss the current strategies for subphenotyping patients with ARDS, including the strengths and limitations, and the future directions of ARDS subphenotyping.

A New Global Definition of Acute Respiratory Distress Syndrome

Background: Since publication of the 2012 Berlin definition of acute respiratory distress syndrome (ARDS), several developments have supported the need for an expansion of the definition, including the use of high-flow nasal oxygen, the expansion of the use of pulse oximetry in place of arterial blood gases, the use of ultrasound for chest imaging, and the need for applicability in resource-limited settings. Methods: A consensus conference of 32 critical care ARDS experts was convened, had six virtual meetings (June 2021 to March 2022), and subsequently obtained input from members of several critical care societies. The goal was to develop a definition that would 1) identify patients with the currently accepted conceptual framework for ARDS, 2) facilitate rapid ARDS diagnosis for clinical care and research, 3) be applicable in resource-limited settings, 4) be useful for testing specific therapies, and 5) be practical for communication to patients and caregivers. Results: The committee made four main recommendations: 1) include high-flow nasal oxygen with a minimum flow rate of ⩾30 L/min; 2) use PaO2:FiO2 ⩽ 300 mm Hg or oxygen saturation as measured by pulse oximetry SpO2:FiO2 ⩽ 315 (if oxygen saturation as measured by pulse oximetry is ⩽97%) to identify hypoxemia; 3) retain bilateral opacities for imaging criteria but add ultrasound as an imaging modality, especially in resource-limited areas; and 4) in resource-limited settings, do not require positive end-expiratory pressure, oxygen flow rate, or specific respiratory support devices. Conclusions: We propose a new global definition of ARDS that builds on the Berlin definition. The recommendations also identify areas for future research, including the need for prospective assessments of the feasibility, reliability, and prognostic validity of the proposed global definition.

Trajectories of Host-Response Subphenotypes in Patients With COVID-19 Across the Spectrum of Respiratory Support

BACKGROUND

Hospitalized patients with severe COVID-19 follow heterogeneous clinical trajectories, requiring different levels of respiratory support and experiencing diverse clinical outcomes. Differences in host immune responses to SARS-CoV-2 infection may account for the heterogeneous clinical course, but we have limited data on the dynamic evolution of systemic biomarkers and related subphenotypes. Improved understanding of the dynamic transitions of host subphenotypes in COVID-19 may allow for improved patient selection for targeted therapies.

RESEARCH QUESTION

We examined the trajectories of host-response profiles in severe COVID-19 and evaluated their prognostic impact on clinical outcomes.

STUDY DESIGN AND METHODS

In this prospective observational study, we enrolled 323 inpatients with COVID-19 receiving different levels of baseline respiratory support: (1) low-flow oxygen (37%), (2) noninvasive ventilation (NIV) or high-flow oxygen (HFO; 29%), (3) invasive mechanical ventilation (27%), and (4) extracorporeal membrane oxygenation (7%). We collected plasma samples on enrollment and at days 5 and 10 to measure host-response biomarkers. We classified patients by inflammatory subphenotypes using two validated predictive models. We examined clinical, biomarker, and subphenotype trajectories and outcomes during hospitalization.

RESULTS

IL-6, procalcitonin, and angiopoietin 2 persistently were elevated in patients receiving higher levels of respiratory support, whereas soluble receptor of advanced glycation end products (sRAGE) levels displayed the inverse pattern. Patients receiving NIV or HFO at baseline showed the most dynamic clinical trajectory, with 24% eventually requiring intubation and exhibiting worse 60-day mortality than patients receiving invasive mechanical ventilation at baseline (67% vs 35%; P < .0001). sRAGE levels predicted NIV failure and worse 60-day mortality for patients receiving NIV or HFO, whereas IL-6 levels were predictive in all patients regardless of level of support (P < .01). Patients classified to a hyperinflammatory subphenotype at baseline (< 10%) showed worse 60-day survival (P < .0001) and 50% of them remained classified as hyperinflammatory at 5 days after enrollment.

INTERPRETATION

Longitudinal study of the systemic host response in COVID-19 revealed substantial and predictive interindividual variability influenced by baseline levels of respiratory support.

Uncovering heterogeneity in sepsis: a comparative analysis of subphenotypes

PURPOSE

The heterogeneity in sepsis is held responsible, in part, for the lack of precision treatment. Many attempts to identify subtypes of sepsis patients identify those with shared underlying biology or outcomes. To date, though, there has been limited effort to determine overlap across these previously identified subtypes. We aimed to determine the concordance of critically ill patients with sepsis classified by four previously described subtype strategies.

METHODS

This secondary analysis of a multicenter prospective observational study included 522 critically ill patients with sepsis assigned to four previously established subtype strategies, primarily based on: (i) clinical data in the electronic health record (α, β, γ, and δ), (ii) biomarker data (hyper- and hypoinflammatory), and (iii-iv) transcriptomic data (Mars1-Mars4 and SRS1-SRS2). Concordance was studied between different subtype labels, clinical characteristics, biological host response aberrations, as well as combinations of subtypes by sepsis ensembles.

RESULTS

All four subtype labels could be adjudicated in this cohort, with the distribution of the clinical subtype varying most from the original cohort. The most common subtypes in each of the four strategies were γ (61%), which is higher compared to the original classification, hypoinflammatory (60%), Mars2 (35%), and SRS2 (54%). There was no clear relationship between any of the subtyping approaches (Cramer's V = 0.086-0.456). Mars2 and SRS1 were most alike in terms of host response biomarkers (p = 0.079-0.424), while other subtype strategies showed no clear relationship. Patients enriched for multiple subtypes revealed that characteristics and outcomes differ dependent on the combination of subtypes made.

CONCLUSION

Among critically ill patients with sepsis, subtype strategies using clinical, biomarker, and transcriptomic data do not identify comparable patient populations and are likely to reflect disparate clinical characteristics and underlying biology.

Identifying molecular phenotypes in sepsis: an analysis of two prospective observational cohorts and secondary analysis of two randomised controlled trials

BACKGROUND

In sepsis and acute respiratory distress syndrome (ARDS), heterogeneity has contributed to difficulty identifying effective pharmacotherapies. In ARDS, two molecular phenotypes (hypoinflammatory and hyperinflammatory) have consistently been identified, with divergent outcomes and treatment responses. In this study, we sought to derive molecular phenotypes in critically ill adults with sepsis, determine their overlap with previous ARDS phenotypes, and evaluate whether they respond differently to treatment in completed sepsis trials.

METHODS

We used clinical data and plasma biomarkers from two prospective sepsis cohorts, the Validating Acute Lung Injury biomarkers for Diagnosis (VALID) study (N=1140) and the Early Assessment of Renal and Lung Injury (EARLI) study (N=818), in latent class analysis (LCA) to identify the optimal number of classes in each cohort independently. We used validated models trained to classify ARDS phenotypes to evaluate concordance of sepsis and ARDS phenotypes. We applied these models retrospectively to the previously published Prospective Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis and Septic Shock (PROWESS-SHOCK) trial and Vasopressin and Septic Shock Trial (VASST) to assign phenotypes and evaluate heterogeneity of treatment effect.

FINDINGS

A two-class model best fit both VALID and EARLI (p<0·0001). In VALID, 804 (70·5%) of the 1140 patients were classified as hypoinflammatory and 336 (29·5%) as hyperinflammatory; in EARLI, 530 (64·8%) of 818 were hypoinflammatory and 288 (35·2%) hyperinflammatory. We observed higher plasma pro-inflammatory cytokines, more vasopressor use, more bacteraemia, lower protein C, and higher mortality in the hyperinflammatory than in the hypoinflammatory phenotype (p<0·0001 for all). Classifier models indicated strong concordance between sepsis phenotypes and previously identified ARDS phenotypes (area under the curve 0·87-0·96, depending on the model). Findings were similar excluding participants with both sepsis and ARDS. In PROWESS-SHOCK, 1142 (68·0%) of 1680 patients had the hypoinflammatory phenotype and 538 (32·0%) had the hyperinflammatory phenotype, and response to activated protein C differed by phenotype (p=0·0043). In VASST, phenotype proportions were similar to other cohorts; however, no treatment interaction with the type of vasopressor was observed (p=0·72).

INTERPRETATION

Molecular phenotypes previously identified in ARDS are also identifiable in multiple sepsis cohorts and respond differently to activated protein C. Molecular phenotypes could represent a treatable trait in critical illness beyond the patient's syndromic diagnosis.

FUNDING

US National Institutes of Health.

Identification and Prediction of Clinical Phenotypes in Hospitalized Patients With COVID-19: Machine Learning From Medical Records

BACKGROUND

There is significant heterogeneity in disease progression among hospitalized patients with COVID-19. The pathogenesis of SARS-CoV-2 infection is attributed to a complex interplay between virus and host immune response that in some patients unpredictably and rapidly leads to "hyperinflammation" associated with increased risk of mortality. The early identification of patients at risk of progression to hyperinflammation may help inform timely therapeutic decisions and lead to improved outcomes.

OBJECTIVE

The primary objective of this study was to use machine learning to reproducibly identify specific risk-stratifying clinical phenotypes across hospitalized patients with COVID-19 and compare treatment response characteristics and outcomes. A secondary objective was to derive a predictive phenotype classification model using routinely available early encounter data that may be useful in informing optimal COVID-19 bedside clinical management.

METHODS

This was a retrospective analysis of electronic health record data of adult patients (N=4379) who were admitted to a Johns Hopkins Health System hospital for COVID-19 treatment from 2020 to 2021. Phenotypes were identified by clustering 38 routine clinical observations recorded during inpatient care. To examine the reproducibility and validity of the derived phenotypes, patient data were randomly divided into 2 cohorts, and clustering analysis was performed independently for each cohort. A predictive phenotype classifier using the gradient-boosting machine method was derived using routine clinical observations recorded during the first 6 hours following admission.

RESULTS

A total of 2 phenotypes (designated as phenotype 1 and phenotype 2) were identified in patients admitted for COVID-19 in both the training and validation cohorts with similar distributions of features, correlations with biomarkers, treatments, comorbidities, and outcomes. In both the training and validation cohorts, phenotype-2 patients were older; had elevated markers of inflammation; and were at an increased risk of requiring intensive care unit-level care, developing sepsis, and mortality compared with phenotype-1 patients. The gradient-boosting machine phenotype prediction model yielded an area under the curve of 0.89 and a positive predictive value of 0.83.

CONCLUSIONS

Using machine learning clustering, we identified and internally validated 2 clinical COVID-19 phenotypes with distinct treatment or response characteristics consistent with similar 2-phenotype models derived from other hospitalized populations with COVID-19, supporting the reliability and generalizability of these findings. COVID-19 phenotypes can be accurately identified using machine learning models based on readily available early encounter clinical data. A phenotype prediction model based on early encounter data may be clinically useful for timely bedside risk stratification and treatment personalization.

The application of artificial intelligence in the management of sepsis

Sepsis is a complex and heterogeneous syndrome that remains a serious challenge to healthcare worldwide. Patients afflicted by severe sepsis or septic shock are customarily placed under intensive care unit (ICU) supervision, where a multitude of apparatus is poised to produce high-granularity data. This reservoir of high-quality data forms the cornerstone for the integration of AI into clinical practice. However, existing reviews currently lack the inclusion of the latest advancements. This review examines the evolving integration of artificial intelligence (AI) in sepsis management. Applications of artificial intelligence include early detection, subtyping analysis, precise treatment and prognosis assessment. AI-driven early warning systems provide enhanced recognition and intervention capabilities, while profiling analyzes elucidate distinct sepsis manifestations for targeted therapy. Precision medicine harnesses the potential of artificial intelligence for pathogen identification, antibiotic selection, and fluid optimization. In conclusion, the seamless amalgamation of artificial intelligence into the domain of sepsis management heralds a transformative shift, ushering in novel prospects to elevate diagnostic precision, therapeutic efficacy, and prognostic acumen. As AI technologies develop, their impact on shaping the future of sepsis care warrants ongoing research and thoughtful implementation.

Leveraging Data Science and Novel Technologies to Develop and Implement Precision Medicine Strategies in Critical Care

Precision medicine aims to identify treatments that are most likely to result in favorable outcomes for subgroups of patients with similar clinical and biological characteristics. The gaps for the development and implementation of precision medicine strategies in the critical care setting are many, but the advent of data science and multi-omics approaches, combined with the rich data ecosystem in the intensive care unit, offer unprecedented opportunities to realize the promise of precision critical care. In this article, the authors review the data-driven and technology-based approaches being leveraged to discover and implement precision medicine strategies in the critical care setting.

Long short-term memory model identifies ARDS and in-hospital mortality in both non-COVID-19 and COVID-19 cohort

OBJECTIVE

To identify the risk of acute respiratory distress syndrome (ARDS) and in-hospital mortality using long short-term memory (LSTM) framework in a mechanically ventilated (MV) non-COVID-19 cohort and a COVID-19 cohort.

METHODS

We included MV ICU patients between 2017 and 2018 and reviewed patient records for ARDS and death. Using active learning, we enriched this cohort with MV patients from 2016 to 2019 (MV non-COVID-19, n=3905). We collected a second validation cohort of hospitalised patients with COVID-19 in 2020 (COVID+, n=5672). We trained an LSTM model using 132 structured features on the MV non-COVID-19 training cohort and validated on the MV non-COVID-19 validation and COVID-19 cohorts.

RESULTS

Applying LSTM (model score 0.9) on the MV non-COVID-19 validation cohort had a sensitivity of 86% and specificity of 57%. The model identified the risk of ARDS 10 hours before ARDS and 9.4 days before death. The sensitivity (70%) and specificity (84%) of the model on the COVID-19 cohort are lower than MV non-COVID-19 cohort. For the COVID-19 + cohort and MV COVID-19 + patients, the model identified the risk of in-hospital mortality 2.4 days and 1.54 days before death, respectively.

DISCUSSION

Our LSTM algorithm accurately and timely identified the risk of ARDS or death in MV non-COVID-19 and COVID+ patients. By alerting the risk of ARDS or death, we can improve the implementation of evidence-based ARDS management and facilitate goals-of-care discussions in high-risk patients.

CONCLUSION

Using the LSTM algorithm in hospitalised patients identifies the risk of ARDS or death.

Advances of mesenchymal stem cells and their derived extracellular vesicles as a promising therapy for acute respiratory distress syndrome: from bench to clinic

Acute respiratory distress syndrome (ARDS) is an acute inflammatory lung injury characterized by diffuse alveolar damage. The period prevalence of ARDS was 10.4% of ICU admissions in 50 countries. Although great progress has been made in supportive care, the hospital mortality rate of severe ARDS is still up to 46.1%. Moreover, up to now, there is no effective pharmacotherapy for ARDS and most clinical trials focusing on consistently effective drugs have met disappointing results. Mesenchymal stem cells (MSCs) and their derived extracellular vesicles (EVs) have spawned intense interest of a wide range of researchers and clinicians due to their robust anti-inflammatory, anti-apoptotic and tissue regeneration properties. A growing body of evidence from preclinical studies confirmed the promising therapeutic potential of MSCs and their EVs in the treatment of ARDS. Based on the inspiring experimental results, clinical trials have been designed to evaluate safety and efficacy of MSCs and their EVs in ARDS patients. Moreover, trials exploring their optimal time window and regimen of drug administration are ongoing. Therefore, this review aims to present an overview of the characteristics of mesenchymal stem cells and their derived EVs, therapeutic mechanisms for ARDS and research progress that has been made over the past 5 years.

Clinical Sepsis Phenotypes in Critically Ill Patients

Sepsis, defined as the life-threatening dysregulated host response to an infection leading to organ dysfunction, is considered as one of the leading causes of mortality worldwide, especially in intensive care units (ICU). Moreover, sepsis remains an enigmatic clinical syndrome, with complex pathophysiology incompletely understood and a great heterogeneity both in terms of clinical expression, patient response to currently available therapeutic interventions and outcomes. This heterogeneity proves to be a major obstacle in our quest to deliver improved treatment in septic critical care patients; thus, identification of clinical phenotypes is absolutely necessary. Although this might be seen as an extremely difficult task, nowadays, artificial intelligence and machine learning techniques can be recruited to quantify similarities between individuals within sepsis population and differentiate them into distinct phenotypes regarding not only temperature, hemodynamics or type of organ dysfunction, but also fluid status/responsiveness, trajectories in ICU and outcome. Hopefully, we will eventually manage to determine both the subgroup of septic patients that will benefit from a therapeutic intervention and the correct timing of applying the intervention during the disease process.

Anti-inflammatory therapies for acute respiratory distress syndrome

INTRODUCTION

Treatments for the acute respiratory distress syndrome (ARDS) are mainly supportive, and ventilatory management represents a key approach in these patients. Despite progress in pharmacotherapy, anti-inflammatory strategies for the treatment of ARDS have shown controversial results. Positive outcomes with pharmacologic and nonpharmacologic treatments have been found in two different biological subphenotypes of ARDS, suggesting that, with a personalized medicine approach, pharmacotherapy for ARDS can be effective.

AREAS COVERED

This article reviews the literature concerning anti-inflammatory therapies for ARDS, focusing on pharmacological and stem-cell therapies, including extracellular vesicles.

EXPERT OPINION

Despite advances, ARDS treatments remain primarily supportive. Ventilatory and fluid management are important strategies in these patients that have demonstrated significant impacts on outcome. Anti-inflammatory drugs have shown some benefits, primarily in preclinical research and in specific clinical scenarios, but no recommendations are available from guidelines to support their use in patients with ARDS, except in particular settings such as different subphenotypes, specific etiologies, or clinical trials. Personalized medicine seems promising insofar as it may identify specific subgroups of patients with ARDS who may benefit from anti-inflammatory treatment. However, additional efforts are needed to move subphenotype characterization from bench to bedside.

ESICM guidelines on acute respiratory distress syndrome: definition, phenotyping and respiratory support strategies

The aim of these guidelines is to update the 2017 clinical practice guideline (CPG) of the European Society of Intensive Care Medicine (ESICM). The scope of this CPG is limited to adult patients and to non-pharmacological respiratory support strategies across different aspects of acute respiratory distress syndrome (ARDS), including ARDS due to coronavirus disease 2019 (COVID-19). These guidelines were formulated by an international panel of clinical experts, one methodologist and patients' representatives on behalf of the ESICM. The review was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement recommendations. We followed the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach to assess the certainty of evidence and grade recommendations and the quality of reporting of each study based on the EQUATOR (Enhancing the QUAlity and Transparency Of health Research) network guidelines. The CPG addressed 21 questions and formulates 21 recommendations on the following domains: (1) definition; (2) phenotyping, and respiratory support strategies including (3) high-flow nasal cannula oxygen (HFNO); (4) non-invasive ventilation (NIV); (5) tidal volume setting; (6) positive end-expiratory pressure (PEEP) and recruitment maneuvers (RM); (7) prone positioning; (8) neuromuscular blockade, and (9) extracorporeal life support (ECLS). In addition, the CPG includes expert opinion on clinical practice and identifies the areas of future research.