Host transcriptomics and machine learning for secondary bacterial infections in patients with COVID-19: a prospective, observational cohort study.

BACKGROUND: Viral respiratory tract infections are frequently complicated by secondary bacterial infections. This study aimed to use machine learning to predict the risk of bacterial superinfection in SARS-CoV-2-positive individuals. METHODS: In this prospective, multicentre, observational cohort study done in nine centres in six countries (Australia, Indonesia, Singapore, Italy, Czechia, and France) blood samples and RNA sequencing were used to develop a robust model of predicting secondary bacterial infections in the respiratory tract of patients with COVID-19. Eligible participants were older than 18 years, had known or suspected COVID-19, and symptoms of a recent respiratory infection. A control cohort of participants without COVID-19 who were older than 18 years and with no infection symptoms was also recruited from one Australian centre. In the pre-analysis phase, data were filtered to include only individuals with complete blood transcriptomics and patient data (ie, age, sex, location, and WHO severity score at the time of sample collection). The dataset was then divided randomly (4:1) into a training set (80%) and a test set (20%). Gene expression data in the training set and control cohort were used for differential expression analysis. Differentially expressed genes, along with WHO severity score, location, age, and sex, were used for feature selection with least absolute shrinkage and selection operator (LASSO) in the training set. For LASSO analysis, samples were excluded if gene expression data were not obtained at study admission, no longitudinal clinical information was available, a bacterial infection at the time of study admission was present, or a fungal infection in the absence of a bacterial infection was detected. LASSO regression was performed using three subsets of predictor variables: patient data alone, gene expression data alone, or a combination of patient data and gene expression data. The accuracy of the resultant models was tested on data from the test set. FINDINGS: Between March, 2020, and October, 2021, we recruited 536 SARS-CoV-2-positive individuals and between June, 2013, and January, 2020, we recruited 74 participants into the control cohort. After prefiltering analysis and other exclusions, samples from 158 individuals were analysed in the training set and 47 in the test set. The expression of seven host genes (DAPP1, CST3, FGL2, GCH1, CIITA, UPP1, and RN7SL1) in the blood at the time of study admission was identified by LASSO as predictive of the risk of developing a secondary bacterial infection of the respiratory tract more than 24 h after study admission. Specifically, the expression of these genes in combination with a patient's WHO severity score at the time of study enrolment resulted in an area under the curve of 0·98 (95% CI 0·89-1·00), a true positive rate (sensitivity) of 1·00 (95% CI 1·00-1·00), and a true negative rate (specificity) of 0·94 (95% CI 0·89-1·00) in the test cohort. The combination of patient data and host transcriptomics at hospital admission identified all seven individuals in the training and test sets who developed a bacterial infection of the respiratory tract 5-9 days after hospital admission. INTERPRETATION: These data raise the possibility that host transcriptomics at the time of clinical presentation, together with machine learning, can forward predict the risk of secondary bacterial infections and allow for the more targeted use of antibiotics in viral infection. FUNDING: Snow Medical Research Foundation, the National Health and Medical Research Council, the Jack Ma Foundation, the Helmholtz-Association, the A2 Milk Company, National Institute of Allergy and Infectious Disease, and the Fondazione AIRC Associazione Italiana per la Ricerca contro il Cancro.

The Role of Transcriptomics in Redefining Critical Illness.

This article is one of ten reviews selected from the Annual Update in Intensive Care and Emergency Medicine 2023. Other selected articles can be found online at  https://www.biomedcentral.com/collections/annualupdate2023 . Further information about the Annual Update in Intensive Care and Emergency Medicine is available from  https://link.springer.com/bookseries/8901 .

Pathway and Network Analyses Identify Growth Factor Signaling and MMP9 as Potential Mediators of Mitochondrial Dysfunction in Severe COVID-19.

Patients with preexisting metabolic disorders such as diabetes are at a higher risk of developing severe coronavirus disease 2019 (COVID-19). Mitochondrion, the very organelle that controls cellular metabolism, holds the key to understanding disease progression at the cellular level. Our current study aimed to understand how cellular metabolism contributes to COVID-19 outcomes. Metacore pathway enrichment analyses on differentially expressed genes (encoded by both mitochondrial and nuclear deoxyribonucleic acid (DNA)) involved in cellular metabolism, regulation of mitochondrial respiration and organization, and apoptosis, was performed on RNA sequencing (RNASeq) data from blood samples collected from healthy controls and patients with mild/moderate or severe COVID-19. Genes from the enriched pathways were analyzed by network analysis to uncover interactions among them and up- or downstream genes within each pathway. Compared to the mild/moderate COVID-19, the upregulation of a myriad of growth factor and cell cycle signaling pathways, with concomitant downregulation of interferon signaling pathways, were observed in the severe group. Matrix metallopeptidase 9 (MMP9) was found in five of the top 10 upregulated pathways, indicating its potential as therapeutic target against COVID-19. In summary, our data demonstrates aberrant activation of endocrine signaling in severe COVID-19, and its implication in immune and metabolic dysfunction.

The Sequential Organ Failure Assessment (SOFA) Score: has the time come for an update?

The Sequential Organ Failure Assessment (SOFA) score was developed more than 25 years ago to provide a simple method of assessing and monitoring organ dysfunction in critically ill patients. Changes in clinical practice over the last few decades, with new interventions and a greater focus on non-invasive monitoring systems, mean it is time to update the SOFA score. As a first step in this process, we propose some possible new variables that could be included in a SOFA 2.0. By so doing, we hope to stimulate debate and discussion to move toward a new, properly validated score that will be fit for modern practice.

Multisite validation of a host response signature for predicting likelihood of bacterial and viral infections in patients with suspected influenza.

BACKGROUND: Indiscriminate use of antimicrobials and antimicrobial resistance is a public health threat. IMX-BVN-1, a 29-host mRNA classifier, provides two separate scores that predict likelihoods of bacterial and viral infections in patients with suspected acute infections. We validated the performance of IMX-BVN-1 in adults attending acute health care settings with suspected influenza. METHOD: We amplified 29-host response genes in RNA extracted from blood by NanoString nCounter. IMX-BVN-1 calculated two scores to predict probabilities of bacterial and viral infections. Results were compared against the infection status (no infection; highly probable/possible infection; confirmed infection) determined by clinical adjudication. RESULTS: Amongst 602 adult patients (74.9% ED, 16.9% ICU, 8.1% outpatients), 7.6% showed in-hospital mortality and 15.5% immunosuppression. Median IMX-BVN-1 bacterial and viral scores were higher in patients with confirmed bacterial (0.27) and viral (0.62) infections than in those without bacterial (0.08) or viral (0.21) infection, respectively. The AUROC distinguishing bacterial from nonbacterial illness was 0.81 and 0.87 when distinguishing viral from nonviral illness. The bacterial top quartile's positive likelihood ratio (LR) was 4.38 with a rule-in specificity of 88%; the bacterial bottom quartile's negative LR was 0.13 with a rule-out sensitivity of 96%. Similarly, the viral top quartile showed an infinite LR with rule-in specificity of 100%; the viral bottom quartile had a LR of 0.22 and a rule-out sensitivity of 85%. CONCLUSION: IMX-BVN-1 showed high accuracy for differentiating bacterial and viral infections from noninfectious illness in patients with suspected influenza. Clinical utility of IMX-BVN will be validated following integration into a point of care system.

Site-tropism of streptococci in the oral microbiome.

A detailed understanding of where bacteria localize is necessary to advance microbial ecology and microbiome-based therapeutics. The site-specialist hypothesis predicts that most microbes in the human oral cavity have a primary habitat type within the mouth where they are most abundant. We asked whether this hypothesis accurately describes the distribution of the members of the genus Streptococcus, a clinically relevant taxon that dominates most oral sites. Prior analysis of 16S rRNA gene sequencing data indicated that some oral Streptococcus clades are site-specialists while others may be generalists. However, within complex microbial populations composed of numerous closely related species and strains, such as the oral streptococci, genome-scale analysis is necessary to provide the resolution to discriminate closely related taxa with distinct functional roles. Here, we assess whether individual species within this genus are specialists using publicly available genomic sequence data that provide species-level resolution. We chose a set of high-quality representative genomes for human oral Streptococcus species. Onto these genomes, we mapped shotgun metagenomic sequencing reads from supragingival plaque, tongue dorsum, and other sites in the oral cavity. We found that every abundant Streptococcus species in the healthy human oral cavity showed strong site-tropism and that even closely related species such as S. mitis, S. oralis, and S. infantis specialized in different sites. These findings indicate that closely related bacteria can have distinct habitat distributions in the absence of dispersal limitation and under similar environmental conditions and immune regimes. Substantial overlap between the core genes of these three species suggests that site-specialization is determined by subtle differences in genomic content.

Bedside assessment of left atrial pressure in critical care: a multifaceted gem.

Evaluating left atrial pressure (LAP) solely from the left ventricular preload perspective is a restrained approach. Accurate assessment of LAP is particularly relevant when pulmonary congestion and/or right heart dysfunction are present since it is the pressure most closely related to pulmonary venous pressure and thus pulmonary haemodynamic load. Amalgamation of LAP measurement into assessment of the 'transpulmonary circuit' may have a particular role in differentiating cardiac failure phenotypes in critical care. Most of the literature in this area involves cardiology patients, and gaps of knowledge in application to the bedside of the critically ill patient remain significant. Explored in this review is an overview of left atrial physiology, invasive and non-invasive methods of LAP measurement and their potential clinical application.

Redefining critical illness.

Research and practice in critical care medicine have long been defined by syndromes, which, despite being clinically recognizable entities, are, in fact, loose amalgams of heterogeneous states that may respond differently to therapy. Mounting translational evidence-supported by research on respiratory failure due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection-suggests that the current syndrome-based framework of critical illness should be reconsidered. Here we discuss recent findings from basic science and clinical research in critical care and explore how these might inform a new conceptual model of critical illness. De-emphasizing syndromes, we focus on the underlying biological changes that underpin critical illness states and that may be amenable to treatment. We hypothesize that such an approach will accelerate critical care research, leading to a richer understanding of the pathobiology of critical illness and of the key determinants of patient outcomes. This, in turn, will support the design of more effective clinical trials and inform a more precise and more effective practice at the bedside.

Echocardiography findings in COVID-19 patients admitted to intensive care units: a multi-national observational study (the ECHO-COVID study).

PURPOSE: Severely ill patients affected by coronavirus disease 2019 (COVID-19) develop circulatory failure. We aimed to report patterns of left and right ventricular dysfunction in the first echocardiography following admission to intensive care unit (ICU). METHODS: Retrospective, descriptive study that collected echocardiographic and clinical information from severely ill COVID-19 patients admitted to 14 ICUs in 8 countries. Patients admitted to ICU who received at least one echocardiography between 1st February 2020 and 30th June 2021 were included. Clinical and echocardiographic data were uploaded using a secured web-based electronic database (REDCap). RESULTS: Six hundred and seventy-seven patients were included and the first echo was performed 2 [1, 4] days after ICU admission. The median age was 65 [56, 73] years, and 71% were male. Left ventricle (LV) and/or right ventricle (RV) systolic dysfunction were found in 234 (34.5%) patients. 149 (22%) patients had LV systolic dysfunction (with or without RV dysfunction) without LV dilatation and no elevation in filling pressure. 152 (22.5%) had RV systolic dysfunction. In 517 patients with information on both paradoxical septal motion and quantitative RV size, 90 (17.4%) had acute cor pulmonale (ACP). ACP was associated with mechanical ventilation (OR > 4), pulmonary embolism (OR > 5) and increased PaCO2. Exploratory analyses showed that patients with ACP and older age were more likely to die in hospital (including ICU). CONCLUSION: Almost one-third of this cohort of critically ill COVID-19 patients exhibited abnormal LV and/or RV systolic function in their first echocardiography assessment. While LV systolic dysfunction appears similar to septic cardiomyopathy, RV systolic dysfunction was related to pressure overload due to positive pressure ventilation, hypercapnia and pulmonary embolism. ACP and age seemed to be associated with mortality in this cohort.

The Role of Mitochondria in the Immune Response in Critical Illness.

This article is one of ten reviews selected from the Annual Update in Intensive Care and Emergency Medicine 2022. Other selected articles can be found online at https://www.biomedcentral.com/collections/annualupdate2022 . Further information about the Annual Update in Intensive Care and Emergency Medicine is available from https://link.springer.com/bookseries/8901 .

Blood transcriptome responses in patients correlate with severity of COVID-19 disease.

Background: Coronavirus disease 2019 (COVID-19) is an infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Infected individuals display a wide spectrum of disease severity, as defined by the World Health Organization (WHO). One of the main factors underlying this heterogeneity is the host immune response, with severe COVID-19 often associated with a hyperinflammatory state. Aim: Our current study aimed to pinpoint the specific genes and pathways underlying differences in the disease spectrum and outcomes observed, through in-depth analyses of whole blood transcriptomics in a large cohort of COVID-19 participants. Results: All WHO severity levels were well represented and mild and severe disease displaying distinct gene expression profiles. WHO severity levels 1-4 were grouped as mild disease, and signatures from these participants were different from those with WHO severity levels 6-9 classified as severe disease. Severity level 5 (moderate cases) presented a unique transitional gene signature between severity levels 2-4 (mild/moderate) and 6-9 (severe) and hence might represent the turning point for better or worse disease outcome. Gene expression changes are very distinct when comparing mild/moderate or severe cases to healthy controls. In particular, we demonstrated the hallmark down-regulation of adaptive immune response pathways and activation of neutrophil pathways in severe compared to mild/moderate cases, as well as activation of blood coagulation pathways. Conclusions: Our data revealed discrete gene signatures associated with mild, moderate, and severe COVID-19 identifying valuable candidates for future biomarker discovery.

Mechanisms underlying interactions between two abundant oral commensal bacteria.

Complex polymicrobial biofilm communities are abundant in nature particularly in the human oral cavity where their composition and fitness can affect health. While the study of these communities during disease is essential and prevalent, little is known about interactions within the healthy plaque community. Here we describe interactions between two of the most abundant species in this healthy microbiome, Haemophilus parainfluenzae and Streptococcus mitis. We discovered that H. parainfluenzae typically exists adjacent to mitis group streptococci in vivo with which it is also positively correlated based on microbiome data. By comparing in vitro coculture data to ex vivo microscopy we revealed that this co-occurrence is density dependent and further influenced by H2O2 production. We discovered that H. parainfluenzae utilizes a more redundant, multifactorial response to H2O2 than related microorganisms and that this system's integrity enhances streptococcal fitness. Our results indicate that mitis group streptococci are likely the in vivo source of NAD for H. parainfluenzae and also evoke patterns of carbon utilization in vitro for H. parainfluenzae similar to those observed in vivo. Our findings describe mechanistic interactions between two of the most abundant and prevalent members of healthy supragingival plaque that contribute to their in vivo survival.

Association Between Premorbid Beta-Blocker Exposure and Sepsis Outcomes-The Beta-Blockers in European and Australian/American Septic Patients (BEAST) Study.

OBJECTIVES: To examine the effect of premorbid ß-blocker exposure on mortality and organ dysfunction in sepsis. DESIGN: Retrospective observational study. SETTING: ICUs in Australia, the Czech Republic, and the United States. PATIENTS: Total of 4,086 critical care patients above 18 years old with sepsis between January 2014 and December 2018. INTERVENTION: Premorbid beta-blocker exposure. MEASUREMENTS AND MAIN RESULTS: One thousand five hundred fifty-six patients (38%) with premorbid ß-blocker exposure were identified. Overall ICU mortality rate was 15.1%. In adjusted models, premorbid ß-blocker exposure was associated with decreased ICU (adjusted odds ratio, 0.80; 95% CI, 0.66-0.97; p = 0.025) and hospital (adjusted odds ratio, 0.83; 95% CI, 0.71-0.99; p = 0.033) mortality. The risk reduction in ICU mortality of 16% was significant (hazard ratio, 0.84, 95% CI, 0.71-0.99; p = 0.037). In particular, exposure to noncardioselective ß-blocker before septic episode was associated with decreased mortality. Sequential Organ Failure Assessment score analysis showed that premorbid ß-blocker exposure had potential benefits in reducing respiratory and neurologic dysfunction. CONCLUSIONS: This study suggests that ß-blocker exposure prior to sepsis, especially to noncardioselective ß blockers, may be associated with better outcome. The findings suggest prospective evaluation of ß-blocker use in the management of sepsis.

Immunophenotyping of Peripheral Blood Mononuclear Cells in Septic Shock Patients With High-Dimensional Flow Cytometry Analysis Reveals Two Subgroups With Differential Responses to Immunostimulant Drugs.

Sepsis is associated with a dysregulated inflammatory response to infection. Despite the activation of inflammation, an immune suppression is often observed, predisposing patients to secondary infections. Therapies directed at restoration of immunity may be considered but should be guided by the immune status of the patients. In this paper, we described the use of a high-dimensional flow cytometry (HDCyto) panel to assess the immunophenotype of patients with sepsis. We then isolated peripheral blood mononuclear cells (PBMCs) from patients with septic shock and mimicked a secondary infection by stimulating PBMCs for 4 h in vitro with lipopolysaccharide (LPS) with or without prior exposure to either IFN-γ, or LAG-3Ig. We evaluated the response by means of flow cytometry and high-resolution clustering cum differential analysis and compared the results to PBMCs from healthy donors. We observed a heterogeneous immune response in septic patients and identified two major subgroups: one characterized by hypo-responsiveness (Hypo) and another one by hyper-responsiveness (Hyper). Hypo and Hyper groups showed significant differences in the production of cytokines/chemokine and surface human leukocyte antigen-DR (HLA-DR) expression in response to LPS stimulation, which were observed across all cell types. When pre-treated with either interferon gamma (IFN-γ) or lymphocyte-activation gene 3 (LAG)-3 recombinant fusion protein (LAG-3Ig) prior to LPS stimulation, cells from the Hypo group were shown to be more responsive to both immunostimulants than cells from the Hyper group. Our results demonstrate the importance of patient stratification based on their immune status prior to any immune therapies. Once sufficiently scaled, this approach may be useful for prescribing the right immune therapy for the right patient at the right time, the key to the success of any therapy.

Blood transcriptome analysis of patients with uncomplicated bacterial infection and sepsis.

OBJECTIVES: Hospitalized patients who presented within the last 24 h with a bacterial infection were recruited. Participants were assigned into sepsis and uncomplicated infection groups. In addition, healthy volunteers were recruited as controls. RNA was prepared from whole blood, depleted from beta-globin mRNA and sequenced. This dataset represents a highly valuable resource to better understand the biology of sepsis and to identify biomarkers for severe sepsis in humans. DATA DESCRIPTION: The data presented here consists of raw and processed transcriptome data obtained by next generation RNA sequencing from 105 peripheral blood samples from patients with uncomplicated infections, patients who developed sepsis, septic shock patients, and healthy controls. It is provided as raw sequenced reads and as normalized log2 transformed relative expression levels. This data will allow performing detailed analyses of gene expression changes between uncomplicated infections and sepsis patients, such as identification of differentially expressed genes, co-regulated modules as well as pathway activation studies.

Prospective validation study of prognostic biomarkers to predict adverse outcomes in patients with COVID-19: a study protocol.

INTRODUCTION: Accurate triage is an important first step to effectively manage the clinical treatment of severe cases in a pandemic outbreak. In the current COVID-19 global pandemic, there is a lack of reliable clinical tools to assist clinicians to perform accurate triage. Host response biomarkers have recently shown promise in risk stratification of disease progression; however, the role of these biomarkers in predicting disease progression in patients with COVID-19 is unknown. Here, we present a protocol outlining a prospective validation study to evaluate the biomarkers' performance in predicting clinical outcomes of patients with COVID-19. METHODS AND ANALYSIS: This prospective validation study assesses patients infected with COVID-19, in whom blood samples are prospectively collected. Recruited patients include a range of infection severity from asymptomatic to critically ill patients, recruited from the community, outpatient clinics, emergency departments and hospitals. Study samples consist of peripheral blood samples collected into RNA-preserving (PAXgene/Tempus) tubes on patient presentation or immediately on study enrolment. Real-time PCR (RT-PCR) will be performed on total RNA extracted from collected blood samples using primers specific to host response gene expression biomarkers that have been previously identified in studies of respiratory viral infections. The RT-PCR data will be analysed to assess the diagnostic performance of individual biomarkers in predicting COVID-19-related outcomes, such as viral pneumonia, acute respiratory distress syndrome or bacterial pneumonia. Biomarker performance will be evaluated using sensitivity, specificity, positive and negative predictive values, likelihood ratios and area under the receiver operating characteristic curve. ETHICS AND DISSEMINATION: This research protocol aims to study the host response gene expression biomarkers in severe respiratory viral infections with a pandemic potential (COVID-19). It has been approved by the local ethics committee with approval number 2020/ETH00886. The results of this project will be disseminated in international peer-reviewed scientific journals.

Evaluating the influence of data collector training for predictive risk of death models: an observational study.

BACKGROUND: Severity-of-illness scoring systems are widely used for quality assurance and research. Although validated by trained data collectors, there is little data on the accuracy of real-world data collection practices. OBJECTIVE: To evaluate the influence of formal data collection training on the accuracy of scoring system data in intensive care units (ICUs). STUDY DESIGN AND METHODS: Quality assurance audit conducted using survey methodology principles. Between June and December 2018, an electronic document with details of three fictitious ICU patients was emailed to staff from 19 Australian ICUs who voluntarily submitted data on a web-based data entry form. Their entries were used to generate severity-of-illness scores and risks of death (RoDs) for four scoring systems. The primary outcome was the variation of severity-of-illness scores and RoDs from a reference standard. RESULTS: 50/83 staff (60.3%) submitted data. Using Bayesian multilevel analysis, severity-of-illness scores and RoDs were found to be significantly higher for untrained staff. The mean (95% high-density interval) overestimation in RoD due to training effect for patients 1, 2 and 3, respectively, were 0.24 (0.16, 0.31), 0.19 (0.09, 0.29) and 0.24 (0.1, 0.38) respectively (Bayesian factor >300, decisive evidence). Both groups (trained and untrained) had wide coefficients of variation up to 38.1%, indicating wide variability. Untrained staff made more errors in interpreting scoring system definitions. INTERPRETATION: In a fictitious patient dataset, data collection staff without formal training significantly overestimated the severity-of-illness scores and RoDs compared with trained staff. Both groups exhibited wide variability. Strategies to improve practice may include providing adequate training for all data collection staff, refresher training for previously trained staff and auditing the raw data submitted by individual ICUs. The results of this simulated study need revalidation on real patients.

The PRICES statement: an ESICM expert consensus on methodology for conducting and reporting critical care echocardiography research studies.

PURPOSE: Echocardiography is a common tool for cardiac and hemodynamic assessments in critical care research. However, interpretation (and applications) of results and between-study comparisons are often difficult due to the lack of certain important details in the studies. PRICES (Preferred Reporting Items for Critical care Echocardiography Studies) is a project endorsed by the European Society of Intensive Care Medicine and conducted by the Echocardiography Working Group, aiming at producing recommendations for standardized reporting of critical care echocardiography (CCE) research studies. METHODS: The PRICE panel identified lists of clinical and echocardiographic parameters (the "items") deemed important in four main areas of CCE research: left ventricular systolic and diastolic functions, right ventricular function and fluid management. Each item was graded using a critical index (CI) that combined the relative importance of each item and the fraction of studies that did not report it, also taking experts' opinion into account. RESULTS: A list of items in each area that deemed essential for the proper interpretation and application of research results is recommended. Additional items which aid interpretation were also proposed. CONCLUSION: The PRICES recommendations reported in this document, as a checklist, represent an international consensus of experts as to which parameters and information should be included in the design of echocardiography research studies. PRICES recommendations provide guidance to scientists in the field of CCE with the objective of providing a recommended framework for reporting of CCE methodology and results.