Dynamic associations between the respiratory tract and gut antibiotic resistome of patients with COVID-19 and its prediction power for disease severity.

The antibiotic resistome is the collection of all antibiotic resistance genes (ARGs) present in an individual. Whether an individual's susceptibility to infection and the eventual severity of coronavirus disease 2019 (COVID-19) is influenced by their respiratory tract antibiotic resistome is unknown. Additionally, whether a relationship exists between the respiratory tract and gut ARGs composition has not been fully explored. We recruited 66 patients with COVID-19 at three disease stages (admission, progression, and recovery) and conducted a metagenome sequencing analysis of 143 sputum and 97 fecal samples obtained from them. Respiratory tract, gut metagenomes, and peripheral blood mononuclear cell (PBMC) transcriptomes are analyzed to compare the gut and respiratory tract ARGs of intensive care unit (ICU) and non-ICU (nICU) patients and determine relationships between ARGs and immune response. Among the respiratory tract ARGs, we found that Aminoglycoside, Multidrug, and Vancomycin are increased in ICU patients compared with nICU patients. In the gut, we found that Multidrug, Vancomycin, and Fosmidomycin were increased in ICU patients. We discovered that the relative abundances of Multidrug were significantly correlated with clinical indices, and there was a significantly positive correlation between ARGs and microbiota in the respiratory tract and gut. We found that immune-related pathways in PBMC were enhanced, and they were correlated with Multidrug, Vancomycin, and Tetracycline ARGs. Based on the ARG types, we built a respiratory tract-gut ARG combined random-forest classifier to distinguish ICU COVID-19 patients from nICU patients with an AUC of 0.969. Cumulatively, our findings provide some of the first insights into the dynamic alterations of respiratory tract and gut antibiotic resistome in the progression of COVID-19 and disease severity. They also provide a better understanding of how this disease affects different cohorts of patients. As such, these findings should contribute to better diagnosis and treatment scenarios.

A multistage multimodal deep learning model for disease severity assessment and early warnings of high-risk patients of COVID-19.

The outbreak of coronavirus disease 2019 (COVID-19) has caused massive infections and large death tolls worldwide. Despite many studies on the clinical characteristics and the treatment plans of COVID-19, they rarely conduct in-depth prognostic research on leveraging consecutive rounds of multimodal clinical examination and laboratory test data to facilitate clinical decision-making for the treatment of COVID-19. To address this issue, we propose a multistage multimodal deep learning (MMDL) model to (1) first assess the patient's current condition (i.e., the mild and severe symptoms), then (2) give early warnings to patients with mild symptoms who are at high risk to develop severe illness. In MMDL, we build a sequential stage-wise learning architecture whose design philosophy embodies the model's predicted outcome and does not only depend on the current situation but also the history. Concretely, we meticulously combine the latest round of multimodal clinical data and the decayed past information to make assessments and predictions. In each round (stage), we design a two-layer multimodal feature extractor to extract the latent feature representation across different modalities of clinical data, including patient demographics, clinical manifestation, and 11 modalities of laboratory test results. We conduct experiments on a clinical dataset consisting of 216 COVID-19 patients that have passed the ethical review of the medical ethics committee. Experimental results validate our assumption that sequential stage-wise learning outperforms single-stage learning, but history long ago has little influence on the learning outcome. Also, comparison tests show the advantage of multimodal learning. MMDL with multimodal inputs can beat any reduced model with single-modal inputs only. In addition, we have deployed the prototype of MMDL in a hospital for clinical comparison tests and to assist doctors in clinical diagnosis.

A multistage multimodal deep learning model for disease severity assessment and early warnings of high-risk patients of COVID-19

The outbreak of coronavirus disease 2019 (COVID-19) has caused massive infections and large death tolls worldwide. Despite many studies on the clinical characteristics and the treatment plans of COVID-19, they rarely conduct in-depth prognostic research on leveraging consecutive rounds of multimodal clinical examination and laboratory test data to facilitate clinical decision-making for the treatment of COVID-19. To address this issue, we propose a multistage multimodal deep learning (MMDL) model to (1) first assess the patient's current condition (i.e., the mild and severe symptoms), then (2) give early warnings to patients with mild symptoms who are at high risk to develop severe illness. In MMDL, we build a sequential stage-wise learning architecture whose design philosophy embodies the model's predicted outcome and does not only depend on the current situation but also the history. Concretely, we meticulously combine the latest round of multimodal clinical data and the decayed past information to make assessments and predictions. In each round (stage), we design a two-layer multimodal feature extractor to extract the latent feature representation across different modalities of clinical data, including patient demographics, clinical manifestation, and 11 modalities of laboratory test results. We conduct experiments on a clinical dataset consisting of 216 COVID-19 patients that have passed the ethical review of the medical ethics committee. Experimental results validate our assumption that sequential stage-wise learning outperforms single-stage learning, but history long ago has little influence on the learning outcome. Also, comparison tests show the advantage of multimodal learning. MMDL with multimodal inputs can beat any reduced model with single-modal inputs only. In addition, we have deployed the prototype of MMDL in a hospital for clinical comparison tests and to assist doctors in clinical diagnosis.

Dynamic Alterations in the Respiratory Tract Microbiota of Patients with COVID-19 and its Association with Microbiota in the Gut.

The role of respiratory tract microbes and the relationship between respiratory tract and gut microbiomes in coronavirus disease 2019 (COVID-19) remain uncertain. Here, the metagenomes of sputum and fecal samples from 66 patients with COVID-19 at three stages of disease progression are sequenced. Respiratory tract, gut microbiome, and peripheral blood mononuclear cell (PBMC) samples are analyzed to compare the gut and respiratory tract microbiota of intensive care unit (ICU) and non-ICU (nICU) patients and determine relationships between respiratory tract microbiome and immune response. In the respiratory tract, significantly fewer Streptococcus, Actinomyces, Atopobium, and Bacteroides are found in ICU than in nICU patients, while Enterococcus and Candida increase. In the gut, significantly fewer Bacteroides are found in ICU patients, while Enterococcus increases. Significant positive correlations exist between relative microbiota abundances in the respiratory tract and gut. Defensin-related pathways in PBMCs are enhanced, and respiratory tract Streptococcus is reduced in patients with COVID-19. A respiratory tract-gut microbiota model identifies respiratory tract Streptococcus and Atopobium as the most prominent biomarkers distinguishing between ICU and nICU patients. The findings provide insight into the respiratory tract and gut microbial dynamics during COVID-19 progression, considering disease severity, potentially contributing to diagnosis, and treatment strategies.

Serum level of testosterone predicts disease severity of male COVID-19 patients and is related to T-cell immune modulation by transcriptome analysis.

BACKGROUND: Severe disease of COVID-19 and mortality occur more frequently in male patients than that in female patients may be related to testosterone level. However, the diagnostic value of changes in the level of testosterone in predicting severe disease of male COVID-19 patients has not been determined yet. METHODS: Sixty-one male COVID-19 patients admitted to the First Affiliated Hospital of Zhejiang University School of Medicine were enrolled. Serum samples at different stages of the patients after admission were collected and testosterone levels were detected to analyze the correlation between testosterone level and disease severity. Transcriptome analysis of PBMC was performed in 34 patients. RESULTS: Testosterone levels at admission in male non-ICU COVID-19 patients (3.7 nmol/L, IQR: 1.5 âˆ¼ 4.7) were significantly lower than those in male ICU COVID-19 patients (6.7 nmol/L, IQR: 4.2 âˆ¼ 8.7). Testosterone levels in the non-ICU group increased gradually during the progression of the disease, while those in the ICU group remained low. In addition, testosterone level of enrolled patients in the second week after onset was significantly correlated with the severity of pneumonia, and ROC curve showed that testosterone level in the second week after onset was highly effective in predicting the severity of COVID-19. Transcriptome studies have found that testosterone levels of COVID-19 patients were associated with immune response, including T cell activation and regulation of lymphocyte activation. In addition, CD28 and Inositol Polyphosphate-4-Phosphatase Type II B (INPP4B) were found positively correlated with testosterone. CONCLUSIONS: Serum testosterone is an independent risk factor for predicting the severity of COVID-19 in male patients, and the level of serum testosterone in the second week after onset is valuable for evaluating the severity of COVID-19. Testosterone level is associated with T cell immune activation. The monitoring of serum testosterone should be highlighted in clinical treatment and the related mechanism should be further studied.