Lymphocyte-C-reactive protein ratio can differentiate disease severity of COVID-19 patients and serve as an assistant screening tool for hospital and ICU admission (preprint)

In this study, we aimed to explore whether Lymphocyte-C-reactive protein ratio (LCR) can differentiate disease severity of Coronavirus disease 2019 (COVID-19) patients and its value as an assistant screening tool for admission to the hospital and the intensive care unit (ICU). A total of 184 adult COVID-19 patients from the COVID-19 Treatment Center in Heilongjiang Province at the First Affiliated Hospital of Harbin Medical University between January 2020 and March 2021 were included in this study. Patients were divided into asymptomatic infection group, mild group, moderate group, severe group, and critical group according to the Diagnosis and Treatment of New Coronavirus Pneumonia (9th edition). Demographic and clinical data including gender, age, comorbidities, severity of COVID-19, white blood cell count (WBC), neutrophil proportion (NEUT%), lymphocyte count (LYMPH), lymphocyte percentage (LYM%), red blood cell distribution width (RDW), platelet (PLT), C-reaction protein (CRP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), serum creatinine (SCr), albumin (ALB), total bilirubin (TB), direct bilirubin (DBIL), indirect bilirubin (IBIL), and D-Dimer were obtained and collated from medical records at admission, from which sequential organ failure assessment (SOFA) score and LCR were calculated, and all above indicators were compared among groups. Multiple clinical parameters, including LYMPH, CRP and LCR, showed significant differences among groups. The related factors to classify COVID-19 patients into moderate, severe and critical groups included age, number of comorbidities, WBC, LCR, and AST. Among these factors, number of comorbidities showed the greatest effect, and only WBC and LCR were protective factors. The area under the receiver operating characteristic (ROC) curve of LCR to classify COVID-19 patients into moderate, severe and critical groups was 0.176. The cut-off value of LCR, and the sensitivity and specificity of ROC curve were 1780.7050, 84.6% and 66.2%, respectively. The related factors to classify COVID-19 patients into severe and critical groups included number of comorbidities, PLT, LCR, and SOFA score. Among these factors, SOFA score showed the greatest effect, and LCR was the only protective factor. The area under ROC curve of LCR to classify COVID-19 patients into severe and critical groups was 0.106. The cut-off value of LCR and the sensitivity and specificity of ROC curve were 571.2200, 81.3% and 90.0%, respectively. In summary, LCR can differentiate disease severity of COVID-19 patients and serve as a simple and objective assistant screening tool for hospital and ICU admission.

On the pragmatism of using binary classifiers over data intensive neural network classifiers for detection of COVID-19 from voice (preprint)

Lately, there has been a global effort by multiple research groups to detect COVID-19 from voice. Different researchers use different kinds of information from the voice signal to achieve this. Various types of phonated sounds and the sound of cough and breath have all been used with varying degree of success in automated voice-based COVID-19 detection apps. In this paper, we show that detecting COVID-19 from voice does not require custom-made non-standard features or complicated neural network classifiers rather it can be successfully done with just standard features and simple binary classifiers. In fact, we show that the latter is not only more accurate and interpretable but also more computationally efficient in that they can be run locally on small devices. We demonstrate this on a human-curated dataset of over 1000 subjects, collected and calibrated in clinical settings.

Continuous Renal Replacement Therapy with oXiris Filter May Not Be an Effective Resolution to Alleviate Cytokine Release Syndrome in non-AKI Patients with Severe and Critical COVID-19 (preprint)

Background: In this study, we aimed to determine whether continuous renal replacement therapy (CRRT) with oXiris filter may alleviate cytokine release syndrome (CRS) in non-AKI patients with severe and critical coronavirus disease 2019 (COVID-19). Methods: Non-AKI patients with severe and critical COVID-19 treated between February 14 and March 26, 2020 were included and randomly divided into intervention group and control group according to the random number table. Patients in the intervention group received CRRT with oXiris filter plus conventional treatment, while those in the control group only received conventional treatment. Demographic data were collected and collated at admission. During ICU hospitalization, the serum levels of cytokine and inflammatory chemokines, including IL-2, IL-4, IL-6, IL-10, TNF-α, and IFN-γ, were measured daily to reflect the degree of CRS induced by SARS-CoV-2 infection. Clinical data, including white blood cell count (WBC), neutrophil proportion (NEUT%), lymphocyte count (LYMPH), lymphocyte percentage (LYM%), platelet (PLT), C-reaction protein (CRP), high sensitivity C-reactive protein (hs-CRP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TB), albumin (ALB), serum creatinine (SCr), D-Dimer, fibrinogen (FIB), IL-2, IL-4, IL-6, IL-10, TNF-α, IFN-γ, number of hospital days and sequential organ failure assessment (SOFA) score were obtained and collated from medical records during hospitalization, and then compared between the two groups. Results: Age, and SCr significantly differed between the two groups. Besides the IL-2 level that was significantly lower on day 2 than that on day 1 in the intervention group, and the IL-6 levels that were significantly higher on day 1, and day 2 in the intervention group compared to the control group, similar to the IL-10 level on day 5, there were no significant differences between the groups. Conclusion: CRRT with oXiris filter may not effectively alleviate CRS in non-AKI patients with severe and critical COVID-19. Thus, its application in these patients should be considered with caution to avoid increasing the unnecessary burden on society and individuals and making the already overwhelmed medical system even more strained (IRB number: IRB-AF/SC-04).

Spatial Distribution and Mechanism of Urban Occupation Mixture in Guangzhou: An Optimized GeoDetector-Based Index to Compare Individual and Interactive Effects

Numerous studies have been devoted to uncovering the characteristics of resident density and urban mobility with multisource geospatial big data. However, little attention has been paid to the internal diversity of residents such as their occupations, which is a crucial aspect of urban vibrancy. This study aims to investigate the variation between individual and interactive influences of built environment factors on occupation mixture index (OMI) with a novel GeoDetector-based indicator. This study first integrated application (App) use and mobility patterns from cellphone data to portray residents’ occupations and evaluate the OMI in Guangzhou. Then, the mechanism of OMI distribution was analyzed with the GeoDetector model. Next, an optimized GeoDetector-based index, interactive effect variation ratio (IEVR) was proposed to quantify the variation between individual and interactive effects of factors. The results showed that land use mixture was the dominating factor, and that land use mixture, building density, floor area ratio, road density affected the OMI distribution directly. Some interesting findings were uncovered by IEVR. The influences of cultural inclusiveness and metro accessibility were less important in factor detector result, while they were found to be the most influential in an indirect way interacting with other built environment factors. The results suggested that both “hardware facilities” (land use mixture, accessibility) and “soft facilities” (cultural inclusiveness) should be considered in planning a harmonious urban employment space and sustainable city.

Twelve-month specific IgG response to SARS-CoV-2 receptor-binding domain among COVID-19 convalescent plasma donors in Wuhan (preprint)

To investigate the duration of humoral immune response in convalescent coronavirus disease 2019 (COVID-19) patients, we conducted a 12-month longitudinal study through collecting a total of 1,782 plasma samples from 869 convalescent plasma donors in Wuhan, China and tested specific antibody response. The results show that positive rate of IgG antibody against receptor-binding domain of spike protein (RBD-IgG) to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in the COVID-19 convalescent plasma donors exceeded 70% for 12 months post diagnosis. RBD-IgG kinetics displayed a gradually downward trend, the titer started to stabilize after 9 months and decreased by 68.1% compared with the 1st month. Moreover, male plasma donors produced more RBD-IgG than female plasma donors and patient age positively correlated with the RBD-IgG titer. A strong positive correlation between RBD-IgG and neutralizing antibody titers was also identified. This study is essential for understanding SARS-CoV-2-induced immune memory to develop vaccine and therapeutics.

A novel multiple instance learning framework for COVID-19 severity assessment via data augmentation and self-supervised learning (preprint)

How to fast and accurately assess the severity level of COVID-19 is an essential problem, when millions of people are suffering from the pandemic around the world. Currently, the chest CT is regarded as a popular and informative imaging tool for COVID-19 diagnosis. However, we observe that there are two issues -- weak annotation and insufficient data that may obstruct automatic COVID-19 severity assessment with CT images. To address these challenges, we propose a novel three-component method, i.e., 1) a deep multiple instance learning component with instance-level attention to jointly classify the bag and also weigh the instances, 2) a bag-level data augmentation component to generate virtual bags by reorganizing high confidential instances, and 3) a self-supervised pretext component to aid the learning process. We have systematically evaluated our method on the CT images of 229 COVID-19 cases, including 50 severe and 179 non-severe cases. Our method could obtain an average accuracy of 95.8%, with 93.6% sensitivity and 96.4% specificity, which outperformed previous works.

Pulmonary Dysfunction in Patients Recovered from COVID-19 Pneumonia: A 6-Month Follow-up Study (preprint)

Objectives: This study investigates the clinical features and pulmonary functions of COVID-19 pneumonia survivors at 3 or 6 months after diagnosis in the Heilongjiang Province, China.Methods: Forty-six patients with COVID-19 pneumonia diagnosed since February 2020 were enrolled in this study for follow-up in July 2020. These patients were categorized into three groups: Group A (n=24) and Group B (n=11) who were diagnosed with moderate or severe pneumonia and followed up at three months after diagnosis; Group C (n=11) who were diagnosed with severe pneumonia and followed up at six months after diagnosis. Data on pulmonary function, arterial blood gas analysis, chest CT, blood test, antibody test, and health-related quality of life during hospitalization and at the follow-up visits were collected and analyzed. Results: Abnormal PO2 (A-a) was more prevalent in severe cases (Group B and C) than in moderate cases (Group A). Pulmonary dysfunction was common in this cohort. Abnormal CT scores of severe cases (Group B and C) were significantly higher than that of moderate cases (Group A). During the follow-up, lung abnormalities gradually resolved in the first 3 months (Group A and B), however, further resolution was not significant from 3 months to 6 months (Group B and C). Conclusion: Although pulmonary interstitial changes due to COVID-19 pneumonia gradually reverse over time, pulmonary dysfunction is common and appears to persist at least up to 6 months in patients recovered from COVID-19 pneumonia.

An in-depth investigation of the safety and immunogenicity of an inactivated SARS-CoV-2 vaccine (preprint)

BACKGROUND In-depth investigations of the safety and immunogenicity of inactivated SARS-CoV-2 vaccines are needed. METHOD In a phase I randomized, double-blinded, and placebo-controlled trial involving 192 healthy adults 18-59 years of age, two injections of three different doses (50 EU, 100 EU and 150 EU) of an inactivated SARS-CoV-2 vaccine or the placebo were administered intramuscularly with a 2- or 4-week interval between the injections. The safety and immunogenicity of the vaccine were evaluated within 28 days. FINDING In this study, 191 subjects assigned to three doses groups or the placebo group completed the 28-day trial. There were 44 adverse reactions within the 28 days, most commonly mild pain and redness at the injection site or slight fatigue, and no abnormal variations were observed in 48 cytokines in the serum samples of immunized subjects. The serum samples diluted from 1:32 to 1:4096 and incubated with the virus did not show antibody-dependent enhancement effects (ADEs) with regard to human natural killer cells, macrophages or dendritic cells. At day 14, the seroconversion rates had reached 92%, 100% and 96% with geometric mean titers (GMTs) of 18.0, 54.5 and 37.1, and at day 28, the seroconversion rates had reached 80%, 96% and 92% with GMTs of 10.6, 15.4 and 19.6in 0, 14 and 0, 28 procedures, respectively. Seroconversion was associated with the synchronous upregulation of ELISA antibodies against the S protein, N protein and virion and a cytotoxic T lymphocyte (CTL) response. Transcriptome analysis shaped the genetic diversity of immune response induced by the vaccine. INTERPRETATION In a population aged 18-59 years, this inactivated SARS-CoV-2 vaccine was safe and immunogenic.

Systematic discovery and functional interrogation of SARS-CoV-2 viral RNA-host protein interactions during infection (preprint)

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the cause of a pandemic with growing global mortality. There is an urgent need to understand the molecular pathways required for host infection and anti-viral immunity. Using comprehensive identification of RNA-binding proteins by mass spectrometry (ChIRP-MS), we identified 309 host proteins that bind the SARS-CoV-2 RNA during active infection. Integration of this data with viral ChIRP-MS data from three other positive-sense RNA viruses defined pan-viral and SARS-CoV-2-specific host interactions. Functional interrogation of these factors with a genome-wide CRISPR screen revealed that the vast majority of viral RNA-binding proteins protect the host from virus-induced cell death, and we identified known and novel anti-viral proteins that regulate SARS-CoV-2 pathogenicity. Finally, our RNA-centric approach demonstrated a physical connection between SARS-CoV-2 RNA and host mitochondria, which we validated with functional and electron microscopy data, providing new insights into a more general virus-specific protein logic for mitochondrial interactions. Altogether, these data provide a comprehensive catalogue of SARS-CoV-2 RNA-host protein interactions, which may inform future studies to understand the mechanisms of viral pathogenesis, as well as nominate host pathways that could be targeted for therapeutic benefit.

Autoproteolytic Products of the SARS-CoV-2 Nucleocapsid Protein are Primed for Antibody Evasion and Virus Proliferation (preprint)

The SARS-CoV-2 nucleocapsid (N) protein is the most immunogenic of the structural proteins and plays essential roles in several stages of the virus lifecycle. It is comprised of two major structural domains: the RNA binding domain, which interacts with viral and host RNA, and the oligomerization domain which assembles to form the viral core. Here, we investigate the assembly state and RNA binding properties of the full-length nucleocapsid protein using native mass spectrometry. We find that dimers, and not monomers, of full-length N protein bind RNA, implying that dimers are the functional unit of ribonucleoprotein assembly. In addition, we find that N protein binds RNA with a preference for GGG motifs which are known to form short stem loop structures. Unexpectedly, we found that N undergoes autoproteolytic processing within the linker region, separating the two major domains. This process results in the formation of at least five proteoforms that we sequenced using electron transfer dissociation, higher-energy collision induced dissociation and corroborated by peptide mapping. The cleavage sites identified are in highly conserved regions leading us to consider the potential roles of the resulting proteoforms. We found that monomers of N-terminal proteoforms bind RNA with the same preference for GGG motifs and that the oligomeric state of a C-terminal proteoform (N156-419) is sensitive to pH. We used mass spectrometry to show that N binds to a monoclonal antibody raised against full-length N. No antibody interactions were detected for N proteoforms without C-terminal residues, therefore locating antigenic regions towards the C-terminus. We then tested interactions of the proteoforms with the immunophilin cyclophilin A, a key component in coronavirus replication. We found that N1-209 and N1-273 bind directly to cyclophilin A, an interaction that is abolished by the approved immunosuppressant drug cyclosporin A. We propose that the proteoforms generated via autoproteolysis evade antibody detection through removal of the antigenic C-terminus and facilitate interactions with structured RNA or cyclophilin thereby enabling the virus to proliferate.

The strength of a NES motif in the nucleocapsid protein of human coronaviruses is related to genus, but not to pathogenic capacity (preprint)

Seven members of the Coronaviridae family infect humans, but only three (SARS-CoV, SARS-CoV-2 and MERS-CoV) cause severe disease with a high case fatality rate. Using in silico analyses (machine learning techniques and comparative genomics), several features associated to coronavirus pathogenicity have been recently proposed, including a potential increase in the strength of a predicted novel nuclear export signal (NES) motif in the nucleocapsid protein. Here, we have used a well-established nuclear export assay to experimentally establish whether the recently proposed nucleocapsid NESs are capable of mediating nuclear export, and to evaluate if their activity correlates with coronavirus pathogenicity. The six NES motifs tested were functional in our assay, but displayed wide differences in export activity. Importantly, these differences in NES strength were not related to strain pathogenicity. Rather, we found that the NESs of the strains belonging to the genus Alphacoronavirus were markedly stronger than the NESs of the strains belonging to the genus Betacoronavirus. We conclude that, while some of the genomic features recently identified in silico could be crucial contributors to coronavirus pathogenicity, this does not appear to be the case of nucleocapsid NES activity, as it is more closely related to coronavirus genus than to pathogenic capacity.

Structural basis for repurposing a 100-years-old drug suramin for treating COVID-19 (preprint)

The COVID-19 pandemic by non-stop infections of SARS-CoV-2 has continued to ravage many countries worldwide. Here we report the discovery of suramin, a 100-year-old drug, as a potent inhibitor of the SARS-CoV-2 RNA dependent RNA polymerase (RdRp) through blocking the binding of RNA to the enzyme. In biochemical assays, suramin and its derivatives are at least 20-fold more potent than remdesivir, the currently approved nucleotide drug for COVID-19. The 2.6 [A] cryo-EM structure of the viral RdRp bound to suramin reveals two binding sites of suramin, with one site directly blocking the binding of the RNA template strand and the other site clash with the RNA primer strand near the RdRp catalytic active site, therefore inhibiting the viral RNA replication. Furthermore, suramin potently inhibits SARS-CoV-2 duplication in Vero E6 cells. These results provide a structural mechanism for the first non-nucleotide inhibitor of the SARS-CoV-2 RdRp and a rationale for repurposing suramin for treating COVID-19.

SARS-CoV-2 infection damages airway motile cilia and impairs mucociliary clearance (preprint)

Understanding how SARS-CoV-2 spreads within the respiratory tract is important to define the parameters controlling the severity of COVID-19. We examined the functional and structural consequences of SARS-CoV-2 infection in a reconstituted human bronchial epithelium model. SARS-CoV-2 replication caused a transient decrease in epithelial barrier function and disruption of tight junctions, though viral particle crossing remained limited. Rather, SARS-CoV-2 replication led to a rapid loss of the ciliary layer, characterized at the ultrastructural level by axoneme loss and misorientation of remaining basal bodies. The motile cilia function was compromised, as measured in a mucociliary clearance assay. Epithelial defense mechanisms, including basal cell mobilization and interferon-lambda induction, ramped up only after the initiation of cilia damage. Analysis of SARS-CoV-2 infection in Syrian hamsters further demonstrated the loss of motile cilia in vivo. This study identifies cilia damage as a pathogenic mechanism that could facilitate SARS-CoV-2 spread to the deeper lung parenchyma.

Sub-second heat inactivation of coronavirus (preprint)

Heat treatment denatures viral proteins that comprise the virion, making virus incapable of infecting a host. Coronavirus (CoV) virions contain single-stranded RNA genomes with a lipid envelope and 4 proteins, 3 of which are associated with the lipid envelope and thus are thought to be easily denatured by heat or surfactant-type chemicals. Prior studies have shown that a temperature of as low as 75 oC and treatment duration of 15 min can effectively inactivate CoV. The applicability of a CoV heat inactivation method greatly depends on the length of time of a heat treatment and the temperature needed to inactivate the virus. With the goal of finding conditions where sub-second heat exposure of CoV can sufficiently inactivate CoV, we designed and developed a simple system that can measure sub-second heat inactivation of CoV. The system is composed of capillary stainless-steel tubing immersed in a temperature-controlled oil bath followed by an ice bath, through which virus solution can be flowed at various speeds. Flowing virus solution at different speeds, along with a real-time temperature monitoring system, allows the virus to be accurately exposed to a desired temperature for various durations of time. Using mouse hepatitis virus (MHV), a beta-coronavirus, as a model system, we identified that 85.2 oC for 0.48 s exposure is sufficient to obtain > 5 Log10 reduction in viral titer (starting titer: 5 x 107 PFU/mL), and that when exposed to 83.4 oC for 0.95 s, the virus was completely inactivated (zero titer, > 6 Log10 reduction).

A genetic variant protective for COVID-19 is inherited from Neanderthals (preprint)

It was recently shown that the major genetic risk factor associated with becoming severely ill with COVID-19 when infected by SARS-CoV-2 is inherited from Neandertals. Thanks to new genetic association studies additional risk factors are now being discovered. Using data from a recent genome-wide associations from the Genetics of Mortality in Critical Care (GenOMICC) consortium, we show that a haplotype at a region associated with requiring intensive care is inherited from Neandertals. It encodes proteins that activate enzymes that are important during infections with RNA viruses. As compared to the previously described Neandertal risk haplotype, this Neandertal haplotype is protective against severe COVID-19, is of more moderate effect, and is found at substantial frequencies in all regions of the world outside Africa.

GNS561 exhibits potent in vitro antiviral activity against SARS-CoV-2 through autophagy inhibition (preprint)

Since December 2019, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2/2019-nCoV) has spread quickly worldwide, with more than 29 million cases and 920,000 deaths. Interestingly, coronaviruses were found to subvert and hijack the autophagic process to allow their viral replication. One of the spotlights had been focused on the autophagy inhibitors as a target mechanism effective in the inhibition of SARS-CoV-2 infection. Consequently, chloroquine (CQ) and hydroxychloroquine (HCQ), a derivative of CQ, was suggested as the first potentially be therapeutic strategies as they are known to be autophagy inhibitors. Then, they were used as therapeutics in SARS-CoV-2 infection along with remdesivir, for which the FDA approved emergency use authorization. Here, we investigated the antiviral activity and associated mechanism of GNS561, a small basic lipophilic molecule inhibitor of late-stage autophagy, against SARS-CoV-2. Our data indicated that GNS561 showed the highest antiviral effect for two SARS-CoV-2 strains compared to CQ and remdesivir. Focusing on the autophagy mechanism, we showed that GNS561, located in LAMP2-positive lysosomes, together with SARS-CoV-2, blocked autophagy by increasing the size of LC3-II spots and the accumulation of autophagic vacuoles in the cytoplasm with the presence of multilamellar bodies characteristic of a complexed autophagy. Finally, our study revealed that the combination of GNS561 and remdesivir was associated with a strong synergistic antiviral effect against SARS-CoV-2. Overall, our study highlights GNS561 as a powerful drug in SARS-CoV-2 infection and supports that the hypothesis that autophagy inhibitors could be an alternative strategy for SARS-CoV-2 infection.

Multi-Clonal Live SARS-CoV-2 In Vitro Neutralization by Antibodies Isolated from Severe COVID-19 Convalescent Donors (preprint)

The interactions between antibodies, SARS-CoV-2 and immune cells contribute to the pathogenesis of COVID-19 and protective immunity. To understand the differences between antibody responses in mild versus severe cases of COVID-19, we analyzed the B cell responses in patients 1.5 months post SARS-CoV-2 infection. Severe and not mild infection correlated with high titers of IgG against Spike receptor binding domain (RBD) that were capable of viral inhibition. B cell receptor (BCR) sequencing revealed two VH genes, VH3-38 and VH3-53, that were enriched during severe infection. Of the 22 antibodies cloned from two severe donors, six exhibited potent neutralization against live SARS-CoV-2, and inhibited syncytia formation. Using peptide libraries, competition ELISA and RBD mutagenesis, we mapped the epitopes of the neutralizing antibodies (nAbs) to three different sites on the Spike. Finally, we used combinations of nAbs targeting different immune-sites to efficiently block SARS-CoV-2 infection. Analysis of 49 healthy BCR repertoires revealed that the nAbs germline VHJH precursors comprise up to 2.7% of all VHJHs. We demonstrate that severe COVID-19 is associated with unique BCR signatures and multi-clonal neutralizing responses that are relatively frequent in the population. Moreover, our data support the use of combination antibody therapy to prevent and treat COVID-19.

A Composite Index Predicts Disease Progression in Early Stages of COVID-19: a Propensity Score-matched Cohort Study (preprint)

Background: Thus far, studies on COVID-19 have focused on the epidemiology of the disease and clinical characteristics of patients (14-19), as well as on the risk factors associated with mortality during hospitalization in critical COVID-19 cases. However, no research has been performed on the prediction of progression in patients in the early stages of the disease. The aim of this work was to identify the early predictors of COVID-19 progression.Methods: The study included 338 patients with COVID-19 treated at two hospitals in Wuhan, Chian, from December, 2019 to March, 2020. Predictors of the progression of COVID-19 from mild to severe stages were selected by the logistic regression analysis. The predictive accuracy was evaluated further in the propensity score-matched cohort.Results: COVID-19 progression to severe and critical stages was confirmed in 78(23.1%) patients. The average value of the neutrophil-to-lymphocyte ratio (NLR) was higher in patients in the disease progression group than in the improvement group. Multivariable logistic regression analysis revealed that elevated NLR, LDH, and IL-10, were independent predictors of disease progression. The optimal cut-off value of NLR for predicting the progression of COVID-19 was 3.75. In the propensity score-matched cohort, NLR ≥ 3.75 was still an independent predictor of COVID-19 progression after multivariate analysis.Conclusions: The performed analysis demonstrates that NLR qualifies as an independent predictor of disease progression in COVID-19 patients at the early stage of the disease. The combined evaluation of NLR and LDH improved the accuracy of the prediction of COVID-19 progression. Assessment of predictors might facilitate early identification of COVID-19 patients at high risk for disease progression and ensure timely administration of appropriate treatment to prevent mild cases from becoming severe.

Synergistic Learning of Lung Lobe Segmentation and Hierarchical Multi-Instance Classification for Automated Severity Assessment of COVID-19 in CT Images (preprint)

Understanding chest CT imaging of the coronavirus disease 2019 (COVID-19) will help detect infections early and assess the disease progression. Especially, automated severity assessment of COVID-19 in CT images plays an essential role in identifying cases that are in great need of intensive clinical care. However, it is often challenging to accurately assess the severity of this disease in CT images, due to variable infection regions in the lungs, similar imaging biomarkers, and large inter-case variations. To this end, we propose a synergistic learning framework for automated severity assessment of COVID-19 in 3D CT images, by jointly performing lung lobe segmentation and multi-instance classification. Considering that only a few infection regions in a CT image are related to the severity assessment, we first represent each input image by a bag that contains a set of 2D image patches (with each cropped from a specific slice). A multi-task multi-instance deep network (called M$^2$UNet) is then developed to assess the severity of COVID-19 patients and also segment the lung lobe simultaneously. Our M$^2$UNet consists of a patch-level encoder, a segmentation sub-network for lung lobe segmentation, and a classification sub-network for severity assessment (with a unique hierarchical multi-instance learning strategy). Here, the context information provided by segmentation can be implicitly employed to improve the performance of severity assessment. Extensive experiments were performed on a real COVID-19 CT image dataset consisting of 666 chest CT images, with results suggesting the effectiveness of our proposed method compared to several state-of-the-art methods.

Potential influence of COVID-19/ACE2 on the female reproductive system.

The 2019 novel coronavirus (2019-nCoV) appeared in December 2019 and then spread throughout the world rapidly. The virus invades the target cell by binding to angiotensin-converting enzyme (ACE) 2 and modulates the expression of ACE2 in host cells. ACE2, a pivotal component of the renin-angiotensin system, exerts its physiological functions by modulating the levels of angiotensin II (Ang II) and Ang-(1-7). We reviewed the literature that reported the distribution and function of ACE2 in the female reproductive system, hoping to clarify the potential harm of 2019-nCoV to female fertility. The available evidence suggests that ACE2 is widely expressed in the ovary, uterus, vagina and placenta. Therefore, we believe that apart from droplets and contact transmission, the possibility of mother-to-child and sexual transmission also exists. Ang II, ACE2 and Ang-(1-7) regulate follicle development and ovulation, modulate luteal angiogenesis and degeneration, and also influence the regular changes in endometrial tissue and embryo development. Taking these functions into account, 2019-nCoV may disturb the female reproductive functions through regulating ACE2.