Machine Learning-Based Scoring System for Early Prognosis Evaluation of Patients with Coronavirus Disease 2019

Background The global spread of coronavirus disease 2019 (COVID-19) continues to threaten human health security, exerting considerable pressure on healthcare systems worldwide. While prognostic models for COVID-19 hospitalized or intensive care patients are currently available, prognostic models developed for large cohorts of thousands of individuals are still lacking. Methods Between February 4 and April 16, 2020, we enrolled 3,974 patients admitted with COVID-19 disease in the Wuhan Huo-Shen-Shan Hospital and the Maternal and Child Hospital, Hubei Province, China. (1) Screening of key prognostic factors: A univariate Cox regression analysis was performed on 2,649 patients in the training set, and factors affecting prognosis were initially screened. Subsequently, a random survival forest model was established through machine analysis to further screen for factors that are important for prognosis. Finally, multivariate Cox regression analysis was used to determine the synergy among various factors related to prognosis. (2) Establishment of a scoring system: The nomogram algorithm established a COVID-19 patient death risk assessment scoring system for the nine selected key prognostic factors, calculated the C index, drew calibration curves and drew training set patient survival curves. (3) Verification of the scoring system: The scoring system assessed 1,325 patients in the test set, splitting them into high- and low-risk groups, calculated the C-index, and drew calibration and survival curves. Results The cross-sectional study found that age, clinical classification, sex, pulmonary insufficiency, hypoproteinemia, and four other factors (underlying diseases: blood diseases, malignant tumor;complications: digestive tract bleeding, heart dysfunction) have important significance for the prognosis of the enrolled patients with COVID-19. Herein, we report the discovery of the effects of hypoproteinemia and hematological diseases on the prognosis of COVID-19. Meanwhile, the scoring system established here can effectively evaluate objective scores for the early prognoses of patients with COVID-19 and can divide them into high- and low-risk groups (using a scoring threshold of 117.77, a score below which is considered low risk). The efficacy of the system was better than that of clinical classification using the current COVID-19 guidelines (C indexes, 0.95 vs. 0.89). Conclusions Age, clinical typing, sex, pulmonary insufficiency, hypoproteinemia, and four other factors were important for COVID-19 survival. Compared with general statistical methods, this method can quickly and accurately screen out the relevant factors affecting prognosis, provide an order of importance, and establish a scoring system based on the nomogram model, which is of great clinical significance.

The optimal vaccination strategy to control COVID-19: a modeling study in Wuhan City, China.

BACKGROUND: Reaching optimal vaccination rates is an essential public health strategy to control the coronavirus disease 2019 (COVID-19) pandemic. This study aimed to simulate the optimal vaccination strategy to control the disease by developing an age-specific model based on the current transmission patterns of COVID-19 in Wuhan City, China. METHODS: We collected two indicators of COVID-19, including illness onset data and age of confirmed case in Wuhan City, from December 2, 2019, to March 16, 2020. The reported cases were divided into four age groups: group 1, ≤ 14 years old; group 2, 15 to 44 years old; group 3, 44 to 64 years old; and group 4, ≥ 65 years old. An age-specific susceptible-exposed-symptomatic-asymptomatic-recovered/removed model was developed to estimate the transmissibility and simulate the optimal vaccination strategy. The effective reproduction number (Reff) was used to estimate the transmission interaction in different age groups. RESULTS: A total of 47 722 new cases were reported in Wuhan City from December 2, 2019, to March 16, 2020. Before the travel ban of Wuhan City, the highest transmissibility was observed among age group 2 (Reff = 4.28), followed by group 2 to 3 (Reff = 2.61), and group 2 to 4 (Reff = 1.69). China should vaccinate at least 85% of the total population to interrupt transmission. The priority for controlling transmission should be to vaccinate 5% to 8% of individuals in age group 2 per day (ultimately vaccinated 90% of age group 2), followed by 10% of age group 3 per day (ultimately vaccinated 90% age group 3). However, the optimal vaccination strategy for reducing the disease severity identified individuals ≥ 65 years old as a priority group, followed by those 45-64 years old. CONCLUSIONS: Approximately 85% of the total population (nearly 1.2 billion people) should be vaccinated to build an immune barrier in China to safely consider removing border restrictions. Based on these results, we concluded that 90% of adults aged 15-64 years should first be vaccinated to prevent transmission in China.

Novel neutralization mechanism of a human antibody with pan-coronavirus reactivity (preprint)

The recurrent outbreak of coronaviruses and variants underscores the need for broadly reactive antivirals and vaccines. Here, a novel broad-spectrum human antibody named 76E1 was isolated from a COVID-19 convalescent patient and showed broad neutralization activity against multiple α- and β-coronaviruses, including the SARS-CoV-2 variants and also exhibited the binding breath to peptides containing the epitope from γ- and δ- coronaviruses. 76E1 cross-protects mice from SARS-CoV-2 and HCoV-OC43 infection in both prophylactic and treatment models. The epitope including the fusion peptide and S2’ cleavage site recognized by 76E1 was significantly conserved among α-, β-, γ- and δ- coronaviruses. We uncovered a novel mechanism of antibody neutralization that the epitope of 76E1 was proportionally less exposed in the prefusion trimeric structure of spike protein but could be unmasked by binding to the receptor ACE2. Once the epitope exposed, 76E1 inhibited S2’ cleavage, thus blocked the membrane fusion process. Our data demonstrate a key epitope targeted by broadly-neutralizing antibodies and will guide next-generation epitope-based pan-coronavirus vaccine design.

Comprehensive Mapping of Binding Hot Spots of SARS-CoV-2 RBD-specific Neutralizing Antibodies for Tracking Immune Escape Variants (preprint)

The receptor-binding domain (RBD) variants of SARS-CoV-2 could impair antibody-mediated neutralization of the virus by host immunity; thus, prospective surveillance for such antibody escape mutants is urgently needed. Here, we comprehensively profiled four antigenic sites of the RBD and mapped the binding hot spots for a panel of RBD-specific monoclonal antibodies isolated from COVID-19 convalescents, especially dominant VH3-53/3–66 antibodies, which are valuable indicators of antigenic changes in the RBD. We further demonstrated that several natural mutations, namely, K417N, F486L, N450K, L452R, E484K, F490S and R346S, significantly decreased the neutralizing activity of multiple human monoclonal antibodies and of human convalescent plasma obtained in the early stage of the COVID-19 pandemic. Of note, among the natural escape mutations, L452R enhanced ACE2 binding affinity, indicating that it potentially increased virulence. Overall, the in-depth maps may have far-reaching value for surveillance of SARS-CoV-2 immune escape variants and guidance of vaccine design.

Effectiveness of potential antiviral treatments in COVID-19 transmission control: a modelling study.

BACKGROUND: Novel coronavirus disease 2019 (COVID-19) causes an immense disease burden. Although public health countermeasures effectively controlled the epidemic in China, non-pharmaceutical interventions can neither be maintained indefinitely nor conveniently implemented globally. Vaccination is mainly used to prevent COVID-19, and most current antiviral treatment evaluations focus on clinical efficacy. Therefore, we conducted population-based simulations to assess antiviral treatment effectiveness among different age groups based on its clinical efficacy. METHODS: We collected COVID-19 data of Wuhan City from published literature and established a database (from 2 December 2019 to 16 March 2020). We developed an age-specific model to evaluate the effectiveness of antiviral treatment in patients with COVID-19. Efficacy was divided into three types: (1) viral activity reduction, reflected as transmission rate decrease [reduction was set as v (0-0.8) to simulate hypothetical antiviral treatments]; (2) reduction in the duration time from symptom onset to patient recovery/removal, reflected as a 1/γ decrease (reduction was set as 1-3 days to simulate hypothetical or real-life antiviral treatments, and the time of asymptomatic was reduced by the same proportion); (3) fatality rate reduction in severely ill patients (fc) [reduction (z) was set as 0.3 to simulate real-life antiviral treatments]. The population was divided into four age groups (groups 1, 2, 3 and 4), which included those aged ≤ 14; 15-44; 45-64; and ≥ 65 years, respectively. Evaluation indices were based on outbreak duration, cumulative number of cases, total attack rate (TAR), peak date, number of peak cases, and case fatality rate (f). RESULTS: Comparing the simulation results of combination and single medication therapy s, all four age groups showed better results with combination medication. When 1/γ = 2 and v = 0.4, age group 2 had the highest TAR reduction rate (98.48%, 56.01-0.85%). When 1/γ = 2, z = 0.3, and v = 0.1, age group 1 had the highest reduction rate of f (83.08%, 0.71-0.12%). CONCLUSIONS: Antiviral treatments are more effective in COVID-19 transmission control than in mortality reduction. Overall, antiviral treatments were more effective in younger age groups, while older age groups showed higher COVID-19 prevalence and mortality. Therefore, physicians should pay more attention to prevention of viral spread and patients deaths when providing antiviral treatments to patients of older age groups.

Effectiveness and Safety of Remdesivir in Patients with COVID-19: A Systematic Review and Meta-Analysis (preprint)

Background: Remdesivir, a nucleoside analogue antiviral drug developed for Ebola, is approved by the US Food and Drug Administration for the treatment of COVID-19. However, the findings of randomised controlled trials (RCTs) and observational studies vary regarding the effectiveness of remdesivir. We aimed to comprehensively review the available evidence identify the effectiveness and safety of remdesivir in patients with COVID-19.Methods: Seven databases (PubMed, Web of Science, Embase, Wanfang database, SinoMed, Chinese National Knowledge Infrastructure and Chinese Science Journal Database) were searched for literatures published until November 2020.Following the PRISMA flow diagram, we included RCTs and prospective observational studies that reported the effectiveness and safety of remdesivir in patients with COVID-19. With extracting study details, as well as patient characteristics and outcomes, data were meta-analyzed by using Review Manager software version 5.4.1. Meta-analyses were conducted with fixed-effect model or random-effect model to calculate risk ratio (RR).Results: Four studies involving 2,279 patients were included in this meta-analysis. Compared with placebo, 10-day remdesivir was associated with significant increased clinical improvement on days 14 and 28 with RR 1.19 (95%CI 1.09-1.30) and RR 1.09 (95%CI 1.03-1.16). The clinical improvement of 5-day remdesivir was better than 10-day remdesivir on days 7 with RR 1.20 (95%CI 1.02-1.41), but the efficacy advantage of 5-day remdesivir disappeared on days 14 (RR 1.08; 95%CI 0.90-1.29). Remdesivir was associated with lower serious adverse events rates and grade 3 or 4 adverse events rates as compared with placebo with RR 0.75(95%CI 0.63-0.89) and RR 0.89(95%CI 0.80-0.99). Compared with 10-day remdesivir, 5-day remdesivir for patients with COVID-19 decreased the risk of serious adverse events rates and grade 3 or 4 adverse events rates with RR 0.65(95%CI 0.47-0.88) and RR 0.74 (95%CI 0.58-0.95). Conclusions: Our meta-analysis suggested that remdesivir would increase clinical improvement conditions and decrease serious adverse events on patients with COVID-19. 5-day remdesivir had the similar clinical effectiveness and mortality with 10-day remdesivir, and had lower serious adverse events rate. Comprehensive considering the cost and benefit, 5-day remdesivir may be a better therapeutic option if available medical resources are limited.