key: cord-0761588-kcdixwrt
authors: Liao, Sih-Han; Hung, Chien-Ching; Chen, Chiung-Nien; Yen, Jui-Yi; Hsu, Chen-Yang; Yen, Amy Ming-Fang; Chen, Chi-Ling
title: Assessing Efficacy of Antiviral Therapy for COVID-19 Patients: A Case Study on Remdesivir with Bayesian Synthesis Design and Multistate Analysis
date: 2021-05-04
journal: J Formos Med Assoc
DOI: 10.1016/j.jfma.2021.04.026
sha: d4f2219e5b2104110657c57516545728a5f0f8b1
doc_id: 761588
cord_uid: kcdixwrt

Background/Purpose A synthesis design and multistate analysis is required for assessing the clinical efficacy of antiviral therapy on dynamics of multistate disease progression and in reducing the mortality and enhancing the recovery of patients with COVID-19. A case study on remdesivir was illustrated for the clinical application of such a novel design and analysis. Methods A Bayesian synthesis design was applied to integrating the empirical evidence on the one-arm compassion study and the two-arm ACTT-1 trial for COVID-19 patients treated with remdesivir. A multistate model was developed to model the dynamics of hospitalized COVID-19 patients from three transient states of low, medium-, and high-risk until the two outcomes of recovery and death. The outcome measures for clinical efficacy comprised high-risk state, death, and discharge. Results The efficacy of remdesivir in reducing the risk of death and enhancing the odds of recovery were estimated as 31% (95% CI, 18-44%) and 10% (95% CI, 1-18%), respectively. Remdesivir therapy for patients with low-risk state showed the efficacy in reducing subsequent progression to high-risk state and death by 26% (relative rate (RR), 0.74; 95% CI, 0.55-0.93) and 62% (RR, 0.38; 95% CI, 0.29-0.48), respectively. Less but still statistically significant efficacy in mortality reduction was noted for the medium- and high-risk patients. Remdesivir treated patients had a significantly shorter period of hospitalization (9.9 days) compared with standard care group (12.9 days). Conclusions The clinical efficacy of remdesvir therapy in reducing mortality and accelerating discharge has been proved by the Bayesian synthesis design and multistate analysis.

The clustered cases of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were first reported in Wuhan in December, 2019, 1,2 which has resulted in COVID-19 pandemic from March, 2020 until April, 2021, which has led to more than 140 million cases and claimed more than 3 million deaths. 3 This soaring number of COVID-19 cases, in spite of great efforts made to put on non-pharmaceutical interventions (NPIs) in most of countries and regions, has stressed medical care systems and compromised the critical care capacity around the world. [4] [5] [6] [7] Although several vaccines have been developed and showed promising effects to prevent SARS-CoV-2 infection, [8] [9] [10] [11] [12] antiviral therapy has been shown to play the key role for the hospitalized patients to reduce the risk of disease progression to its severe form such as acute respiratory distress syndrome (ARDS) and to shorten the length of hospital stay. [13] [14] [15] [16] [17] To prove evidence-based efficacy, several randomized controlled trials (RCTs) have been conducted to investigate the effects of selected compounds such as chloroquine, hydroxychloroquine, lopinavir/ritonavir, ivermectin, interferon, steroids and remdesivir since the identification of SARS-CoV-2 in 2020. [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] While there is a strong recommendation supporting the systematic use of steroids in patients with severe and critical COVID-19, there are uncertainties regarding the suggestions for the clinical use of other therapeutic compounds. Among the candidate compounds for COVID -19, hydroxychloroquine, chloroquine, and lopinavir/ritonavir have been excluded from the recommendation lists of the treatment guidelines of the World Health Organization J o u r n a l P r e -p r o o f (WHO) and National Institutes of Health (NIH) due to the lack of clinical efficacy in reducing mortality and severe disease requiring invasive ventilation and in accelerating recovery and discharge from COVID-19. [27] [28] [29] [30] Currently, ivermectin and antibody therapies are also not recommended due to insufficient evidence supporting the beneficial effect of their clinical use. 16, 17, 31, 32 Remdesivir is another promising compound for treating COVID-19 patients and have been approved by the US FDA for such a purpose. 33 Remdesivir is a monophosphoramidate adenosine analogue prodrug which could be metabolized to active tri-phosphate form to inhibit the synthesis of viral RNA. 16, [34] [35] [36] Several clinical trials and observational studies have been conducted to evaluate the clinical efficacy in treating COVID-19 patients with various disease severities. [18] [19] [20] [21] [22] However, discrepancies in the efficacy exist for different studies. The clinical use of this antiviral therapy thus remains controversial. More importantly, current evidence on its clinical efficacy in reducing mortality for COVID-19 patients still remains inconclusive even after several one-arm studies and two-arm RCTs. [18] [19] [20] [21] [22] Based on the inconsistent evidence, the current WHO guideline suggests against the administration of remdesivir in addition to usual care for treating hospitalized COVID-19 patients. 16, 32 However, remdesivir has been recommended by the NIH guideline for the hospitalized COVID-19 patients requiring supplementary oxygen without the necessity for oxygen delivery through a high-flow device, non-invasive or invasive ventilation, or extracorporeal membrane oxygenation (ECMO). 17 To prove an evidence-based clinical efficacy, the best way is to conduct a RCT.

However, the conduction of an RCT for assessing a new or a variant of the existing J o u r n a l P r e -p r o o f antiviral therapy can be very challenging in the era of COVID-19 pandemic. Moreover, even a two-arm RCT may be underpowered and may also have ethical and feasibility concerns. 33, [37] [38] [39] [40] While waiting for a well-powered RCT trial for validating the efficacy of any new antiviral therapy, it is worthwhile to elucidate how antiviral therapy alters the mechanism of COVID-19 disease progression from mild to severe stage and until death based on available information from one-arm study without the control group and also from the two-arm RCT.

The aim of this study was to estimate the clinical efficacy of remdesivir by using a novel synthesis sequential design and analysis to integrate the empirical information provided by the one-arm compassion study and two-arm RCT with the consideration of the dynamic of COVID-19.

J o u r n a l P r e -p r o o f

The antiviral therapy of remdesivir has been proposed as a candidate and been provided to hospitalized COVID-19 patients in the early stage of first pandemic period.

Although the early results suggested the possible benefit of remdesivir therapy, the lack of comparator makes it difficult to quantify the clinical efficacy. The use of RCT design in the latter studies provide the ground of evidence-based evaluation for remdesivir.

However, the heterogeneity in the clinical evolution of COVID-19 patients render the results controversial.

Given these scenario, we thus used the Bayesian synthesis sequential design to integrate the information from two studies on the use of remedisivir taking into account the temporal sequence. Specifically, the period of enrollment for the one-arm compassionate between January 25 and March 7, 2020 was taken as the prior study before the RCT. Following the conduction of one-arm compassionate use for remdesivir, the two-arm ACTT-1 trial was performed with the enrollment of study participants between February 12 and April 19, 2020.

The information on the clinical evolution of COVID-19 patients with remdesivir was first obtained on the basis of the empirical data provided by the one-arm compassionate use study. 18 The aggregated data listed in the report of ACTT-1 19 was then used as the main study to update the information derived from the prior study sequentially. Namely, prior information on the daily progression of COVID-19 disease J o u r n a l P r e -p r o o f states were first estimated from the data of one-arm study, which was then updated by using the data derived from the two-arm ACTT-1 trial. 19

As indicated in the synthesis design, two empirical data sets, the one-arm compassionate remdesivir use study 18 For the one-arm compassionate remdesivir use study, data on the transitions for COVID-19 patients were abstracted, which provides the empirical information on the daily change of the COVID-19 risk states for 53 patients. This detailed patient-level information recorded on a daily basis through the 28-day period gives a clear profile of disease evolution across the risk states of COVID-19 from the date of the initiation of remdesivir therapy until discharge, death, or the end of study. 18 Regarding the two-arm ACTT-1 trial, the original article provides the baseline distribution of COVID-19 risk states by the two treatment groups of remdesivir and standard care and the aggregated information on the transition of COVID-19 states from J o u r n a l P r e -p r o o f each of the baseline risk state during the 14-day study period. 19 The aforementioned data with the information on transition from baseline risk states to that observed in the 14-day period were used as the basis for multistate analysis. 42 

To depict the evolution of COVID-19 through the three transient risk states (low-, medium-, and high-risk state) to the two events of discharge and death we applied a fivestate Markov model (Figure 1 ) that have been proposed to assess the efficacy of antiviral therapy. 42 In brief, the hospitalized COVID-19 patients can progress and regress between each of the low-, medium-, and high-risk states (low-risk ⇆ medium-risk ⇆ high-risk transitions, Figure 1 ). Patients at each of the risk state are possible to recover and discharge at a higher rate from low-risk state followed by that of medium-and high-risk state. For patients with unfavorable outcome, the COVID-19 disease state may progress to the high-risk state followed by the terminal outcome of death, which is also captured by a daily event rate. The kernel consisting of eight transition rates is thus required for the full specification of the five-state Markov model for COVID-19 evolution.

The proposed COVID-19 transition model not only models forward progression but also allows for backward regression from medium-to low risk and from high-to medium risk. The mechanisms in the benefit of interventions such as antiviral therapy can thus be captured by both the acceleration in the regression between risk states and the enhancement of the discharge from different risk states.

The continuous-time five-state Markov models were applied to estimate the daily transition rates of movements between risk states, discharge, and death regarding the J o u r n a l P r e -p r o o f disease evolution of COVID-19 patients with the remdesivir-treated group and the standard-care group. [41] [42] [43] Bayesian Markov Chain Monte Carlo (MCMC) simulation was used to estimate these daily transition rates in light of likelihood functions based on aggregated data on the remdesivir-treated and the standard-care groups abstracted from the original articles. 18, 19, 42 The dynamic curve depicting the evolution of COVID-19 across three risk states and two events of discharge and death in 28-day period was derived from the transition probability matrix for the five defined states given the estimated results on eight daily transitions rates for the five-state Markov model (Figure 1) . Based on the predicted 28day probabilities to the outcomes of discharge and death for the remdesivir-treated and the standard care groups, we were able to assess the efficacy of remdesivir therapy in accelerating discharge and in decreasing subsequent risk of death for hospitalized COVID-19 patients. We further evaluated the clinical efficacy of remdesivir in reducing the risk of high-risk state and the composite outcome of death and high-risk state. The effect of remdesivir on the prognosis of hospitalized COVID-19 patients was evaluated by using relative risk derived by comparing the probability distributions of patients receiving remdesivir with that of patients with standard care. Table 1 shows the total of 53 patients on repeated data featuring the change of risk states of COVID-19 after receiving remdesivir. 18 Table 1 also shows the transition of 53 patients across three risk states and the final destination of discharge during a onemonth study period. The data of Table 1 are used for deriving the rate of COVID-19 evolution in the light of the five-state disease transition models. Table 2 show the estimated results on the daily transition rates of progression and regression, discharge, and death for three risk states by treatment groups of remdesivir versus standard care based on the five-state COVID-19 progression model. Both groups show significant higher discharge rates for the low-risk state compared with those of medium-and high-risk patients. Figure 2 shows that the discharge was mainly from the low-risk state of COVID-19 patients with the orders of 0.1678 (95% CI, 0.1456-0.1917, Table 2 ) and 0.1396 (95% CI, 0.1178-0.1618, Table 2 ) for the remdesivir-treated group and the standard-care group, respectively. Considering the dynamics between the progression and regression for patients at medium risk by using the net force of regression (regression rate -progression rate, 

The efficacy of remdesivir therapy was further elucidated by the dynamics of COVID-19 across three risk states and the outcomes of discharge and death based on the estimated results on daily transition rates with the application of the two disease transition modes to detailed empirical data. Figure 3 (a) shows the daily progression on the dynamic of COVID-19 for the hospitalized patients with low-risk state at baseline who received remdesivir and those who received standard care. In line with the estimated results listed in Table 2 , the discharge rate for COVID-19 patients were accelerated by remdesivir therapy. The probability of discharge (green line) was uniformly higher compared with that of the control group. For the medium-and high-risk state, the benefit of regression to a lower risk states resulting from remdesivir therapy (Figure 3 (a) ) was demonstrated by the uniformly lower probability for medium-(orange line) and high-risk (gray line) compared with the control group (Figure 3 (b) ). The risk of death (red line) J o u r n a l P r e -p r o o f after one-month follow-up period was also lower for COVID-19 patients receiving remdesivir compared with those receiving standard care.

Similar trends regarding the probabilities of discharge, death, and high-risk state can be observed in Figure 3 (c) vs (d) and Figure 3 (e) vs (f) , showing the dynamic of COVID-19 by two treatments for patients at medium-and high-risk state at enrollment, respectively.

Based on the estimated results on the daily rates for clinical evolution of COVID-19 patients by using the Bayesian synthesis sequential design and analysis, we further assessed the efficacy of remdesivir therapy in reducing the risk of death and increasing the odds of recovery and discharge. Table 3 lists the estimated results on the clinical efficacy in terms of discharge, high-risk states, and death given the 28-day period of follow-up derived by comparing the 28-day probability for each of the defined outcome for two groups (S- Table 1 Table 3 ) followed by the medium-risk group (RR, 1.11; 95% CI, 1.07-1.16) and the low-risk group (RR, 1.11; 95% CI, 1.07-1.14). For low-risk patients at baseline, remdesivir therapy led to the reduction of subsequent progression to high-risk state by 26% (RR, 0.74; 95% CI, 0.55-0.93) and to final death by 62% (RR, 0.38; 95% CI, 0.29-0.48). For the medium-risk patients, less but still statistically significant efficacy results were noted in reducing progression to death by 39% (RR, 0.61; 95% CI, 0.56-0.67). Patients at high-risk state treated with remdesivir also led to a 35% reduction in death from COVID-19 (RR, 0.65; 95% CI, 0.62-0.69).

The median days to discharge for hospitalized COVID-19 patients at low-, medium-, and high-risk receiving remdeisvir treatment was estimated as 4. Reported from the one-arm study, remdesivir has been proposed for compassionated use in 53 patients and demonstrated 68% clinical improvement. 18 However, the one-arm study has been argued with a lacking of control group. It requires a two-arm RCT to demonstrate its evidence-based efficacy. [19] [20] [21] [22] However, in the era of COVID-19 pandemic, identifying a new potential antiviral therapy with a RCT design is fraught with the difficulty of logistics in implementation and ethical concerns. An underpowered RCT without consideration of the dynamic of COVID-19 further results in the controversial evidence. Alternative methods for evaluating evidence-based antiviral therapy with efficiency is therefore urgently needed.

The proposed synthesis sequential design analysis can be a solution to this dilemma. By making use of the information derived from two studies, we were able to derive precise estimates on the daily rate of COVID-19 evolution altered by remdesivir therapy. This approach not only takes into account the temporal sequence on the clinical use of remdesivir for COVID-19 patients but also provides a framework for the Such an efficacy of antiviral therapy in the prophylaxis and treatment has been demonstrated in the management of influenza. 49, 50 In conclusion, we propose a Bayesian synthesis sequential design with multi-state analysis to evaluate evidence-based antiviral therapy with efficiency. The illustrated results based on the proposed approach not only provide an even precise estimate of efficacy in reducing death and time-to-discharge of COVID-19 patients but also shed light on the underlying mechanism for the potential benefit of antiviral therapy, which can enlighten the clinical management of COVID-19 patients with precision and timeliness.

This work was supported by Ministry of Science and Technology, Taiwan (MOST 108-2118-M-038-001-MY3, MOST 109-2327-B-002 -009).

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Authors declare no competing interests J o u r n a l P r e -p r o o f 

High 403

Discharge 1

Death 6 a The total patient numbers of the low-, medium-, and high-risk states at baseline were 12, 7, and 34, respectively . b Low-risk state: no and low oxygen supplement; medium-risk state: non-invasive ventilator and high-flow oxygen supplement; high-risk state: ECMO and invasive ventilator. 

Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

An interactive web-based dashboard to track COVID-19 in real time

Transforming ORs into ICUs

Intensive care management of coronavirus disease 2019 (COVID-19): challenges and recommendations

Critical Supply Shortages -The Need for Ventilators and Personal Protective Equipment during the Covid-19 Pandemic

The Toughest Triage -Allocating Ventilators in a Pandemic

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine

Interim Results of a Phase 1-2a Trial of Ad26.COV2.S Covid-19 Vaccine

Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK

BNT162b2 mRNA Covid-19 Vaccine in a Nationwide Mass Vaccination Setting

Efficacy and safety of antiviral treatment for COVID-19 from evidence in studies of SARS-CoV-2 and other acute viral infections: a systematic review and meta-analysis

Current Therapeutic Approaches for COVID-19: A Systematic Review and a Metaanalysis

Mild or Moderate Covid-19

World Health Organization

COVID-19 Treatment guidelines

Compassionate Use of Remdesivir for Patients with Severe Covid-19

Remdesivir for the Treatment of Covid-19-Preliminary Report

Remdesivir for the Treatment of Covid-19 -Final Report

Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19

Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial

Hydroxychloroquine in the Treatment of COVID-19: A Multicenter Randomized Controlled Study

A Trial of Lopinavir-Ritonavir in Adults Hospitalized with Severe Covid-19

Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19

A multicenter, randomized, open-label, controlled trial to evaluate the efficacy and tolerability of hydroxychloroquine and a retrospective study in adult patients with mild to moderate coronavirus disease 2019 (COVID-19)

Effect of Hydroxychloroquine in Hospitalized Patients with Covid-19

Hydroxychloroquine for Early Treatment of Adults with Mild Covid-19: A Randomized-Controlled Trial

Randomized double-blinded placebo-controlled trial of hydroxychloroquine with or without azithromycin for virologic cure of non-severe Covid-19

Lopinavir-ritonavir in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial

A living WHO guideline on drugs to prevent covid-19

A living WHO guideline on drugs for covid-19

FDA Approval of Remdesivir -A Step in the Right Direction

Structure of the RNAdependent RNA polymerase from COVID-19 virus

The antiviral compound remdesivir potently inhibits RNA-dependent RNA polymerase from Middle East respiratory syndrome coronavirus

Remdesivir: Review of Pharmacology, Preclinical Data, and Emerging Clinical Experience for COVID-19

Remdesivir for COVID-19: challenges of underpowered studies

Efficacy of Remdesivir in COVID-19

A very, very bad look" for remdesivir

Covid-19: Remdesivir has little or no impact on survival, WHO trial shows

World Health Organization

Evaluating Clinical Efficacy of Antiviral Therapy for COVID-19: A Surrogate Endpoint Approach

Risk assessment of multistate progression of breast tumor with state-dependent genetic and environmental covariates

Addressing challenges to rolling out COVID-19 vaccines in African countries

Building public trust: a response to COVID-19 vaccine hesitancy predicament

Planning for a COVID-19 Vaccination Program

Vaccine hesitancy: the next challenge in the fight against COVID-19

Resurgence of COVID-19 in Manaus, Brazil, despite high seroprevalence

Strategies for containing an emerging influenza pandemic in Southeast Asia

Neuraminidase inhibitors for influenza

Amy Ming-Fang Yen

Manuscript title:AUTHORSHIP All persons who meet authorship criteria are listed as authors, and all authors certify that they have participated sufÀ ciently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript. Furthermore, each author certiÀ es that this material or similar material has not been and will not be submitted to or published in any other publication.Please indicate the speciÀ c contributions made by each author (list the authors' initials followed by their surnames, e.g., Y.L. Chang). The name of each author must appear at least once in each of the three categories below.

Conception and design of study:, , , ;acquisition of data: , , , ;analysis and/or interpretation of data: , , , .

Drafting the manuscript: , , , ;revising the manuscript critically for important intellectual content: , , , .

Approval of the version of the manuscript to be published (the names of all authors must be listed):

All persons who have made substantial contributions to the work reported in the manuscript (e.g., technical help, writing and editing assistance, general support), but who do not meet the criteria for authorship, are named in the Acknowledgments and have given us their written permission to be named. If we have not included an Acknowledgments, then that indicates that we have not received substantial contributions from non-authors. 

A conÁ ict of interest occurs when an individual's objectivity is potentially compromised by a desire for À nancial gain, prominence, professional advancement or a successful outcome. JFMA Editors strive to ensure that what is published in the Journal is as balanced, objective and evidence-based as possible. Since it can be difÀ cult to distinguish between an actual conÁ ict of interest and a perceived conÁ ict of interest, the Journal requires authors to disclose all and any potential conÁ icts of interest.

The authors whose names are listed immediately below certify that they have NO afÀ liations with or involvement in any organization or entity with any À nancial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-À nancial interest (such as personal or professional relationships, afÀ liations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.Author names:

The authors whose names are listed immediately below report the following details of afÀ liation or involvement in an organization or entity with a À nancial or non-À nancial interest in the subject matter or materials discussed in this manuscript. Please specify the nature of the conÁ ict on a separate sheet of paper if the space below is inadequate.