Risk/Benefit Profiles of Currently Approved Oral Antivirals for Treatment of COVID-19: Similarities and Differences

Round 1

Reviewer 1 Report

In this paper, the authors review three orally available drugs available for the treatment of COVID-19: Favipiravir Molnupiravir and Paxlovid. They discuss the mechanism of action of these drugs and protein target, potential for drug resistance, PK studies in animals, and clinical profiles. While there are several reviews of these drugs individually, a comparison of the three drugs together is a nice addition to current literature. However, the paper could be improved in terms of overall clarity and order, and references and statements should be carefully checked.

General comments

While the English is good, and sentences are clearly written and easy to understand, the paper could be made clearer in regards to the general layout and organization. A lot of information is repeated, and/or is not relevant to the section. There are often jumps between concepts and drugs, which makes it harder to follow

Some examples -

1. The ‘drug mechanisms of action’ section covers not only the MOA, but also potential concerns (e.g potential for mutation of host genome, Ames test), EC50/CC50 results from infected cells (more relevant to antiviral potency section) and the pharmacologically active trough concentrations in COVID-19 patients for Favipiravir. This does not fit in this section, and furthermore is not consistently mentioned for all three drugs.

2. The ‘antiviral potency section’ briefly mentions that all compounds are effective, then talks mostly about potential combination therapies.

3. Under ‘animal pharmokinetics’ there is reference to the use of Favipiravir use in COVID-19 patients with end-stage renal disease on dialysis (lines 242-243), and adverse events with Influenza treatment, then finishes with potential DDIs, which should be a separate section. Potential DDIs is then repeated in section 9, clinical pharmacology section.

4. ‘Animal toxicology’ discusses hyperuricemia (lines 307-311), but it is unclear if this is in regards to animals or humans. The problem of hyperuricemia is then repeated later in section 9 in 3 different locations (line 444, 456 and 469-472).

   ‘Animal toxicology’ also discusses the authorization of Paxlovid in individuals younger than 12 (lines 320- etc), which is then also repeated later in section 9.

5. Rather than discussing the drugs in order, there is often a jump back and forth between drugs. For example, lines 77-90 jump from Favipiravir, to Molnupiravir, to Paxlovid, back to Molnupiravir and then finally back to Paxlovid. Moving lines 82-84 after respective drugs would be a simple solution. This happens throughout the manuscript (especially throughout section 9) and should be rectified for clarity.

Major comments

1) A more extensive literature search should be conducted to check for missing references. 

For example, https://doi.org/10.1016/j.bmcl.2022.128629 - discussing Mpro varients and showing Nirmatrelvir has similar potency across all SARS-CoV varients and https://doi.org/10.1016/j.jbc.2021.100770 one of the first publications showing a detailed mechanism of action for molnupiravir, neither of which are referenced in the paper. Under sections discussing DDIs (see for example lines 376-383), there is a single reference for all three drugs. References are also not always related or correct, or could be more appropriate. For example Lines 10-105 then explain that favipiravir can cause termination of Influenza RNA with 2 or more consecutive nucleotides were incorporated. While this is correct, it is more relevant to mention that for SARS-CoV-2, it has been shown that chain termination only occurs after several consecutive incorporations which is very unlikely to be biologically relevant (Shannon, Nat comm, 2020). Occasionally, conclusions are drawn from several papers, but the relationship or how the conclusion was drawn is not clear. For example lines 368-371, references 55 is a modelling paper and 56 is experimental treatment of ebola virus, but the conclusion is that the antiviral efficacy of favipiravir is not impacted.

2) Figures, particularly of the drugs in their prodrug and active metabolite forms, and tables summarizing clinical trials DDI etc would greatly aid in overall clarity

3) A general overview of the activation pathway and related drug names (from the prodrug through to the active triphosphate form) should be described more clearly for the two NAs. Mentioning the original chemical names of the prodrugs and active metabolites would also benefit the reader (e.g T-705; MK-4482/EIDD-2801). NHC as the active metabolite is described in one of the first sections, but Favipiravir triphosphorylated form is only discussed in section 6. The authors should also be careful that they refer to the correct molecule. For example, in lines 115-116 stating that Favipiravir and Molnupiravir act by inducing lethal mutations in the genomes of target viruses – this is technically incorrect, as it is the active metabolite that is recognized by the viral RdRp and introduced into the genome where it causes increases in mutation rates and lethal mutagenesis.

4) The risk of DDIs for Paxlovid, is mentioned briefly in one or two sentences, and cites only reference 17 – the paxlovid fact sheet. Yet this has been quite well discussed in DOI: 10.1002/cpt.2646 (and other missing references). Given that Ritonavir has a high potential for DDIs – this must be expanded.

5) Line 102-103 - stating that drugs are chain terminators is false and needs to be changed (especially for Molnupiravir, see https://doi.org/10.1016/j.jbc.2021.100770 and Kabinger et al), and the reference is incorrect. This statement is also contradictory to lines 97-99 stating (correctly) that these drugs are mutagens. Line 115 then states that the drugs ‘additionally act by a separate mechanism of action inducing lethal mutations’, but this is a repeat of lines 97-99 and not a separate mechanism.

Minor comments

1) Line 73-74 states that all three drugs were orignaly developed for other diseases, but it is not clearly stated which diseases, or what stages of approval/use was obtained for the treatment of these other diseases

2) Line 93 - Favipiravir in prodrug form isn't a nucleoside analogue – it is a prodrug converted into a nucleotide analogue

3) Line 101-102 ‘their effects are modulated by the status of intracellular nucleoside pools’ – this phrase is unclear - they are in competition with intracellular nucleoside pools, but their activity is modulated by nucleoside biosynthesis enzymes. The reference is also not entirely relevant here, but could be mentioned later in a combined therapy section.

4) Lines 109-111: While it is true that Ritonavir is a HIV protease inhibitor, it is no longer used specifically for its activity against the HIV protease due to its toxicity at high concentrations. It is now only used in combination therapy at low doses as a pharmacokinetic enhancer to boost activity of other drugs. Is there a reference for Ritonavir-resistant HIV-1 cases to justify this statement?

5) Lines 124-125 – Worth mentioning that the host mutagenic capacity of Molnupiravir is not clear-cut. Sheahan et al (DOI: 10.1126/scitranslmed.abb5883) did not observe mutagenesis of host mRNA

6) Lines 125-127 – It is not clear to me what is the relationship between host mutagenic risk and immunocompromised patients who shed virus longer. Mutagenic potential of the drug is related to the concentration and duration of drug treatment.

7) Line 142 – I assume the authors mean non-mutagenic to the host. Needs to be clarified, because favipiravir triphosphate is mutagenic to the virus.

8) Lines 368-371 – unclear. If viral infection impacts Favipiravir exposure, how would this not impact the antiviral efficacy? What is the relationship with references 55 and 56?

9) Lines 372-375 – reference missing.

10) Lines 401-404 – repeated information from 320-327.

11) many abbreviations (e.g RTP, DDI, SAD/MAD) are not defined, defined twice or more, or used and then defined later in the text.

Author Response

Dear Dr. Gu,

Thank you for the opportunity to revise our manuscript “Risk/benefit Profiles of Currently Approved Oral Antivirals for Treatment of COVID-19: Similarities and Differences.”

We thank the reviewers for their time and effort to evaluate this manuscript. In response to comments from Reviewer 1, the manuscript was revised extensively. The overall organization was modified to improve clarity, a figure showing drug structures was added as were tables showing drug-drug interactions for Paxlovid and a summary of completed Phase 2/3 trials for all three drugs. New literature citations were added according to the reviewer’s suggestion and our scan of the most recent publications. Several critical points were clarified.

Below, we detail individual responses to comments from Reviewer 1. The requested modifications have improved the overall clarity and content of our manuscript and we hope it will now be ready for publication.

REVIEWER 1 (author responses are italicized)

General comments

While the English is good, and sentences are clearly written and easy to understand, the paper could be made clearer in regard to the general layout and organization. A lot of information is repeated, and/or is not relevant to the section. There are often jumps between concepts and drugs, which makes it harder to follow.

 The text was revised extensively to reduce confusion. Our goal was to compare drugs for individual characteristics, a head-to-head approach. This was not always possible based on available data and/or relevance to a particular compound. We hope the revised version reduces confusion and presents drug comparisons in a better way.

Major comments

1) A more extensive literature search should be conducted to check for missing references.

We appreciated the suggestion and especially appreciate specific citations recommended by the reviewer. All of the reviewer’s highlighted references have been added to the revised manuscript. In particular, we cited https://doi.org/10.1016/j.jbc.2021.100770 in reference to SARS-COV-2 variants and susceptibility to Nirmatrelvir. This is a worthy addition but speaks only to naturally occurring variants that drug susceptibility. The more important question of acquired resistance remains to be elucidated. We have included a reference to the Gordon paper (reference 19) and noted the conclusion that lethal mutagenesis is the primary mechanism of action for Favipiravir and Molnupiravir.

We also cited Marzolini and added additional references concerning Nimatrelvir DDI. The issue of naturally occurring Mpro gene variants was discussed, along with a citation to Ullrich, et al. (reference 76) regarding susceptibility to Nirmatrelvir.

2) Figures, particularly of the drugs in their prodrug and active metabolite forms, and tables summarizing clinical trials DDI etc would greatly aid in overall clarity.

A figure of prodrug/drug structures was added (Fig. 1). Two tables were included to summarize DDI for Paxlovid and Phase 2/3 clinical trials completed for the three drugs of interest.

3) A general overview of the activation pathway and related drug names (from the prodrug through to the active triphosphate form) should be described more clearly for the two NAs. Mentioning the original chemical names of the prodrugs and active metabolites would also benefit the reader (e.g T-705; MK-4482/EIDD-2801).

The text was modified to provide additional drug identifiers as well as adding Fig. 1.

4) The risk of DDIs for Paxlovid, is mentioned briefly in one or two sentences, and cites only reference 17 – the paxlovid fact sheet. Yet this has been quite well discussed in DOI: 10.1002/cpt.2646 Marzolini, 2022(and other missing references). Given that Ritonavir has a high potential for DDIs – this must be expanded.

We expanded the section n DDI for Paxlovid and included a table showing drug classes of interest. The Marzolini paper was cited and additional information was included.

5) Line 102-103 - stating that drugs are chain terminators is false and needs to be changed (especially for Molnupiravir, see https://doi.org/10.1016/j.jbc.2021.100770 Gordon and Kabinger et al), and the reference is incorrect.

The text was revised to emphasize lethal mutagenesis as the relevant mechanism of action and the Gordon paper was cited.

Minor comments

1) Line 73-74 states that all three drugs were orignaly developed for other diseases, but it is not clearly stated which diseases, or what stages of approval/use was obtained for the treatment of these other diseases..

Additional information was added to describe earlier development of these drugs.

2) Line 93 - Favipiravir in prodrug form isn't a nucleoside analogue – it is a prodrug converted into a nucleotide analogue.

This was sloppy use of terms and has been corrected.

3) Line 101-102 ‘their effects are modulated by the status of intracellular nucleoside pools’ – this phrase is unclear - they are in competition with intracellular nucleoside pools, but their activity is modulated by nucleoside biosynthesis enzymes.

Thank you for the correction. We made appropriate revisions.

4) Lines 109-111: While it is true that Ritonavir is a HIV protease inhibitor, it is no longer used specifically for its activity against the HIV protease due to its toxicity at high concentrations. It is now only used in combination therapy at low doses as a pharmacokinetic enhancer to boost activity of other drugs. Is there a reference for Ritonavir-resistant HIV-1 cases to justify this statement?

This comment was removed as it was difficult to identify a credible reference. When Ritonavir was used in first-line HIV therapy, there were multiple publications on drug resistance. However, as pointed out by the reviewer, these events are 30 years in the past and no longer of clinical relevance.

5) Lines 124-125 – Worth mentioning that the host mutagenic capacity of Molnupiravir is not clear-cut. Sheahan et al (DOI: 10.1126/scitranslmed.abb5883 did not observe mutagenesis of host mRNA.

We included the Sheahan reference and appreciate the experimental design and conclusions of that work. Our concern is about searching for apples-to-apples comparisons. Sheahan tests for Molnupiravir mutagenicity of host RNA and finds no evidence for changes due to drug treatment. However, that mechanism involves DNA-dependent RNA polymerase and it is unclear to what extent the findings are related to the viral RNA-dependent RNA polymerase of coronavirus or impacts on DNA-dependent DNA polymerases studied in Zhou, et al., 2021. We elected to highlight these important contributions and to comment on the disparate mechanisms studied by each group.

6) Lines 125-127 – It is not clear to me what is the relationship between host mutagenic risk and immunocompromised patients who shed virus longer. Mutagenic potential of the drug is related to the concentration and duration of drug treatment.

This section was removed due to lack of strong, primary literature to explain or support the contention.

7) Line 142 – I assume the authors mean non-mutagenic to the host. Needs to be clarified, because favipiravir triphosphate is mutagenic to the virus.

The reviewer is correct and the text was revised accordingly.

8) Lines 368-371 – unclear. If viral infection impacts Favipiravir exposure, how would this not impact the antiviral efficacy.

The text was revised to reflect current knowledge in the field. In Irie, K., et al., Pharmacokinetics of Favipiravir in Critically Ill Patients With COVID-19. Clin Transl Sci, 2020. 13(5): p. 880-885, the authors note the peculiarities of drug PK in critically ill patients. They comment on the complex medical status of ICU patients who are receiving many medications, frequently including opioids and that altered PK in this context is expected.

9) Lines 372-375 – reference missing.

The references were added.

10) Lines 401-404 – repeated information from 320-327.

We revised the text to minimize repetitions.

11) many abbreviations (e.g RTP, DDI, SAD/MAD) are not defined, defined twice or more, or used and then defined later in the text.

We revised to provide definitions for acronyms.

REVIEWER 2

Comments and Suggestions for Authors

This is an interesting comprehensive review demonstrating the clinical and pharmacological profiles of the currently approved oral antivirals for treatment of COVID-19.

Thank you.

Reviewer 2 Report

This is an interesting comprehensive review demonstrating the clinical and pharmacological profiles of the currently approved oral antivirals for 2 treatment of COVID-19.

Author Response

Dear Dr. Gu,

Thank you for the opportunity to revise our manuscript “Risk/benefit Profiles of Currently Approved Oral Antivirals for Treatment of COVID-19: Similarities and Differences.”

We thank the reviewers for their time and effort to evaluate this manuscript. In response to comments from Reviewer 1, the manuscript was revised extensively. The overall organization was modified to improve clarity, a figure showing drug structures was added as were tables showing drug-drug interactions for Paxlovid and a summary of completed Phase 2/3 trials for all three drugs. New literature citations were added according to the reviewer’s suggestion and our scan of the most recent publications. Several critical points were clarified.

Below, we detail individual responses to comments from Reviewer 1. The requested modifications have improved the overall clarity and content of our manuscript and we hope it will now be ready for publication.

REVIEWER 1 (author responses are italicized)

General comments

While the English is good, and sentences are clearly written and easy to understand, the paper could be made clearer in regard to the general layout and organization. A lot of information is repeated, and/or is not relevant to the section. There are often jumps between concepts and drugs, which makes it harder to follow.

 The text was revised extensively to reduce confusion. Our goal was to compare drugs for individual characteristics, a head-to-head approach. This was not always possible based on available data and/or relevance to a particular compound. We hope the revised version reduces confusion and presents drug comparisons in a better way.

Major comments

1) A more extensive literature search should be conducted to check for missing references.

We appreciated the suggestion and especially appreciate specific citations recommended by the reviewer. All of the reviewer’s highlighted references have been added to the revised manuscript. In particular, we cited https://doi.org/10.1016/j.jbc.2021.100770 in reference to SARS-COV-2 variants and susceptibility to Nirmatrelvir. This is a worthy addition but speaks only to naturally occurring variants that drug susceptibility. The more important question of acquired resistance remains to be elucidated. We have included a reference to the Gordon paper (reference 19) and noted the conclusion that lethal mutagenesis is the primary mechanism of action for Favipiravir and Molnupiravir.

We also cited Marzolini and added additional references concerning Nimatrelvir DDI. The issue of naturally occurring Mpro gene variants was discussed, along with a citation to Ullrich, et al. (reference 76) regarding susceptibility to Nirmatrelvir.

2) Figures, particularly of the drugs in their prodrug and active metabolite forms, and tables summarizing clinical trials DDI etc would greatly aid in overall clarity.

A figure of prodrug/drug structures was added (Fig. 1). Two tables were included to summarize DDI for Paxlovid and Phase 2/3 clinical trials completed for the three drugs of interest.

3) A general overview of the activation pathway and related drug names (from the prodrug through to the active triphosphate form) should be described more clearly for the two NAs. Mentioning the original chemical names of the prodrugs and active metabolites would also benefit the reader (e.g T-705; MK-4482/EIDD-2801).

The text was modified to provide additional drug identifiers as well as adding Fig. 1.

4) The risk of DDIs for Paxlovid, is mentioned briefly in one or two sentences, and cites only reference 17 – the paxlovid fact sheet. Yet this has been quite well discussed in DOI: 10.1002/cpt.2646 Marzolini, 2022(and other missing references). Given that Ritonavir has a high potential for DDIs – this must be expanded.

We expanded the section n DDI for Paxlovid and included a table showing drug classes of interest. The Marzolini paper was cited and additional information was included.

5) Line 102-103 - stating that drugs are chain terminators is false and needs to be changed (especially for Molnupiravir, see https://doi.org/10.1016/j.jbc.2021.100770 Gordon and Kabinger et al), and the reference is incorrect.

The text was revised to emphasize lethal mutagenesis as the relevant mechanism of action and the Gordon paper was cited.

Minor comments

1) Line 73-74 states that all three drugs were orignaly developed for other diseases, but it is not clearly stated which diseases, or what stages of approval/use was obtained for the treatment of these other diseases..

Additional information was added to describe earlier development of these drugs.

2) Line 93 - Favipiravir in prodrug form isn't a nucleoside analogue – it is a prodrug converted into a nucleotide analogue.

This was sloppy use of terms and has been corrected.

3) Line 101-102 ‘their effects are modulated by the status of intracellular nucleoside pools’ – this phrase is unclear - they are in competition with intracellular nucleoside pools, but their activity is modulated by nucleoside biosynthesis enzymes.

Thank you for the correction. We made appropriate revisions.

4) Lines 109-111: While it is true that Ritonavir is a HIV protease inhibitor, it is no longer used specifically for its activity against the HIV protease due to its toxicity at high concentrations. It is now only used in combination therapy at low doses as a pharmacokinetic enhancer to boost activity of other drugs. Is there a reference for Ritonavir-resistant HIV-1 cases to justify this statement?

This comment was removed as it was difficult to identify a credible reference. When Ritonavir was used in first-line HIV therapy, there were multiple publications on drug resistance. However, as pointed out by the reviewer, these events are 30 years in the past and no longer of clinical relevance.

5) Lines 124-125 – Worth mentioning that the host mutagenic capacity of Molnupiravir is not clear-cut. Sheahan et al (DOI: 10.1126/scitranslmed.abb5883 did not observe mutagenesis of host mRNA.

We included the Sheahan reference and appreciate the experimental design and conclusions of that work. Our concern is about searching for apples-to-apples comparisons. Sheahan tests for Molnupiravir mutagenicity of host RNA and finds no evidence for changes due to drug treatment. However, that mechanism involves DNA-dependent RNA polymerase and it is unclear to what extent the findings are related to the viral RNA-dependent RNA polymerase of coronavirus or impacts on DNA-dependent DNA polymerases studied in Zhou, et al., 2021. We elected to highlight these important contributions and to comment on the disparate mechanisms studied by each group.

6) Lines 125-127 – It is not clear to me what is the relationship between host mutagenic risk and immunocompromised patients who shed virus longer. Mutagenic potential of the drug is related to the concentration and duration of drug treatment.

This section was removed due to lack of strong, primary literature to explain or support the contention.

7) Line 142 – I assume the authors mean non-mutagenic to the host. Needs to be clarified, because favipiravir triphosphate is mutagenic to the virus.

The reviewer is correct and the text was revised accordingly.

8) Lines 368-371 – unclear. If viral infection impacts Favipiravir exposure, how would this not impact the antiviral efficacy.

The text was revised to reflect current knowledge in the field. In Irie, K., et al., Pharmacokinetics of Favipiravir in Critically Ill Patients With COVID-19. Clin Transl Sci, 2020. 13(5): p. 880-885, the authors note the peculiarities of drug PK in critically ill patients. They comment on the complex medical status of ICU patients who are receiving many medications, frequently including opioids and that altered PK in this context is expected.

9) Lines 372-375 – reference missing.

The references were added.

10) Lines 401-404 – repeated information from 320-327.

We revised the text to minimize repetitions.

11) many abbreviations (e.g RTP, DDI, SAD/MAD) are not defined, defined twice or more, or used and then defined later in the text.

We revised to provide definitions for acronyms.

REVIEWER 2

Comments and Suggestions for Authors

This is an interesting comprehensive review demonstrating the clinical and pharmacological profiles of the currently approved oral antivirals for treatment of COVID-19.

Thank you.

Round 2

Reviewer 1 Report

The authors have done a good job of revising the manuscript, which is now clearer and easier to follow. They have responded well to all comments, with the exception of the following:

Line 103 – the mechanism of action of NHCTP for chain termination is not controversial – there is no evidence of chain termination and it is generally accepted as a viral mutagen. NHCTP should be removed from this phrase. Line 106 should be modified to state clearly that chain termination does not occur following NHCTP incorporation.

Author Response

Thank you for your comments regarding the manuscript.  Lines 103-108 were revised in response to the reviewer's comments. For convenience, the revised text is copied just below:

Whether FTP acts as a terminator of RNA chain elongation is controversial. In vitro studies demonstrated chain termination when multiple, consecutive FTP were incorporated into a growing RNA [14, 15]. However, consecutive incorporation of FTP is unlikely in the cellular environment where guanosine triphosphate (GTP) and adenosine triphosphate (ATP) are present and compete with FTP [14].  NHCTP incorporation does not lead to chain termination [17, 19]. Lethal mutagenesis, also called error catastrophe, is the most critical mechanism of action for both Favipiravir and Molnupiravir.