Thread by @michaelzlin

Would you prefer (1) we get back to normal activities sometime, or (2) we make new vaccine-evading coronaviruses continuously, suffer widespread breakthrough waves, and wear masks forever?

If you chose #1, then know this: Merck's molnupiravir should not be approved.

Molnupiravir is, to put it in clear terms, a potentially dangerous and virus-enhancing drug. It is not an effective antiviral medication outside of the confined conditions of cell culture and hamster cages. I'll explain below.

Well known independent voices have brought this up, such as @CT_Bergstrom (renowned skeptic of bad research) @WmHaseltine (HIV pioneer) and @JamesEKHildreth (FDA advisor). Others such as @chasewnelson have done more in-depth analysis. I'm here to try to explain the issue simply.

The problem with molnupiravir (MOV) is its mechanism of action: It creates mutations in the coronavirus genome as the virus replicates. This is bad to people at low levels, because mutations are what creates variants such as Delta or Omicron.

MOV can be bad to the virus in particular conditions: If you can trap the virus in an environment with high enough MOV for long enough time, then all copies of the virus in that environment will accumulate mutations that break some important function of a viral protein.

The key thing to appreciate is you can create the conditions for MOV to work in a petri dish and in lab animals in a cage, but you cannot do it reliably in human patients. And then when you don't, what you get are viruses that are mutated but not dead.

And then from there, any mutations that benefit the virus, e.g. by evading antibodies from vaccines or prior infection or by making it more contagious, will expand, no matter how rare they are. That's natural selection.

I'll give you one piece of evidence that viral escape happens, then we'll discuss more details later. The evidence is Merck's, and it's that MOV only works 30% of the time to prevent hospitalizations in at-risk patients, while the antiviral drug Paxlovid works 89% of the time.

This means that if you maximally inhibit the virus, you prevent 89% of hospitalizations. Since MOV only prevents 30%, then in 59% of cases, the virus is still replicating for some time in people despite being mutated. And these mutated viruses can hop to other people at any time.

I'll also point out that the above numbers mean most of the time Merck's drug doesn't work. Not only is 30% a minority of 100%, but Paxlovid shows a good antiviral can get to 89%. So Merck's drug isn't even half as good as the competition either.

Okay let's step back a bit now. Yesterday we heard FDA is on the verge of approving MOV. I don't know if this is true. I hope to heavens it's not. Most of the scientists I've spoken with oppose approval.

So how did we get here?

news.yahoo.com/u-fda-set-auth…

The idea behind MOV was a longstanding observation that coronaviruses needed proofreading to be able to reliably replicate their genomes without too many errors that they lose function. The idea is then, if we can make virus mutate more, then it will lose function, right?

MOV is a mimic of one unit of RNA, it gets used in place of a RNA by the viral RNA replicase. Experiments on coronavirus in cells showed that at high enough concentrations over several days, all virus was eliminated. The same was shown in some animals.

That was promising enough to try in humans, and that's where Merck announced 50% efficacy, getting everyone excited as this would be a pill you can take at home. Later this was revised to 30%, but by then Merck had already established a favorable first impression in the press.

Press reports seemed to only quote people who worked on the drug or were involved in formulating the interesting error catastrophe hypothesis, people who have clearly good intentions in hoping their research can lead to a treatment.

Haseltine was the first to raise the alarm, pointing out that mutating viruses in people is different from doing it in petri dishes or animal cages, because any mutated virus with better fitness/evasion will get selected in the patient and amplified outside the patient.

Everyone seemed to obsessed with Delta and vaccines to notice (it's tiring all the news, really). For non-academics, it can be hard to understand. That's why I'm explaining it here. People trusted that Merck couldn't possibly do something with a chance of making worse viruses.

When FDA asked their AMDAC advisors to consider benefit/risk ratio, the question of escaped mutated virus was not on the agenda. There was instead a long discussion on mutations to the patient genome that confused nonspecialists as to the major issue.

FDA did raise the issue of viral mutagenesis, and presented data showing viruses did pick up mutations (as expected) in humans. Merck said no virus remained after the 5-day treatment, but this in healthy Phase 2 volunteers, 89% of whom clear virus anyway. medrxiv.org/content/10.110…

No data was presented about whether live mutated virus can be isolated from those people who we expect to harbor them: those in the larger Phase 3 with weaker immunity who ended up in the hospital despite the drug

Merck also showed data from hamsters that the drug only dropped viral titers 100-fold. That sounds impressive until you realize there are 1 billion viral particles per gram of mucous in infected people. 100-fold less is still 10 million particles per gram.

In the end, when FDA asked their AMDAC advisors to vote, it was to consider the benefit/risk ratio to the patient, not to the world at large. And even this vote was close, 13-10. I wonder what the vote would be now, after Paxlovid has been approved.

Even though they weren't asked to comment on it, 6 advisors raised the issue of mutant virus escape. One was Dr. James Hildreth, who correctly identified it as the most pressing issue:

https://twitter.com/JamesEKHildreth/status/1465806607909064705

I go through this because you can see that Merck is checking off the boxes: ☑️MOV looked good in prelinical work. ☑️MOV has some non-zero efficacy. ☑️At the time it was being evaluated by AMDAC there were no other oral options approved.☑️MOV isn't terribly toxic to the patient.

Normally that would be fine, but MOV is not like any other drug. It is the first drug designed to mutate a virus without inhibiting its replication directly. Until the virus is completely eliminated from the body, then a treated person is a breeding ground for mutated live virus.

The problem with legal checkboxes is this: FDA and Merck aren't legally mandated to look out for public health, but it's everybody's obligation, whether written in Congressional statute or not, not to deliberately create new virus variants that can prolong a deadly epidemic.

Someone with leadership needs to step in and impose sanity. France did it (yet another reason to love France). They rejected MOV, citing correctly the poor 30% efficacy vs 80% for mAbs, and lack of proof mutated virus has been eliminated from patients.
has-sante.fr/jcms/p_3304161…

Now there are a two more things that really need mentioning. One, simple off-patent antidepressants such as Prozac appear similarly or *more* effective than MOV in preventing hospitalization (32% in the study below) at low drug costs and no mutated virus!
thelancet.com/journals/langl…

Second, there are already concerns about mutant virus escape even if everyone is compliant with the drug because it takes days for MOV to eliminate the virus even in the best case, but on top of that, many people will certainly only take a partial course of MOV.

Patient compliance is a major problem with antimicrobials, because many stop taking drugs the moment they feel better. This is rational; who wants to take drugs with side effects if they're getting better? But in the case of COVID19, >90% will get better in a few days anyway.

Alternatively some may feel worse in some way that they attribute to side effects from the drug, and discontinue it. This is also more likely with COVID19 because many people know they have a high chance of recovering anyway.

So the circumstances of COVID19 hospitalization prevention — most people will get better quickly — sets up MOV for poor patient compliance. And if MOV is given for just 2 or 3 days, then the virus is still there, mutated and ready to hop to someone else.
medrxiv.org/content/10.110…

So I know this is a long thread, and thank you for bearing with me. MOV approval is most momentous decision the FDA will make in this epidemic, maybe in FDA history, maybe even in the entire history of pharmaceuticals.

The worst-case scenario that we haven't ruled out is MOV will lead to years of new variants, with people desperately taking it to fight the new variants that it spawns, creating a vicious positive feedback loop while causing countless suffering and deaths.

Isn't that grim enough to delay this decision until Merck convincingly tells us how many functional mutants are created and transmitted per every 1000 patients to their family members for example?

Even better, we as a society should tell our leaders to reject this viral mutagen and put resources into safer drugs, such as the equally efficacious off-patent antidepressants (we already have millions of doses) or the more efficacious Paxlovid or sotrovimab.

You might wonder, if it's so grim, why have I not heard about this? That's a good question. Most of my colleagues share my concerns. And I've been in contact with journalists who have heard these concerns. But so far they've hesitated to write about them.

Reporters don't know who to trust, so they by default have been deferring to FDA to make the right decision, or just assume Merck must know what they are doing. But circumstances may be leading them to make risky decisions, and the fast pace is not giving people time to evaluate.

I had hoped this would get covered by the press, or that FDA would make the right decision. I prefer not to have to argue against a drug approval. But with this week's news that approval is imminent in the US, there is no more time to wait, and there is an obligation to speak up.

So please think about what I wrote. Ask questions to people you trust. Retweet if you agree or if you are not certain. It needs to be discussed, and not just within the FDA but in public, as it may affect all of us.

😭Dismayed that FDA has now made the worst decision in its history. We cannot give up on raising awareness of the dangers of molnupiravir, and its poor efficacy. We must limit its use while we work on a worldwide campaign to reverse this.
fda.gov/news-events/pr…

Great {sarcasm} virus is still present in 1.9% of people taking the drug for 3 days. That's already multiple replication cycles, so mutations have accumulated. Will 100% of patients take the drug long enough to kill all virus? Might some people stop at 3 days? h/t @EricTopol

@EricTopol Topol thinks the abstract, aka advertisement, is good. We must think more critically and imaginatively. Evaluate the benefit (3% in intent-to-treat population). Think about worst-case scenarios (new evasive variants). Then decide whether it's worth it.

https://twitter.com/EricTopol/status/1474099769265975303

@EricTopol We know, FDA knows, Merck knows (it's their data) that at 3 days there are mutated viable virus in patients taking molnupiravir. How are we going to prevent patients from stopping the drug, or forgetting a dose, or simply coughing on a family member around that time? We can't.

@EricTopol says he has spoken out on the risks of molnupiravir. Thought I'd mention that, although it's impossible to know it from his post above.

Fact sheet for prescribers says isolation and completion of the therapy are important to minimize viral spread. That confirms FDA acknolwedges a non-zero risk of mutated virus being released. But how many are going to truly isolate (wear masks, sleep and eat in a separate room)?

But I'm glad FDA wrote "Completion of the full 5-day treatment course and continued isolation in accordance with public health recommendations are important to maximize viral clearance and minimize transmission of SARS-CoV-2." It refutes all those who say escape is impossible.

And FDA knew this drug has PH risks: "changes in the spike protein occurred at positions targeted by monoclonal antibodies and vaccines. The clinical and public health significance of these changes are unknown."

Actually the danger of mutating antibody binding sites is known!

And both drug-specific reasons for granting EUA are incomplete or incorrect
#2: This again sounds like talking about benefit/risk for the patient, not for everybody else
#3: sotrovimab, Paxlovid, remdesivir are better, and fluvoxamine is as good, cheaper, and more available.

The Washington Post @wapo has published my thoughts on how FDA's ill-considered approval of molnupiravir may result in more immunoevasive SARS-COV-2 variants. We should enforce isolation of patients using it until Merck describes the mutations it produces: washingtonpost.com/outlook/2021/1…

@wapo I have some time now (by hiding away from my poor family on a holiday) to discuss arguments claiming molnupiravir will be a-okay. I wish there could be rigorous logical arguments for it, but as you might expect from my conclusions above, these are in the end not reassuring.

Some arguments are easy to see as wishful thinking, some require some more deep diving and calculations to see the problem, but I'll do my best to explain them. Before we start, I'll just make one comment.

I want a way out of this pandemic as much as anybody else. I don't want to get COVID19 or for my >70yo parents to get COVID19 or for my two children 6 and 12yo to get COVID19. I don't want anyone else to get it for that matter.

It's why my lab is developing protease inhibitors. Just as I like to buy insurance for my home, we figured we'd start on it in case the pandemic got out of control. As a lab with the appropriate background and good ideas, we felt obligated to try.

We've gotten no funding from any companies. NIH turned our application down in 2020 because the reviewers saw no need with vaccines being rolled out, so we have just private donors. Those donors pay for 1 scientist in my lab, and 0% of my salary

So COVID19 research is not a direct source of income for me. I'd be happy if COVID19 is gone so we don't have to work on it at all. It doesn't have to be my lab who makes a good COVID19 drug.

I'm happy enough Pfizer made a good drug by modifying the HCV protease inhibotr boceprevir, even if they don't acknowledge the intellectual contributions from my and other labs who published that idea. In the end, it's stopping COVID19 that matters.

https://twitter.com/michaelzlin/status/1473727651231764480

It would be dumb for me to argue against a drug I think actually makes a net positive difference to COVID19 without undue risks. I have a few more decades I'd like to live, preferably mask-free, and my kids (and your kids) have many more after we're gone.

I'm aware that posting or publishing a logical argument against molnupiravir will immediately elicit ad hominem attacks on me and not on the arguments, or accusations that I must be paid to lobby against Merck.

It's a sad statement on the state of public debate that many assume every opinion is paid for, or can't believe some people would come to conclusions from data on a case-by-case basis. We got used to notable politicans & CEOs lining their pockets but not everyone works that way

So it doesn't benefit me to speak out against molnupiravir. But as you can see from my pinned post, when I see or hear something wrong (and there have been a lot of wrong things said in this pandemic), I speak up. Just my scientific obsession about doing things right, I guess.

Okay now to the arguments you will hear in support of molnupiravir. This may be a bit disorganized as I am not parsing into tweet-size segments ahead of time.

My intention is to present these arguments accurately, not as straw-men arguments. In the several weeks since I first presented my concerns, below, I have been in contact with other scientists who also perform research on viruses.

https://twitter.com/michaelzlin/status/1464149312540233729

Some are scientists who have made important contributions to public education in this epidemic, some have been involved in molnupiravir research, some have reached out to me directly. I have learned a lot from their expertise and engagement and am grateful for it.

Okay, let's begin. People well versed in RNA virus evolution point out that errors in RNA copying limit either viral fitness or genome size, so that for a given RNA virus of a certain genome size, increasing mutation rates will cause it to lose fitness and eventually viability.

This is an accurate statement AFAIK, and it is the inspiration for using viral mutagens as an antiviral medication. The statement might be made by defenders of MOV primarily as an educational FYI, perhaps to make sure people understand MOV's mechanism of action.

As an argument for MOV safety, the fact that viruses live near the edge of error catastrophe naturally is not reassuring by itself. If you want to use a drug to push viruses into error catastrophe, in people, you can't give it a chance to escape before all copies are defective.

That is, dosing a patient with MOV for 5 days while surrounded by people not on MOV is not the same as shifting an entire species of virus to a higher mutagenesis rate.

Studies of how viruses have evolved over decades can certainly be applied to think of how to make drugs, but one should be careful to assume drug administration can mimic a change in the replicative properties of a virus.

In people there is again that window of time before 5 days when many copies of the viral genome have picked up some mutations but not all copies are nonfunctional (obviously there is an in-between state between no mutations and so many mutations that the virus is dead).

There are also intermediate concentrations of MOV, which might happen in people who skip a dose or stop it early, where virus replication is partially inihbitied. It means there can be mutagenesis but not a big difference to viral fitness or levels.

https://twitter.com/michaelzlin/status/1465796741291122692

So I believe statements such as "MOV works because coronaviruses need accurate replication and live on the edge of error catastrophe" are accurate if interpreted as "MOV can work" but not as "MOV can never create escaped mutant viruses"

Another thing you will hear is "most mutations are deleterious to the virus". Of course. That's true in general for any reproducing entity based on DNA or RNA. And yet species evolve not because most mutations are deleterious, but in spite of it. That's how evolution works.

The diversity and amazing specialization abilities of life forms on earth (yes viruses are alive for purposes of this conversation) occur because those rare mutations that create greater fitness are selected for, allowing progeny to reproduce better and take over their niche.

So we can say both that MOV will kill most viruses because most mutations are deleterious, and that it's possible that early in MOV treatment mildly mutated virus can hop to people not taking MOV and have a reproductive advantage (by chance).

Again the "most mutations are deleterious" is often presented without being clear whether it is an argument for why MOV could work, or why MOV won't have issues, e.g. science.org/content/articl…. It's fine as the former, but I think some people have misinterpreted it as the latter.

Other statements of support for MOV are that "it works 100% in cell culture and animals" and "it worked better than anything else we've tried in cell culture and animals". Here the issue is physical and historical context that doesn't apply so well to people and is now outdated.

These kinds of statements were more common in the press when efficacy was believed to be 50% in an interim analysis, and probably played some role in setting up a favorable view of the drug among commentators.

Those statements may have added some heuristic value when we didn't know how well an oral antiviral could possibly work to prevent hospitalization. MOV was the first oral pill, so it was plausible that the best possible result possible was 50% reduction.

So statements that MOV was 100% effective in cells and animals and that had never been seen before suggested we should approve MOV because there's not going to be anything better. However we know now that that's not true, so these statements are now more misleading than useful.

First, MOV efficacy against hospitalizations dropped from 50% to 30% in the final analysis because there was essentially no benefit in the second half of the trial. So we know now there is some disconnect between efficacy in cell culture or lab animals with efficacy in humans

Second, Paxlovid proved to have 89% efficacy in the same metric. This shows that a more effective antiviral can indeed do at least as well at 89%. So those earlier comments about MOV wouldn't have the same (or any) impact if issued now, but they linger in the discussion.

Ironically the fact that MOV was being tested initially in academic labs meant we heard a lot about its preclinical results, whereas that wasn't true for Pfizer's drug. So more anticipation and premature commentary was built up for MOV than for Pfizer's drug.

MOV was called NHC previously, and it was actively being tested against coronaviruses like MERS at UNC right before COVID19. The below article is from 2019 November. That means academic collaborators were rapidly engaged once COVID19 hit.

pubmed.ncbi.nlm.nih.gov/31578288/

You can see some of this history here, and how a large set of academics were involved. The preclinical work was quite open, which is usually a good thing.
nature.com/articles/d4158…

In fact I've discussed MOV with someone involved in its early testing. This person is someone I've worked with before and greatly admire, whom I consider the best academic antiviral drug developer. He was enthusiastic about MOV because of his own results.

He said regarding escape, they never saw it in lab animals, and they tested for it. Sounds reassuring. But he did admit that the animals don't get to skip doses or discontinue early.

I didn't press beyond that, because the next step anyway is to consider how well it does in humans. We now know it's not as good in humans, even with perfect dosing, so the animal models are different in some way.

Perhaps MOV in humans, despite reaching blood concentrations similar to lab animals, only reaches mutagenic but sublethal concentrations (as in the curves above) in places where coronaviruses replicate, e.g. nasal mucosa. We don't do nasal biopsied to tell.

Humans being bigger or just different from lab animals might mean tissue concentrations may be different even with similar blood concentrations. In any case, animal models didn't predict efficacy and may not predict escape.

This brings up another reason why you might not have heard a lot of academics going public against MOV, although there are concerns in private. Academic research is sweat and tears and we want to be supportive of each other.

The above 3 arguments essentially are "MOV should work great in theory and does work great in model systems" but such arguments are trumped by what we actually see in humans: efficacy far worse than other treatments, and detection of viable mutated virus for several days.

Below I'll discuss two arguments about how mutagenesis is unlikely to make things worse even if not perfect. One is easily seen as clearly incorrect. The other requires some numbers, that we do not know, to accurately address.

The first argument that escape from molnupiravir mutagenesis is nothing to worry about is that SARS-CoV-2 had already fully explored the mutations useful for it. This is clearly not true. The history of variants through Omicron shows mutations increasing without leveling off.

In the extreme case we know for example that SARS-CoV-1 and SARS-CoV-2 are both capable of infecting humans, and these diverge by 20%, or 6000 RNA letters. The reasons for SARS-CoV-1's higher pathogenicity are unclear but reside somewhere in those 6000 changes.

Meanwhile so far SARS-CoV-2 has only picked up dozens of mutations. It may not be possible for SARS-CoV-2 to evolve directly to something like SARS-CoV-1 but I'd rather not find out, and the numbers show there is a lot of functional mutagenesis space still unexplored.

A more accurate version of this idea is that, rather than mutational space being fully explored, that it will be fully explored in the future anyway. I believe that's what people mean when they speak in the present tense, although the time interval being envisioned is not defined

Two responses here. First it will take a while to explore the entire possible diversity of SARS-CoV-2 (say 6000 based). Naturally it would take decades and occur via new variants arising occasionally.

Why speed up this process by artificially introducing more mutations into the viral genome pool, making new breakthrough waves more frequent and making it harder for our vaccines and therapies to keep up?

Second, this argument that molnupiravir may not be much different from no molnupiravir fails to account for the fact that generating 1 genome with several mutations has much higher probability to create an immunoevasive strain than generating several genomes with 1 mutation.

We have seen this directly in Omicron it is the fact it has 12 mutations in the spike protein that make it so immunoevasive. 12 variants each with 1 mutation would present little problem.

But the 12 spike mutations in that first copy of Omicron allowed it to evade existing immunity and then grow exponentially to infect the entire globe in a few months.

Molnupiravir tries to work by introducing a large # of mutations per genome. It won't work unless that happens!

It is thus probable that molnupiravir will give rise to variant-of-concern evolution at a much faster rate than possible with natural mutagenesis rates.

As explained above, any response that "if this were true viruses would evolve with faster mutation rates" would be a mistake in understanding the situation.

Molnupiravir again won't raise mutation rates species wide 100% of the time. It will cause little bursts of mutagenesis in people taking it, and then people not taking it will allow stable non-mutating propagation of better-fit variants.

Another argument you will hear repeatedly from the semi-informed is that "ribavirin also induces mutations, and we've used that safely before." Two major reasons why molnupiravir is not ribavirin:

First, viral mutagenesis is only one of several mechanisms of action of ribavirin. It's even been shown that clinically relevant concentrations of ribavirin do not cause lethal mutagenesis of RNA viruses, so that's clearly not the primary mechanism.

Unfortunately even some scientists are under the misimpression that ribavirin works by viral mutagenesis. They need to google. First hit under ribavirin mechanism of action is
ncbi.nlm.nih.gov/pmc/articles/P…. Conclusion: multiple antiviral mechanisms, and error catastrophe doesn't happen

Second google hit (up-to-date information is not hard to find, if you're willing to look for it) is ncbi.nlm.nih.gov/pmc/articles/P…. It has a nice figure showing 5 possible mechanisms of action of ribavirin (red and green lines)

This one also notes RNA mutagenesis (error catastrophe) is not at all an accepted major mechanism of action of ribavirin.

Rather it concludes that potentiating interferon-mediated activation of antiviral genes is the mechanism with the best evidence.

The second thing about ribavirin is, even if it were a viral mutagen (and the quoted references show it may not be at clinical doses, and its antiviral effect is not dependent on it), it has only been used in viruses with much less explosive potential than SARSCoV2.

Notably it was used for HCV and RSV, but HCV as a bloodborne pathogen like HIV is not transmitted by casual contact, and RSV has a R0 near 1. SARSCoV2 Delta has a R0 of 5-10 and Omicron estimates are 20.

So any enhanced HCV or RSV variant caused by ribavirin is much more likely to die out in that patient, rather than be spread to a family member or caretaker, and from there potentially doubling every 2 or 3 days, reaching millions of people in a matter of weeks, as SARSCoV2 does.

Sorry for the technical discussion, but I warned upthread I would need to get into details to explain why the arguments I've heard trying to prove molnupiravir is safe have not been reassuring upon deeper examination.

Now there is one final argument defending molnupiravir, and this is the only one with enough logic and applicability to SARSCoV2 to be worth debating. The argument is that the lower virus #'s due to molnupiravir at least offsets the higher mutation #'s in the remaining copies.

To assess this argument, we need some numbers. Viruses isolated from patients on molnupiravir show about 5-fold more mutations than placebo, although not clear which day (fda.gov/media/154472/d…).

Presumably these viruses were isolated either on day 1 or day 2. We do know only 1 in 50 patients show culturable viruses at day 3, so not likely those viruses were isolated from day 3.
science.org/doi/10.1126/sc…

SARSCoV2 isolates tend to have about 2 mutations per transmission cycle, with a serial interval of about 3-5 days, so that means naturally SARSCoV2 genomes accumulate about 1 mutation per 2 days in the host.
journals.plos.org/plospathogens/…

So MOV causing 5x more mutations in 2d of drug means 2.5 hits per genome on average on day 1, and 5 on day 2. If it was 5x more mutations in 1d of drug instead, then that's 5 hits on day 1 and 10 hits on day 2.

So we estimate MOV causes mutation at a rate of 2.5-5 hits per day

Merck actually saw very little difference in viral titers after 3 days on drug vs. placebo (people are clearing virus anyway) but 8x fewer had detectable infectious virus, so most treated virus copies are dead at 3d (with 7.5-15 mutations presumably)

The danger zone is the first 2 days of therapy (or people who stop after 1 or 2 days of therapy) when virus is still replicating okay in MOV (if it weren't then you wouldn't see similar titers on day 3 between MOV and placebo) despite having 2.5-10 mutations.

We'll be generous to MOV and say on day 2 of therapy functional virus numbers are actually 10x lower than placebo (seems they can't be lower than that and still produce nearly identical virus levels at day 3).

Then you have 10x lower levels of functional viruses taking MOV than not taking MOV at day 2, but those viruses each have 5-10 times more mutations. Seems about a wash. Can we conclude MOV doesn't increase the potential to generate variants of concern then? I don't think we can.

Recall people are producing viruses with 1-2 mutations per transmission event and we're not getting immunoevasion all the time. That's because the immune response is polyclonal: multiple antibody species are produced that target multiple locations on spike.

It's only when mutations happen to remove more than 1 antibody binding site on spike's receptor binding domain (RBD) that a new variant can overcome existing immunity appreciably. This is exactly what happened with omicron – it has 15 mutations on the RBD
asm.org/Articles/2021/…

Thus when it comes to creating immunoevasive variants, I believe it is much worse to have 10x fewer viruses each with 10 mutations, than it is to have an untreated number of viruses each with 1 mutation.

I believe this is the critical point missed completely by the argument that MOV doesn't increase the chance of immunoevasive variants because it reduces viral load.

Would have been prudent for FDA to have required measurements of infectious virus levels and their complete sequences on treatment days 1 and 2 for > 100s of patients so we can get a better idea of this risk. But they granted EUA without this info. Thus we're now taking a gamble.

Anyway as stated above, I'm grateful to the scientists who have presented these arguments in defense of MOV. Be assured I've been thinking about them ever since I raised my concerns about MOV a month ago. I wouldn't want to be incorrect, obviously, so I did my homework

Certainly long-term followers and those who work with me know I don't like to guess. I also don't try to reach for a particular conclusion and find data to justify it. This case is no different. I wrote this thread and the WaPo opinion piece only after considering all arguments.

Anyway I think that about wraps up this thread. In sum, I have presented my concerns about the potential of MOV to create new immunoevasive or more virulent variants of SARSCoV2, derived from its only mechanism of action being mutagenesis of SARSCoV2.

The low efficacy of only 30% in preventing hospitalization is similar to generic antidepressants and worse then the 89% of other oral antivirals, which also suggests viral clearance is not as fast or complete as other antivirals.

Viable mutated viruses that exist early in MOV treatment, or in patients who skip a dose or discontinue treatment, may then become new variants of concern. The arguments for why this is unlikely to happen are not satisfactory enough.

FDA was aware of the risks of mutations in antibody binding sites and of viral escape, as detailed in their MOV Fact Sheet for Healthcare Providers, but still granted EUA. That's why I wrote we must limit its use until we find out more.

End of thread.

washingtonpost.com/outlook/2021/1…