Is remdesivir the drug that can kill the coronavirus? | Hacker News Reader

Is remdesivir the drug that can kill the coronavirus?

economist.com

26 pointsaluket48 comments

Previous discussion regarding how difficult it is to produce remdesivir:

https://news.ycombinator.com/item?id=22783363

It's remarkable how a non-randomized study of Hydroxychloroquine is anathema, yet a non-randomized study of Remdesivir that's paid for by its manufacturer and written up by that manufacturer's employee is not. Hmm, I wonder why that is. To top it off, it might actually have the same QT prolongation issues, because some other antivirals have them.

That said, politics aside, I hope it works far beyond our wildest expectations so fewer people die.

Why does this have to be political?

One drug has turned out to be mostly ineffective and in some cases dangerous, people have died. Another one seems to work okay-ish in a huge coordinated experiment.

I don't see where politics come into this.

You can't support either of the assertions with available data. No control arm? You know bupkis.

> people have died

Citation needed. And no, aquarium cleaner dude is not what I mean by "citation".

Both drugs are under trial but we have seen some leaks on HN already.

Even without that, an unbiased person would at the very least be equally sceptical to both.

> Even without that, an unbiased person would at the very least be equally sceptical to both.

Unless you know someone who recovered after taking the drug.

Sure, but that's at best anecdotal. At worst, the patient would have recovered anyway and the drug only made it harder.

The official reports will be available in 2 weeks.

> At worst, the patient would have recovered anyway and the drug only made it harder.

They had called the family in to say goodbyes because the guy wasn't supposed to make it through the night. I'd say the drug saved his life.

> I don't see where politics come into this.

Short answer, because Politicians publicly pushed Hydroxychloroquine. You should know that.

I'd like to know if there was any evidence or indication of the value of HQNN before a certain president started spouting it. AFAIK the only reason we're talking about it is because he did. If that's so then there's no reason to pick HQNN as being special over any other drug. I suppose there's that very dodgy french study, perhaps that's what started it all off.

Basically exactly the same kind of "evidence" as what we have now for remdesivir. Some patients got better while on it, but because there was no control arm we don't know whether they'd do worse (or better) without the drug.

Virologists generally agree that it is a plausible drug to treat COVID-19, but will quickly point out that efficacy is unproven and we need to wait for better studies.

Also, a couple of doctors who are regarded as "cranks" started applying it early in the disease and presumably had success with. Trump picked up on that, so now a lot of people have reason to hate on it. Trump poisoned the well.

No "proper" doctor is going to administer a dangerous drug to a non-critical patient to prevent what might turn out to be a harmless course of the diseases. However, that also means if these drugs work, but only in the early stages, then we won't find out any time soon.

> No "proper" doctor is going to administer a dangerous drug to a non-critical patient to prevent what might turn out to be a harmless course of the diseases.

This "dangerous" drug has been used in medicine for 60 years for probably more than a billion or two people. I'm sure that a "proper" doctor faced with a dying patient that is unresponsive to everything else is going to offer this drug. Its a good thing that congress approved the Right To Try law.

Parent explicitly said "...to a non-critical patient..." You're describing the opposite (but I'd probably agree with you there).

> This "dangerous" drug has been used in medicine for 60 years

Note that the doses required here are MUCH higher (one European hospital tried 1000mg doses) which is toxic. It can result in blindness, heart failure and other issues.

If I was dying and nothing else was working, I'd accept the risk.

But what if people died _because_ of the drugs??

https://www.nicematin.com/sante/coronavirus-nous-avons-deja-...

Its all French to me.

Retinal damage is not an issue unless you take it for years. Heart failure is not an issue in a hospital because it's not sudden, so all treatment protocols specify that patients are monitored for QT prolongation. Moreover if you don't get it right away, you're unlikely to get it at all. "Other issues" in this case are mostly nausea and diarrhea. I don't know about you, but I'd rather shit myself than die.

> for probably more than a billion or two people

That's a few orders of magnitude off.

We already have some preliminary data that shows it isn't very effective for critical patients:

https://news.ycombinator.com/item?id=22797185

That's why I wrote non-critical patients. There are doctors administering it in a prophylactic fashion, supposedly with success. That's highly controversial.

The evidence is perhaps shown by the countries that use HQNN to prevent malaria, that also have very low infection rates for the coronavirus.

Are there countries in which a significant fraction of the population takes it on a regular basis? Presumably these are the same places in which an actually large fraction of the population has had malaria and in which there is some naturally evolved resistance to it, so you'd have to control for that...

There aren't any. Most regions with high risk of malaria (like India) have malarias that have chloroquine resistance, and the only places without chloroquine resistant malaria are in Central America and the Caribbean. Hydroxychloroquine is even less effective than chloroquine on those chloroquine resistant malarias. https://pubmed.ncbi.nlm.nih.gov/12837731/

Most Indians, for example, don't take any antimalarials until they actually have malaria.

> enough to supply more than 140,000 patients

> the firm has set an “ambitious goal” of producing more than 500,000 treatment courses by October, and 1m by the end of the year

So no. Even if this is a cure, the pandemic gets worse before it gets better.

Global infection rate has stabilized, it's not getting "worse" in that sense. At that rate, there won't be much more than another doubling.

If 80% of those infected do not need treatment at all, that amount of courses may well be sufficient, or close to it.

I'm still waiting to see what happens in Africa and South America.

There's no 5G there, so I guess nothing.

It's stabilized under the current shelter in place regime, which will almost certainly have to be lifted at least partially for economic-political reasons.

When that happens my guess is there will be new waves of illness.

The places that are growing are too early in the curve to move that needle. The developing world got it first because people travel more, but the virus has reached poor countries and they won't have as many options.

I think this makes the mistake of thinking there is such a thing as a global infection rate. There are instead a couple of hundred separate epidemics, each with its own R0 and trajectory.

It’s also dangerously wrong, but human nature, to look at the current state of epidemics and project their future state. Epidemics that have been suppressed rather than extinguished resemble unstable explosives. Just because they’re sitting there nicely today doesn’t mean they won’t explode 3 seconds from now.

There is a global average infection rate. Of course it's just an average, but is representative of the growth of global cases. Of course you can arbitrarily divide this down from hundreds of epidemics to thousands of outbreaks, but we're talking about the global supply of the drug.

> It’s also dangerously wrong, but human nature, to look at the current state of epidemics and project their future state.

I'm sorry, but in the real world you have to make predictions in order to make decisions. If you never make predictions you can always be right, but then you'll always be useless.

> Epidemics that have been suppressed rather than extinguished resemble unstable explosives. Just because they’re sitting there nicely today doesn’t mean they won’t explode 3 seconds from now.

An "explosion" in that sense will still take weeks to become critical, giving you time to react, assuming you have testing and tracing capacity.

> There is a global average infection rate. Of course it's just an average, but is representative of the growth of global cases

Flattening the curve in NYC and Lombardy does not change the trajectory of Brazil and every other developing country. Even India’s fast response is less likely to work than a rich country’s shelter in place. As long as there is exponential spread out there it will keep showing up wherever you call home.

> I'm sorry, but in the real world you have to make predictions in order to make decisions.

I'll overlook the implication that I don't live in the 'real world'. What the statement means is that just because you're fine today you can ease up, because you really don't like the shutdown. You're still standing in an explosives factory, and now propose to shoot off bottle rockets just because nothing bad has happened today.

In the real world, you look at conditions as they are. In this case, those conditions are that the virus is still epidemic and there are none of the preconditions necessary to contain it present in the US, at least, and also in most of Europe.

> assuming you have testing and tracing capacity

Where is it that you are describing, since you are 'in the real world'?

Here in the US there is no where that I am aware of that has sufficient testing and tracing capacity to deal with a 3-week old exponential outbreak.

Like the article says, there have been some hopeful anecdotes, but that is a long way from an effective cure and rigorous testing phase. Aka to early to tell.

what is dead may never die

Probably not because it was suggested months ago already, so if it was a magical solution we’d know by now.

Actually no we wouldn’t, because real clinical trials with thousands of subjects take a long time to wrap up, even when researchers are working at breakneck speed.

FWIW The anecdotal evidence from doctors and patients who have managed to get a hold of some of it has been very good so far.

Plus, killing off the virus isn't necessarily going to be equivalent to observed outcomes. It is easy to imagine someone having no viruses but their lungs aren't working properly because of damage done so they are still sick.

So even a perfectly efficacious drug would still not necessarily be an obvious miracle cure, depending on how the disease progression works.

This is a good time to wait for clinical trials to finish.

Most drugs aren't "miracle cures". Even a 50% treatment success rate would put it well ahead of many other staple drugs.

yes but, if in a trial, if one arm performs substantially different than the control arm (e.g. 50% or even 100% lower death rate), the large difference between arms can lead to investigators ending the trial early

ironically, this happened in a previous 2019 trial of Remdemsivir, for a different virus, Ebola

in that trial, 4 drugs for Ebola were tested, including Remdemsivir. 2 non-Remdemsivir drugs succeeded substantially--the large difference between those drugs and Remdemsivir, was enough to cause the investigator to end the trial early and drop Remdemsivir

ironically, the investigator for that trial was none-other-than Dr. Fauci. Maybe that's why he hasn't mentioned Remdemsivir much on TV these days, based on his past professional disappointing experience with Remdemsivir

also, double ironically, Remdemsivir was initially designed to treat Ebola virus, not for flu or coronavirus type viruses. The fact that after the Ebola trial, Remdemsivir was declared not that effective for Ebola (the very virus Remdemsivir was created for), means there is even less of a chance Remdemsivir will prove to be effective for Covid-19 in the current trials

https://www.sciencemag.org/news/2019/08/finally-some-good-ne...

as an extreme example

if Remdemsivir cures 100% of all Covid-19 cases, halfway through the trial, they would have enough data to show a net positive benefit. They don't need to finish the whole trial to declare Remdemsivir useful

the fact that they are running the whole trial to finish, most likely means the effect of Remdemsivir is not too noticeable. They need more data to get to statistical significance. Hence, they want the trial to run to original planned finish date.

my guess is

Remdemsivir probably has some positive effect. Say, 10% lower death rate if used during 1st few days of infection. Still great, much much better than our current situation of nothing. It would not cure Covid completely, but it would help substantially, especially in large populations of infected

Looks like you may have been right. Oh well, I hope it at least helps people at the early stages of the disease.

Reply on news.ycombinator.com