I've been looking at the published results, both interim and extended, and trying to make sense of them. The plot thickens.
The interim results from a placebo-controlled phase 3 trial showed that, if given within 5 days of confirmed diagnosis, the drug was pretty effective in preventing hospitalisation or death in patients with at least one medical risk factor for progressing to severe disease.
The prospect of an effective Covid antiviral pill was greeted with much fanfare, and Molnupiravir, aka Lagevrio, was soon given regulatory approval in the UK.
BUT, if we separate the earlier and later stages of this trial, the numbers for hospitalisation or death in each patient group were:
EARLIER: Placebo 53/ 377 (14%), Drug 28/ 385 (7 %) => risk reduction of 50 %
What are the implications of these figures? Are there confounding factors that have not yet come to light? Is the drug any good? Is this trial dependable?
Perhaps the answers to some of these questions lie in the trial design. The patient exclusion criteria don't mention either vaccination status or parallel treatments. In theory, therefore, participants in the Molnupiravir trial could already have accessed another antiviral or even a monoclonal antibody within the 5 day cutoff period.
Note that in both stages of this trial the drug performed much the same, while "placebo" magically improved in the later stage.
I'm just tossing out ideas here, but the figures look very suspect to me. Help!
Written by
bennevisplace
To view profiles and participate in discussions please or .
I know for a fact that if you write to the trial team with your observation, they mostly reply to you (are required to clarify).
The change being at the placebo end certainly points to either a design flaw or parallel treatment differences between the two cohorts.
The fact that only 6-7% of treated were hospitalized is still a good sign.
You might get NHS data about hospitalization rates for general population (that is the defacto placebo arm rate!). If that is also 6-7%, then effectively this medication too is a wash.
But publishing such crappy data just shows bad science or poor design.
Thanks for your reply sujitsur, what you say makes sense to me.
Comparing study hospitalisation rates with NHS stats is problematic, e.g. the trial enrolled only people with one or more medical risk factor from populations with different healthcare systems and no doubt different hospital admission criteria.
I wonder what the FDA will make of this on Tuesday and whether the truth of the matter will emerge.
Molnupiravir in COVID-19: A systematic review of literature
"A systematic literature search was conducted in the electronic database of PubMed, MedRxiv and Google Scholar from inception until October 15, 2021, using MeSH keyword “COVID-19”, “SARS-CoV-2”, AND “Molnupiravir”, “EIDD-2801”, “MK-4482”."
This is great, thanks Seymour. I'm especially looking forward to perusing the FDA briefing document. First I need to recharge my batteries with a long nap - I'm fighting a winter virus right now (PCR negative for C19).
I hope you feel better soon! It's sad that we don't have better general virus diagnostics yet. Viruses are the dark matter of biology.
Seems like every year around Thanksgiving (late November), I get simultaneous disturbances - sinus/ears and gut. No fever this year, though.
It's not that I'm getting out and participating in the urban biome, nor does my immediate family get out much. My wife does the shopping, and also teaches 8th grade. She's well masked, and OCD about hand sanitizer.
I'm beginning to think it's seasonal stress plus seasonal changes in the local fungus to bacteria balance.
Thanks, much better for a zzz and a walk, though this virus is reluctant to fizzle out, it's already made more come-backs than Tiger Woods.
On the FDA website too, a series of slides giving the main points the briefing note fda.gov/media/154473/download - and crucially it does include the later analysis for all 1433 particpants. All the same, the slides make no issue of the anomolous out-performance of placebo in the later group. I have now drawn this to their attention.
The conclusions in slide form are:
MOV reduces the risk of hospitalization or death among adults with mild to
moderate COVID-19 and who are at high-risk for progression to severe
COVID-19
MOV appeared generally safe in adults with mild to moderate COVID-19
Several safety issues were identified based on nonclinical findings
Well, one thing's for sure. This briefing document only takes into account the interim results of the MK-4482-002 Part 2) study with 775 participants, giving a risk reduction of 52%.
It doesn't take into account the later ~650 participants in whom taking M apparently increased the risk of hospitalisation!
IS THERE AN UPDATE TO THIS DOCUMENT?
This briefing also doesn't discuss the potentially confounding issue of parallel/ earlier treatments; and seropositivity appears to have been established by a test for total antibodies, meaning the FDA was unable to assess the drug's benefit to vaccinated versus unvaccinated patients.
While the FDA believes the mutagenicity risk is generally low (though M is likely to be contraindicated in pregnancy), it reached no conclusion about M's ability to drive mutations in the virus.
Addendum to FDA Briefing Document, Antimicrobial Drugs Advisory Committee Meeting, November 30, 2021
Compare Figure 1 in the addendum to Figure 1 in the original FDA brief.
Figure 1 shows only the continent info. The original brief mentions the proportional breakdown and countries:
Only 5% of participants were enrolled from sites in North America; the majority (609/775) of participants were enrolled from sites in Latin America (56%) and Europe (23%). Among 179 participants enrolled from sites in Europe, 72% were enrolled from sites in Russia, 17% from Ukraine, and 11% from countries in Western Europe
Comparing continental info between the 2, we see varying numbers of case differences (pardon the proportional font - best to copy and past to Excel, I think):
A B C D E
37 87 235% 5% 6% North America
421 650 154% 56% 46% Latin America
176 462 263% 23% 33% Europe
--- 35 0% 2% Asia Pacific
117 174 149% 16% 12% Africa
A - original brief
B - addendum
C - % increase
D - brief % of total
E - addendum % of total
So the new cases had dramatically fewer placebo fatalities in the updated analysis. The MK-4482-002 start date was October 19, 2020 - back in the olden pandemic days. Treatment protocols have changed a lot since then. Fewer Latin American and African cases, more European and Asian cases. Studies like these are not good with moving targets of variants and changes in best practices.
I think the whole 30% vs 50% effectiveness issue is like the Hitchhiker's Guide to the Galaxy - harmless vs. mostly harmless. It's way oversummarized. It's like answering "what's the risk of a flood" after noting that 3/5 ths of the planet is water. It's just not helpful. We want to know what our elevation is in relation to water. We want to know how deep the flood might be.
The antivaxxers have already latched onto the difference in the results to spin fantasies of skulduggery - not that there is no skulduggery in international pharmaceuticals. But these numbers are not the evidence we would need to make a case.
The issue for us with CLL/SLL is always that people with hematological cancers, and hypogammaglobulinemia and immune dysregulation are not included in the numbers at all. So we cannot extrapolate anything from these results at all. Maybe it's a boon. Maybe it's a risk. If one is watch and wait, and has good IgA, AgG, and IgM, is not suffering infections or other comorbidities, one may find something in the tables that applies. If one is in treatement, all bets are off.
Thanks, looks like our latest replies "crossed in the post".
Good work.
The regional make-up of study participants did change from early to late, but not dramatically. I'm not sure this explains why the sudden success of placebo, while the drug performed about the same (edit: in both stages).
The earlier stage ended August 5th and the later stage began August 6th 2021.
"harmless vs. mostly harmless" I like that. Does it justify the UK government spending £250 million of taxpayers' money on this drug?
I think we'd have to see the raw data for 8/6/2021 – 10/2/2021 to see where the placebo skew was.
I'm concerned to hear about the bone marrow effects in dogs. I haven't found a paper that mentions it yet. The slide deck mentions dogs in a 28 day dosage, and unclear what dosage level. FDA is recommending 5 days of the drug to minimize risk. I feel a bit queezy about it.
The U.S. govenrment bought even more doses. I think it was back in June, 2021, when Delta was getting serious.
When governments panic, they write checks. When I panic, I can barely write at all! Either way, when we panic, we aren't coherent. This whole pandemic has raised many opportunities to gamble.
One other thought about Molnupiravir - it induces fatal mutations in single strand RNA viruses. So it's probably at least somewhat effective against other ssRNA viruses, such as:
hantaviruses
ebola
marburg
measles
mumps
respiratory synctial virus
rabies
So, I think we can expect to hear more research on Molnupiravir beyond SARS-CoV-2, even if Pfizer's Paxlovid or something else works better for SARS-CoV-2.
Content on HealthUnlocked does not replace the relationship between you and doctors or other healthcare professionals nor the advice you receive from them.
Never delay seeking advice or dialling emergency services because of something that you have read on HealthUnlocked.
Info for AussieNeil
Important SARS-CoV-2 and Covid-19 information for Chronic Lymphocytic Leukaemia/CLL patients
COVID-19 among fit patients with CLL treated with venetoclax-based combinations
