Moderna VRBPAC Data Summary — Deplatform Disease

The short version: Moderna’s phase 3 clinical trial data are available. They show that the vaccine has excellent efficacy at preventing COVID-19, and in particular, preventing severe COVID-19. Serious adverse events were rare in the trial, and in general, are probably not related to the vaccine (but safety monitoring should and will continue). The vaccine itself is very similar to the one made by Pfizer/BioNTech, but is much easier to store because it’s stable at higher temperatures. In general though, this vaccine provokes a strong immune response which can be uncomfortable (patients often experienced pain at the injection site, headache, fatigue, and joint pains; some nausea was also reported and some people did develop temporary swelling of their lymph nodes aka “glands” in their armpits), especially after the second dose, so if possible, I would aim to get it, especially the second dose, right before a period you can rest e.g. a Friday so you have the weekend to lie around.

Firstly, if you prefer to listen instead of reading, Professor F. Perry Wilson MD MSCE has already done a fantastic job going through this data set:

Moderna has now joined Pfizer about 1 week later for consideration for EUA, which is very exciting. VRPAC documents can be found here (FDA briefing document), here (Moderna’s detailed dataset) and there’s an interesting addendum here.

Firstly though, there are a few differences between Moderna’s and Pfizer’s vaccine:

I think the addendum is the most interesting part, so I want to talk about it first.

Protection against Asymptomatic Cases

The addendum deals with the question of whether a single dose of vaccine can prevent asymptomatic infection. In general, vaccines do not prevent infection ( sterilizing immunity), but indirect vaccine protection (herd immunity) still occurs with many vaccines even though this is the case (e.g. measles vaccines are 97–99% effective after 2 doses; but asymptomatic measles infection does occur even though community protection from the vaccine are substantial). Moderna presents the following data as evidence:

Based on these findings it is conceivable that a single dose of vaccine may protect against asymptomatic infection. Maybe. I am not entirely convinced. It’s true that nearly 3 times as many people in the placebo group tested positive for SARS-CoV-2 on PCR (which is, despite everything you may have heard, the gold standard diagnostic tool for COVID-19 diagnosis), which is encouraging and regardless of any statistical analysis, is a fairly large effect (which makes me inclined to think it is real; I thought about trying to see what the p-value would be for this result but honestly there’s not enough information here to reliably inform which statistical test is appropriate). To remind everyone, this trial is double-blinded i.e. when these tests were being done, no one knew in advance which group received the vaccine and which group received placebo, meaning that these results cannot realistically be the part of systematic tampering to coax the desired result. However, there is still much that remains unknown about the value of a single dose of Moderna’s vaccine, just like Pfizer’s, so for the same reasons as I want one for Pfizer’s vaccine, I would like to see a clinical trial initiated of a single-dose regimen of the vaccine.

I do have to say though, when I put on my immunology hat, I become a bit skeptical about these findings (but I should be clear that this doesn’t diminish my optimism for the vaccine). In general, for a pathogen that invades through the upper respiratory tract, to achieve sterilizing immunity (as is being tentatively suggested by the findings here) requires substantial levels of sIgA (secretory IgA) in the upper respiratory tract. This is very challenging for vaccines to do, and in general requires mucosal delivery of the vaccine (i.e. the vaccine would need to be given intranasally), and generally the vaccine has to be a live attenuated one (because it’s hard to make mucosal adjuvants that are both safe and effective). IgA that is present in the plasma does not appear in secretions (see the letter responses), hence it was not measured in Moderna’s early studies to assess vaccine immunogenicity. I do think it is possible that if this vaccine were given intranasally, you could achieve good levels of protection in the upper respiratory tract, but the issue here is that it would be less likely that you could get good protection elsewhere (and prior pharmacokinetic studies of Moderna’s lipid nanoparticles shows that the IM injection is highly local and for the most part the LNPs remain heavily concentrated in muscle rather than distributing extensively e.g. into the upper respiratory tract; I would imagine that Pfizer’s LNPs are similar). mRNA vaccines like Moderna’s are very similar in terms of the immune response they induce to live attenuated viral vaccines (but have the benefit of being safe to give to immunocompromised people because there is no living pathogen within them). For a parenterally administered vaccine (given by injection), it is almost certain that you could not attain high levels of sIgA at the upper respiratory tract. Moderna’s non-human primate study does note that there is an increase in anti-spike protein IgA from the vaccine detected in bronchoalveolar lavage (BAL; basically a washout of the lung) after challenge with SARS-CoV-2, and the titer positively trends with the dose of the vaccine. However, in this same study, after challenging the primates with SARS-CoV-2, sterilizing immunity is not achieved:

Two days after challenge, only one of eight animals in each of the vaccine dose groups had detectable subgenomic RNA in BAL fluid, as compared with eight of eight animals in the control group (Figure 3A). By day 2, none of the eight animals in the 100-μg dose group had detectable subgenomic RNA detected in nasal swab specimens, as compared with five of eight animals in the 10-μg dose group and six of eight animals in the control group (Figure 3B).

Now, it should be noted that the dose of virus that animals receive in these studies tends to be larger and more exaggerated than what would be encountered in “real life” because of the need to reliably cause infection and disease, and correlates of protection, like neutralizing antibodies, can become overwhelmed- if too many viral particles are present, there will not be sufficiently many antibodies to neutralize them. That’s standard for trials of this nature, but it may not correspond well to real-world conditions. Additionally, while we do think of IgA as being the principal protector at the upper respiratory mucosa (note that this is not true of ALL mucosal tissues), IgA is not the only antibody type that is present in the upper respiratory tract, and other types may also contribute to protection, but the proposition that sterilizing immunity can be achieved by these other types of antibodies seems dubious to me simply because sIgA vastly outnumbers all of them combined.

Basically: I want more data on this point. I don’t know whether or not this vaccine grants sterilizing immunity, and to be frank, that doesn’t matter much to me. It would be nice- but you can still have substantial reductions in transmission even if you don’t have sterilizing immunity as shown by basically every single vaccine in routine use with a few exceptions e.g. HPV (which does grant sterilizing immunity). I will reiterate once more that I think that for a vaccine with efficacy this high, there is likely a substantial indirect vaccine effect, but data are needed to substantiate that, so for now, you should get both doses and continue to wear your mask and continue to take appropriate precautions for public health .

The Kaplan-Meier curve to the left shows the difference in the incidence of COVID-19 in the vaccine and placebo groups. I do not think anyone can make an argument in good faith that this vaccine does not prevent COVID-19. As defined in this study, the primary efficacy endpoint was the ability of the vaccine to prevent COVID-19 in individuals who received both doses of the vaccine 14 days or more after developing it (because it takes some time to generate a protective immune response- which is again a reminder that you still have to take precautions after you get your vaccine). Right now, the evidence suggests that people should STILL get both doses of this vaccine. The efficacy across all participants for the vaccine was measured to be 94.1% with a 95%-confidence interval of (89.3%-96.8%), which is excellent. This corresponds to individuals who complete both doses having a risk of developing COVID-19 that is about 20 times less than those who did not receive a vaccine. Unsurprisingly, vaccine efficacy seemed to be lower in the >65 years old group, which is probably related to the immunological changes of advanced age, measured to be 86.4% (61.4%,95.5%); the confidence interval is wider because there were fewer total cases in the > 65 years of age group. This corresponds to a ~ 7.4 times reduced likelihood of contracting COVID-19 in this age group for those receiving both doses compared with those who received none. However, because the confidence intervals overlap for efficacy in those aged 18 to 65 and those older than 65, it is possible that there is not a difference in efficacy across age groups (though I do think it is likely that there is).

Interestingly, the Moderna data also show a vaccine efficacy of 80.2% (55.2%,92.5%) after the first dose, which is substantial. Notably, it is not clear whether or not the vaccine has any efficacy for the first 14 days after the first dose, as the confidence interval crosses 0%. This is not unusual as it takes time for an immune response to mature to the point of getting protective, and it takes time to encounter the pathogens. If you look strictly at the data for >14 days after, there is 92.1% efficacy (68.8%, 99.1%) of the vaccine which is substantial, and also makes me more inclined to believe that Pfizer’s data regarding the protection from one dose are legitimate. But again, for now, Pfizer’s vaccine should be given as 2 doses until data can confirm the efficacy of a single dose. I imagine that Moderna will also not attempt to push for a single dose version of their vaccine on EUA.

However, perhaps the most important point: this vaccine demonstrated 100% efficacy at preventing severe COVID-19, which in this trial was defined as having a diagnosis of COVID-19 and one or more of the following:

It’s important to be explicit about these definitions, especially when they aren’t standardized, and I think most people would agree that the definition of a severe case of COVID-19 here is appropriate. I think that it is abundantly clear that this is a vaccine that can save your life. I also do feel obligated to make the point once more: it is generally more difficult for a vaccine to prevent mild disease than it is for it to prevent severe disease.

Widge AT, Rouphael NG, Jackson LA, Anderson EJ, Roberts PC, Makhene M, Chappell JD, Denison MR, Stevens LJ, Pruijssers AJ, et al. 2020. Durability of responses after SARS-CoV-2 mRNA-1273 vaccination. N Engl J Med.(NEJMc2032195). doi:10.1056/NEJMc2032195.

What is less clear is how long this vaccine retains efficacy. Moderna has published a brief update in which they examined antibody levels in the subjects of their Phase 1 which you can view. In general, antibody responses seem to be stable for about 120 days after the first injection and seem to retain the ability to neutralize spike RBD, which is an excellent sign. I do think it is likely that this vaccine, and Pfizer’s (as they are very similar) will confer protection for at least several years- but this is just a guess on my part. Immunology Professor Deepta Bhattacharya has a great thread on Twitter in which he discusses why he feels protection for at least 1 year is likely. I would like to caution people however against simply examining antibody levels to determine whether or not protection has persisted from a given vaccine, however (though in the event that high levels of long-lived plasma cells are induced, antibody levels can remain persistently high and this would likely correlate to protection). Firstly, we should recall that the panicked reports of rapid waning antibodies that appeared earlier in the pandemic do not correlate well with the very rare instances of reinfection, some of which have even been asymptomatic. There is also the case of hepatitis B vaccines, which demonstrate good protection even after antibodies seem to wane. I also do want to express caution about using the presence or absence of memory cells as a marker of protection. Even though memory T cells specific to SARS-CoV are present 17 years after initial infection, the only way to know if they are actually sufficient to protect is to challenge these people with SARS-CoV (which for ethical reasons we obviously should not do). I think the most rigorous and proper way to determine how long protection lasts is honestly to wait until people who received both doses of the vaccine begin to develop infections again.

Additionally Moderna’s vaccine trial seemed to do a very good job of ensuring that their subject pool was diverse. The distribution of sex was nearly 1:1 for males to females and ages included in the efficacy analysis ranged from 18 to 95 years. Like Pfizer’s trial there does appear to be some under-representation of black people in the data at 9.7% (they account for 13.6 % of the US’s population per the most recent census) and also some under-representation of Asian people at 4.9% (they account for 5.9% of the US population per the most recent census). There is also some slight over-representation of white people in this trial at 79.4% (they account for 76.3% of the US population per the most recent census). Overall though this trial does a fair job of approximating the racial breakdown of the US. In addition, this trial does an excellent job of including healthcare workers who are at higher risk of contracting COVID-19 and account for 25.4% of the study population.

Vaccine Adverse Events

This vaccine is reactogenic- there’s no way around that. It was very common for the vaccine to cause injection site pain (91.6%), fatigue (68.5%), headache (63.0%), muscle pain (59.6%), joint pain (44.8%), and chills (43.4%); severe adverse reactions (i.e. those that are disruptive to daily life) reactions occurred in 0.2% to 9.7% of participants, were more frequent after dose 2 than after dose 1, and were generally less frequent in participants ≥65 years of age as compared to younger participants. There were also some cases of lymphadenopathy (swelling of the lymph nodes) in the armpit, which is normal after a vaccine. This happens because there is rapid division of cells of the immune system and is temporary. In general the individuals older than 65 were LESS likely to have these adverse effects suggesting that they are directly related to activation of the immune system and thus not concerning. You should expect to have discomfort after this vaccine, just like Pfizer’s and I would strongly consider getting the second dose on a Friday or some other day where you can rest in the next few days.

In the trial, there were 3 cases of Bell’s palsy in the vaccine group and 1 in the placebo group. This is not frequent enough to support that this vaccine causes Bell’s palsy. In general, routinely given vaccines are not able to cause Bell’s palsy, and I don’t see why these vaccines should be an exception, but monitoring should continue.

Additionally, there were slightly more hypersensitivity reactions (i.e. allergies) in the vaccine group vs. the placebo group (1.5% vs. 1.1%) but this is not statistically significant. There were NO cases of anaphylaxis or other severe allergic reactions associated with the vaccine. Again, monitoring should continue.

Serious adverse events (i.e. those that would require prolonged hospital stay or could be fatal) were rare and incidence did not differ among the vaccine and placebo groups. There were 2 instances of facial swelling in the vaccine group, which resolved, and one case of intractable nausea and vomiting, which also resolved. I do want to make special mention of two of them.

On the matter of these SAEs, the FDA writes:

In FDA’s opinion following review of the narratives, 3 SAEs are considered likely related, including the one report of intractable nausea/vomiting and 2 reports of facial swelling. The possibility that the vaccine contributed to the SAE reports of rheumatoid arthritis, peripheral edema/dyspnea with exertion, and autonomic dysfunction cannot be excluded. The vaccine was unlikely to have contributed to the other SAEs assessed by the investigator as related. As described in detail in a previous section, there was one report of Bell’s palsy in the vaccine arm which occurred 32 days after vaccination; both the investigator and the Sponsor assessed this event as unrelated to the study vaccine, but in FDA’s assessment a causal relationship cannot be definitively excluded.

Overall, it looks very similar to Pfizer/BioNTech’s vaccine: safe, effective, and unfortunately, reactogenic.

Vaccine Composition

Per the FDA briefing document:

The Moderna COVID-19 Vaccine is a white to off-white, sterile, preservative-free frozen suspension for intramuscular injection. The vaccine contains a synthetic messenger ribonucleic acid (mRNA) encoding the pre-fusion stabilized spike glycoprotein (S) of SARS-CoV-2 virus. The vaccine also contains the following ingredients: lipids (SM-102, 1,2-dimyristoyl-rac-glycero3-methoxypolyethylene glycol-2000 [PEG2000-DMG], cholesterol, and 1,2-distearoyl-snglycero-3-phosphocholine [DSPC]), tromethamine, tromethamine hydrochloride, acetic acid, sodium acetate, and sucrose.

So in essence, the composition is very similar to Pfizer/BioNTech’s.


I write about vaccines here. You can find me on Twitter @enirenberg and at (where I publish the same content without a paywall)