Based on what I have read, the Pfizer, Moderna, AstraZeneca, and Johnson & Johnson COVID-19 vaccines all work by training the immune system to recognize the spike protein on the surface of the SARS-CoV-2 virus.

I understand that the vaccines deliver the genetic instructions for creating the spike protein in different ways - the Pfizer and Moderna vaccines use mRNA, while the AstraZeneca and Johnson & Johnson vaccines use an adenovirus.

But if the end result is the same - a person's immune system being able to recognize the same spike protein no matter which of the 4 vaccines was administered, then shouldn't the vaccines be equally effective? I assume they are not though - why is that?

(It has been reported in Live Science and The Atlantic that the vaccines each have a different efficacy - 95% for Pfizer, 94% for Moderna, 70% for AstraZeneca, and 66% for Johnson & Johnson - but those efficacy numbers cannot be compared since they are based on trials conducted during different stages of the pandemic with different people in different regions.)

  • The 2 links near the bottom are a good catch. They need to be a bit more prominent in my view and provided in full size in some way rather than in superscript as they can be easily missed on a mobile phone with small letters. Commented Mar 4, 2021 at 9:25
  • 1
    I think the takeaway with the Live Science article is that "All three vaccines were 100% effective at preventing severe disease six weeks after the first dose (for Moderna) or seven weeks after the first dose (for Pfizer and Johnson & Johnson, the latter of which requires only one dose). Zero vaccinated people in any of the trials were hospitalized or died of COVID-19 after the vaccines had fully taken effect." I wonder whether there is something comparable with the AstraZeneca vaccine? Commented Mar 4, 2021 at 9:44
  • Well, surely another reason is they're very different technologies. AstraZeneca and J&J are vector vaccines while Pfizer and Moderna are mRNA vaccines.
    – Carey Gregory
    Commented Mar 5, 2021 at 0:23
  • 2
    The question is sort of like asking "These four carpenters all operate the same way, they use hammers and nails to put up a house. If the end result is a house in all cases, why is it not the case that all the carpenters are equally skilled?" By saying "the end result is the same" you're already assuming that there are no differences between one vaccine and another. But the vaccines can differ in how effectively they train the immune system.
    – BrenBarn
    Commented Mar 6, 2021 at 5:13
  • @CareyGregory Yes I had mentioned that in the question. If the delivery mechanism impacts the immune system response, then an explanation of that would be a good answer. Commented Mar 6, 2021 at 7:58

2 Answers 2


Even assuming the same ecological conditions, efficacy of a vaccine is in a (dose-dependent trade-off with side effects. (This was e.g. easily shown in the phase I/II trials of Moderna's vaccine.) Even for the exact same tech/vaccine, manufacturers may choose different points along this curve.

Furthermore, the amount of "training" and thus resulting levels of antibodies etc. depends on how long the injected vaccine lasts in your body. There are numerous factors that affect this, see e.g. prior discussion here what affects that in mRNA vaccines alone.

Changing the delivery vector to something more substantially different, e.g. from a lipid nanoparticle (LNP) to an adenovirus (Janssen, Astra-Zeneca, Sputnik) can have more substantial effects in both of the above regards; i.e. the side-effect vs efficacy profile may be on different curve; the technology that makes a pure mRNA vaccine persist long enough in your body is not the same as that using an adenovirus--by the way, the latter is a DNA virus, so the protein is coded in a slightly different substrate, with a different "decay rate". You can also have prior immunity to some adenoviruses (and also develop it from the vaccine, I think).

Furthermore, some other Covid-19 vaccines (e.g. Novavax) deliver the proteins themselves (in a LNP) rather then mRNA, so they are even more fundamentally different than mRNA/DNA based ones with respect to all of the above. The same mRNA molecule can get reused a good number times in a cell to produce (spike, in this case) proteins, but a protein is more or less of "one use" for the immune system. Generally speaking, protein vaccines (and this include's Novavax) use various adjuvants to get more immune response out of the same dose of proteins, which are fairly involved to manufacture. (Such adjuvants aren't used by current mRNA vaccines that have been approved, probably because they generate [translate to] more than enough proteins.)

Of the vaccines you've listed, Pfizer/Biontech and Moderna are the most similar to each other technology-wise (both mRNA in a PEGylated LNP--their actual formulation differences are somewhat obscure), so somewhat not surprisingly they produced the most similar results.

Additionally, small changes have been made to the genetic sequence of the spike protein; more specifically

  • Astra-Zeneca & Sputnik don't report any modifications
  • Pfizer and Moderna (both) made the same proline substitutions
  • Janssen additionally made two mutations at furin cleavage sites
  • Novavax made three of those

The role of proline substitutions is to keep the (produced) spike protein in the prefusion conformation for a longer time, thus effectively increasing yield. If you wonder how they both got this idea: it's because this was first done on a MERS vaccine. It was observed that most of the naturally produced antibodies target the RBD, which is only exposed in the prefusion conformation.

According to Janssen's pre-clinical research, their furin mutations have an additional stabilizing effect and cause the host to produce a greater ratio of neutralizing to non-neutralizing antibodies (compared to the variant with the two proline substitutions only). Novavax basiscally did the same thing but altered the cleavage site at three points RRAR→QQAQ.

  • Is it actually the case that the spike proteins as delivered by the different vaccines are completely identical?
    – BrenBarn
    Commented Mar 9, 2021 at 4:44
  • @BrenBarn: from what we know about Pfizer and Modena (from publications) it's the same in the coding region (same proline substitutions), although Moderna unlike Pfizer has not published their full sequence (particularly for the non-conding regions), as far as I know. Novavax and Janssen have made some more changes to the coding region itself. These affect protein folding and stability, so the likelihood you'll get (relevant) antibodies. But I don't know how these translate to vaccine efficacy numbers/changes. Commented Mar 9, 2021 at 9:57
  • To find that out you'd have to compare the same tech of vaccine, with and without the coding region changes, which no mfg has done (or at least published), as far I know. BioNTech/Pfizer actually did publish data on both a full spike protein vaccine and just a RBD one. Eventually the former was selected for approval. Commented Mar 9, 2021 at 10:00
  • Also AstraZeneca and Gamaleya's (Sputnik) vaccines don't use any proline substitutions, it seems. Commented Mar 9, 2021 at 10:06
  • I'm just wondering because it seems the most obvious response to the question would be "the reason they are not equally effective is that the spike proteins they deliver are not exactly identical so the end result of the 'training' is not necessarily the same" if indeed that is the case. Your answer doesn't say that so I was thinking maybe you had reason to believe they are identical.
    – BrenBarn
    Commented Mar 9, 2021 at 19:51

As Fizz noted in a comment, from what we know about the vaccines, the Novavax and Janssen vaccines make slightly different modifications to the spike protein than the Moderna and Pfizer vaccines. In other words, they are not delivering the exact same instructions for recognizing the virus.

We don't know if these small differences are what causes the difference in efficacy. But your assumption that the "end result" is the same is not necessarily valid. There may be differences in the nature or degree of the immune system's ability to recognize the virus because, in a manner of speaking, the different vaccines give slightly different instructions for doing so.

It's also worth noting that the different vaccines were tested at different times and in different places. It's possible that differences in their measured efficacy reflect different COVID-19 variants that were prevalent in the populations where the trials took place.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.