Issue 18: From Mice to Monkeys: Evaluating COVID-19 Vaccine Candidates

Sep 16, 2020

By Robert Seder, M.D.

Chief, Cellular Immunology Section, Vaccine Research Center, U.S. National Institute of Allergy and Infectious Diseases


Robert Seder’s research is focused on understanding both innate and adaptive immunity induced by vaccines in various animal models and in humans. He is currently involved in evaluating the mRNA-1273 vaccine candidate being developed by the U.S. National Institute of Allergy and Infectious Diseases and Moderna, which is currently in Phase III human clinical trials. In July, Seder and colleagues reported data in The New England Journal of Medicine showing that this candidate induced rapid neutralization of SARS-CoV-2 in the upper and lower airways of non-human primates, a promising sign, according to Seder. HVP Editor Kristen Jill Abboud recently discussed these results with Seder, as well as the advantages and disadvantages of different animal models for evaluating SARS-CoV-2 infection.

An edited version of the conversation appears below.

What are the advantages of the different animal models for COVID-19?

I’ve only worked with the primate model, which has several advantages. One is that it is useful for clinical translation because you can use the same doses in monkeys that you give to humans, which is not the case with rodents. For the RNA vaccine candidate, we’re testing two 100 microgram doses in humans and in monkeys, whereas mice are protected with only one microgram dose. 

Primates are also more often reflective of the immunogenicity you’d see in humans, and therefore the primate model is likely to be more predictive, especially related to T-cell responses. You can measure antibodies in hamsters, but you can’t measure T cells so well.

The other question, however, is which animal model mimics human COVID-19 disease. It appears that primates have a fairly mild disease that usually resolves around 10-14 days after infection. They don’t get very sick from COVID. If you give hamsters more virus, they will lose weight and show pathology. In mice, it depends on what model you’re using, but they can show pathology. There are many different ways to do it, but I think, in general, people like to see the primate data—both for immunogenicity and protection—to give an idea of what will happen in humans. That is why I’ve focused on the primate model.

One of the many perplexing things about SARS-CoV-2 infection is the variability in pathology. Is that something you can model in non-human primates or any other animal?

You can give animals different amounts of virus to model a heavy infection versus a more modest infection. But with SARS-CoV-2, even at the highest dose, we’re really not seeing much pathology in monkeys beyond 10 days. There have been studies testing whether older monkeys may have a worse disease than younger monkeys. We’re studying that now in terms of modeling COVID-19 pathogenesis.

How old is an “older” monkey?

Probably more than 12 years old. 

Data from several non-human primate studies have been reported for different COVID vaccine candidates. Are these studies comparable? Can you draw any conclusions about which candidate might work better or are there too many variables?

There are many variables and it’s really hard to compare across everybody’s studies. The variables are the number of animals, the ages of the animals, the infectiousness of the challenge virus, the dose of the virus, and the route in which you are administering the virus. Most people are using a Washington virus isolate, so presumably, we’re using a comparable challenge virus, but that’s not even clear. Depending on how many passages and how it was made, there could be differences in the infectivity of the virus. And nobody knows what amount of virus is transmitted in humans, so the right challenge dose is also an open question.

Irrespective of all of that, people then have different ways of measuring immune responses. If you’re looking at antibody responses, most people do standard ELISA assays, but then there are functional neutralization assays, including pseudo-neutralization assays, which are variable across labs, and then there are at least three different viral neutralization assays. Unless you’re sending the serum to the same lab for the same assay, which we’ve been advocating for within Operation Warp Speed, it is difficult to compare everyone’s studies. Only a standardized assay would give you the best idea of the relative potency of the vaccine candidates.

What conclusions can you draw based on the non-human primate data that is available so far?   

In all of the different monkey challenge models, it looks like seven different vaccines all seem to be highly protective against lower airway infection, regardless of dose. It seems to be pretty easy to limit viral replication in the lower airway. We’ve found that the vaccines can really be segregated on whether or not they can control infection in the upper airway. We feel that limiting viral replication in the nasal passages is really the higher bar and would be important for potentially limiting virus transmission. Controlling replication in the lower airway would help limit disease or severe disease, but if a vaccine can control viral replication in the upper airway, it might be able to prevent you from passing it to others. We were the first to show rapid control of the virus in both the upper and lower airway, so that was a big deal.

There also seems to be a hierarchy of neutralization. The best candidate so far is probably the Moderna mRNA candidate, then the Novovax candidate, and then the adenovirus candidates. Antibodies are likely to be the correlate and the primary mechanism of protection. T-cells are more nuanced. Monkeys control infection quickly, so it would be very hard to show a T-cell effect. My guess is that you would have to find a model where you challenge well after vaccination and try to show the role of T cells once antibodies are waning. But that’s a large experiment with a lot of animals and it would be expensive to do.

Are you optimistic about the prospects for vaccines?

Based on all the animal data, I think all of us are optimistic. I don’t think this is very complicated—you have one protein, it hasn’t changed that much, antibodies bind to it, and when they do, you prevent infection. The question is: are some vaccines better at blocking the virus in both the upper and lower airway? And, would more potent vaccines be more durable? That’s where things will shake out a little bit. If you rank the vaccines against each other based on these factors, there might be a hierarchy, but my guess is all of them may have some protective effect.

What are the biggest questions you are trying to answer now?

The biggest questions are: how durable will vaccine protection be? If you don’t have sterilizing protection, will infection boost the vaccine response? And what really is the role of T cells in mediating short- or long-term protection? Those are the things I’m interested in knowing.

Interview by Kristen Jill Abboud


Register for the HVP COVID Report: