Paper demonstrating Omicron's immune escape vs. vaccination and previous infection

Biological researchers in Switzerland released a preprint in late December providing very useful information about the immune escape properties of Omicron vs. previous infection and/or vaccination with respect to neutralizing antibodies titers.

In this blog post, I summarize the key results.

They looked at neutralizing antibody titers (Nabs) for convalescent plasma from people previously infected with Delta, Gamma, Beta, Alpha, and from 2020 "wild type" versions of the virus (non-VOC) vs. each variant plus Omicron (and Zeta).

This table plots ratio of fold-reduction in Nabs relative to the matching variant.

To explain the table, the columns indicate the variants of the convalescent plasma, i.e. the variant of the previous infection, while the rows indicate the variant to which the convalescent plasma was exposed, i.e. the new exposure. The Nabs titer is computed from the laboratory study, and for each column the ratio of the titers for each "mismatched variant" (i.e. the exposed variant differs from the previous infected variant) with the "matched variant" (i.e. the exposed variant is the same as the previous infected variant).

For example, for someone infected with the 2020 pre-VOC variant, their Nabs vs. Alpha were only 2.2x lower, not bad, and for Delta 3.1x lower, also not bad, which should contribute to strong protection vs. reinfection. However, vs. Omicron, the Nabs levels are 45.6x lower, so the previous infection has very little immune protection vs. reinfection based on the Nabs levels.

For someone more recently infected with Delta, if they were cross-exposed to the original wild type (B.1) or Alpha, their Nabs levels would be 2.9x or 4.0x lower against the non-matching variant. However, if exposed to Omicron, their Nabs levels would be 23.8x lower, again not providing much protection vs. re-infection.

We see some reduction of Nabs for non-matching variants, but the reduction was quite mild, <10-fold pre-Omicron, which is strong enough to protect vs. reinfection. One exception is that previous infection with Gamma was not very protective vs. Delta, which may explain how countries in South America who were hit hard by Gamma still experienced such a dramatic Delta wave. But previous infection is not very protective vs. Omicron at all, with 23-56x reduction in Nabs depending on the variant with which you were infected. This type of reduction of Nabs is enough to make your protection vs. reinfection very weak. Omicron does not just have immune escape vs. previous infection, but also vs. 2-dose vaccination.

The left column of the table shows reduction in Nabs for 2-dose vaccines vs. different variants relative to the original 2020 wild type virus (which was in the phase 3 vaccine studies).

Incidentally, Switzerland used 63% Moderna, 36% Pfizer, and 0.4% Johnson & Johnson

We see mild reduction in Nabs vs. all pre-Omicron variants of 3-7x, but that is not enough to substantially reduce vaccine effectiveness (VE)

However, with Omicron, we see 86x reduction in Nabs, enough to reduce VE down near zero.

A quick look at these numbers might make people think vaccines are less protective than previous infection vs. Omicron, but remember these are relative levels, and as has been well documented vaccines produce much higher Nabs than previous infection to begin with.

Note here the absolute levels of Nabs titers, between 100 and 1000 for 2020 wild type variants, and ~100 for the pre-Omicron variants in 2021, with all having measurable Nabs levels above 1. For Omicron, 5/16 did not have measurable Nabs at all, and of the 11/16 who did their titers were centered ~10.

Here are the actual Nabs levels of previous infection with wild type 2020 virus (pre-VOC).

We see that the Nabs levels vs. wild type 2020 (pre-VOC) are between 10 and 100, with all samples showing measurable antibodies. For non-Omicron (and non-Zeta) variants, we see ~10% of samples did not have measurable Nabs, but of the 90% that did, most had titers between 10 and 100. For Omicron, however, 29/34 of the samples had no measurable Nabs levels, and for the 5/34 who did, their levels were ~10, similar to what was seen for vaccination.

When we look at convalescent plasma from those previously infected with Alpha, we see 10/12 Omicron showed no evidence of measurable Nabs, and for the 2/12 with Nabs the levels were ~10.

When we look at those previously infected with Delta, the results are a little bit better, with only 4/10 exposed to Omicron having no measurable Nabs, and out of the 6/10 who had Nabs evident we see levels also ~10.

Thus, with respect to Nabs, the immune escape from previous infection is just as greater or greater than it is for those 2x vaccinated. Greater than 2/3 of the vaccinated samples demonstrated Nabs, with titer levels ~10, while only between 15% and 40% of those previously infected demonstrated measurable Nabs, and their titer levels were also ~10.

If we look at people with "hybrid immunity" from previous infection plus vaccination, we see stronger immune protection remaining vs. Omicron, with all samples showing measurable Nabs, and the titer levels between 10 and 100. This is reduced relative to previous variants, but still stronger than vaccination or previous infection alone.

So we can clearly see that the immune escape properties of Omicron are much higher than anything we've seen in any previous variant, and this immune escape does not just affect immune protection from vaccination but also previous infection just as much or more. From these samples, we see with respect to Nabs, those 2x vaccinated have stronger immune protection than those previously infected, but in both cases the level of protection is greatly reduced relative to previous variants.

This is why we see Omicron spreading out of control wherever it goes -- the population immunity built up from vaccination and previous infection doesn't do much to prevent Omicron, and thus nearly all in the population are susceptible to infection if exposed.

Note that this study only focuses on Nabs, which is only one element of the immune system, and other elements, including T-cells, also play a role in the immune response, which is one of its limitations. It does not contain epidemiological data estimating re-infection or breakthrough infection rates for Omicron.

However, Nabs have been shown to be strongly predictive of protection vs. infection, e.g. in this excellent Nature Medicine paper from Spring 2021, so are a useful laboratory surrogate for assessing immune escape of variants against particular vaccines or previous infection.

Those with "hybrid immunity" who were both vaccinated and previously infected are more strongly protected, but still not nearly as much as for previous variants. Also, although this paper did not look at those 3x vaccinated (i.e. boosted), other studies have shown that this group has much higher Nabs than 2x vaccinated, and levels similar or better than what is seen in those with "hybrid immunity." It is too bad this study did not have data on this group to make direct comparisons.

So we need more information on other elements the immune systems to fully characterize the immune escape properties of Omicron -- but some studies have already been done, with one study here demonstrating T-cell responses induced by both previous infections and vaccines are likely robust,

There are also other papers on immunology of Omicron, with this Nature paper and this second Nature paper showing also that neutralizing antibodies were reduced in previously infected and vaccinated, but that those 3x or 4x vaccinated (boosted) maintained higher levels, and the latter paper also showing that antibody binding to the receptor binding domain (RBD) and N=terminal domain (NTD) were reduced substantially in previously infected but not vaccinated individuals.

Undoubtedly, more will follow in the near future, providing more validation of these insights as well as deeper insights into other aspects of the immune escape properties of Omicron.

Another limitation of this study is that it did not account for any waning, measuring whether time since vaccination or time since infection alters the Nabs levels -- based on what we've seen for other variants, this is very likely a strong effect, and those infected or vaccinated more recently will tend to have higher Nabs levels vs. Omicron than those infected or vaccinated long ago.

Also, a note about severity: While we see Omicron appears to be more mild in terms of proportion of infections leading to hospitalizations and/or deaths, the much higher transmission levels mean that even so the absolute numbers of serious cases requiring hospitalization and/or leading to death can be high. For example, even if the rate of hospitalizations or deaths per infection is 3x lower, if we have 6x more people infected at once then we still have double the number of absolute hospitalizations and deaths.

This means that we should not "whistle by the graveyard" with Omicron, and assume it is "just a cold" with no risk of worst outcomes. For some, they will experience severe disease, and tragically some will die from Omicron-based COVID-19. We certainly should not encourage people to get infected with Omicron to "get immunity" -- this follows the same foolish reasoning driving the Great Barrington Declaration in fall 2020.

We should be vigilant to reduce its transmission, but need not live in fear.

Vigilance, not fear.