26 August 2022 Episode 284 Update on immunity to BA.5 and omicron-adapted boosters

Fri, 08/26/2022 - 20:14

Episode 284 Follow-up on immunity to BA.5 and COVID boosters


Dear colleagues,


The debate on whether/with what/when to boost is all over the scientific and popular media.  People ask my opinion.  Well, I second just planned my second booster with Pfizer, because I think it is still useful to prevent severe infection  and/or long COVID, especially since I have no hybrid immunity.  The invitation is not really clear on whether I will receive the “standard” or “omicron-adapted” (presumably BA.1?) vaccine (see Ep 284-0 highlighted)  


Anyhow, let’s see what the preprint literature tells us today:


  • Par 1 summarizes some recent reviews as a background,
    • Emphasizing the importance of cellular immunity in  protection and pathogenesis.
    • Discussing the possible different roles of “imprinting”


  • In Par 2 contains clinico- and immune-epidemiological evidence on immune escape of BA.5, after previous BA.1/2 (breakthrough) infection and even after 3 or 4 doses of mRNA.   
    • Rapidly waning protection against BA.5 infection by previous BA.1/2 breakthrough infections (i.e. in twice vaccinated subjects (Ep 284-3 Portugal)
    • Unvaccinated or (2dose) vaccinated BA.1/2 subjects have low neutralizing Ab against BA.5 (Ep 284-4  USA)
    • Third Wuhan WA-1 standard mRNA vaccine induces weak and waning neutralizing Ab against BA.5, but hybrid immunity generates better humoral responses against this subvariant  (Ep 284-5 Canada).
    • Fourth dose of the standard mRNA vaccine boosts neut Ab against all variants, but highest (10 X) increase against homologous WA-1 and lowest against BA.5.  Neut against BA.2.75 similar to BA.5.  Clear additional boosting effect of BA.1 breakthrough infection (Ep 284-6 Canada). 


  • In Par 3 the human trial results of bivalent mRNA (Moderna) booster are shown
    • Slightly higher (2X) titers against Omicron BA.1 were observed with Omicron-containing vaccines compared to Prototype (= Wuhan WA-1) mRNA as booster
    • Titers against Omicron BA.4/BA.5 were lower than against BA.1 for all candidate vaccines.


  • In Par 4 Recent data on neutralizing Ab induced by protein (and other vaccine formats):  once again mRNA vaccines are clearly superior.  Novavax is a good alternative for the original D614G variant, but it elicits much weaker neut Ab against all omicron variants, including BA.5.


  • Par 4 presents three opinions on which booster to take when by Heide Ledford, Eric Topol and Juan Lama
    • The advantage of a BA.1 boost (as compared to Wuhan) is a slightly higher induction of neutralizing Ab against omicron BA.1, but the response to BA.5 remains weak (unless hybrid immunity by previous infection is present).  Not clear if that will translate in significant clinical benefit.
    • FDA will make BA.5 specific mRNA booster available in September, based on favorable neutralization data in a limited mouse experiment.  This decision is controversial. 




Par 1 General introduction on cellular + humoral immunity and “imprinting” to SARS-CoV-2  


Two recent reviews, one in a scientific journal (Front Imm) and the second in a popular journal (Wahington Post)


Ep 284-1: Esther Moga argues that the robustness of cellular immunity determines the fate of SARS-CoV-2 Infection




  1. Cross-reactivity, either humoral or cellular, between common cold hCoV and SARS-CoV-2 does not prevent infection but may be associated with less severe COVID-19.


  1. SARS-CoV-2-specific CD4+ Th1 IFN-g-producing cells and CD8+ CTLs cells were associated with reduced disease severity.


  1. T lymphocyte recruitment to infected lung tissues and T lymphocyte apoptosis/necrosis caused by the cytokine storm might be crucial determinants of CD4+ and CD8+ T-cell lymphopenia in severe COVID-19 cases.


  1. Severe /fatal disease presents with excessive hyperactivation of immune function with increased Tregs and Th2 and/or Th17 cell-biased phenotype, leading to T cell exhaustion and subsequently to a state of anergy.


  1. Functional memory B and T cells to SARS-CoV-2 have been detected 12 months after natural infection. SARS-CoV-2-specific T cell memory may be long lasting given that COVID-19 convalescent patients develop SARS-CoV-2-specific T stem cell memory cells (TSCM) that display a non-exhausted phenotype.


  1. The immunogenicity of SARS-CoV-2 vaccines involves the humoral response (number of spike-specific antibodies, neutralizing antibodies, and antibody neutralization capacity) and the cellular response (IFN-g-producing CD4+ and CD8+ T cells). Therefore, a combined analysis of humoral and cellular immunity is necessary for the identification of vaccine responders and the immune protection evolution.





In mild COVID-19:  

  • Early induction of the SARS-CoV-2 specific CD4+ Th1 cell-biased phenotype with IFN-g secretion.
  • SARS-CoV-2-specific CD8+ T cells perform rapid responses, acting as CTLs, secreting cytotoxic granules and high levels of IFN-g.
  • Moreover, activated T follicular helper (Tfh cell in the draining lymph nodes activate naïve B cells that are necessary for the development of long-lived plasma cells and memory B cells.


Cellular immune response in Severe COVID-19.

  • Reduced numbers and functions of dendritic cells (DCs), leading to decreased numbers of CD4+ T cells.
  • In this case, an elevation of Th2 phenotype and/or a dysregulation of the Treg/Th17 cell ratio toward the Th17 phenotype can be seen.
  • Furthermore, decreased numbers of CD8+ T cells with an exhausted phenotype results in reduced CTL functionality
  • T cell-mediated activation of B cells in extrafollicular focus induces their differentiation into short-lived plasma cells.


Ep 284-2: Carolyn Johnson explains how “imprinting” (also called “original antigenic sin”) can be

  • Helpful, because the reactivation of existing memory B cells quickly provides antibodies that at least partly block viral infection, thus “buying time” for generating new antibodies, better adapted to the new variant.
  • Obstructive, if the “old” memory prevents the new more specific memory to arise.
  • Dangerous, if the old memory is not able at all to neutralize the new variant, but rather enhances its infectivity, by infecting additional targets (such as macrophages) that have receptors for the constant Fc part of the immunoglobulins.  This is the concept of “enhancing antibodies, seen in the second Dengue infection with a different serotype from the first.


Now, it seems that imprinting has been rather beneficial up to Delta, but with the successive omicron variants, we are seeing less and less protection by existing antibodies, induced by previous variants.  Provisionally, there are NO indications of “enhancing” effects, probably because SARS-CoV-2 doesn’t grow well in Fc-receptor bearing cells such as macrophages.




Par 2 Are there immune-epidemiological arguments for omicron-specific (BA.1 or BA.5) vaccines


Ep 284-3: Malato medRxiv 17 Aug 2022 Rapid waning of protection by BA.1/BA.2 infection against BA.5 infection


Set-up: individuals infected during the period of BA.1/BA.2 dominance (>90% of sample isolates) divided in successive 15-day intervals and determined the risk of subsequent infection with BA.5 over a fixed period.


Almost universal (98 %) coverage with vaccine → all infections considered as breakthrough


Relative risk compared with people NOT infected with BA.A/BA.2


Main result:  Compared with uninfected people, a previous BA.1/2 infection conferred

  • At 3 months substantial protection against BA.5 re-infection (RR=0.12; 95% CI: 0.11-0.12).
  • At 5 months the protection was reduced by two-fold (RR=0.24; 0.23-0.24).




The table shows that recently BA.1/2 infected (16/2-3/3) had the lowest RR (0.116) for reinfection with BA.5, while previous BA. ½ infection in January had a higher (0.24) RR on BA.5 breakthrough.


Also evident is that infections with previous variants (Wuhan to Delta), which were longer ago carried a higher (0.4-0.5) RR on breakthrough with BA.5.


Authors conclusion: Questionable stability/durability of protection induced by vaccine that would be based on BA.1



Ep 284-4: Lindeman medRxiv 23 August 2022: Comparison of neutralizing antibodies (against live virus) after Delta and Omicron infection


Study between July 2021 and March 2022, but no genotyping.


Delta infected:

  • Unvaccinated: a highly biased neutralizing antibody response towards the infecting Delta strain with slightly  lower responses against the WA1 strain, but with strikingly lower titers against BA.1, BA.2, and BA.5.
  • Vaccinated with Wuhan WA-1: higher and equivalent responses to WA1 and Delta strains, but still very low neutralizing antibody to Omicron strains.


Omicron infected (mostly BA.1 see Suppl Table 1 p. ):

  • Unvaccinated or vaccinated made more balanced responses,  including against BA.5 (but these antibodies still lower than against BA.1 and BA.2) 


Hence: inclusion of BA.5 Spike in vaccine preparation could be beneficial?

 = reasoning “ex absurdo”: no proof provided that BA.5 itself is sufficiently immunogenic.


Ep 284-5: Tauzin medRxiv 4 Aug 2022 Weak and waning neutralization of BA.5 after third mRNA vaccination


Figure 2. Neutralization activity of SARS-CoV-2 Spike variants by plasma from naïve, breakthrough infection and post-infection individuals after the third dose of mRNA vaccine.





(A-E) Neutralization activity was measured by incubating pseudoviruses bearing SARS-CoV-2 S  glycoproteins, with serial dilutions of plasma for 1 h at 37°C before infecting 293T-ACE2 cells. Neutralization half maximal inhibitory serum dilution (ID50) values were determined usingnormalized non-linear regression using GraphPad Prism


(A-C) Plasma samples were  grouped in two different time points (4 weeks and 4 months).

(D-E) Neutralization activity of plasma collected at 4 weeks (D) and 4 months (E) post vaccination were measured.


Naïve, BTI  and PI individuals are represented by red, yellow and black points respectively, undetectable measures are represented as white symbols, and limits of detection are plotted.


Conclusions:  After third (standard) mRNA vaccine

  • In naïve individuals BA.4/5 Spike is markedly less neutralized compared to the D614G 33 and Omicron BA.2, rapidly decline over 4 months and will likely decrease further with future SARS-CoV-2 evolution.


  • Individuals who have been infected before or after vaccination (= BTI) present better humoral responses than SARS-CoV-2 naïve vaccinated individuals, thus indicating that hybrid immunity generates better humoral responses against this subvariant









Ep 284-6: Xuping Xie bioRxiv 8 Aug 2022:  Neutralization of SARS-CoV-2 Omicron sublineages by 4 doses of mRNA vaccine and effect of BA-1 breakthrough infection


Figure 1. Neutralization of Omicron sublineages before and after 4 doses of mRNA vaccine.






Figure 2. Neutralization of Omicron sublineages after 2 or 3 doses of mRNA vaccine and BA.1 infection





  • 4-dose- mRNA vaccine boosts neut Ab, with lowest titers against BA.5;
  • 2-dose-vaccine-plus-BA.1-infection elicits significantly higher GMTs against Omicron sublineages than 4-dose-vaccine;
  • 3-dosevaccine-plus-BA.1-infection elicits slightly higher GMTs (statistically insignificant) than the 2-dose-vaccine-plus-BA.1-infection
  • BA.2.75 similar susceptibility as BA.5 to neutralizing Ab induced by 4 vaccinations






Par 3 Clinical trial data on omicron-specific vaccines


Ep 271-6: Chalkias medRxiv 25 June 2022: Moderna bivalent (Wuhan WA-1/Omicron BA.1) mRNA-1273.214 monovalent Wuhan-mRNA-1273 as a third dose compared:

  1. Higher neutralization against omicron BA.1, but titers lower than against D614G (= First European variant in 2020, immunologically similar to ancestral Wuhan strain).




  1. The bivalent mRNA boost also induces higher neutralization of BA.4/5, but those titers are still lower than against BA.1





Clear effect of “’hybrid immunity” (combination of infection and vaccination) in all cases


Ep 284-7: Branche medRxiv  23 Aug 2022 Preliminary analysis of six different Moderna variant vaccines as second boost in adult volunteers (open phase 2)


Six second booster arms (50μg dose): Prototype Wuhan WA-1 (mRNA-1273), Omicron BA.1+Beta (, Omicron BA.1+Delta, Omicron BA.1 monovalent, and Omicron BA.1+Prototype.





  • Higher Omicron BA.1 titers were observed with Omicron-containing vaccines compared to Prototype
  • Titers against Omicron BA.4/BA.5 were lower than against BA.1 for all candidate vaccines.



Par 3: What about the protein vaccine (Novavax or NVX-COV2373)


Ep 284-8: Hielscher medRxiv 3 Aug 2022  Comparison of humoral and cellular responses after two doses of Novavax protein vaccine versus 2 doses of mRNA 


  1. Lower levels of spike-specific IgG after NVX-COV2373-vaccination with similar neutralizing activity as mRNA-vaccinated individuals.





B = infection-naïve individuals; C = history of infection after the first Novavax dose


  1. Lower levels of spike-specific T cells, lower CTLA-4 expression and lower percentages of polyfunctional CD4 T cells after NVX-COV2373-vaccination.




  1. NVX-CoV2373-induced T cells equally recognize parental SARS-CoV-2 and variants of concern.



Novavax is apparently slightly less potent than mRNA, but it has less local and systemic side effects: no thrombosis or myocarditis.


Ep 284-9: Bowen Science 19 Aug 2022: Interesting comparison between neutralzing activity elicited by various vaccine formats


Fig. 2 Neut after 2 doses


Plasma neutralizing antibody titers elicited by primary COVID-19 vaccination series determined using SARS-CoV-2 S VSV pseudotypes using VeroE6/TMPRSS2 as target cells. 1x infected samples (n=24) were obtained 26-78 days (mean 41 days) post symptom onset, 2x mRNA-1273 samples (n=14) were obtained 6-50 days (mean 13) post second dose, 2x BNT162b2 samples (n=14) were obtained 6-33 days (mean 14) post second dose, 2x NVX-CoV2373 samples (n=10) were obtained 17-168 days (mean 82) post second dose, 1x Ad26.COV2.S samples (n=10) were obtained 9-142 days (mean 79) post first dose, 2x AZD1222 samples (n=16) were obtained approximately 30 days post second dose, 2x Sputnik V samples (n=12) were obtained 60-90 days post second dose, and BBIBP-CorV samples (n=12) were obtained 9-104 days (mean 69) post second dose.

Individual points are representative geometric mean titers from two independent experiments consisting of two replicates each. Bars represent geometric means and error bars represent geometric standard deviations for each group.

Statistical significance between groups of paired data was determined by Wilcoxon rank test and ***p< 0.001, ****p< 0.0001. Patient demographics are shown in table S4. Normalized curves and fits are shown in fig. S5. G614: Wuhan-Hu-1/G614.


As you can read, the timing post vaccination is not perfectly controlled for, but the impression is that:

  • mRNA (Moderna and Pfizer)  slightly better than protein (Novavax)and much better than Adeno (Janssen, Astra-Zeneca and Sputnik) and inactivated (BBIPB)
  • In all cases, the neut against BA.4/5 is very weak


Fig 3 Effect of booster



Plasma neutralizing antibody titers elicited by COVID-19 vaccine boosters determined using SARS-CoV-2 S VSV pseudotypes and VeroE6/TMPRSS2 as target cells. 3x mRNA-1273/BNT162b2 samples (n=13) were donated 13-97 days (mean 30) post third dose, 2x NVX-CoV2373 1x mRNA-1273/BNT162b2/Ad26.COV2.S samples (n=7) were donated 5-20 days (mean 13) post third dose, 1x Ad26.COV2.S 1x Ad26.COV2.S/BNT162b2 samples (n=14) were donated 12-16 days (mean 14) post second dose, 2x AZD1222 1x BNT162b2/mRNA-1273 samples (n=18) were donated 30-123 days (mean 87) post third dose, 2x Sputnik V 1x AZD1222/BNT162b2 samples (n=14) were donated 45-60 days post third dose, and 2x BBIBP-CorV 1x BNT162b2/mRNA-1273 samples (n=18) were donated 29-89 days (mean 50) post third dose.


Again, no perfect matching with regard to time and kind of booster, but  mRNA clearly superior to Novavax especially with regard to neut titers against omicron BA.1 up to BA.5

Par 4: Opinions on which booster to take when?


Ep 284-10: Heidi Ledford Nature Briefing 19 Augustus 2022:  Which COVID boosters to take and when: a guide for the perplexed.


UK bets on bivalent ancestral/BA.1 second booster, US on BA.5 booster

  • BA.1 sequences (in mRNA vaccines) boosts neutralization of Omicron by about twofold, but it’s unclear how much, if any, extra protection against illness this will produce.
  • Human data on BA.5-specific vaccine only expected by September


Is it useful to wait with your booster for omicron-specific vaccines?


John Moore is not expecting much of Omicron specific responses by variant vaccines: The immune system has been trained to remember this (original) variant — and a dose of vaccine, even one with Omicron-specific components, will tend to boost those earlier immunological memories. This sentence clearly refers to “imprinting” (see Ep 284-2)


People should factor in whether they’re at risk of serious disease, their community’s infection rate and how well they can shield themselves from SARS-CoV-2


“There’s so little potential advantage to having an Omicron booster,” says Moore. “Why bother, when you can use the existing booster sooner?


How long between two boosters:


Consensus on at least 4 months: if you take it sooner circulating antibodies might clear the spike protein in the vaccine before the immune response has a chance to be boosted.


Can you get too many boosters?


From an individual standpoint: as long as the boosters are sensibly spaced, there’s really no such thing as “too many” says Moore.

From a public-health standpoint, however, a focus on boosting everyone could shift attention and resources away from the people who most need boosters: those over 50 years old, and people with pre-existing health conditions.


Ep 284-11: Eric Topol 25 August 2022:  The imminent BA.5 booster (in US)?


A critical opinion on the readiness and usefulness of this booster.


  1. The BA.1 booster induces limited neutralization against BA.5




  1. BA.5 is genetically and immunologically quite distinct from BA.1






  1. Pros and cons for BA.5 booster




  1. Only Immunogenicity data of BA.5 booster in mice




Booster with BA.5 specific mRNA in a limited number (n =8) of mice of one strain, who had been given 2 shots of the original Pfizer vaccine previously, resulted in a solid BA.5 neutralizing antibody response which was 2-fold higher for the monovalent (OMI, BA.5 only) booster. That’s encouraging compared with the BA.1 shot, but it’s not the same as human data.  


The extrapolation from limited mice immunogenicity data to approve a next version of a human vaccine is based on the experience with Influenza, but there is no proof that this will also apply to SARS-CoV-2. 


Eric Topol then goes on to discuss the possible drawback of a bivalent vaccine: the “ancestral component” is likely to induce the wrong (already escaped from) antibodies (see “imprinting”). 


Up to Delta there was a very robust and prolonged protection by vaccination against severe disease (even if protection against infection and transmission waned with time).  In the case of omicron, the protection, even by additional boosters, against infection and transmission is weaker and short-lived.  Moreover BA.5 is also know to block interferon responses more than previous variants. 


Topol seems to be cautiously optimistic that a BA.5 specific booster can be beneficial on the short run, but he would like to see human data first.


Topol is convinced, however, that this is, at best, a temporary solution and that we need a “pansarbeco” vaccine for nasal application for a broader and more robust protection against the variants to come. (For extensive discussion see Episode 278).


Ep 284-12:  Juan Lama in Virus World What's Behind the FDA's Controversial Strategy for Evaluating New COVID Boosters:  offers further arguments pro and con the “hasted” BA.5 strategy in the US.


That’s it for this week…..


Best wishes,