10 Sept 2021 Episode 171 Mu variant and heterologous prime-boost vaccination

Dear colleagues,

As usual, I received some input and questions that triggered my curiosity.  Herewith you find  an elective reading of recent papers on variants and heterologous prime boost vaccination.

The µ variant put into perspective

Ep 171-1:  The Mu (µ) or B.1.621 was first found in South-America: it  made up 69% of cases in early July in Colombia. It also popped up in the US, where in July it accounted for 9 % of 101 sequenced cases in Florida, but thereafter it was “beaten” by Delta.

Ep 171-2:  The majority of Mu variants harbors the following eight mutations in spike:

T95I, YY144-145TSN, R346K, E484K, N501Y, D614G, P681H, and D950N

Of which:

E484K (shared with Beta, Gamma),

N501Y (shared with Alpha, Beta, Gamma),

P681H (shared with Alpha) (with R in Delta)

D950N (shared with Delta)

Importantly, Mu pseudoviruses are relatively more  resistant to neutralization convalescent sera (first wave) and Pfizer-vaccinated sera than any other variant (including recalcitrant beta and delta). 

Ep 171-3: Mu is clearly another example of “convergent” evolution, a concept that was nicely illustrated already in a bioRxiv by Zahradník  in May.  See Table 1 p. 13.

Ep 171-4: Just a few days ago Zhou et al provide another clue to the evolution: evidence of coinfections of various lineages in some US patients.  Obviously, in this way several SARS-CoV-2 lineages can create more “fit” recombinants, with higher affinity for the receptor, more efficient fusion, better proliferative capacity,  and/or escape from infection- or vaccine induced antibodies.   

Ep 171-5:   Although Mu has an “favorable” mutation profile, it doesn’t seem to conquer the world, according to the most recent weekly WHO report.

As of 29 August, over 4500 sequences (3794 sequences of B.1.621 and 856 sequences of B.1.621.1) have been uploaded to GISAID from 39 countries. Although the global prevalence of the Mu variant among sequenced cases has declined and is currently below 0.1%,

To me this situation is reminiscent of the “South-African” beta variant.  As is shown again in Ep 171-2, Beta is clearly an “immune escape” variant with also favorable receptor recognition (N501Y), it has competed out the 2020 D614G variant in South-Africa.  All vaccines tested in South-Africa were less effective (with Astra-Zeneca performing very poorly) than elsewhere, where only the D614G or Alpha were circulating. Beta started to spread around the world, but it was quickly overthrown by Delta.  So provisionally Delta (and its descendants) remain the “world champion”.  Let’s see how long this statement will hold (?)  

More on heterologous prime-boost

This vaccination strategy has been used extensively in experimental vaccination.  Most of the time, the “prime” is DNA and the “boost” either a viral vector or a protein, all focused on the envelope or spike.  In animals and humans, this strategy usually results in “better” immune responses: broader and “deeper” (= recognizing more mutants), including both antibodies and T cells etc and in animal models, protection against challenge was often (but not always) more robust than in “homologous” controls. But this strategy is NOT in clinical use, presumably, because it requires additional costly trial comparing homologous and heterologous strategies.

In case of COVID, it is being used since a few months for two main reasons:

• The unexpected side effects of Adeno-vectors (VITT) in high income European countries : hence a dose of Astra-Zeneca was completed with an mRNA boost
• The suspicion that inactivated vaccines (used in low-middle income countries) were not so powerful: hence additional boost with Adenovirus.
• Some other cases: mistakes, small trials in immune-compromised patients

Clearly, most of these studies are “incidental”, small and/or not properly controlled, especially not for vaccine efficacy.

1. Combination of mRNA with Adeno vector

Ep 171-6: a systematic review of 4 early immunogenicity and tolerability trials

- Homologous BNT162b2 vaccinations and heterologous AstraZeneca ChAd/ Pfizer-BNT had the highest antibody titers.

- Heterologous ChAd/BNT had the highest T cellular responses.

- Higher neutralizing activities against variants of concern B.1.1.7 (alpha), B.1.351 (beta), and B.1.617 (delta) were observed in heterologous ChAd/BNT vaccinations.

- Although adverse events were more commonly reported in the BNT- boosted participants, reactogenicity was tolerable in all combinations.

Ep 171-7: Large population-wide study in Denmark, where the subjects first vaccinated with AZ-ChAd were offered a second dose of Pfizer-BNT.  The authors find a very high efficacy of 88 % against infection and 100 % protection against hospitalization and death (up to 23 June 2021).  The efficacy against infection was much higher than in the minority who got only one shot of ChAd (but those were also 90 % protected against hospitalization) and is “similar” to homologous twice Pfizer, but no direct comparison in the same population is presented.

Ep 171-8: Hillus et al in Berlin confirm what is stated in Ep 171-6: the combination ChAd-BNT is clearly superior to homologous BNT and also homologous ChAd with regard to neutralization of alpha and beta variants and IFN-g production by T cells.  It is however a relatively small study of 380 participants.

Ep 171-9: Bonelli et al reported on 8 Sept in medRxiv on a comparison between ChAd or mRNA (either Pfizer or Moderna) as a third dose in patients with severe auto-immune disease, treated with the B-cell depleting anti-CD20 rituximab, who failed to seroconvert after 2 doses of mRNA vaccine. 

• Seroconversion at 4 wks was comparable: 6/27 (22%) after ChAd and 9/28 (32 %) after mRNA
• T cell responses in 100 % of ChAd and 81 % of mRNA booster.

Clearly, in this specific situation a third dose is important, but it doesn’t matter if it is homologous mRNA or heterologous ChAd.

Ep 171-10: Along the same lines Lysky reported on 7 Sept on 2 patients with chronic lymphocytic leukemia (malignant B cell proliferation with dysfunctional B cells).  Both had failed to seroconvert after 2 doses of Pfizer and received Ad26 from Janssen. One fully seroconverted, the other not.  The non-responsive patient was older and had undergone immune suppressive treatments. 

Ep 171-11:  The real world data by Powell in Eurosurveillance confirm that reactogenicity (local or systemic):

• is very common with either Pfizer or ChAd vaccines.
• higher in younger adults, women and after the first dose of ChAd in any schedule.

Importantly:  Previously-uninfected individuals who received heterologous prime-boost schedules were 2.4 times (27.8% vs. 11.6%) more likely to report severe reactogenicity, including increased requirement for medical attention, after their second dose than those receiving homologous schedules. These findings were irrespective of the reason for receiving a heterologous schedule.

Advice:  adults who have a severe reaction to their first dose should be advised that their risk of a severe reaction after their second dose will be higher with a heterologous schedule than

a homologous schedule.

Therefore, completion with the same vaccine brand should be considered unless there is clear evidence of anaphylaxis or other contraindications such as VITTs.(thrombosis and thrombocytopenia

2. Combining an inactivated vaccine with an Adenovector

Ep 171-12: Kant on a “serendipitous” finding: 18 people received the Covishield (the Indian ChAd-Ox1-S) as a prime and  inactivated Indian Covaxin  as a boost and were compared with 40 subjects with either of both homologous regimens.   Antibody levels and especially neutralizing  titers against the VOC alpha, beta and delta were higher in the heterologous group.  No data on vaccine efficacy.

Ep 171-13: In this Sept medRxiv paper by Jingxin Li either  1 or 2 doses of CoronoVac (inactivated Sinovac) as prime, followed by either Convidecia (Chinese Adeno-5 vector expressing SARS-CoV-2 Spike) or CoronaVac boost. It is very evident that:

• Antibody responses, including neutralization against the original Wuhan strain, are higher with Adeno boost.  No clear difference in T cell responses. No data on variants.  
• Local and systemic side effects are more pronounced with the Ad5 boost

Ep 171-14: Another Sept medRxiv by Yorsaeng in Thailand is very preliminary: compares homologous CoronaVac or Astra-Zeneca with CoronaVac as prime and Astra-Zeneca as boost:  antibody titers in the heterologous scheme are equivalent with homologous Astra-Zeneca and higher than homologous CoronaVac, but no data on neutralization, T cells or side effects.

Ep 171-15: A Chinese study in medRxiv on seroconversion and reversion after inactivated vaccines. It is a very complicated paper, but Fig 1 on p. 12 clearly shows that antibody titers wane rather quickly…. Hence boosters may indeed be required.   

3. The Sputnik vaccine : as we know the Russian Sputnik V is in fact and heterologous prime with Ad26-COV2-S (similar to Janssen), followed by an Ad5 boost (similar to the Chinese Convidecia).

Ep 171-16:   a very remarkable preprint from Argentina in medRxiv, following up a total of 1800 sera from vaccinated subjects between January and August shows for Spike-binding and neutralizing antibodies:

• A slight, but significant waning of IgG anti-Spike, but a stable neutralizing titer between day 45 and Day120, resulting in an increase in “neutralization potency index (NPI)” (Fig 1 p. 5).
• Even more intriguing:  a relative increase of NPI for the beta, gamma and delta variants as compared to alpha (Fig 2C p. 7)

This remarkable phenomenon of “temporal increase in neutralization potency” and “reduced viral variant escape” is explained as “maturation” of the immune response.  A phenomenon that has also been observed by Michel Nussenzweig’s group after natural infection.  In both cases, we should presume that the virus (either SARS-CoV-2 itself or the Spike-recombinant Adenovectors) somehow persist for many months in the immune system (e.g. in follicular dendritic cells of the lymph nodes) to induce this maturation process of the antibody-producing .     

4. Mouse studies

Ep 171-17: Qian He in Emerg Microb Infect comparing various homologous and heterologous prime-boost regimens with (Chinese) inactivated, Adeno, recombinant protein and mRNA candidates.  Unfortunately, not all combinations are really consistently compared for the various parameters, hence a final conclusion is not possible.  To me, the most significant result are in Fig1 and 2 on neutralizing Ab.

• From Fig 1, the impression is that the best combinations are: homologous protein, or heterologous Adeno5 with  either protein or inactivated virus (irrespective of the sequence). Comparison with RNA is lacking here.  
• From Fig 2 it is pretty clear that prime with Adeno, followed by RNA boost is superior to the reverse heterologous scheme and to homologous schemes, but here protein and inactivated virus are lacking from the comparison.  

Clearly, no results on variants

Ep 171-18: Spencer et al in Nat Comm on a novel strategy with a self-amplifying RNA (saRNA based on an alpha-viral replicase and the target S antigent)  and an adenoviral vectored vaccine (ChAdOx1 nCoV- Astra-Zeneca).   The brief summary:

• Antibody responses are higher in two-dose heterologous vaccination.
• Neutralizing titres after heterologous prime-boost comparable with homologous saRNA and higher than after homologous viral vectors.
• Importantly, the cellular immune response after a heterologous regimen is dominated by cytotoxic T cells and Th1+ CD4 T cells, which is superior to the response induced in homologous vaccination regimens in mice.

To me, it is clear that saRNA is a powerful vaccine platform, but the question remains whether the subtle advantage of a boost with a viral vector will outweigh the fear for the rare VITT side effects of the Adenovectors.

PRELIMINARY CONCLUSIONS on heterologous prime-boost:

• The expected benefits on higher/broader/deeper immunogenicity of several heterologous schemes are confirmed. They may also enhance immune memory, but that needs to be studied further.
• We have NO comparative data on vaccine efficacy of heterologous versus homologous prime-boost, but those data will become available in the coming months
• There is evidence that the already high reactogenicity of either mRNA or viral vector is even increased by combining them in an heterologous scheme.
• Overall there is no convincing benefit of  either vector/RNA or vector/inactivated combinations over homologous RNA, but there are certainly possible indications to be explored (e.g. in immune compromised subjects or those with specific adverse reactions to a particular vaccine type).
• The fact that heterologous prime-boosts are finally being used in humans, is a positive evolution and opens perspectives as several different non-Adeno-based platforms become available for COVID (S protein, DNA, saRNA etc).

That’s it for this week.  I still have quite a few interesting papers, suggested by some of you, but there are some other duties to be done.   Will be for med-late next week.

Have a nice weekend

Guido