RR:C19 Evidence Scale rating by reviewer:
Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.
The authors address perhaps the most important public health question about protective immunity against COVID-19 at the present time: the durability of immune protection in the context of inevitable viral mutations, after natural SARS-CoV-2 infection or vaccination. This is an experimental study on human subjects (n=24). Levels of antibodies by ELISA and frequencies of memory B-cells (MBC) against SARS-CoV-2 spike protein receptor-binding domain (RBD) were measured serially for parental strain or variants of concern (VoC) from the blood of subjects. In addition, MBC was measured for reactivity to the wild-type and variant protein ex vivo. The authors conclude that there is a reason for optimism because MBC frequencies against spike RBD are stable over almost 1 year (consistent with previous reports), do not have a lower ability to react to VoCs than wild-type RBD and are observed in seronegative PCR-positive subjects. These are important results, particularly the finding that MBC reacts similarly to parental strain and to VoC spike RBD, and the methods are technically solid. There are several important limitations to the work presented, however, that make the optimistic conclusion premature.
Extrapolation to vaccines from natural infection is not warranted. The authors present and analyze only data from natural infection but state “there is a reason for optimism regarding the capacity of vaccination, prior infection, and/or both, to limit disease severity and transmission of VoC “ (pg. 12) and “reason for optimism regarding the capacity of vaccination, prior infection, and/or both, to limit disease severity and transmission of variants of concern” (Abstract). Extrapolation from natural infection to mRNA vaccines is not warranted, for many reasons. The main public health concern today relates to vaccine-induced protection, in that far more people will be vaccinated than will have been naturally infected. This manuscript does not advance our understanding of this question. Statements extrapolating their results to vaccines should be removed from the manuscript and are not justified by their data.
The key parameter (potential neutralizing titers against VoC produced by affinitymatured MBC) is not characterized here and immune protection from potentially neutralizing titers based on MBC frequencies is not established. Some VoCs, such as the 484 variant, have been reported to be 10 – 100-fold less sensitive to neutralizing antibodies than wild-type virus. The authors’ data here show similar MBC frequencies against the parental strain and VoC RBD but do not report the neutralizing capacity against VoC compared to the parental strain(wild-type) virus. Nor do the authors show direct evidence of affinity maturation/ somatic hypermutation by the MBC that they detect. This is technically possible by sequencing but is not shown here. More specifically, what they show is “detectable” MBC per 106 PBMC. How should this be interpreted in terms of immune protection? Can we extrapolate to capacity to produce neutralizing titers on re-exposure? The authors do not present data Submission Click here to access/download;Review;Review of MBC -621.pdf or literature support to contextualize the potential neutralizing capacity of these MBC frequencies.
For the circulating antibodies, the data are not presented optimally. Several previous reports have shown at ½ of 3-4 months for antibodies after natural infection. Due to the rule of exponentials, t ½ of 3 – 4 months will result in 8 – 16-fold lower titers after a year. The authors do not present t ½ of their circulating antiVoC antibodies here. Data for durability (t ½) of ELISA titers should be presented for the different clinical subgroups. Moreover, detectable antibody levels are not the main issue for clinical immune protection but are what the authors show (Fig. 1), rather than the more relevant neutralizing titers.
Some laboratory outcomes presented are difficult to interpret for immune protection. What does the correlation between wild–type and VoC-specific serum antibody titers mean? If these represent the same antibodies from the same plasma cells, then the finding is trivial. The authors do not explore this correlation in sufficient detail.
Because of the fundamental limitations of the work presented, it is premature to adopt much evidence-based “optimism” based on their data for long-term MBC-based serologic protection against variant SARS-CoV-2 exposure. The possibility remains that affinity maturation of MBC in germinal centers will provide better protection against reexposure to SARS-CoV-2 mutant viruses than circulating antibodies provide after vaccination or natural COVID-19 infection. The data here add some useful information, but the work does not show affinity maturation or neutralizing capacity, which represent the key information needed for this very important public health question.