This week, when the CDC’s vaccine advisory committee considers approving the Moderna and Pfizer/BioNTech mRNA Covid-19 vaccines for infants and toddlers, the issue of imprinting may not be on the agenda. But it should be, given lessons from the Russian pandemic of 1889, the Spanish flu pandemic of 1918, the Hong Kong flu of 1968, the swine flu pandemic of 1957, and the 2009 H1N1 pandemic.
Immune imprinting results from exposure to proteins or other biological structures of viruses, like those found in SARS-CoV-2, that allow the virus to penetrate host cells and cause infection. The process, also called original antigenic sin, refers to the preference of the immune system to recall existing memory cells (immune cells that “remember” the same pathogen for faster future antibody production, such as memory B cells and memory T cells) rather than stimulating de novo responses when encountering a novel but closely related antigen.
Imprinting may come directly from an acute infection or indirectly through vaccination. It can result in reduced — or enhanced — responses to future variants with unknown clinical consequences. The former is beneficial, the latter is not.
The immune systems of infants and toddlers — the targets of the latest Covid-19 vaccination approval — are immature and developing. If an immature immune system is immunologically imprinted, either by acute infection from the currently circulating viral variant or by a Covid-19 vaccine based on the original, wild-type variant that is no longer in circulation, it may fail to develop appropriate defenses when confronted — even years later — by a Covid variant or another totally different pathogen.
Immune system imprinting is not a new concept, but it is one that has garnered relatively little attention outside of academic circles. Historical precedents suggest the negative effects of imprinting. Infants who survived the Russian pandemic of 1889 were more likely to experience excess mortality as adults during the Spanish flu pandemic of 1918. This association was described in a 2013 PLOS One paper by a team of researchers with expertise in microbiology, sociology, anthropology, and medical statistics led by Alain Gagnon, professor of demography at the University of Montreal.
The group explored the fact that a statistically inordinate number of healthy 28-year-olds, whose mothers had been pregnant with them during Russian pandemic of 1889, experienced excess mortality during the Spanish flu pandemic of 1918. The team wrote, “We posit that in specific instances, development of immunological memory to an influenza virus strain in early life may lead to a dysregulated immune response to novel strains encountered in later life.”
Infants who survived the Spanish flu were more likely to experience excess mortality as adults during the Hong Kong flu of 1968. More recently, research links exposure of infants to the swine flu pandemic of 1957 with excess mortality as adults during the 2009 H1N1 pandemic in Mexico.
Such analysis of past epidemics suggests, but does not prove, that imprinting can lead to worse outcomes when new pathogens arise — and they will arise. What has been the core strength of early vaccination against disease also has the potential, especially in the case of SARS-CoV-2 vaccines, to have unintended, delayed, and significant adverse consequences.
A second concern about the administration of Covid-19 vaccines to infants and toddlers is the potential for the unintended development of a runaway inflammatory process called cytokine storm. Cytokines play a key role in the normal immune response to infection. With infections such as Covid-19, however, the virus can cause an immune system dysfunction leading to an excess release of cytokines, triggering uncontrolled inflammation, blood clots, and organ damage. One of us (S.B.) formed a company in 2006 called StormBio that focused on developing treatments that would modulate the inflammation caused by the excess release of cytokines in influenza and other inflammatory conditions.
Infants and toddlers have a minimal risk of suffering adverse events from vaccines. The absence of short-term adverse events to vaccination, however, is no justification for future health risks.
The FDA’s Vaccines and Related Biological Products Advisory Committee (VRBPAC) is taking on a herculean task. Facing considerable political pressure, the committee is deciding whether to approve vaccinating infants and toddlers (children 6 months to 5 years old) using 18 months of data extrapolated from an older population using an analysis known as immunobridging.
Immunobridging is a well-established method for predicting the clinical safety and efficacy of vaccines in the absence of rigorous randomized clinical trial data. Bridging trials are controlled studies, but they enroll fewer subjects than traditional double-blind, randomized, controlled studies. To be useful, bridging methodology requires a validated and accepted correlate of protection to provide the bridge.
In the case of Covid-19 vaccines, analysts compare the safety and efficacy of extensive adult studies with smaller studies of infants and toddlers to see if there is a bridge — namely, is the safety and efficacy found in adults applicable to infants and toddlers?
The Covid-19 bridging study uses neutralizing antisera titer — the number of antibodies that can be detected — as the bridge. Yet Covid-19 neutralizing antisera provides protection from infection only for a finite time, a reason why vaccinated individuals may be infected with SARS-CoV-2, or reinfected later.
A still-unexplained phenomenon is that even in the absence of neutralizing antisera, vaccination affords protection from severe disease, hospitalization, and death. This protection must lie in another part of immune system. T cells are one possibility, though there may be others.
Neutralizing antisera is not the best parameter by which to measure long-term benefit from vaccination, yet that is the metric the VRBPAC is using to justify administration of the Covid-19 vaccine to infants and toddlers.
The data from immunobridging neutralizing titer levels from adult Phase 3 trials are flawed for another reason. The adult trials demonstrated clinical efficacy against a strain of SARS-CoV-2 that is no longer dominant, having been supplanted by multiple variants. That strain has become largely undetectable and continuous data demonstrate that currently circulating strains exhibit reduced cross-reactive neutralizing activity. That means the newer strains of SARS-CoV-2 are already evading the protections afforded neutralizing antibodies derived from the current vaccines.
It is unreasonable to assume that a two-day meeting will afford adequate time for reasoned analysis, debate, and discussion of so many unknowns. The panel has been given just 18 months of vaccine data, all based on the original, or ancestral, strain of SARS-CoV-2. Even Omicron has been replaced by fourth and fifth generations of variants. Asking the panel to extend approval for the current vaccines to a younger age group is not unlike advocating for administration of the 2019 influenza vaccine for the 2022 flu season.
The almost total reliance on vaccination to dealing with the pandemic has left the U.S. with one main therapeutic: mRNA vaccines. Unfortunately, the benefits wane not only with time but with each booster administered. SARS-CoV-2 is not the flu, yet success against it is measured using influenza statistical modeling and therapies. This is little more than confirmation bias which reinforces preconceived notions that Covid-19 is just another flu; reality is contradicting that reasoning.
There’s no question that over the course of modern medical practice, vaccines have saved more lives than any other medical intervention. That said, perfect vaccines do not exist. Pursuing a therapeutic course to treat the Covid-19 pandemic based solely on these vaccines will not only fail to deal effectively with the pandemic but will also erode the credibility of the response to Covid-19 and with it credibility in the health care system.
Our message is that the VRBPAC must exercise extreme caution before authorizing administration of vaccines to the incredibly young until there is better assurance that a current health benefit will not produce a negative effect later in life. Immunobridging data are insufficient justification for administering Covid-19 vaccines to infants and young children. The VRBPAC panel must take into consideration a Centers for Disease Control and Prevention report that stated, as of February, 75% of all infants, children, and unvaccinated adolescents had been infected with Covid-19, and most experienced no symptoms.
Mortality data from previous pandemics suggest that the imprinting that took place in utero, or at the earliest of ages, appears to have had the strongest effect, which would be expected from naïve immune systems. Expanding the vaccination pool to infants and exposing them to elements from a stronger and noncirculating virus is doing something that had not been possible before.
Rather than hastening to extend emergency use authorization of Covid-19 vaccines to infants and toddlers, the VRBPAC should pause and demand more than the two hours allotted for discussion and voting. The panel should also call for a three- to four-year study comparing vaccinated and unvaccinated control cohorts of infants and toddlers.
The vote on this vaccine for this vulnerable sector of the population is not inexorable. The availability of a therapy is not a justification for its use when benefits of such use are so poorly justified and no data on future consequences for this population to specifically include imprinting even exists.
The VRBPAC should say “no” to vaccinating infants and toddlers with the Moderna or Pfizer/BioNTech vaccines. “First do no harm” has never been a more important dictum.