It Is Urgent That We Address Declining Vaccine Efficacy Against Variants
(AP Photo/Charles Krupa, File)
It Is Urgent That We Address Declining Vaccine Efficacy Against Variants
(AP Photo/Charles Krupa, File)
Story Stream
recent articles

News reports indicate that demand for both initial Covid vaccination as well as booster shots continue to fall precipitously. Yet compelled vaccination continue to be mandated by many employers, including the United States military, which has discharged thousands of service members unwilling to get the vaccine. Still other employers, notably medical facilities, have imposed booster mandates on their employees. Many colleges and universities imposed similar mandates on their students at the beginning of the Spring semester.

Notably, all of these shots continue to use the original, long-extinct wild-type variant contained in the first generation of vaccines. This repeated administration of vaccines raises a risk of effects known as original antigenic sin (OAS) and antigenic seniority that, aside from a few exceptions, public health officials have not addressed.

But before widespread boosters are administered, especially for younger people, it's essential that doubts about this phenomenon be resolved. The concern is especially urgent because, as will be seen, there are already alarming signals that OAS and antigenic seniority might already be operative with respect to the rapidly spreading Omicron variant.

Although the data are rapidly emerging and any conclusions to be drawn are tentative, theory and multiple different layers of evidence developed since the onset of the pandemic raise alarming signals that the long-term negative implications of mass vaccination against SARS-CoV-2, especially in children and young adults such as college students, could be profound and must be monitored closely. Caution is warranted.

What Is OAS and Antigenic Seniority?

OAS refers to a well-known phenomenon of immunology, which is the concept that “humoral memory responses generated against one set of antigens can affect the nature or antibody responses elicited to challenge infections or vaccinations containing a similar but not identical array of antigens.” Thus, when first exposed to a new pathogen, the immune system essentially makes an imprint of the virus and remembers it by storing it in long-term humoral memory so the body can rapidly recall and produce appropriate antibodies when reexposed to the same virus later in life.

Antigenic seniority refers to a closely related phenomenon, which is the evolution of “dominant antibody responses as a consequence of repeat exposures to the same antigen(s) rather than to the first antigenic exposure or imprinting that is the core tenet of OAS.” Antigenic seniority, therefore, effectively specializes the immune system to address a recurrent threat in one’s environment, creating even ability to fight that threat. These two phenomena are not exclusive—one can have initial imprinting via OAS and then have that reinforced by repeat exposure creating antigenic seniority.

For purposes of the discussion here, this technical difference between the two is irrelevant, and thus for simplicity of presentation I will refer to just OAS throughout the discussion, but the reader should be aware that repeated exposure—such as booster shots—exacerbates the concern of OAS with antigenic seniority.

These are very efficient immune mechanisms, especially in our evolutionary past where we lived in relatively small, stable communities, where viral evolution was slow and individuals could be expected to encounter the same or similar viral variants repeatedly during their lifespan. This ability to rapidly identify a virus upon subsequent reexposure can be particularly problematic in the modern world, where variation can occur rapidly due to living in large, densely populated, global communities, especially when placed under the intense evolutionary selection pressure (pdf) of widespread administration of leaky vaccines that promote immune escape.

What Are the Concerns?

OAS raises the concern that as a virus evolves—a phenomenon referred to as “antigenic drift”—over time the immune system will respond in a suboptimal fashion to neutralize the new variant. As noted by Vatti, et al. in a review prior to COVID-19, OAS “implies that when the epitope varies slightly, then the immune system relies on memory of the earlier infection, rather than mount another primary or secondary response to the new epitope which would allow faster and stronger responses.” This could result in the immunological response being “inadequate against the new strain, because the immune system does not adapt and instead relies on its memory to mount a response.” [Emphasis added.]

Counterintuitively, OAS can result in a secondary immunological response to the new variant that is inferior to the primary response generated to an entirely novel virus. While OAS is relevant to both naturally acquired immunity and vaccine-acquired immunity, this concern is especially pronounced in the case of the current generation COVID-19 vaccines as they're designed to narrowly target just the spike protein of the original wild-type virus and not the full complement of 29 proteins that comprise the SARS-CoV-2 virus; thus, the initial imprinting of the spike protein excludes those other proteins (immunity derived from natural infection, by contrast, imprints the full complement of proteins). In addition, antibodies generated by natural infection continue to evolve for at least a year to increase their breadth and potency of response to mutations, whereas antibodies elicited by vaccination remain static and similar to their initial response.

The relevance of OAS to COVID-19 vaccination in not just theoretical. A study by Horndler, et al., from October 2021, drew blood from volunteers in Spain who had received partial vaccination (one shot) or full vaccination (two shots) and examined the reactivity of the antibody titers against the original Wuhan, Alpha, Delta, and Kappa variants. The authors found that when the vaccines were applied to the Wuhan variant, the administration of two shots was highly effective at binding to the virus. But they found that with respect to the other three variants, although the application of the second shot increased antibody titers (and better neutralized the variant for which it was designed) “there is a relative loss of reactivity with the three VOCs [varients of concern] compared to the Wuhan strain occurring upon administration of the booster dose of vaccine.”

In other words, they note that while repeated exposure to the same variant (via two vaccine doses) created a stronger and deeper response to that variant, this came at the expense of reduced “breadth” of the antibodies, “that is, their capacity to bind to epitopes that differ slightly from those of the immugen.” Moreover, this “deeper but narrower” immune response emerged after only two shots of the vaccine, raising concern about the potential effect of adding a third or fourth shot in reinforcing this. Given that the current generation of vaccines target a now-extinct variant of the SARS-CoV2 virus, this result is troubling.

Field Evidence

Although far from definitive, early, tentative field evidence associated with vaccine breakthrough infections is consistent with the hypothesis of OAS. For example, in Moderna’s published Phase 3 vaccine trial results, 66 percent of those who suffered natural infections during the trials later exhibited evidence of anti-nucleocapsid antibodies in the blood; only 23 percent of those who suffered vaccine breakthrough infections did so. This unfortunately suggests that while the vaccine showed some efficacy at preventing infection in the trial, it did so at the expense of the long-term ability of the immune system to create a similar degree of robust, broad antibodies as those naively infected to recognize the broader complement of proteins in the SARS-CoV-2 virus.

Other studies have similarly found that presence of anti-nucleocapsid antibodies in the blood is markedly lower for those who suffer vaccine breakthrough infections than those who are naively infected.

Writing in the Journal of Infection in October 2021, Allen, et al., found that only 6 of the 23 health care workers who suffered breakthrough infections (26 percent) had detectable anti-nucleocapsid antibodies following their infection, compared to 82 percent of those who were infected without being vaccinated earlier.

Moreover, Whitaker, et al., found that when anti-nucleocapsid seroconversion occurred, the levels were lower for vaccine breakthrough infections than for unvaccinated individuals. In the case of the Alpha variant, for example, anti-nucleocapsid antibody levels were almost eight times higher in unvaccinated individuals than in fully vaccinated. In addition, they found that anti-nucleocapsid antibody levels following breakthrough infections were much lower for those who were fully vaccinated than for those who received a single dose, which suggests that repeat vaccination (such as a second, third, or fourth shot) further reduces the immune system’s response to produce broader antibodies that can protect against reinfection by future variants.

Clinical Evidence

These concerns have begun to manifest themselves in clinical evidence regarding the breadth and durability of protection afforded by so-called “hybrid” immunity, which refers to a situation in which a vaccinated individual suffers a breakthrough infection (or alternatively, an individual with natural immunity who is later vaccinated and then suffers a reinfection). Examining Israeli data, Goldberg, et al., however, found that when an individual was vaccinated prior to suffering a breakthrough infection, the risk of a subsequent reinfection was higher than for either natural immunity alone or natural immunity followed by vaccination. Moreover, they found the gap in the relative risk of infection between those who were vaccinated first compared to the other groups grew with the passage of time (i.e., six to eight months versus four to six months).

Notably, in Israel, “full vaccination” of a COVID-recovered individual was defined as receiving only one shot following recovery; thus, any OAS effects for those who recovered from infection prior to vaccination might have been dampened. In the United States, by contrast, employer and government mandates have required a full two-dose course of treatment even for those with natural immunity. Some researchers have expressed concern this could give rise to an elevated risk of OAS in those individuals. Adding a third "booster" shot with the same spike-protein variant elevates those concerns. In particular, while one vaccine shot post-recovery has been shown to increase preexisting levels of functional and specific antibodies, a second dose resulted in either zero or even a reduction in virus-neutralizing antibodies, as reported by Semanovic, et al., Lozano-Ojalvo, et al. (pdf), and Mazzoni, et al. (pdf)

Even more worrisome, the authors found that the risk of severe cases upon reinfection was approximately twice as high for those who were vaccinated before being infected compared to the other relevant groups (natural immunity alone or natural immunity followed by vaccination), which further suggests a misalignment between the immune system as primed by the vaccines and the evolving SARS-CoV-2 variant.

The Arrival of Omicron

The arrival of the Omicron variant has highlighted the risk of OAS with respect to emergent viral variants. Omicron is a highly mutated strain of SARS-CoV-2, possessing roughly 50 mutations compared to earlier strains, of which over 30 are found on the spike protein. If OAS is present with respect to Omicron, this large number of mutations would be reflected in highly reduced vaccine efficacy in protection against infection (and potentially serious illness) against more heavily mutated strains. Available evidence is consistent with that hypothesiswhich suggests that current-generation vaccines are not merely ineffective against Omicron but might actually be enabling its rapid spread.

A study from Toronto (pdf) concluded that compared to unvaccinated individuals “receipt of 2 doses of COVID-19 vaccines was not protective against Omicron infection at any point in time” starting at 6 percent efficacy 7 to 59 days after initial completion of the second dose, becoming negative at minus-13 percent by 60 to 119 days (i.e., two months after receipt of vaccination) and declined steadily to minus-42 percent by 180 to 239 days.

A study from Denmark (pdf) reported positive efficacy for Pfizer’s vaccine up to 61 to 90 days after vaccination (9.8 percent vaccine effectiveness for days 61 to 90), but then a staggering plummet in efficacy to minus-76.5 percent at the 91 to 150-day range. Moderna reported an initial vaccine efficacy of only 36.7 percent during the first 1 to 30 days, declining to 4.2 percent at 61 to 90 days and minus-39.3 percent after that.

Another study (pdf) focused on the protective effect provided by the Moderna vaccine specifically against Omicron also found negative vaccine efficacy after six months (which the authors reported as zero rather than negative because they arbitrarily imposed a lower bound of zero as reported vaccine effectiveness, which obscures the actual effect).

A study by Accorsi, et al., published in JAMA (the Journal of the American Medical Association) of the protection provided by the Pfizer and Moderna mRNA vaccines also found with respect to omicron overall vaccine efficacy from two doses started at only approximately 40% relative risk reduction compared to unvaccinated individuals but quickly fell to near zero by three months finally turning negative at about six months. The collapse was especially pronounced with respect to the Pfizer vaccine.

In addition, early reports from South Africa suggested that vaccination also has reduced efficacy against serious illness and hospitalization. Whereas protection against hospitalization was reported as 93 percent for earlier variants, protection against hospitalization declined to only 70 percent in the Guateng Province during the period of high Omicron prevalence.

Findings from Qatar (pdf) regarding vaccination of individuals who previously recovered from COVID-19 infection (i.e., those with naturally acquired immunity) raises further alarming signals about the potential presence of OAS among those who receive vaccination. The central findings of the report are reassuring—natural immunity continues to provide protection against reinfection by the Omicron variant (56 percent overall) and extremely robust protection against severe illness and death (87.8 percent). More troubling, however, is that vaccination of those with natural immunity was found to actually decrease protection against reinfection from 62 percent to 56 percent, resulting in negative vaccine efficacy for those with natural immunity.

Risks of Boosters

Repeated booster shots raise the risk of further exacerbating the effects of OAS for recipients. Although at the current time there are no clinical studies with respect to the medium- to long-term efficacy, with widespread administration of booster shots in Israel (where 46 percent of the population has received three shots and half a million have received a fourth dose), early reports suggest that any modest initial protective effect that booster shots provides against the Omicron variant is likely to be short-lived and in all likelihood to quickly become negative as well. 

Research in non-human primates published last month raises concern that the presence of OAS effects might frustrate the efficacy of administering variant-specific booster shots against emergent variants. The authors first administered an original two dose course of vaccine to several macaques. Then several months later they administered an omicron-specific booster shot to see if the variant-specific booster provided better protection than the mediocre degree of protection afforded by a third shot of the original variant. The researchers found the immune system responded similarly to an omicron-specific booster than a third dose of the vaccine with the original variant. The authors note that this likely reflects the influence of OAS, as exposure to the omicron variant of the virus via vaccination instead calls forth the neutralizing antibodies related to the original variant. As the authors note, “The observation that boosting with either mRNA-1273 or mRNA-Omicron resulted in an expansion of a similarly high frequency of cross-reactive B cells likely stems from the principle of original antigenic sin, otherwise termed antigenic imprinting, whereby prior immune memory is recalled by a related antigenic encounter.”

It should be stressed that the data remains in flux and could be overturned by future findings, but it's also clear that a pattern may be emerging that provides confirmatory evidence beyond one anomalous finding. Moreover, any conclusion that OAS and antigenic seniority explain these findings is inherently tentative. Nevertheless, as discussed below, OAS provides the most plausible explanation for the observed clinical data than alternative explanations.

If OAS is at work, therefore, this raises especially acute concern about the risk that repeated vaccination with the same (extinct) viral strain via booster could undermine the effectiveness of future vaccinations targeted against future variants. Studies of the immunological response in the context of influenza vaccination have found that a new vaccination series calls forth not only an antibody response aimed at the new variant but also a strong antibody response to the original variant, creating a sort of competition between the two sources of immunological response. As Petras and Lesna observed in the context of SARS-CoV-2 vaccination, “This gives rise to a situation whereby the targeted and desirable response to new variants of the influenza virus types and subtypes is suppressed” [emphasis added], whereas a response to the original variant that shares the same antigenic determinants with the new ones is preferred.

The extremely high levels of specific anti-SARS-CoV-2 antibodies generated by vaccination (compared to the levels generated by natural immunity) are particularly alarming with respect to the risk of OAS occurring. One way to reduce the risk of OAS might have been to include a larger array of protein components beyond just the spike protein, such as the nucleocapsid or envelope proteins.

Lurking in the data, however, is one possible surprising piece of solace—the Toronto study (pdf) found that after the dramatic downward descent into negative efficacy between two and six months, at six months and beyond, while vaccine efficacy remains negative, it rebounds slightly from minus-42 percent at three to six months to only minus-16 percent beyond six months. Similarly, the Accorsi, et al., JAMA study found that although VE turned negative around six months following vaccination, by about 10-11 months vaccinated individuals were no longer at greater risk than unvaccinated individuals of infection.

This suggests that abstaining from repeat vaccination for some period of time might provide some opportunity for the immune system to begin to heal itself. This possibility is especially tentative given the thinness of the data but provides some hope of avoiding long-term catastrophe from repeat vaccination, at least with respect to this concern. Nor is it clear why one would observe this, but one speculative possibility is that just as vaccination’s protection against illness wanes over time, perhaps negative OAS-type effects decline as well. If this is true, then ceasing booster shots for many people might enable them to still be protected against serious illness while avoiding the risks of long-term repeat exposure to the extinct SARS-CoV-2 variant. Moreover, although Omicron has obviously achieved complete immune-escape from vaccines (and partial escape from natural immunity) in terms of protection from infection, it appears that T-cells generated by vaccination still provide some protection against serious symptoms from Covid. 

It should be stressed yet again that the concerns raised here are tentative. But given the well-established bases of the theory along with emerging consistent data, it's urgent that public health officials turn attention to this issue immediately. More importantly, they shouldn't be dissuaded by hand-waving “experts” who dismiss the concerns as invalid without good cause. More evidence will be needed before one can raise definitive concern about the risk of OAS with respect to the SARS-CoV-2 vaccines. But theory, laboratory evidence, and clinical analysis all point to this as a rapidly emerging risk of the current SARS-CoV-2 vaccines, and one that could be exacerbated by widespread application of booster shots using the extinct variant, which could still further narrow the ability of the immune system to adapt to future variants.

Consideration of the risk of OAS may be of less of a concern to older Americans, who are at high short-term risk from COVID-19 relative to the longer-term risk of OAS. But for lower-risk younger Americans who can be expected to reencounter the virus in multiple different variants over their long lifespans, the risks associated with OAS suggests caution in administering repeated vaccination using the original SARS-CoV-2 spike protein variant. This is especially urgent in the case of thousands of young, healthy, already-vaccinated college students that are being coerced to receive booster shots before returning to spring classes. Not only is this pointless and raises serious risks of short-term side-effects, the consequences could last a lifetime. Given the trivial if any benefit to further shots and the uncertain but tangible risks of OAS that are emerging, the evidence weighs heavily against forcing further shots on most Americans, but especially young people.

Addendum: Addressing the Critics

Some vaccine advocates have tried to dismiss the findings of negative vaccine efficacy against Omicron, asserting the evidence can be explained by alleged behavioral differences between vaccinated and unvaccinated individuals that bias the data, such as that vaccinated people feel “safer” and thus are more willing to engage in group social activities or less likely to take alternative mitigation precautions such as wearing masks and social distancing. They provide no evidence to support these suppositions.

Moreover, reams of data and studies demonstrate that not only are these suppositions groundless, they are also exactly opposite to reality. Studies demonstrate what everyday experience during the pandemic tells us—that vaccinated individuals are much more likely to fear SARS-CoV-2 more than unvaccinated and more likely to take precautions against potential COVID-19 infection, such as by wearing masks or avoiding large gatherings of people. In addition, unvaccinated individuals are much less likely to be concerned about contracting COVID-19 or its severity and more likely to express belief that the pandemic has been exaggerated. Those who aren't vaccinated are also more likely to trust in their neighbors than those who are vaccinated, which partly explains why they don't take these additional precautions. It's a curious claim that those who are least fearful of contracting COVID-19 are more likely to get vaccinated and take precautions than those who are more fearful, yet this is what these claims would have us believe. In short, while selection bias as a result of mitigating behavior may be present, it likely means that vaccine efficacy is overstated, not understated. Nor is it particularly plausible that the massive size of the coefficients found in these studies—such as minus-76 percent in one study and minus-42 percent in another—can be explained by these sorts of behavioral biases.

Moreover, these ad hoc behavioral explanations for the vaccines’ collapse of protection against the Omicron variant ignores the fact that no similar effect is seen with respect to protection against other variants, such as Delta. The Toronto study (pdf), for example, compares data for Omicron and Delta and finds that while protection against Delta slid only from 84 percent during the first two months to 71 percent at six months, vaccine efficacy for Omicron fell from 6 percent to minus-42 percent over the course of six months. Similarly, the Danish study (pdf) reported that while vaccine efficacy for Pfizer against Delta declined from 86 percent to 53 percent over three to six months, protection against Omicron fell from 55 percent during the first month to minus-76 percent. The authors provide no explanation for why the findings about Omicron are biased by these alleged behavioral biases but not Delta. Moreover, the authors have never before expressed concern that where vaccines have suggested high levels of protection that those findings were ever questioned because of assertions of behavioral bias. These unsupported explanations simply lack any scientific plausibility.

Explanations analogous to OAS, such as “immune exhaustion” and High Zone Tolerance, also flounder on similar grounds. Although these explanations are more plausible than claims of data biases, if these concerns were operative they would predict significantly diminished protection against all variants across the board, not just Omicron. But as noted, although studies show waning protection by vaccines against all variants over time, the waning of protection against the highly mutated Omicron is orders of magnitude higher than legacy variants, which suggests OAS may be the most viable interpretation.

Todd Zywicki is George Mason University Foundation Professor of Law at Antonin Scalia Law School and was the plaintiff in a lawsuit challenging his employer’s vaccine mandate.

Show comments Hide Comments