The evolution of SARS-CoV-2 could undermine the efficacy of vaccines currently under development


Similar to bacteria evolving resistance to antibiotics, viruses can evolve resistance to vaccines, and the evolution of SARS-CoV-2 could undermine the effectiveness of vaccines that are currently under development, according to a paper published November 9 in the open-access journal PLOS Biology by David Kennedy and Andrew Read from Pennsylvania State University, U.S. The authors also offer recommendations to vaccine developers for minimizing the likelihood of this outcome.

“A COVID-19 vaccine is urgently needed to save lives and help society return to its pre-pandemic normal,” said David Kennedy, assistant professor of biology.

“As we have seen with other diseases, such as pneumonia, the evolution of resistance can quickly render vaccines ineffective. By learning from these previous challenges and by implementing this knowledge during vaccine design, we may be able to maximize the long-term impact of COVID-19 vaccines.”

The researchers specifically suggest that the standard blood and nasal-swab samples taken during clinical trials to quantify individuals’ responses to vaccination may also be used to assess the likelihood that the vaccines being tested will drive resistance evolution.

For example, the team proposes that blood samples can be used to assess the redundancy of immune protection generated by candidate vaccines by measuring the types and amounts of antibodies and T-cells that are present.

“Much like how combination antibiotic therapy delays the evolution of antibiotic resistance, vaccines that are designed to induce a redundant immune response – or one in which the immune system is encouraged to target multiple sites, called epitopes – on the virus’s surface, can delay the evolution of vaccine resistance,” said Andrew Read, Evan Pugh Professor of Biology and Entomology and director of the Huck Institutes of the Life Sciences.

“That’s because the virus would have to acquire several mutations, as opposed to just one, in order to survive the host immune system’s attack.”

The researchers also recommend that nasal swabs typically collected during clinical trials may be used to determine the viral titer, or amount of virus present, which can be considered a proxy for transmission potential.

They noted that strongly suppressing virus transmission through vaccinated hosts is key to slowing the evolution of resistance, since it minimizes opportunities for mutations to arise and reduces opportunities for natural selection to act on those mutations that do arise.

In addition, the team suggests that the genetic data acquired through nasal swabs can be used to examine whether vaccine-driven selection has occurred.

For example, differences in alleles, or forms of genes that arise from mutations, between the viral genomes collected from vaccinated versus unvaccinated individuals would indicate that selection has taken place.

“According to the World Health Organization, at least 198 COVID-19 vaccines are in the development pipeline, with 44 currently undergoing clinical evaluation,” said Kennedy. “We suggest that the risk of resistance be used to prioritize investment among otherwise similarly promising vaccine candidates.”

Vaccination is a primary tool for controlling and eradicating infectious diseases. Although research on vaccines has traditionally been the purview of medical scientists and virologists, evolution- ary biologists have, in recent years, made significant empirical and theoretical contributions to this field. Many of these contributions have stemmed from a growing realization that evolutionary biol- ogy can offer important insights into a variety of issues related to human health and disease [1,2].

Pathogens often have the potential to evolve very rapidly, because of their short generation times, large population sizes and high rates of mutation. It is now commonly believed that the use of vaccines will typically result in novel selective pressure on pathogen populations, often resulting in the emergence of resistant genotypes.

The purpose of this Special Supplement is to evaluate the current state of knowledge of vaccine-driven evolution, and to consider important potential areas of future research on this topic. The research reported in this Special Supplement originated from a workshop on the Evolutionary Considerations of Vaccine Use held at Rutgers University’s Center for Discrete Mathematics and Theo- retical Computer Science (DIMACS) in June 2005.

The contrasting evolutionary outcomes of vaccination for measles versus influenza demonstrate that developing a thorough understanding of the evolutionary consequences of vaccination is crucial for designing successful vaccination programs. Influenza displays a well-characterized pattern of continual antigenic evolution (see articles by Boni and by Gog, in this Special Supplement), whereas measles undergoes relatively little evolutionary change in this regard.

As a result, influenza vaccines must be continually updated to maintain their effectiveness, while measles vaccines do not. This makes vaccination a more effective control strategy for measles than for influenza, because influenza can, in effect, evolve to circumvent this control measure. Similar differences in evolu- tionary outcomes have also been identified and analyzed for other pathogens ([3]; see articles by Gandon and Day and by Poolman et al., this Special Supplement).

These simple comparisons highlight a clear need for the devel- opment of a predictive evolutionary framework, based on the use of quantitative models, to help in the design of optimal vaccination

strategies. While some progress towards this goal is being made (as evidenced by the contributions to this Supplement), many impor- tant issues still remain to be explored. These include:

  • Conflicts between epidemiology and evolution. Vaccination strategies that are optimal from an evolutionary standpoint need not be optimal from an epidemiological standpoint. For example, perhaps the strategy that is most likely to be success- ful in the absence of evolutionary change is also the one that is most likely to lead to adverse evolutionary outcomes. Can we predict when conflicts between evolutionary and epidemiolog- ical processes are likely? If there are conflicts for vaccination strategies, how can we weight the relative importance of evolu- tionary and epidemiological issues in order to make informed decisions? The optimal balance between epidemiological and evolutionary outcomes depends on the timescales over which these outcomes occur, and the level of discounting of the future relative to present. Evolutionary processes generally occur over a much shorter time scale than epidemiological processes. Con- sequently, the lower the discounting of the future relative to the present, the more important the evolutionary repercussions.
  • Vaccination and virulence. What is the expected relationship between vaccine use and the evolution of pathogen virulence, and how do different vaccination strategies affect the expected virulence of a pathogen? Examples of virulence include the prevalence of non-toxigenic diphtheria in highly vaccinated populations and the classical example of myoxma virus/rabbit studies of Australia. To date, most work on the evolution- ary effects of vaccination has focused on escape mutants, but recent innovative research has addressed virulence evolution as well (see articles by Gimeno and by Mackinnon et al., this issue).
  • Modes of vaccine action. Vaccines work in different ways. Some block transmission, some reduce pathogen replication, while others might slow the progression of disease (see article by Mackinnon et al., this Special Supplement). Each of these modes has its own epidemiological advantage, but how does the mode of vaccine action affect the evolutionary response in the pathogen population? Are some types of vaccine more apt to result in evolutionary change than others? For example, are escape mutants more likely to occur, and to be evolu- tionary successful, in individuals that are vaccinated with transmission-blocking vaccines or replication-inhibiting vac- cines? Are some types of vaccine more likely to result in beneficial evolutionary responses than others in terms of disease control? For example, some vaccines may select for increased virulence while others may select for reduced vir- ulence [4,5]. Furthermore, for live vaccines, what conditions could promote reversion to the virulent strain of the pathogen?
  • Multiple levels of natural selection. Evolutionary change in pathogen populations takes place on at least two distinct scales [6]. First, evolutionary change in pathogen sub-populations within a host can occur. This within-host level of selection is exemplified by HIV, which rapidly evolves resistance against antiviral drugs. Second, evolutionary change in the pathogen population can also occur at the community level if some strains are more effective at being transmitted from person- to-person than others. This between-host level of selection  is exemplified by influenza, which evolves over the course of an epidemic season. These different levels of selection are reflected in the phylogenetic trees of antigenic evolution in these pathogens. Evolutionary biologists have long been inter- ested in such phylogenetic patterns and ‘‘levels of selection’’. It is clearly important that these issues be incorporated into any theory that deals with the evolutionary consequences of vaccination. How do different types of vaccines and/or vaccina- tion strategies affect evolutionary change at these two levels? Is evolutionary change at one level often expected to oppose evolutionary change at the other? For example, does vaccina- tion tend to result in the evolution of escape mutants within vaccinated individuals, but these escape mutants are never- theless selected against at the population level because they do not transmit well? If so, when might we expect there to be sufficient time for compensatory evolution to occur within an individual that allows for efficient transmission between hosts? Are there vaccination protocols that minimize the prob- ability of such detrimental outcomes?
  • Mechanisms of vaccine delivery. Modern vaccines are comprised of purified, inactivated microorganisms typically administered by a sterile injection. Today’s vaccines generally introduce a weakened version of an antigen that stimulates the produc- tion of specific antibodies. In a new and promising approach, DNA vaccination, genes encoding an antigen are delivered to cells that then produce the antigen and display it on their sur- face. New drive systems are at the heart of the new delivery mechanisms. They can include, among others, genetic vacci- nation using plasmid DNA, microparticle-based DNA delivery (in which the genes are encapsulated within or immobilized on a spherical polymer particle), and live attenuated transgene vectors. The new delivery mechanisms can improve vaccine potency by targeting the genes to appropriate cells of the immune system, and by allowing for the expression of anti- gens in synchrony with the life cycle of white blood cells and pathogen life-cycle stages. As of yet, however, very little is known about how such novel vaccination approaches are likely to affect the evolution of pathogen populations.
  • Epidemiologic and surveillance methods for the study of the evolutionary repercussions of vaccination. Many ecological and evolutionary consequences of long-term vaccination programs in populations will become obvious only on time scales longer than those of vaccine trials. The formalism of epidemiologic methods, which have proved useful in identifying risk factors in chronic and infectious diseases as well, still require further developments in order to address key questions in this con- text. It is expected that large-scale use of vaccines will alter the number and virulence of pathogen genotypes, either by reduc- ing the force of infection or by directly altering the population dynamics of subsets of the circulating genotypes. Therefore, the licensing of new vaccines as well as the surveillance of new and current vaccines already in use must take into account the possible consequences of this for public health. Analogously to current vaccine trials that provide the necessary empirical background to assess the efficacy of a vaccine, informed public health decisions in this area must rely on study designs, sam- pling mechanisms and epidemiologic parameters specifically conceived with this end in mind.

The objective of this Special Supplement is to assemble leading experts who are working on the aforementioned problems. To date, most work has concentrated on the issues raised in (i)–(iii) and has resulted in two largely separate bodies of research [7,8]. The first focuses on so-called ‘escape’ mutants, and is directed towards understanding how vaccination and natural host immunity select for antigenic evolution resulting in strains that are able to evade the protective effects of the vaccine [7,9–13].

The second focuses on so- called ‘virulence’ or ‘life-history’ mutants, and is directed towards understanding how vaccination causes evolutionary change in the extent to which a parasite harms its host (i.e., evolutionary changes in its virulence; [4,7,14–16]). Recent work has also begun to draw these two areas into a single, comprehensive theory [17] but this division is still useful for categorizing much current research. As such, this volume is organized along these lines.

In the first article, Gandon and Day review both epidemiological and experimental evidence for vaccine-driven evolution in a vari- ety of pathogens, including both escape mutant evolution, subtype replacement and virulence evolution. They present evidence that vaccination can increase the virulence of diseases, such as malaria, as well decrease virulence in other cases, for example diphtheria. Gandon and Day highlight the need for more empirical quantifi- cation of the costs of vaccine escape mutants in order to more accurately predict the evolutionary consequences of vaccination.

The next three articles focus on  escape  mutants  and  anti- genic evolution. Boni analyzes data on the relationship between vaccination and antigenic  evolution  in  influenza,  explaining  how intermediate levels of vaccination may generate the most antigenic influenza of pathogens.

Gog presents a modeling framework that takes into account  functional  constraints  that  may limit antigenic evolution, and she applies this to explain patterns  of  antigenic  evolution  in  influenza.  Poolman  et  al. then present a novel theoretical framework  for  understanding  and predicting antigenic evolution in Human Papilloma virus. They demonstrate that, depending on the degree of cross- immunity elicited by the vaccine, vaccination may either expand  or  contract  the  niche  of  HPV,  but  that  the  latter  is  more likely.

The last two articles focus on vaccine-driven virulence evolu- tion. Gimeno presents a review of the extremely interesting case of vaccine-driven evolution in the infectivity and virulence of Marek’s Disease Virus in chickens. MacKinnon et al. review and analyze the fascinating empirical research that has been done on virulence evolution and vaccination in malaria.


  1. Stearns SC, Koella J, editors. Evolution in health and disease. Oxford: Oxford University Press; 2007.
  2. Dieckmann U, Metz JAJ, et al., editors. Adaptive dynamics of infectious diseases: in pursuit of virulence management. Cambridge, U.K.: Cambridge University Press; 2002.
  3. Lipsitch M. Bacterial vaccines and serotype replacement: lessons from Haemophilis influenzae and prospects for Streptococcus pneumoniae. Emerg Infect Dis 1999;5:336–45.
  4. Gandon S, Mackinnon MJ, Nee S, Read AF. Imperfect vaccines and the evolution of pathogen virulence. Nature 2001;414:751–6.
  5. Galvani AP. Epidemiology meets evolutionary ecology. Trends Ecol Evol 2003;18:132–9.
  6. Frank SA. Models of parasite virulence. Q Rev Biol 1996;71:37–78.
  7. van BovenF M., Mooi FR, Schellekens JFP, de Melker HE, Kretzschmar M. Pathogen adaptation under imperfect vaccination: implications for pertussis. Proc R Soc B 2005;272:1617–24.
  8. Restif O, Grenfell B. Integrating life-history and cross-immunity into the evolu- tionary dynamics of pathogens. Proc R Soc B 2006;273:409–16.
  9. McLean AR. Vaccination, evolution and changes in the efficacy of vaccines: a theoretical framework. Proc R Soc B 1995;261:389–93.
  10. McLean AR. Vaccines and their impact on the control of disease. Br Med Bull 1999;54:545–56.
  11. Scherer A, McLean A. Mathematical models of vaccination. Br Med Bull 2002;62:187–99.
  12. Wilson JN, Nokes DJ, Carman WF. Current status of HBV vaccine escape variants—a mathematical model of their epidemiology. J Viral Hepatitis 1998;s2:25–39.
  13. Wilson JN, Nokes DJ, Carman WF. The predicted pattern of emergence of vaccine-resistant hepatitis B: a cause for concern? Vaccine 1999;17:973–8.
  14. Gandon S, Mackinnon MJ, Nee S, Read AF. Imperfect vaccination: some epidemi- ological and evolutionary consequences. Proc R Soc B 2003;270:1129–36.
  15. Andre´  JB, Gandon S. Vaccination, within-host dynamics and virulence evolu- tion. Evolution 2006;60:13–23.
  16. Ganusov VV, Antia R. Imperfect vaccines and the evolution of pathogens causing acute infections in invertebrates. Evolution 2006;60:957–69.
  17. Gandon S, Day T. The evolutionary epidemiology of vaccination. J R Soc Interface 2007;4:803–17.

More information: PLOS Biology (2020). DOI: 10.1371/journal.pbio.3001000


Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.