The Vaccine Handbook: A Practical Guide for Clinicians, Twelfth Edition

CHAPTER 1

Introduction to Vaccinology

Immunization

Immunization is the process of protecting individuals from disease by making them immune. This is most often accomplished actively through vaccination, the delivery of antigens (substances that are foreign to the host) contained in vaccines for purposes of stimulating an immune response (immunization can also be accomplished passively by the administration of antibodies). While not technically correct in all instances, the terms vaccination and immunization are often used interchangeably.

Table 1.1 shows one approach to classifying vaccines used in humans. Broadly speaking, vaccines are either live or non-live; Table 1.2 lists generalizations about vaccines by type, which have practical consequences for storage, scheduling, efficacy, contraindications, and the potential for adverse reactions.

■ Live Vaccines

Live vaccines replicate in the host, stimulating immune responses that mimic those induced by natural infection. They are generally attenuated, or weakened, such that they cause subclinical infection with little risk of disease. There are several approaches to attenuation.

•Serial passage—This is the classical method of attenuating viruses, dating back to the 1930s when Thieler weakened yellow fever virus by serial passage in eggs. Serial passage causes the accumulation of mutations that, while adapting the virus to growth in vitro, render it less fit in vivo. Sabin developed the modern prototype of live attenuated virus vaccines by passaging poliovirus serially through monkey cells and demonstrating that oral administration of the attenuated virus protected against polio. Serial passage also was used to attenuate measles, mumps, and rubella viruses for use in vaccines. The virus used to make VAR was originally isolated from a child in Japan in the early 1970s; it was then serially passaged in human embryonic lung, embryonic guinea pig, and WI-38 (human diploid) cells to achieve attenuation. More recently, a strain of rotavirus that circulated in Cincinnati in the late 1980s was used to make RV1 by serial passage in Vero (African green monkey kidney) cells.

Attenuation of bacteria dates to the mid-1800s, when Pasteur protected animals from anthrax using a form of the bacterium that had been weakened using chemicals. In vitro passage also has been used to attenuate bacteria. For example, Bacille Calmette-Guérin (BCG), a vaccine that protects against disseminated tuberculosis and is used outside the US, was a strain of Mycobacterium bovis (which is related to M tuberculosis) originally isolated from a cow in 1908 and passaged over 200 times in culture.

•Heterologous host—This method dates to the late 1700s, when Jenner used an animal poxvirus to protect humans from smallpox. Many animal viruses are naturally attenuated for humans. For example, RV5 was derived from a bovine strain of rotavirus (WC3) that replicates in humans but does not cause disease. The wild-type virus does not induce sufficient protective antibody against human strains, so it was altered to express immunogenic surface proteins of human rotaviruses. This was accomplished through reassortment, whereby the parental strain was co-cultured with natural human strains—bovine viruses that accidentally packaged genes for the human G or P proteins (dominant protective antigens) were selected and propagated. The vaccine strains consist of viruses that in every way are identical to the naturally attenuated bovine virus, except that each one expresses an immunogenic human protein instead of the corresponding bovine protein.

•Engineered attenuation—The attenuated phenotype can be purposefully introduced into an organism through mutagenesis and selection. An example of this is the oral typhoid vaccine Ty21a, which was derived from Salmonella typhi strain Ty2 after treatment with a mutagenic agent and selection for attenuation. Attenuation can also be introduced through genetic manipulation. An example of this is the Ebola vaccine that was approved in 2019. This replication-competent viral vector vaccine consists of vesicular stomatitis virus (an animal pathogen that can infect humans but usually does not cause disease) in which the native surface glycoprotein gene has been replaced by the gene for the immunodominant surface glycoprotein of Ebola virus. The substitution confers definitive attenuation on the virus, and the vector not only expresses the Ebola antigen of interest but also amplifies in the host.

•Altered site of infection—The attenuated phenotype can be achieved by something as simple as using an unnatural route of inoculation. The adenovirus vaccine, which is used in the military, consists of enteric-coated tablets, one containing live (non-attenuated) adenovirus type 4, and the other, type 7. When these viruses are delivered to the gastrointestinal tract, they replicate and stimulate an immune response without causing disease.

■ Non-live Vaccines

Non-live vaccines have been referred to as inactivated, but that term is problematic. It is most accurate when referring to toxoids, which are inactivated toxins, although even here there is a problem—their toxicity is inactivated by chemical means, but toxoids are still active immunologically. Likewise, inactivated polio vaccine is noninfectious but still antigenic; in this context, killed might be a better descriptor. Some vaccines like HepB and HPV are made by in vitro expression of pathogen-derived genes; here, the vaccine antigens are not inactivated (they are natural proteins) or killed (they were not live to being with). For simplicity’s sake, the term non-live is used herein to mean a vaccine that does not replicate in the same sense that a live virus or bacterium does.

•Whole agent—Non-live whole agent vaccines date back to the late 1800s, when Pasteur used killed rabies virus to protect animals and, eventually, humans from rabies. Salk developed the modern prototype of non-live whole-virus vaccines by growing poliovirus in cell culture, purifying it, inactivating it with formaldehyde, and demonstrating that intramuscular injection of the inactivated virus protects against polio. HepA, JEV, RAB, and TBE vaccine are made in much the same way. The modern prototype of non-live whole bacterial vaccines is whole-cell pertussis, which was made from suspensions of cultured Bordetella pertussis organisms that were killed and detoxified. Because it contained every antigen from the live organism, this vaccine was both effective and reactogenic.

•Subunit—These vaccines use only a part of the pathogen instead of the whole organism.

–Toxoids: These are protein toxins elaborated by bacteria that have been chemically modified to reduce pathogenicity while maintaining immunogenicity. The only current toxoid vaccines are those for diphtheria and tetanus (inactivated pertussis toxin is generally not referred to as a toxoid, although it is pretty much the same thing).

–Purified subunits: In this case, an immunogenic component of the organism is physically purified and used (essentially in unmodified form) as a vaccine. Examples include the polysaccharide vaccines for S pneumoniae and S typhi, where the capsular polysaccharide is stripped from the cell surface and purified.

–Engineered subunits: Polysaccharide vaccines induce only short-term immunity, do not produce memory, and are not immunogenic in young infants. These problems were overcome by chemically conjugating the polysaccharides to proteins (see below).

–In vitro-expressed subunits: Subunits can be produced in vitro using recombinant DNA technology. The prototype here is recombinant-derived HepB, where the gene for hepatitis B surface antigen (HBsAg) was inserted into yeast cells, which produce large quantities of the protein for purification. A similar method was used to produce HPV. In this case, the gene for the L1 protein was expressed in either yeast (HPV4 and HPV9) or insect (HPV2) cells. L1 spontaneously aggregates into virus-like particles that look like viruses on the outside, are immunogenic, carry no genetic material and are incapable of replicating. In the case of COV-aPS (Novavax), the S-protein from SARS-CoV-2 is expressed in insect cells and formulated into a nanoparticle that stabilizes the antigen and presents it in similar fashion to a viral particle.1 Genetically engineered Chinese Hamster Ovary cells are used to express the antigens that go into HepB3, RSV (GSK), RSV (Pfizer), and RZV. Recombinant DNA technology has also been used to make vaccines for N meningitidis serogroup B and influenza.

–In vivo-expressed subunits: Subunits can also be expressed in vivo. mRNA-based technologies had been under development for 25 years when the COVID-19 pandemic hit in 2020, and they were used by Pfizer-BioNTech and Moderna to rapidly develop vaccines. Here, synthetically produced mRNA enters host cells through transfection at the site of inoculation or downstream in regional lymph nodes, directing cellular ribosomes to produce the protein of interest—in this case, the S-protein of SARS-CoV-2 (the virus that causes COVID-19), which mediates attachment and fusion and carries neutralizing epitopes. The protein expressed by host cells, especially dendritic cells and other professional antigen-presenting cells in lymph nodes and possibly even liver and spleen,2 is then presented to immune effectors, which go on to generate specific adaptive immune responses. Hurdles had to be overcome before mRNA vaccines were ready for widespread use. For example, mRNA is rapidly degraded by ubiquitous ribonucleases and therefore had to be protected in some way. Moreover, there had to be a way to reliably get the mRNA into the cytoplasm. One solution to both problems was to package the mRNA in lipid nanoparticles; this stabilizes the molecule and facilitates endocytosis. Native mRNA can stimulate strong innate immune responses, and thus had to be modified to be less reactogenic. The mRNA sequences were also optimized for protein expression; this, combined with stimulation of innate immunity by the lipid package and the modified mRNA, leads to strong antibody responses.

Proteins made by mRNA vaccines undergo post-translational modification, folding, intracellular transport, and surface expression just like proteins made during natural infection; the similarity of vaccine-encoded antigen to native antigen ensures robust and relevant antibody responses. Proteins expressed in transfected cells are taken up by antigen-presenting cells (APCs), degraded into peptides, and presented to helper T lymphocytes (Th-cells) in the context of MHC class II molecules (MHC-II) (see below). APCs themselves may also be transfected and produce the protein internally, in which case fragments of the nascent polypeptide chain are loaded into MHC class I molecules (MHC-I) molecules for presentation to cytotoxic T-cells (Tc-cells). mRNA vaccines are thus capable of stimulating strong antibody responses as well as cellular immunity. mRNA vaccines are non-infectious, cannot cause the target disease, carry no risk of integrating into host genetic material, and are not affected by pre-existing immunity, as might be the case for viral expression vectors (see below). All that is needed to make a vaccine is the sequence of the target protein; this means that development can be expeditious (it also means that the vaccines can be modified quickly if new strains of the virus evolve). The SARS-CoV-2 genetic sequence was made available on January 11, 2020; by January 13, Moderna had finalized the mRNA sequence for its vaccine and by February 7 a clinical batch was ready for analytical testing. mRNA vaccines are also rapidly scalable—by July 2021, Phase 3 studies involving tens of thousands of subjects were underway, and by years’ end millions of doses were ready for delivery.

Another way to express subunits in vivo is to deliver the instructions for making them through a replication-incompetent viral vector. As with mRNA vaccines, this technology had been under development when COVID-19 hit. The Janssen COVID-19 vaccine, which was no longer used in the US as of June 2023, is a recombinant adenovirus type 26 in which the E1 and E3 genes are deleted, rendering it incapable of replicating, and into which the S-protein gene is inserted. Host cells are infected and produce the S-protein, conferring the same advantages as for mRNA vaccines. The vaccine virus itself does not replicate; for production, the vaccine is grown in cells that complement the function of the deleted genes.

In classical vaccinology, the immunogenic subunits of a pathogen are identified, physically purified, and injected into animals to measure the immune response. Reverse vaccinology starts with the genotype (the genes) rather than the phenotype (the antigens), looking for genes that code for molecules that are likely to be important in immunity (ie, they are expressed on the cell surface and have a critical role in infectivity). The genes are then expressed, and the resultant antigens screened for immunogenicity in animals. Going forward, the identification of optimal antigens for use in vaccines may depend on systems immunology, a holistic picture of the immune response that is now possible with techniques like next-generation sequencing, protein and peptide microarrrays, flow and mass cytometry, and metabolomics, as well as advances in data processing and analysis.3

■ Passive Immunization

Passive immunization is the process by which short-term protection from disease is conferred by administration of antibodies. This process occurs naturally during the last 2 months of pregnancy, when large quantities of IgG are transferred across the placenta to the fetus, and it explains the relative protection that newborns have against invasive S pneumoniae and H influenzae type b infections, among others. Passive immunization is necessary for patients with humoral immune defects who cannot synthesize their own antibody. Polyclonal immune globulin is used to prevent specific infections such as measles and hepatitis A in vulnerable hosts, as the level of antibody against these viruses in pooled blood donations is sufficiently high. In some cases, hyperimmune globulin is used; this is derived from donors with high antibody levels to the pathogen (examples include varicella zoster immune globulin and hepatitis B immune globulin). Hyperimmune globulins contain antibodies to agents in addition to the one they target; while selected for their high specific antibody levels, the donors also have antibodies to other organisms. Antibodies also can be engineered for prevention of specific diseases, as in the case of the monoclonal antibody products against RSV and SARS-CoV-2. Antitoxins, also known as heterologous hyperimmune sera, are also used for passive immunization. These are produced in animals like horses and target toxins such as diphtheria, botulinum, and tetanus.

Passively acquired antibodies can inactivate live vaccines. MMR and VAR are not routinely given in the first year of life to avoid inactivation by maternal antibodies (these degrade sufficiently by 12 months of age). Likewise, administration of MMR and VAR is deferred in persons who receive blood products (which contain antibodies; see Table 5.2). Maternal antibodies also can interfere with the response to some non-live vaccines4; this is one reason why HepA is usually given in the second year of life. Passively acquired antibodies do not appear to interfere with vaccines administered at mucosal surfaces, which is why RV can be given as early as 6 weeks of age.

Basic Vaccine Immunology

Vaccines are designed to generate pathogen-specific antibodies and T-cells by stimulating the adaptive immune system, which recognizes and remembers specific pathogens and learns to respond to them more strongly after each exposure. What follows is a simplified version of the immune mechanisms that underpin vaccination.5 These concepts shed light on the differences between various types of vaccines, the duration of protection, dosing schedules, and other aspects of vaccine practice.

■ Antibodies

Antibodies are proteins that bind to conformational, 3-dimensional patterns called epitopes that are present on antigens. They constitute the humoral, or soluble, arm of the adaptive immune system and are produced as different immunoglobulin isotypes (IgG, IgA, IgM, IgE, IgD) that have different functions. A given antibody with a given antigenic specificity can be produced as one of several different isotypes. Antibody binding is specific in that each antibody molecule binds best to one epitope; any given antigen may express many different epitopes, and any given organism may have hundreds of different antigens. The possible consequences of antibody binding to bacteria are illustrated in Figure 1.1. Virus-infected cells may also be flagged for destruction by phagocytes or natural killer cells, a process known as antibody-dependent cell-mediated cytotoxicity (ADCC). The Holy Grail of vaccine development is the identification of epitopes expressed on pathogens that elicit broadly neutralizing antibodies—those that can take out many different strains of the organism.6

Antibody is the only element of the adaptive immune system that can prevent viral infection because it can neutralize a virus before it has a chance to replicate in cells. It is also the mainstay of protection against bacterial invasion since it can facilitate destruction of the organism before it has a chance to replicate. The battle between antibodies and pathogens takes place at different sites. At mucosal surfaces, where most pathogens attempt to gain entry, secretory IgA and serum-derived IgG antibodies are important. Live vaccines that replicate at mucosal surfaces (eg, RV and LAIV) have the advantage of inducing strong local IgA responses that can neutralize pathogens before attachment. Vaccines given parenterally are not as good at generating mucosal IgA, although many do. A vaccine that induces high levels of serum IgG not only protects against bloodstream invasion and infection of extravascular spaces, but also blocks infection at mucosal sites before the organism gains a foothold.

Antibodies are produced by plasma cells, which are derived from B-lymphocytes or B-cells. B-cells have immunoglobulins on their surface (mostly IgM, also referred to as the B-cell receptor) that express one and only one antibody specificity. During ontogeny, genetic rearrangements lead to a diverse repertoire of B-cell clones, each with its own unique antibody specificity (B-cells that emerge from this process with receptors that recognize self-antigens are deleted). B-cells that leave the bone marrow and do not encounter their cognate antigen are short-lived, but new B-cells with new specificities are generated every day; in this way, the B-cell repertoire is continuously refreshed. People are, therefore, walking around with millions of B-cells, each pre-committed to recognizing one and only one epitope. Theoretically, humans even have B-cells capable of recognizing the Andromeda strain,7 should it happen to (again) fall to earth. The point is that the B-cell repertoire theoretically covers every possible antigen—extant or imagined—that is not self. When a given B-cell finds its antigenic match, it is activated and proliferates; this is known as clonal expansion. The daughter cells eventually differentiate into plasma cells, which secrete large amounts of antibody.

There are two basic pathways through which antibody production occurs, and the differences between them are important to understanding how vaccines work.

■ The Extrafollicular Reaction

The extrafollicular reaction is best exemplified by the immune response to polysaccharides (Figure 1.2, Panel A). A pre-committed B-cell encounters its polysaccharide match (at the site of vaccination or in a lymph node to which the vaccine antigen was transported). This leads to activation, proliferation, and differentiation into plasma cells that migrate predominantly to the red pulp of the spleen and intramedullary areas of lymph nodes, where they begin producing antibodies. It is important to understand that these are germline antibodies, ie, antibodies transcribed from genes that were present in the B-cell to begin with. Germline antibodies have low affinity for their corresponding antigens because the genes encoding them are created by a random process during the B-cell’s development. These B-cells (and their genes) are selected during their development because they do not recognize self antigens, not because they will someday bind especially tightly to foreign antigens. Furthermore, most of the antibodies produced in the extrafollicular reaction are of the same isotype that was present on the B-cell surface, namely IgM, which does not offer the functional benefits of other isotypes like IgG. Finally, plasma cells produced through the extrafollicular reaction ultimately die out—no more antibodies produced, no more protection, no ability to remember the encounter and respond more quickly or decisively the next time.

This is called the extrafollicular reaction because it takes place outside of the germinal centers of lymph nodes. Antigens that elicit this response are called T-cell independent because the responding B-cells differentiate without much T-cell interaction. The characteristics of T-cell independent responses—rapid production (days to weeks) of short-lived, low-affinity, predominantly IgM antibodies without induction of memory—are hallmark features of the response to polysaccharide antigens, such as those in PPSV23. Children <2 years of age do not mount robust T-cell independent responses, making polysaccharide vaccines poor immunogens in that age group.

Another problem with polysaccharides is hyporesponsiveness—individuals who initially receive polysaccharide vaccines respond less well to subsequent polysaccharide challenge.8 It appears that the initial exposure to antigen uses up some of the pre-existing antigen-specific B-cell pool.9 Interestingly, infants receiving PCV7 have decreased responses to the S pneumoniae serotypes with which they are colonized at the time they are immunized, suggesting that colonization—a natural form of exposure to capsular polysaccharide—also induces hyporesponsiveness.10

It is important to point out that there is some crossover between immunologic pathways. For example, some degree of T-cell help (see below) may be available to extrafollicular B-cells, such that some isotype switching occurs and some memory may be generated.

■ The Germinal Center Reaction

Underpinning the adaptive immune system is the more primitive innate immune system, which initiates the battle against invading microorganisms in a nonspecific fashion. Cells of the innate immune system—most notably dendritic cells and monocytes—carry receptors that recognize conserved patterns among pathogens (pathogen-associated molecular patterns, or PAMPs) that are not found in self-tissues. Among these are Toll-like receptors (TLRs), each of which recognizes a different PAMP. TLR3, for example, recognizes double-stranded viral RNA; TLR4—endotoxin; TLR5—bacterial flagellins; TLR7—single-stranded RNA; TLR9—double-stranded DNA.11 Engagement of pattern-recognition receptors activates the cell, which then secretes cytokines (intercellular communication molecules) that activate and recruit other cells, setting up an inflammatory reaction. The result might be destruction of the invader through processes such as phagocytosis. Importantly, this is a one-time occurrence—once the pathogen is destroyed, there is no pathogen-specific memory of the encounter that might facilitate a response the next time around (although there is evidence that innate immune cells can be trained to be better killers and to more effectively trigger adaptive responses12).

The innate immune system is critically important because its activation can trigger and augment the adaptive immune response. The key link is provided by APCs, the most important of which are dendritic cells. Immature APCs circulate through the body or reside in tissues (immature dendritic cells in the dermis are called Langerhans cells). When a PAMP is encountered—in the form, for example, of a vaccine—the APCs begin to mature, express new receptors on their surface, and migrate through lymphatic vessels to regional lymph nodes. Some of them also engulf the vaccine antigens, degrade the proteins into small peptides, load the peptides into the groove of MHC-II, and express those molecules on their surface.

The mature APCs are now activated (secreting proinflammatory cytokines and expressing co-stimulatory molecules), flagged (by surface expression of antigen-derived peptides in the context of MHC-II), and have migrated to the follicular region of the lymph node. At this stage, the APCs are ready to engage immature Th-cells by interacting with the T-cell receptor (TCR) and the CD4 co-receptor (both are on the Th-cell). Each Th-cell is predetermined to recognize a particular peptide:MHC-II flag by virtue of its TCR and CD4 molecule. Each TCR has a unique antigenic specificity that was generated during development by a random process, much like the generation of germline antibodies (T-cells that emerge from this process with receptors that recognize self-antigens are deleted or inactivated). Through soluble cytokines and co-stimulatory receptor-ligand interactions, contact with the right APC results in activation and maturation of the Th-cell into antigen-specific subtypes that differ in their cytokine profiles and functions. The most important of these are Th1-cells, which help establish cellular immunity, and Th2-cells, which help establish humoral immunity (others include Th17-cells, which are involved in mucosal immunity, and Th follicular-cells, which help in affinity maturation). Several factors determine whether a Th-cell differentiates into a Th1- or Th2-cell during immunization; among these are the dose of antigen, route of administration, and the effect of adjuvants (see below). These factors, therefore, also determine whether humoral or cellular responses predominate.

Th1-cells have some direct antimicrobial functions. More importantly, though, Th1-cells seek out Tc-cells and macrophages, coaxing them into performing cytotoxic functions (see below). Th2-cells also have some direct antimicrobial functions, particularly against parasites. More importantly, though, Th2-cells seek out their unique B-cell matches to help them make antibody (to the antigen from which the peptide in the MHC groove was originally derived). B-cells are primed to be helped by engulfing some of the bound antigen, digesting the proteins, and presenting the peptides on their surface in the context of MHC-II, much like dendritic cells do (Figure 1.2, Panel B). As further evidence of the interaction between the innate and adaptive immune systems, there is evidence that neutrophils (which are part of the innate immune system) are involved in stimulating antibody responses to vaccines.13

The interaction between B-cells and Th-cells takes place in the germinal centers of lymph nodes. The signals provided by Th2-cells, as well as the persistence of antigen shuttled there by APCs, drive B-cells to undergo massive clonal proliferation—producing millions of daughter cells capable of making the same antibodies. As they proliferate under these conditions, the B-cells undergo a process of somatic hypermutation, whereby the genes encoding the immunoglobulin-combining region mutate at exceptional rates, randomly producing a spectrum of antibodies with varying affinities for the antigen. Those B-cells expressing high-affinity antibodies are selected for and clonally expanded. The result is a set of dominant B-cell clones that produce high-affinity antibodies, ones that bind the antigen better than the germline antibodies (this is referred to as affinity maturation). This process takes a week or two.

Interestingly, the genes encoding TCRs do not undergo somatic hypermutation as do the genes encoding antibodies. One reason for this is the central role that T-cells play in adaptive immunity—they are involved in both humoral and cellular responses, and constraining their diversity helps limit autoimmunity. While we want TCR diversity so we can respond to a variety of foreign antigens, TCRs must also recognize MHC molecules as part of the antigen presenting complex; too much diversity could result in T-cells that are prone to self-recognition, leading to autoimmunity. On the other hand, we do not want T-cells that fail to recognize the MHC portions of the antigen-presenting complex, because then adaptive immune responses would be suboptimal.

In the germinal center, signals from Th2-cells drive isotype switching (from IgM to other isotypes) as B-cells transform into plasma cells. The result is large numbers of plasma cells that migrate to the bone marrow and produce high-affinity IgG, as well as other isotypes. Th2-cells also drive the evolution of memory B-cells, which migrate to the spleen and lymph nodes and wait there—sometimes for decades—until they again encounter their cognate antigen, at which time they rapidly proliferate and differentiate into antibody-producing plasma cells. This is called the anamnestic response. Memory Th-cells are also generated, but their numbers wane with time.

The germinal center reaction has implications for vaccination.

•Stimulating innate immunity—The more innate immunity is stimulated, the more robust the adaptive immune response will be. For example, intradermal vaccination may stimulate more robust responses than intramuscular injection because there are more immature dendritic cells in the dermis.14 As another example, live viral vaccines are more immunogenic than non-live ones because they can disseminate and encounter dendritic cells at many sites, ultimately establishing multiple foci of germinal center reactions. Live vaccines also express PAMPs that can activate innate immune cells.

•Adjuvants—Adjuvants are substances that potentiate the immune response to vaccine antigens (Figure 1.3). They are necessary for some non-live subunit vaccines, but they are not necessary for live vaccines because these stimulate innate immune responses on their own. For many decades, the only adjuvant used in human vaccines was alum, but the last decade has seen growth in the number of available adjuvants, improvement in potency, development of adjuvant systems (combinations of immune stimulants) and targeting of specific elements of the immune system (Table 1.3). Adjuvants may allow for antigen sparing; for example, an adjuvanted version of IIV is just as immunogenic as a non-adjuvanted version, even though it contains one-fourth the amount of antigen.15 Adjuvants also may cause epitope spread, a broadening of the antibody repertoire such that more epitopes on a given antigen are recognized.16

•Priming and boosting—Ultimately, the magnitude and duration of the antibody response to non-live vaccines depends on the immunologic set point following primary immunization. In other words, the more germinal centers that are formed, the more long-lived plasma cells and memory B-cells there will be. In addition to optimizing antigen dose and using adjuvants, primary immunization is enhanced by multiple doses of the vaccine given in succession, classically separated by 1 or 2 months (as in the 2-, 4-, and 6-month schedule for PCV). These doses are given too soon to exploit memory responses, but they do drive the process of affinity maturation. Following priming, vaccine schedules take advantage of anamnestic responses, which boost the level of high-quality antibody (Figure 1.4)17; in general, the longer the interval to boosting, the better the antibody response.18 The need to wait until the germinal center reaction is complete explains the longer interval between the primary series of a vaccine and the booster doses (as in the 12- to 15-month dose of PCV). The response to booster doses of vaccine mimics what happens when the natural pathogen is encountered.

A good example of a booster is Tdap, which boosts the immunity to tetanus, diphtheria, and pertussis that has waned since the childhood series was completed (see Chapter 14: Diphtheria, Tetanus and Pertussis). One might notice that the package insert for Infanrix (a brand of DTaP) states that the doses given at 2, 4, and 6 months of age constitute a primary series for pertussis and the toddler and school-aged doses are boosters. The Daptacel (another brand of DTaP) package insert states that the primary series for pertussis is 4 doses (2, 4, 6, and 15 to 20 months), and only the school-aged dose is considered a booster (similar language is contained in the package insert for Pentacel and Vaxelis, vaccines that are based on Daptacel). The reasons behind this have more to do with regulatory precedent and interpretation than they do with the biology of immune responses; arguably, the long interval between Doses 3 and 4 of Daptacel takes advantage of priming, B- and T-cell memory, and anamnestic responses, all characteristics of boosters.

There may be a downside to the immunological set point when it comes to protection against pathogens that evolve antigenically over time. It was known as far back as the 1940s that prior exposure to, or vaccination with, one strain of influenza virus can impair the response to a new related but antigenically distinct strain, a phenomenon referred to as original antigenic sin.19 It’s as if the first exposure to antigen imprints on the immune system what the antibody repertoire ought to be in response to subsequent exposure to a similar antigen; however, if the new antigen is sufficiently different, the antibody response, which is biased towards the previous antigen, might be less effective against the new version, at least when compared to the response of naïve individuals. The key to this phenomenon lies in the germinal center and the continuous mutation and selection process that the B-cell receptor undergoes.20 Imprinting could in part explain why individuals vaccinated against the ancestral strain of SARS-CoV-2 early in the pandemic were so easily infected with later variants21; at the same time, vaccinated people were protected against severe disease, because T-cells, unlike B-cells, are long-lived, do not undergo somatic mutation after they leave the thymus, and are broadly reactive.

•Specificity—Hypermutated, high-affinity antibodies target single epitopes. For this reason, most non-live vaccines are very specific in the protection they afford. For example, IPV contains formalin-inactivated, disrupted viral particles from three different serotypes of poliovirus because the antigens from any given serotype do not induce antibodies that will neutralize the other serotypes. Germinal center reactions by and large generate homotypic responses, ie, responses directed against the antigen used as the vaccine. Heterotypic, or cross-protective, responses occur only if there is sufficient similarity between the antigens of different strains, or if there is a common antigen between them. Tc-cells offer much more potential for cross-strain protection (see below). One of the biggest questions about the first COVID-19 vaccines, which were based on antigens derived from the original strain of SARS-CoV-2, is whether they would protect against emergent strains. Those strains emerge under the immunologic pressure represented by populations previously infected with and/or immunized to the original strain, among other evolutionary forces22; the extent to which emergent strains are antigenically different from the original—yet still fit for replication and transmission—is called antigenic escape.

•Persistence of antibody—The germinal center reaction peaks in several weeks, after which it is terminated. The mechanisms by which serum antibodies persist for long periods of time—something so critical to maintaining protection—are not yet fully understood.23 Some models suggest that there is constitutive differentiation of memory B-cells into plasma cells, stimulated by persistent antigen, reinfection, or exposure to cross-reactive antigens. Other models propose nonspecific, or so-called bystander, activation of memory B-cells. These models have important implications for vaccine programs. For example, early studies of VAR showed persistent if not rising titers of antibody over time. But these studies were done at a time when the wild-type virus was still circulating; therefore, immunized individuals might have experienced repeated subclinical reinfections with natural virus, coaxing memory B-cells to differentiate into plasma cells and boosting antibody production. As transmission of natural varicella decreased, some studies began to show waning immunity over time (this observation made it less likely that antibody persisted because of boosting from periodic reactivation of latent vaccine virus). Although initially intended to immunize those who failed to seroconvert after the first dose (so-called primary vaccine failures), the second dose of VAR also serves to boost immunity in those individuals whose antibody levels have fallen low enough to allow take, or replication of the vaccine virus.

Other models suggest that plasma cells derived from the germinal center reaction can live for a very long time. Either way, for some vaccines, it is necessary to periodically stimulate an anamnestic response through booster vaccination. One thing is clear—to prevent serious infections that have a short incubation period (invasive meningococcal disease is a good example), one must have enough circulating antibody when the pathogen is encountered. Anamnestic responses, while brisk, are not fast enough to be of much help when dealing with rapidly replicating bacteria. They may be sufficient, however, for pathogens with longer incubation periods. Thus, for example, although antibodies to HBV may wane with time, protection against disease does not wane—there is plenty of time for memory responses to kick in before the virus can do damage.

•Making better immunogens—As mentioned above, polysaccharides are poor immunogens. Vaccinologists have learned, however, to harness the power of the germinal center reaction to enhance polysaccharide immunogenicity (Figure 1.2, Panel B). The polysaccharides are chemically attached to protein carriers—among the variety of carriers used are an outer-membrane protein from N meningitidis (Hib-OMP [PedvaxHIB]); the mutant diphtheria toxin CRM (PCV13-CRM [Prevnar 13], PCV15-CRM [Vaxneuvance], PCV20-CRM [Prevnar 20], and MenACWY-CRM [Menveo]); tetanus toxoid (Hib-T [ActHIB, Hiberix], MenACWY-T [MenQuadfi]); and diphtheria toxoid (MenACWY-D [Menactra]). B-cells that are pre-committed to producing anti-polysaccharide antibody are stimulated by their encounter with the antigen (stimulation occurs through cross-linking of the B-cell receptor on the cell surface). They also engulf the bound antigen, digest it, and present peptides from the protein portion of the vaccine on their surface in the context of MHC-II (there is some evidence that B-cells and other APCs can also present polysaccharide antigens that are attached to peptides24). So, what you have is a B-cell committed to making anti-polysaccharide antibodies that is displaying a peptide:MHC-II flag on its surface. All it takes now is for APCs to display the same peptides in the context of MHC-II, stimulating Th2-cells, which then find their B-cell matches and help them make antibody through the germinal center reaction. In essence, the Th2-cells think they are helping B-cells make antibodies to the protein antigen from which the peptides were derived, when, in fact, the flagged B-cells make polysaccharide antibody. By converting a T-cell independent response to a T-cell dependent one, the antigenic shortcomings of polysaccharides are overcome (Table 1.4).

■ Cytotoxic T Cells

Tc-cells are the main effectors of the adaptive cellular immune response. Like Th-cells, they express the TCR on their surface, which has a unique antigenic specificity encoded in the germline. Unlike Th-cells, which express CD4, Tc-cells express the CD8 co-receptor, which directs engagement with MHC-I. Like MHC-II, MHC-I loads pathogen-derived peptides into its groove and presents those peptides on the cell surface. However, instead of coming from the digestion of exogenous proteins that were engulfed by the cell, the peptides loaded into MHC-I are, for the most part, made inside the cell—by infecting viruses or other intracellular pathogens.

MHC-I is expressed on all nucleated host cells, including APCs. Naïve Tc-cells, predetermined to recognize a particular peptide:MHC-I flag, engage infected APCs that express those peptides in the context of MHC-I (Figure 1.5). While this leads to activation of the Tc-cell, the Tc-cell does not become a killer until it receives additional signals—those coming in the form of cytokines from Th1-cells, which in turn have been activated by engagement of APCs expressing pathogen-derived peptides in the context of MHC-II (Th1-cells have other functions as well, such as activation of macrophages). The respective, critical role of the two different types of Th-cells is obvious—Th2-cells help B-cells produce high-quality antibodies and memory cells, and Th1-cells help Tc-cells become killers (and memory cells).

Tc-cells find and destroy infected cells that express pathogen-derived peptides on their surface in the context of MHC-I. Killing occurs through the release of cytotoxins that create holes in the cell membrane, and by induction of apoptosis, or programmed cell death, through receptor-ligand interactions and the release of certain enzymes.

The unique characteristics of Tc-cells and their generation have several implications for vaccination.

•Viruses vs bacteria—Tc-cells are more important in controlling viral rather than bacterial infections. This is because most bacteria replicate outside of cells, and therefore do not generate cells flagged with endogenously derived peptide:MHC-I molecules. Viruses, on the other hand, only replicate inside cells by usurping the machinery for protein synthesis; infected somatic cells routinely carry endogenously derived peptide:MHC-I flags on their surface and are therefore recognizable by Tc-cells. Certain APCs, like dendritic cells, can cull peptides from the extracellular environment and present them in the context of MHC-I, enabling them to stimulate Tc-cells without themselves being infected; this process is called cross-presentation25 (dendritic cells can even acquire the entire peptide:MHC complex from other APCs in a process called cross-dressing26).

Vaccines against bacteria are designed to maximize high-quality antibody production, partly through the generation of large pools of Th2-cells. The ideal vaccine for a virus would maximize high-quality antibody production (to inactivate the inoculum, thereby preventing infection), as well as generate Tc-cells (to kill infected cells, thereby controlling infection).

•Preventing infection vs limiting disease—Tc-cells operate after infection has taken place; they limit but do not prevent infection. Unlike antibody, a Tc-cell cannot kill a free virion—it must wait until a cell is infected and expressing viral peptides. If a vaccinated person is exposed to a virus, the first line of defense is the antibody that resides in the immediate local environment. That antibody can neutralize the virus and prevent infection. If it does not, the virus enters cells and expresses proteins; this activates memory Tc-cells, which limit disease by destroying those cells (in some cases, however, the inflammation caused by activated Tc-cells may contribute to disease manifestations). Antibody may also play a role in the destruction of infected cells through ADCC.

Certain differences between varicella and zoster vaccines are instructive. VAR is designed to stimulate antibodies that can prevent primary infection (varicella, or chickenpox). Herpes zoster, or shingles, results from reactivation of endogenous VZV that has been latent since the person had varicella. As such, zoster vaccine is designed to expand the pool of memory Tc-cells that can be stimulated by cells expressing viral peptides on their surface as reactivation begins. RZV, a non-live vaccine containing one surface glycoprotein from the virus, expands memory Tc-cells through cross-presentation (see below) and the guiding power of a strong adjuvant.

•Live vs non-live vaccines—Live viral vaccines infect cells and direct the synthesis of virus-specific proteins; peptides from nascent polypeptide chains are loaded into MHC-I molecules and expressed on the cell surface in the MHC-I context, stimulating strong T-cell responses. Live vaccines are also able to amplify and disseminate antigens, and they stimulate strong innate immune responses. Peptides from non-live protein vaccines can be presented on the cell surface in the context of MHC-I through cross-presentation. Whereas this process is less efficient, it nevertheless explains how some non-live vaccines can induce Tc-cell responses. mRNA vaccines mimic live vaccines in that they direct endogenous protein synthesis and stimulate strong cellular responses,27 but they do not replicate. Non-replicating vectored viral vaccines act similarly.

•Broad protection—Tc-cells recognize peptides that are derived from any protein made by the pathogen, some of which are well-conserved between strains. Thus, Tc-cells stimulated by vaccination might be able to limit disease caused by a variety of strains, even when the elicited antibodies are strain-specific.

■ Correlates of Protection

Ideally, vaccines are shown to be effective in randomized, blinded, placebo-controlled trials, where one group of subjects gets vaccine, another gets placebo, and the measured outcome is efficacy, or ability to prevent disease (or infection). This works if the disease is prevalent, such that the number of subjects needed in a clinical trial is within reach. For diseases that are relatively rare, demonstrating efficacy may not be feasible.28 In these situations, immunogenicity, or the ability to generate an immune response, is relied upon as a predictor of efficacy. The concept of correlates of protection (CoP, also known as correlated immune marker) connects the measured immune response to the predicted efficacy.29 CoP may be mechanistic, ie, biologically responsible for protection (also referred to as a protective immune marker), or non-mechanistic, ie, statistically related to protection but not biologically responsible (also referred to as a surrogate immune marker). A good example of a mechanistic CoP is bactericidal antibody to N meningitidis (the antibody being measured is the one that kills the bacterium). However, for many