Vaccine-induced immunologic memory and pace of pathogenesis: predicting the need for boosters

Pediatric and adult vaccine scientists, public health authorities, vaccine companies and others are considering the question of the need for booster vaccinations to maintain the efficacy of recommended vaccines [1]. This is a complicated question, due to gaps in our knowledge and often a lack of specific evidence in humans. This article will examine the basic knowledge of B- and T-cell memory and data for specific vaccines and specific diseases that serve as examples of important differences among vaccines and differences in the pace of pathogenesis.

B-cell memory

For diseases where vaccine-induced protection is mediated by antibody, the immune memory is characterized by:

• • A more rapid production of antibodies than that which follows primary antigen exposure

• • A switch in antibody class from predominantely IgM to predominantely IgG

• • A higher antibody level following re-exposure to an antigen than that which follows primary exposure

An additional feature of a memory response is the production of higher-affinity antibody (and an antibody population of higher avidity) for the antigen, as a result of a process known as affinity maturation. The demonstration of higher avidity antibody has been used in some vaccine studies as a surrogate marker of immune memory.

Kinetics of B-cell memory response

With respect to the rapidity of the memory antibody response available data for tetanus [2], Haemophilus influenzae type b (Hib) [3] and other vaccines support the notion that a detectable response occurs within 2–7 days following antigen exposure, with some variability in that range depending on the host and perhaps the antigen. In the window of time from antigen exposure to detectable antibody presence in serum, the antigen is taken up and processed by antigen-presenting cells (APCs; e.g., dendritic cells and macrophages), interaction of APCs with B cells and T cells occurs in peripheral lymphoid tissues, B cells proliferate and mature to plasma cells, and then the plasma cells produce and release circulatory antibodies.

There are several in vitro assays to evaluate the duration of immunity to a pathogen, but the best tool is to perform challenge experiments. To simulate what might happen when exposure to a vaccine-targeted pathogen occurs in nature, the current challenge experiment strategy involves the use of a reduced antigen dose and subsequent characterization of the immune response [4]. A reduced dose is intended to resemble the amount of antigen that might be experienced by a host. However, this is an imperfect model because little is known about the actual pathogen-specific antigen dose for most organisms, and administration of vaccine by the parenteral route is clearly not the same as the reality in nature, where exposures mostly occur via the mucosal route (respiratory, genitourinary and GI tracts).

In addition, the impact of immune modulation as might occur with concurrent coinfections (e.g., a viral upper respiratory infection occurring simultaneously to exposure to a bacterial pathogen, such as Hib) is not simulated.

Affinity maturation

After primary exposure to a vaccine antigen, the availability of antigen decreases over time as it is taken up by APCs and/or degrades. When antigen becomes more limited, those B cells exposing surface IgG that has a better fit with the immunogenic epitopes of the antigen will be stimulated to divide and persist. Subsequent exposure to the antigen will stimulate these B cells to produce higher-affinity antibody [5]. In general, higher-affinity antibodies are more bactericidal and virucidal than lower-affinity antibodies. This explains why measures of antibody affinity (or avidity) have more recently been included in the evaluation of immune memory. Affinity maturation takes time and higher-affinity antibodies are produced when vaccine antigens are given at wider-spaced intervals. This is because as time passes and antigen becomes less available, natural selection favors cells that express the highest-affinity IgG receptor molecules.

Higher memory antibody levels

A higher secondary memory antibody response compared with primary antibody responses is characteristic. However, for some vaccines, the antibody levels are very high following the primary series and booster vaccinations do not produce significantly higher levels. This is because the immune system controls the robustness of an immune response so that excessive responses do not occur (i.e., progression to multiple myeloma). Therefore, the absence of a multifold, higher memory antibody response compared with a primary response is sometimes to be expected.

Persistence of antibody over time

Antibody levels decline after vaccination. The half-life of immunoglobulins is approximately 30 days, and this antibody decay results in levels that often decline to undetectable amounts over time if antigen stimulation does not occur. This opens the debate about the necessity for a circulating level of antibody to prevent infection in vaccinated individuals versus the notion that immune memory will be triggered on exposure to a vaccine-targeted pathogen and antibody will be produced quickly to provide protection against infection. An examination of actual antibody measurements for some vaccine antigens validates the principle of a decline in levels over time, sometimes to undetectable amounts. But for some vaccine-induced antibodies, persistence in levels occurs well beyond the time when they should have disappeared (i.e., a 30-day half-life). Several mechanisms have been hypothesized to explain the persistence of low levels of circulatory antibodies long after antigen exposure has occurred.

The three hypotheses are called:

• • Treadmill of antigen

• • Sequestration in bone marrow of B memory cells

• • Polyclonal B-cell activation [5–8]

In the treadmill hypothesis, the notion is that small amounts of antigen are retained inside APCs of peripheral lymph nodes and spleen. The antigen is periodically released as those cells die, only to be taken up by other APCs, re-presented to B and T cells (like a treadmill) and those B and T cells cooperate in the production of new antibody. Not much antigen is retained, so the numbers of newly stimulated B cells are few, and the antibody amounts are small. The second hypothesis suggest that memory B cells migrate to the bone marrow and may remain sequestered there. In the bone marrow sanctuary, even without further antigen stimulation, the memory B cells can continue to periodically divide and replenish their small pool. Some of these new memory B cells go on to mature to plasma cells, leave the bone marrow and produce small amounts of antibody in the serum. The third hypothesis suggests that the maintenance of a small circulatory pool of specific memory B cells capable of maturity to plasma cells does not require ongoing exposure or processing of antigen. Instead, the notion is that polyclonal B cells activators maintain a small memory B-cell pool.

In addition, natural boosting can occur. A good example involves Hib. Before Hib vaccines became available, it was well established that children developed ‘natural antibody’ to the Hib polysaccharide over time, such that by 5 years of age virtually all children had detectable protective antibody levels. (Even today, that is still true and accounts for the fact that boosters of Hib conjugate vaccine are not recommended beyond 5 years of age). This rise in natural antibody occurs partly due to asymptomatic nasopharyngeal and oropharyngeal colonization with Hib, but this is not common. Instead, most natural antibodies to the Hib polysaccharide are induced by colonization of the gut by a strain of Escherichia coli (K1) that exposes a polysaccharide capsule nearly identical to the Hib capsule structure [9]. E. coli K1 colonization is the source of most antigen stimulating anti-Hib capsule antibody. Thus, a combination of antigen sources may stimulate circulation of antibody by:

• • Asymptomatic colonization to the organism

• • Exposure to an organism expressing a cross-reactive antigen

• • Subclinical reinfection

Natural boosting may decrease over time as a pathogen circulates less widely in a population. This is an ongoing issue relative to several vaccines, notably varicella, since the absence of natural boosting among vaccinated children could lead to susceptible adolescents or adults.

T-cell memory

Effector memory T cells reside in peripheral tissues whereas another pool, termed central memory T cells, reside in lymphoid organs [10]. Phenotypically, human-naive T cells are CCR⁺, DC27⁺, CD28⁺, CD45RA⁺; central memory cells are CCR⁺, DC27⁺, CD28⁺, CD45RA^-, and effector memory T cells are CCR^-, DC27⁺, CD28⁺, CD45RA^-[11].

Kinetics & cytokine response of memory T cells

The continuous recirculation of effector memory T lymphocytes from the blood into tissue is a fundamental feature of the immune system that ensures the dissemination of memory T cells of particular specificities throughout the body. For memory T cells, the capacity to enter tissues from the blood is integral to the process of immune surveillance for previously encountered antigens. It is well established that memory CD4 and CD8 T cells respond more quickly, and with a wider array of cytokines (CD4) and responses observed in cytotoxic molecules (CD8), than a primary response. This faster and broader response, in association with antibody, has the potential to control infection quickly on re-exposure to a pathogen. These characteristics of the memory T-cell response are a consequence of clonal expansion of antigen-specific T cells during a primary infection and reprogramming of the T-cell gene-expression profile. In this way, larger quantities of cytokines and cytotoxic molecules (e.g., perforin and granzyme B) are produced [12–14]. In both CD4 and CD8 in vitro models, the generation of T-cell effector functions are evident after approximately 2 days of stimulation [15–18].

Maintenance of T memory

Is antigen needed for the maintenance of T-cell memory? This issue has been investigated for at least the past 30 years. The answer appears to be that different antigens have quite different half-lives; peptide versus replicating virus represents the extremes. Several studies have argued that antigen is needed [19–22], while others say it is not [23–26]. Also, some experiments demonstrate that CD8 T-cell memory is subject to exhaustion if exposed to too much antigen [27].

Pace of pathogenesis

The likelihood that a memory response will be protective in the absence of circulatory antibody must be assessed in light of the pace of pathogenesis for different organisms. Where evidence is available, an argument will be proposed to support the hypothesis that, in some cases, memory might be enough and, in other cases, it will not be.

Three examples as models

Haemophilus influenzae type b

The pace of pathogenesis for Hib is rapid. In an animal model, it was shown that, in a matter of hours, single Hib bacteria can adhere to the nasopharynx, gain entry to the bloodstream, replicate and seed the meninges [28]. Clinically, most children are ill with Hib invasive disease for 1–3 days before they seek medical attention. Hib infection is prevented by the presence of antibody directed to the Hib polysaccharide capsule. When the Hib polysaccharide vaccine was first introduced, several articles appeared reporting an increase in invasive disease several days after vaccination [29,30]. This increased susceptibility was attributed to the formation of immune complexes of the vaccine polysaccharide with pre-existing natural antibody. Before an active immune response occurred, the temporary depletion of antibody resulted in an increased incidence of disease. Subsequently, the Hib conjugate vaccines were developed and introduced in the developed world. Whereas the Hib polysaccharide vaccines did not induce immunologic memory, the Hib conjugate vaccines did. When the Hib conjugate vaccines were combined with diphtheria, tetanus and acellular pertussis (DTaP) vaccines, a diminution in antibody levels was observed [31,32]. There followed a debate about the necessity for a minimum circulatory level of antipolysaccharide antibody versus the potential for immune memory to respond with antibody production in time to prevent infection after exposure occurred [33]. The emergence of invasive Hib disease in children from the UK, who had low antibody responses and no boosters, settled the debate [34–36].

Hepatitis B

The pace for pathogenesis of hepatitis B is slow. Clinically, before hepatitis B vaccines became available, the administration of passive antibody in the form of intramuscular γ-globulin could be carried out 2 weeks following exposure and still be effective. In the postvaccine era, it has been observed that many vaccinated individuals who develop high levels of antibody have their antibody levels fall to undetectable levels over time. Nevertheless, there is no disease breakthrough. Memory antibody responses occur following exposure in sufficient time to afford protection [37].

Human papillomavirus

Controversy now exists on the possible need for human papillomavirus (HPV) boosters as antibody levels wane. It is well recognized that the mechanism of protection following HPV vaccination is the production of neutralizing antibody. Animal models suggest that the pace of pathogenesis for HPV is a matter of a few hours for the virus to exist in the vicinity of basal epithelial cells before it enters the cell and becomes largely inaccessible to antibody. On this basis, one would expect that it would be necessary for a minimum level of antibody to be present in mucus (whether locally produced or a transudate of serum) at the time of exposure in order to prevent infection. Immune memory responses occurring 2–5 days later would be ineffective. However, there are issues with the animal model and with the assays performed by both current manufacturers of HPV vaccines. The ELISA used by GlaxoSmithKline measures all binding antibody and not neutralizing-specific antibody. The monoclonal capture assay used by Merck measures the displacement of a monoclonal antibody with a good fit for the major neutralizing antigen epitope for HPV serotype 16, but with only a fair fit for the major neutralizing antigen epitope for HPV serotypes 6, 11 and 18. Therefore, currently there are no definite answers.

Conclusion

B-cell and T-cell memory responses provide potential protection from disease approximately 2–5 days after exposure. In the time interval before these effective responses occur, the innate immune system and pre-existing immune protection (largely reflected in circulating antibody levels) must hold the infection in check. If the pace of pathogenesis is a matter of hours to less than 2 days, the memory response alone will probably not be sufficient in some individuals to prevent onset of diseases, such as Hib and meningococcal infection. If the pace of pathogenesis is slow (>2–5 days), immune memory responses should occur in time to prevent disease; for example, hepatitis B. For some vaccines and diseases, these issues are not currently well understood. Monitoring for breakthrough infections will provide answers.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.