The correlates of malaria immunity
Vaccines are probably the single most cost-effective public health intervention, past, present and possibly future, to control emerging and re-emerging infectious diseases. A vaccine is given in order to stimulate the development of adaptive immune responses to fight a particular pathogen against which the vaccine has been developed. Regardless of the success in the development of malaria vaccines, there is still a lack of understanding of individual immunity against malaria. Since the work of Koch on Java Island at the end of the 19th Century, which showed that adults who survive malaria infection acquire a highly effective immunity, the mechanisms involved and how they operate remains partly unknown, although the antibody that blocks the invasion of merozoites into erythrocytes appears to play a crucial role [
11].
Some work has been done to determine which protective antigens or epitopes can be used in the construction of recombinant, subunit or synthetic malaria vaccines [
12‐
14]. Surrogate markers of antibody efficacy currently rely on in vitro assays that are laborious and difficult to reproduce, and it remains unclear if such in vitro assays are predictive of functional immunity in humans due to the lack of suitable animal model permissive for
P. falciparum[
15]. The work leading to antibody-dependent cellular inhibition (ADCI) is an important cue in the case presented in this commentary on the need of more human evidence based hypothesis. This approach led to studies to correlate clinical protection in an endemic population with the immune responses to malaria vaccine candidates in development and gave insights into the relative proportion of cytophilic antibodies to non-cytophilic antibodies (the C:NC ratio) as being the most important surrogate marker of protection to date[
16‐
18].
Currently, there are no clear correlates of immunity against pre-erythrocytic and blood-stage parasites. Immuno-assays can be validated only once a vaccine demonstrates efficacy in a clinical trial. Once an immune correlate for protection is identified, it can be used for decision making in clinical development. Immunoepidemiological studies have demonstrated that immunity against blood stage
Plasmodium falciparum is associated with the acquisition of anti-parasite antibodies of the cytophilic subclasses [
19], and in particular IgG3 [
18,
20‐
26].
Recently, it has been shown that, there is an association between the frequency of RTS,S/AS01E induced (circumsporozoite protein) CSP-specific CD4+ T cells and protection from clinical malaria, most strongly seen for IFNc-IL2-TNF + CD4+ T cells. Furthermore, there were significant interactions between CSP-specific TNF + CD4+ T cell responses and anti-CSP antibodies induced by RTS,S/AS01E vaccination. This interaction suggests that the protection afforded by the combination of CD4+ T cells and anti-CSP antibodies is greater than would be predicted by their sum [
27].
RTS,S vaccine candidate induces high concentrations and frequencies of antibodies and CD4+ T cells, respectively, specific for circumsporozoite protein (CSP) [
28,
29]. Anti-CSP antibodies correlate with protection against infection in malaria naïve adult challenge studies [
28] and field studies in young children [
30], against clinical malaria in trials with young children in Kenya/Tanzania [
31] and in Gabon/Ghana/Tanzania [
32], but anti-CSP antibodies did not correlate with protection against clinical malaria in a trial with older children in Mozambique [
33]. Anti-CSP antibodies could protect by a variety of mechanisms including complement activation, antibody dependent cellular cytotoxicity, sporozoite neutralization, and/or FccR mediated phagocytosis [
34].
Immunologic analyses indicate that high titre anti-CS IgG are most strongly associated with RTS,S-mediated protection, with an important additive component from CS-specific Th1 cells. One recent study highlighted a correlation between CS-specific TNFa(+) CD4 (+) T cells and reduced morbidity, which requires confirmation in other studies [
31]. The first results from the Phase 3 trial were published and were in line with expectations from the Phase 2 trials [
7,
35], expect for the young age group which provided modest protection against malaria [
8]. A likely explanation for the lower vaccine efficacy among infants is an age-dependent differential immune response to the vaccine. This concept is supported by the lower anti-circumsporozoite antibody titers observed in infants as compared with titers in older children reported previously [
7].
Infants may have mounted a lower immune response than older children owing to coadministration of RTS,S/AS01 with routine EPI vaccines, an inhibitory effect of maternally derived anti-circumsporozoite antibodies, an absence of priming with hepatitis B vaccine or with
P. falciparum infection, or the infant’s immature immune system. Coadministration of RTS,S/AS01 with the pentavalent vaccine and the oral poliovirus vaccine might have resulted in immune interference and contributed to the lower anti-circumsporozoite antibody titers in the younger infants [
8].
There are many lessons to be learned from the RTS,S trials including the major contribution of sporozoite challenge trials, the importance of adjuvant, dose and schedule optimization [
36].
In a Phase 2a experimental sporozoite challenge trial in malaria non-immune Caucasian volunteers, vaccine related partial but modest protection against sporozoite challenge was observed in terms of a delay in time to parasitaemia [
37], although no sterile protection was observed. A recently completed Phase 1b vaccine trial in semi-immune Tanzanian adults and children confirmed the safety and immunogenicity of the platform. In addition, an exploratory analysis showed a reduced incidence of clinical episodes of malaria. As interesting as it is, this requires confirmation in field efficacy studies [
38].
Other vaccines based on irradiated sporozoites or genetically modified attenuated sporozoites have provided protection in challenge models [
39]. Such whole organism attenuated vaccines may provide effective protection against malaria and significantly reduce parasite transmission. Over 1,000 bites by the irradiated mosquitoes per volunteer were required for consistent protection against challenge. Importantly, there have been no breakthrough
P. falciparum infections in volunteers immunized by sporozoites irradiated with > 120 Gray units. Protection against challenge lasted for at least 42 weeks (10 months) after the last immunization. Furthermore, studies have shown that volunteers who are exposed to infected non-irradiated mosquitoes while taking chloroquine develop durable protection [
40,
41]. However, considerable technological challenges in terms of manufacturing, formulation, and delivery of such attenuated sporozoite vaccines need to be overcome [
5].
As for vaccines that target the sexual stage of the parasite, they do not aim to prevent illness or infection in the vaccinated individual, but to prevent or decrease transmission of the parasite to new hosts. This ‘transmission-blocking’ vaccine can be seen as a true altruistic vaccine [
12]. Previous clinical trials of sexual stage vaccines that have been discontinued involve ookinete antigens Pfs25 from
P. falciparum and Pvs25 from
P. vivax. In both studies and in pre-clinical work by the same group there is a consistent correlation between titre of anti-Pfs25 antibody and membrane-feeding assay (MFA) activity [
42].
Nonetheless, in the absence of surrogate measures of protection conferred by these vaccines, probably the best way to assess the effectiveness of a developed vaccine is by conducting clinical trials in natural conditions. For a vaccine to be effective, it must elicit the appropriate immune responses that will protect the individual from future infections or disease.
The understanding of the immune correlates will provide the missing piece of puzzle to improve the performance of RTS,S and to fully optimize other vaccine candidates. The ongoing phase III RTS,S vaccine trial is a unique opportunity, with African scientists playing a central role in assessing not only efficacy of the vaccine, but also its mode of action and correlate(s) of protection. Exploration of factors that might affect vaccine efficacy, including the effect of maternal antibodies, the role of immune interference by EPI vaccines, the effect of the RTS,S/AS01 booster, and status with respect to previous exposure to
P. falciparum parasites, will provide crucial information for the further development of this vaccine and for other malaria vaccines under development [
9].
Tackling such questions leads to additional research capacity (resources both human and infrastructure) development in the field of basic human immunology and systems biology. It will further foster the GCP-ICH and good clinical laboratory practice (GCLP) approaches through standardization of approaches, methods and procedures.
Clinical trial end points
Suitable choice of the primary end points in the controlled trials is critical for each phase of clinical development of the vaccine. For example, the ongoing phase III RTS,S malaria vaccine trial has well-defined and harmonized end points. This allows comparability of the performance of the same intervention in different locations, age groups and over time at the same location [
43].
The investigation of relationship between parasite density and likelihood of clinical disease can help to develop a model of specificity and sensitivity for end point definition. Standardization of one method that might have been developed during previous studies is desirable for accuracy, precision and key to comparisons across trials and intra-trial (different sites). Based on this, laboratory procedures for malaria parasite quantification in RTS,S trial has been harmonized across sites to ensure both accuracy and to allow for comparability throughout the trial [
44]. This includes, for instance, similar standard operation procedures (SOPs) for slide reading and interpretation, a key determinant of study end point in a malaria intervention trial [
45].
Selection of the appropriate clinical trial efficacy end points (e g, risk of developing a disease or severe disease) is of crucial importance and calls for standardized protocols of malaria case definition for current and future trials or studies adapted to the different levels of care where the protocols will be applied (e g, dispensary versus hospital). Similarly, phase IV trials will also require full standardization of other key end points, such as health system factors or cost-effectiveness and cost-benefit assessments.
Monitoring of malaria transmission intensity
The malaria control strategies implemented in many endemic areas have resulted in a decrease of malaria transmission in many parts of Africa and elsewhere [
46]. Despite these efforts, vulnerable children, under five years of age, are still at risk of dying from malaria disease. This calls for the continued search for a malaria vaccine, which could complement the existing integrated control strategies. Being able to measure accurately malaria transmission is a key factor for any control programme, as well as measuring the impact of new control tools as identified by the MalERA-process [
47]. Concurrent assessment of malaria transmission intensity during malaria vaccine trails is important in the interpretation of efficacy results by providing accurate information on the endemicity and seasonality of malaria transmission. Accurate data on transmission will also help in the design of phase IV studies, deployment of the vaccine and subsequently effective surveillance-response systems in order to be able to evaluate and understand the relationship between transmission intensity and health outcomes in a given area. Unfortunately there is still a lack of appropriate methodological approaches.
Data generated through microsimulation [
48,
49] proposed an algorithm estimation of human infection rates from the entomological inoculation rate (EIR). Infant mortality rates decrease markedly when the EIR is reduced, probably largely because of prevention of indirect mortality. It was also observed that reduction of exposure to malaria during infancy is not reflected in increased mortality at older ages, a concern of many who think that good childhood protection programmes may predispose to later susceptibility [
48]. Modeling is only one of the avenues to be pursued for quick and better capturing transmission with changing levels of endemicity; new approaches involving biomarkers and particularly serology needs to be explored [
47]. However, modeling cannot replace real-time monitoring of transmission during studies.