By Patricia Kahn
While AIDS researchers often say that HIV is one of the most formidable pathogens ever targeted for a vaccine, the same holds true for the parasites that cause malaria—a disease that claimed 1 to 2 million lives every year, 75% of them children under five.
One of the biggest obstacles for malaria vaccine developers is that they express between 5,000 and 6,000 proteins, compared to only 9 for HIV—vastly complicating the task of determining which antigens to include in a vaccine. The parasites also show a high degree of genetic diversity, which—given their huge genome size—has defied classification into groups analogous to HIV subtypes. And its complex life cycle encompasses three very different forms (see figure) and a highly sophisticated strategy for evading the immune system through frequent switching of surface antigens.
Yet despite these immense challenges, two lines of evidence indicate that it is indeed possible to generate protective immunity against malaria parasites, which fall into four separate species of Plasmodium (of which two are responsible for most disease). One is that highly-exposed people who survive multiple bouts of malaria in childhood gradually develop partial (“semi”-) immunity that reduces the severity of disease during later infections. Another is that infection can be blocked by immunization with irradiated sporozoites—the parasite form that invades liver cells immediately after infection, where it replicates intracellularly for 1-2 weeks without inducing symptoms. While this latter protection requires cellular immunity (and can be transferred in mice via CD8 cells), natural semi-immunity appears to depend mainly on antibodies to the merozoite, the form that enters the bloodstream and invades erythrocytes after infected liver cells finally burst.
Attempting to exploit these findings, malaria researchers have developed vaccines based on both sporozoite and merozoite antigens, using some of the same new technologies used for HIV vaccines. And, thanks largely to support from Malaria Vaccine Initiative (Washington, DC), and the Wellcome Trust (UK) three of these candidates are now in clinical trials in Africa: two protein-based vaccines, including one that already showed significant but short-lived protection in an efficacy trial in The Gambia, and another based on a DNA/MVA prime-boost approach, a widely-used strategy in the HIV field.
Some developers expect that these candidates may not be instant “home runs” conferring highly effective, long-lasting immunity. Rather, they could be contributors to a multi-component vaccine, seen by many as the more likely formula for success. “Most people in the field expect that an effective malaria vaccine will require targeting multiple stages and antigens,” says Dan Carucci, director of the Malaria Program at the US Naval Medical Research Center. While immune responses directed at sporozoite-infected liver cells could reduce the number of parasites entering the bloodstream, he adds, others aimed at the merozoite stage would blunt the severity of disease. In babies, the overall effect could be “to catapult their malaria immune status into that of adolescents,” says malaria researcher Hermann Bujard of the University of Heidelberg.
Testing Malaria Vaccines
Determining vaccine efficacy is simpler for malaria than for HIV. Once a candidate’s safety is established, first indications of effectiveness may be gathered by challenging small numbers of vaccinated volunteers (at experienced research centers) under carefully controlled conditions where they are bitten by infected mosquitoes; volunteers who become infected are treated immediately with anti-malarial drugs.
Gray Heppner, who heads the Malaria Vaccine Program at the Walter Reed Army Institute of Research (WRAIR, Rockville) where many of these challenge studies have been done, points out that it’s a somewhat artificial model. The challenge uses a single strain of parasite at up to 10-fold the natural dose, to ensure that all volunteers are exposed to an infectious dose. What’s more, volunteers at these US or European centers have no prior exposure to malaria, unlike people in endemic regions where a vaccine is most urgently needed. But prevention or delay of infection is nonetheless a useful, albeit highly stringent, hint of efficacy, says Heppner.
Then comes the real test in endemic regions among semi-immune populations normally exposed to a wide diversity of circulating strains. In high-incidence regions, vaccine efficacy can be measured in small, short trials (compared to those needed for testing HIV vaccines). For example, in The Gambia, where malaria occurs only in the July-to-November rainy season, about 60% of adults become infected during a single season, and efficacy trials require only a few hundred volunteers and about six months time.
GSK’s Protein Subunit Vaccine
As the most advanced candidate now in African trials, GlaxoSmithKline’s so-called RTS,S vaccine contains about half of the major sporozoite coat protein fused to the Hepatitis B surface antigen. When formulated with GSK’s AS02 adjuvant, challenge studies done collaboratively with the WRAIR found RTS,S to be about 50% effective in blocking infection with homologous sporozoites (i.e., of the same strain as the vaccine protein), and to protect one in five volunteers upon re-challenge 6 months later. Immune analysis detected strong antibody responses and some cellular responses, including “modest” CD8 levels, according to Joe Cohen, who directs GSK’s program on vaccines against emerging diseases.
Moving to the field, safety studies in The Gambia were followed by a collaborative efficacy trial in 306 highly exposed male volunteers (Lancet 358:1927;2001). The results: significant delay in time to infection, with about 70% protection in the first two months, but waning to zero by week 15. Re-vaccination of 158 volunteers the next year showed about 47% protection over 9 weeks. Protection extended beyond the vaccine strain to other, “unmatched” circulating strains.
The vaccine was also tested in Gambian children, looking first at safety and dosage in 6-11-year olds, then in 1-4 year olds. Based on these results, Phase I studies are now underway in children in Mozambique, where malaria is transmitted year-round rather than seasonally; by the end of 2002, a Phase IIb pediatric study will begin to gather preliminary efficacy data. That trial will also introduce a new element, says Cohen: Rather than measuring only sterilizing immunity, it will also look at endpoints reflecting severity of disease at the time children become ill and are brought to the clinic for treatment.
In the meantime, further work aims at improving the levels of CD8 T-cell responses, for example through the use of other adjuvants and immunization schedules. And, in keeping with the goal of developing multi-component vaccines, small clinical studies are looking at prime-boost combinations with different candidates, including one in Oxford with the MVA-based vaccine described below.
Sporozoite Antigens in DNA/MVA Vaccines
Another strategy now in efficacy studies is a prime-boost combination developed by Adrian Hill’s group at Oxford University. The two vaccines encode a complete TRAP protein (Thrombospondin-Related Adhesion Protein), one of the main sporozoite antigens, downstream from 20 individual peptides containing mostly CD8 T-cell epitopes from six sporozoite or liver-stage antigens.
Multiple clinical studies resulted in an immunization regimen generating a 10-fold boost in T-cell Elispot numbers with DNA/MVA compared DNA or MVA alone, to levels in the 1,000 spots per million PBMC range; the better regimes (using DNA/MVA or fowlpox plus DNA) yielded broad strain cross-reactive results. About 100 volunteers in Oxford have now been challenged-in this case with a heterologous strain, resulting in delay of average time to infection that corresponds to “substantial” reduction in the estimated numbers of parasites emerging from the liver. In addition, about 50-60 volunteers have been safely immunized in Phase I studies at the Medical Research Council Laboratories in Banjul, including 20 children who were given MVA alone. Results will be known in Spring 2003.
Also in development: the same vaccine antigens (TRAP and sporozoite coat protein) in fowlpox, which Hill says so far looks “as good, and maybe much better.”
Merozoite Protein-Based Vaccines
Will people vaccinated with one of two main strains of merozoite coat protein (MSP-1) also recognize the other one? That’s the question being studied in WRAIR’s ongoing 60-person trial, launched in Kenya in April 2002 in collaboration with the Kenya Medical Research Institute, MVI and USAID. The vaccine contains a portion of the MSP-1 protein formulated with GSK’s AS02 adjuvant. Also in the pipeline: the first full-length MSP-1-based vaccines, made by Hermann Bujard’s group at the University of Heidelberg, which synthesized the two main strains of this notoriously unclonable protein from scratch. The first clinical tests will be conducted in TŸbingen, Germany and WRAIR before moving into trials in Burkina Faso. In addition to the protein-based MSP-1 vaccines, Bujard’s group is developing MVA-based versions.
Finding the Right Antigens: Future Directions
Since 1993, the US Navy’s malaria program has been developing DNA-based candidates, working towards the strategy of using cocktails containing plasmids with different antigens. With three Phase I trials under their belts, their current vaccine contains 5 different antigens (3 from sporozoites, 2 from merozoite); in parallel they are developing MVA- and adenovirus-based vectors as possible boosts, following encouraging protection results in monkeys. Other groups, including WRAIR and the malaria program at NIAID (the US National Institute of Allergy and Infectious Diseases) are studying additional antigens for clinical evaluation.
But as they work to improve immunogenicity of DNA-based vaccines, the Navy program is also seeking better ways to identify which antigens from the huge number of potential candidates might really matter, says Carucci. So far, studies of protected versus unprotected volunteers (from vaccination with irradiated sporozoites) have proved frustratingly inconclusive: the researchers find only low, although detectable, cellular responses to peptide pools from the few antigens they’ve looked at. “Maybe we’re missing something big,” says Carucci.
But the completion of the Plasmodium falciparum genome sequence offers a way to approach the problem on a large scale rather than antigen-by-antigen, he adds. Looking at the total pool of encoded proteins, the researchers have identified over 1,000 new proteins, and are developing assays to screen the blood of protected volunteers for responses to any of them—so they can determine whether there is a dominant response that correlates with protection and also identify potential new antigenic targets. So far, they’ve found six new parasite proteins on the surfaces of infected erythrocytes. Still, says Heppner, “we need more antigens, more adjuvants, to increase the magnitude and duration of protection. And we need more money. So many good concepts are still restricted by lack of funds.”