The malaria parasite Plasmodium falciparum kills more children under the age of five than any other infectious agent. The need for an effective vaccine is more urgent than ever: Plasmodium parasites are increasingly resistant to drugs, as are the Anopheles mosquito vectors to insecticides.

There are many successful vaccines against viruses, such as smallpox and polio, and bacteria, such as Haemophilus influenzae, but there are no commercially available vaccines for human parasites. These present a far greater challenge because they are more complex.

First, they have much larger genomes coding for more proteins. Second, they have multi-stage life cycles in which they express many different proteins at different times (see Figure, below). As a result, protective immune responses against the extracellular sporozoites that enter with the bite of a mosquito, for example, may have no effect on the asexual ‘erythrocytic stage’ merozoites that later emerge from the liver and infect red blood cells. Third, P. falciparum in particular has enormous variability in its proteins. This is critical to the parasite's survival, enabling it to evade host immune defences. It also means that a vaccine containing a single sequence of a single protein, or just a few, may fail to have a large, sustainable impact.

There is little precedent for a successful vaccine being created in any way other than through the production of whole organisms, generally in laboratory culture, which for malaria has so far proved impossible. All commercially available vaccines but one consist of material from whole viruses or bacteria, or purified components. The only successful recombinant-protein vaccine is the hepatitis B surface-antigen vaccine. Almost all malaria vaccine candidates are based on individual components (generally proteins or parts of proteins) that have been created in the lab using recombinant proteins, synthetic peptides, recombinant viruses and bacteria, or DNA or RNA plasmids.

Finally, the most effective ‘subunit’ vaccine may need to induce both antibody and T-cell responses. Antibodies could block sporozoites as they enter the body, but have to act within minutes to block entry into the liver. They can also prevent infection of red blood cells, help destroy those already infected and prevent infection of mosquitoes. T cells have the potential to kill infected liver cells, thereby controlling and even eliminating infection. Both types of response may have to be directed against multiple proteins, at different stages of the life cycle, and at the same time. If so, vaccine developers face a technical problem that has never been solved.

Scientists who are attempting to develop an effective vaccine against malaria keep two observations in mind. The first is that most malaria deaths and severe disease in sub-Saharan Africa occur in infants, young children and pregnant women. Adolescents and adults rarely develop severe disease or die after repeated infection with P. falciparum. They have presumably developed natural immunity that limits parasite replication and severe forms of malaria, but does not prevent infection resulting in milder symptoms. Pregnant women, especially with their first child, seem to lose this immunity. This observation has led to the idea that a vaccine would be worthwhile even if it only limited the severity of disease for those most at risk, without preventing infection or moderate disease. Such a vaccine would probably not be very useful for tourists, but would be beneficial in most parts of sub-Saharan Africa.

A thousand bites

The second observation is that when volunteers are exposed to more than a thousand bites from P. falciparum-infected Anopheles mosquitoes that have been irradiated to weaken the sporozoites they carry, they develop protective immunity against multiple strains of P. falciparum. If these volunteers are exposed to normal sporozoites, more than 93% are completely protected against developing erythrocytic stage infection1. This is the strongest evidence that a highly effective vaccine is possible. So some vaccinologists, including this author, are focusing on achieving a more powerful vaccine that prevents all infections with P. falciparum in more than 85% of recipients2.

Vaccinologists have adopted three main strategies. The first is to create vaccines that counter sporozoites as they enter the body and invade and reproduce in the liver (pre-erythrocytic stage vaccines). These have the potential to prevent infection altogether. The second is to limit invasion of erythrocytes and subsequent multiplication and pathological effects (asexual erythrocytic stage vaccines). Such vaccines would only limit severe disease — they would not prevent infection or mild disease. The third strategy is to prevent the spread of viable parasites to other people with ‘transmission-blocking vaccines’. These stimulate the production of antibodies that are ingested when the parasite is sucked up by a mosquito. The antibodies destroy the parasite within the vector's gut.

Vaccinologists may need to combine all three strategies to have the best chance of success. Current vaccine candidates in clinical trials, however, contain just one or a few proteins. In contrast, the protective immune responses elicited by natural exposure to malaria or by immunization with radiation-attenuated sporozoites could be directed at many, perhaps hundreds or even thousands, of the proteins encoded by the 5,300 genes in the P. falciparum genome.

On test

Children may benefit most from a vaccine — even if it only limited the severity of the disease. Credit: S. VAN ZUYDAM/AP

All three strategies are now being pursued, thanks to recent increases in funding. About US$65 million was invested in malaria vaccine research in 2003, and some $85 million is being invested in 2004. The US National Institutes of Health has provided the largest amount ($33 million in 2003), with the Malaria Vaccine Initiative in second place, giving $14 million in 2003. According to the World Health Organization, there are 25 candidates in phase Ia testing (safety and immunogenicity), six in phase IIa testing (efficacy against experimental challenge), eight in phase Ib testing (safety and immunogenicity in a disease-endemic country) and two in phase IIb testing (efficacy in a disease-endemic country).

Despite all of these efforts, only one P. falciparum protein, the circumsporozoite protein (PfCSP), has been repeatedly evaluated in clinical trials and shown to provide complete protection in a portion of volunteers. The lead candidate based on this protein is called RTS,S/AS02A, which GlaxoSmithKline Biologicals and the Walter Reed Army Institute of Research in the United States initially developed, and which the Malaria Vaccine Initiative is now supporting in large-scale trials. In its first trial, the vaccine protected six out of seven volunteers against P. falciparum challenge three weeks after the last immunization3, but subsequent tests found that it protected only 40–50% of volunteers within 2–3 weeks4. It protected 70% of semi-immune Gambian adults for two months, but no longer5. RTS,S/AS02A is now being tested in a phase IIb study in 2,000 children in Mozambique, with results expected this autumn.

Delayed onset

Just one other vaccine candidate, developed at the University of Oxford, has reached phase IIb trials. It includes sequences from several pre-erythrocytic stage proteins, and the entire coding sequence of PfTRAP/ PfSSP2. It is delivered using a ‘heterologous prime boost’ strategy that involves initial doses containing a DNA plasmid or recombinant pox virus expressing P. falciparum epitopes and whole proteins, followed by later injections of a recombinant modified vaccinia virus encoding the same proteins. This approach reproducibly delays the onset of parasitaemia in volunteers, but has prevented infection entirely in only a few recipients6.

With large numbers of candidates in preclinical and clinical development, many more, especially asexual erythrocytic stage vaccines, are likely to enter field trials in the next five years. Furthermore, emerging genomic and proteomic studies of P. falciparum7 will lead to the development of even more candidate vaccines.

Until recently, most groups working on pre-erythrocytic subunit vaccines aimed for complete protection. But results from clinical trials have indicated that the current vaccines are unlikely to accomplish this in the majority of recipients3,4,5,6. Many investigators are nonetheless continuing in the hope of at least achieving a product that limits severe disease. Others, including this author, have since re-examined the model system on which the strategy was based: the immunization of volunteers with radiation-attenuated P. falciparum sporozoites, followed by parasite challenge through the bites of infected mosquitoes3. We have founded a company, Sanaria, to develop a radiation-attenuated P. falciparum sporozoite vaccine. The manufacturing process is currently being optimized, with clinical trials planned for within 18–24 months.

When can we expect a vaccine to be widely available to infants in Africa? GlaxoSmithKline estimates that RTS,S/AS02A could be licensed as soon as 2010 (ref. 8 and personal communication). But gearing up to produce enough vaccine for widespread use, and having it deployed as part of the Expanded Programme for Immunization, will take longer. The time taken since the first PfCSP-based vaccine entered clinical trials in 1986 demonstrates how long and arduous the development process is.

There can be no doubt that an effective vaccine would have an enormous impact on the appalling toll of malaria. The dedication of many scientists and increased investment by government and non-profit institutions have led to the recent explosion of new candidates. We must work hard to sustain and increase both the enthusiasm and investment if we are to realize the dream of a vaccine for those whose lives are devastated by malaria.