More than 500,000 people die from malaria each year, with millions more infected and falling ill. While anti-malarial medications and widespread preventative initiatives have made strides in decreasing that number, there is still room for improvement, and a malaria vaccine has long been a goal. Such a vaccine is now in the clinic, and a recent study by an international team led by researchers at the Broad Institute of MIT and Harvard, the Harvard T.H. Chan School of Public Health and Fred Hutchinson Cancer Research Center has made a discovery that could help pinpoint where and when to use the vaccine most effectively: when the allele used in a vaccine matches that of a strain of malaria, the vaccine is infinitely more effective than it is against a mismatched strain. Details of this work were published in The New England Journal of Medicine.
GlaxoSmithKline began developing RTS,S, the vaccine in question, in 1987, and it is the first malaria vaccine candidate that has made it to a Phase 3 trial. The RTS,S vaccine contains a fragment of a protein found on the surface of the malaria parasite capable of eliciting an immune response. However, this protein is genetically diverse across different strains, and RTS,S includes only one form, or allele, of the protein. Final trial results for RTS,S, which were reported in April, showed roughly one-third fewer episodes of clinical and severe malaria in young children who received three doses and a booster, though protection waned over time.
A team from the Broad Institute analyzed blood samples from nearly 5,000 of about 15,000 infants and children who participated in the Phase 3 clinical trial of RTS,S. Quantitative scientists at Fred Hutch designed a statistical analysis plan for the data, settling on sieve analysis, a statistical method honed at Fred Hutchinson for evaluating vaccines against HIV. They compared parasite strains found in trial participants who received the vaccine but contracted malaria anyway with strains found in individuals who received a placebo and contracted malaria and found that the placebo group was more varied. In the vaccinated group, some strains were blocked while some slipped through “holes” in the vaccine.
“This is an example of the benefits of applying genomics to a real world problem of global health importance,” said Dr. Dyann Wirth, Broad senior associate member and a top malaria researcher at Harvard. Wirth, along with Fred Hutchinson biostatistician Dr. Peter Gilbert, led the study.
“This is the first time a malaria vaccine has had good enough protection to do a sieve analysis, and there were enough samples and enough malaria [cases],” said Gilbert, a member of the Vaccine and Infectious Disease Division and director of the statistical center for the Hutch-based HIV Vaccine Trials Network. “And the Broad and Harvard went to the trouble to measure all these sequences. Those three things were incredibly useful.”
Analysis of this data required new statistical methods in order to deal with the complexities of how the malaria parasite evolves and is transmitted as well as verification of their accuracy in simulations and the use of parallel computing, according to Gilbert. Dr. Michal Juraska, another biostatistician from Fre Hutchinson and co-first author of the paper, headed up the development of the reproducible statistical code and data analysis with help from Fred Hutch bioinformatics expert Ted Holzman. Juraska's analysis discovered that while RTS,S offered at least partial protection against all strains of malaria, it was markedly more protective when the parasite matched the allele in the vaccine—protection against parasites with matched alleles was 50.3 percent one year after the vaccination series, compared to 33.4 percent against mismatched malaria.
Protection from the vaccine was higher at six months than at one year, 70.2 percent against matched malaria and 56.3 percent against mismatched. Immediately after the vaccination series, protection from infection by matched malaria was 95 percent.
“Now that we know that [the difference] exists, it contributes to our understanding of why the vaccine is not effective in all cases and informs future vaccine development efforts,” said Dr. Dan Neafsey, associate director of the Genomic Center for Infectious Diseases at the Broad and co-first author of the study.
The knowledge that a vaccine's efficiency is dependent on the genetics of a pathogen could enable a manufacturer to add additional strains to increase the protectiveness of the vaccine, but it can also allow public health officials to deploy RTS,S more effectively as-is by choosing when and where to use it.
“You might expect the vaccine to work better in regions where the circulating parasites are similar to what’s in the vaccine,” commented David Benkeser, a biostatistics doctoral student at the University of Washington under Gilbert who contributed a novel statistical method for the analysis. “And if you can give the vaccine ideally right before a high-intensity [malaria] season, you might expect it’s going to protect very well in that window.”
A decision from the World Health Organization on whether to recommend use of the vaccine in countries with high malaria rates is expected before the end of the year.