High antibody titer against apical membrane antigen-1 is required to protect against malaria in the Aotus model.

A Plasmodium falciparum 3D7 strain Apical Membrane Antigen-1 (AMA1) vaccine, formulated with AS02A adjuvant, slowed parasite growth in a recent Phase 1/2a trial, however sterile protection was not observed. We tested this AS02A, and a Montanide ISA720 (ISA) formulation of 3D7 AMA1 in Aotus monkeys. The 3D7 parasite does not invade Aotus erythrocytes, hence two heterologous strains, FCH/4 and FVO, were used for challenge, FCH/4 AMA1 being more homologous to 3D7 than FVO AMA1. Following three vaccinations, the monkeys were challenged with 50,000 FCH/4 or 10,000 FVO parasites. Three of the six animals in the AMA+ISA group were protected against FCH/4 challenge. One monkey did not become parasitemic, another showed only a short period of low level parasitemia that self-cured, and a third animal showed a delay before exhibiting its parasitemic phase. This is the first protection shown in primates with a recombinant P. falciparum AMA1 without formulation in Freund's complete adjuvant. No animals in the AMA+AS02A group were protected, but this group exhibited a trend towards reduced growth rate. A second group of monkeys vaccinated with AMA+ISA vaccine was not protected against FVO challenge, suggesting strain-specificity of AMA1-based protection. Protection against FCH/4 strain correlated with the quantity of induced antibodies, as the protected animals were the only ones to have in vitro parasite growth inhibitory activity of >70% at 1∶10 serum dilution; immuno-fluorescence titers >8,000; ELISA titers against full-length AMA1 >300,000 and ELISA titer against AMA1 domains1+2 >100,000. A negative correlation between log ELISA titer and day 11 cumulative parasitemia (Spearman rank r = −0.780, p value = 0.0001), further confirmed the relationship between antibody titer and protection. High titers of cross-strain inhibitory antibodies against AMA1 are therefore critical to confer solid protection, and the Aotus model can be used to down-select future AMA1 formulations, prior to advanced human trials.