Lack of allele-specific efficacy of a bivalent AMA1 malaria vaccine

Abstract

Background: Extensive genetic diversity in vaccine antigens may contribute to the lack of efficacy of blood stage malaria vaccines. Apical membrane antigen-1 (AMA1) is a leading blood stage malaria vaccine candidate with extreme diversity, potentially limiting its efficacy against infection and disease caused by Plasmodium falciparum parasites with diverse forms of AMA1.

Methods: Three hundred Malian children participated in a Phase 2 clinical trial of a bivalent malaria vaccine that found no protective efficacy. The vaccine consists of recombinant AMA1 based on the 3D7 and FVO strains of P. falciparum adjuvanted with aluminum hydroxide (AMA1-C1). The gene encoding AMA1 was sequenced from P. falciparum infections experienced before and after immunization with the study vaccine or a control vaccine. Sequences of ama1 from infections in the malaria vaccine and control groups were compared with regard to similarity to the vaccine antigens using several measures of genetic diversity. Time to infection with parasites carrying AMA1 haplotypes similar to the vaccine strains with respect to immunologically important polymorphisms and the risk of infection with vaccine strain haplotypes were compared.

Results: Based on 62 polymorphic AMA1 residues, 186 unique ama1 haplotypes were identified among 315 ama1 sequences that were included in the analysis. Eight infections had ama1 sequences identical to 3D7 while none were identical to FVO. Several measures of genetic diversity showed that ama1 sequences in the malaria vaccine and control groups were comparable both at baseline and during follow up period. Pre- and post-immunization ama1 sequences in both groups all had a similar degree of genetic distance from FVO and 3D7 ama1. No differences were found in the time of first clinical episode or risk of infection with an AMA1 haplotype similar to 3D7 or FVO with respect to a limited set of immunologically important polymorphisms found in the cluster 1 loop of domain I of AMA1.

Conclusion: This Phase 2 trial of a bivalent AMA1 malaria vaccine found no evidence of vaccine selection or strain-specific efficacy, suggesting that the extreme genetic diversity of AMA1 did not account for failure of the vaccine to provide protection.

Figure 1

Flow chart showing the samples and apical membrane antigen-1 (AMA1) ectodomain sequences used in the analysis.

Figure 2

Frequency of 3D7 and FVO apical membrane antigen-1 (AMA1) cluster 1 loop (c1L) haplotypes observed before and after immunization in the AMA1-C1 vaccine and control groups.

Figure 3

Population structure of Plasmodium falciparum apical membrane antigen-1 (AMA1) sequences using DISTRUCT program [24]. The population structure of AMA1 was assessed using a Bayesian clustering method implemented in Structure 2.2 with k = 6. Each haplotype is represented by a vertical bar, and each color represents a population. Six populations are shown: Group 1 in red (3D7 group, haplotype 29), group 2 in green (FVO group, haplotype 49), group 3 in dark blue, group 4 in yellow, group 5 in pink and group 6 in light blue.

Figure 4

Distribution of six Plasmodium falciparum apical membrane antigen-1 (AMA1) sequence groups before and after vaccination with the malaria vaccine and control vaccine. Group 1 (red) includes AMA1 sequence corresponding to the 3D7 vaccine strain of P. falciparum and Group 2 (green) includes sequence corresponding to the FVO vaccine strain.

Figure 5

Kaplan-Meier plot of survival curve without malaria clinical episode with a 3D7 or FVO c1L type allele following vaccination with apical membrane antigen-1 (AMA1) malaria vaccine or a control vaccine. Top panel, 3D7. Log rank statistic = 0.02, p = 0.88 and Wilcoxon statistic = 0.11 p = 0.73. Bottom panel, FVO. Log rank statistic = 0.35, p = 0.55 and Wilcoxon statistic = 1.70 p = 0.19.