The effect of Plasmodium falciparum exposure and maternal anti-circumsporozoite protein antibodies on responses to RTS,S/AS01E vaccination in infants and children: an ancillary observational immunological study to a phase 3, randomised clinical trial

Background The RTS,S/AS01E malaria vaccine showed lower antibody response and protective efficacy in infants aged 6-12 weeks compared with children aged 5-17 months (for whom this vaccine is recommended). We aimed to study the effect of previous Plasmodium falciparum exposure on the antibody responses to RTS,S/AS01E vaccination in infants and children, and the mediating effect of baseline (including maternal) anti-circumsporozoite protein (CSP) antibodies. Methods In this observational study, we included children and infants from six African countries (Burkina Faso, Gabon, Ghana, Kenya, Mozambique, and Tanzania) enrolled in the MAL067 immunology ancillary study of the RTS,S/ AS01E phase 3 clinical trial from March 27, 2009, to Jan 21, 2011. We used comparator-vaccinated infants and children to identify antibody-based signatures of previous P falciparum exposure, which were later applied to RTS,S/AS01E- vaccinated infants and children. In these participants, we explored the relationship between vaccine antibody immunoglobulin G (IgG) responses measured by ELISA and pre-vaccination serological markers of malaria exposure by assessing the IgG levels against 1000 P falciparum antigens using partial proteome microarrays. Findings We included 718 comparator-vaccinated infants (348 [48%]) and children (370 [52%]) and 606 RTS,S/AS01E- vaccinated infants (329 [54%]) and children (277 [46%]). Anti-CSP IgG responses to primary vaccination did not correlate with a baseline signature of previous exposure in children, suggesting that prior P falciparum exposure does not significantly affect antibody immunogenicity in children (Pearson's r=-0<middle dot>02 [95% CI -0<middle dot>13 to 0<middle dot>10]). By contrast, high P falciparum exposure signature levels at the time of vaccination in infants, presumably driven by maternally transferred antibodies and declining within the initial 6-12 months of life, correlated with reduced RTS,S/AS01E responses (r=-0<middle dot>17 [-0<middle dot>27 to -0<middle dot>06]). This negative correlation was stronger for anti-CSP IgG than for the exposure signature or any other more immunogenic blood stage P falciparum antigens (r=-0<middle dot>42 [-0<middle dot>50 to -0<middle dot>33]), persisted after adjustment by baseline levels of the exposure signature (semi-partial correlation r=-0<middle dot>44 [-0<middle dot>55 to -0<middle dot>33]), and involved antibodies to the central NANP region (r=-0<middle dot>39 [-0<middle dot>49 to -0<middle dot>28]) but not the C-terminal region (r=0<middle dot>02 [-0<middle dot>10 to 0<middle dot>15]) of CSP. The negative effect of maternal anti-CSP IgG in infants did not appear to be confounded by other malaria transmission-dependent variables.