Toward a Surrogate Marker of Malaria Exposure: Modeling Longitudinal Antibody Measurements under Outbreak Conditions

BACKGROUND: Biomarkers of exposure to Plasmodium falciparum would be a useful tool for the assessment of malaria burden and analysis of intervention and epidemiological studies. Antibodies to pre-erythrocytic antigens represent potential surrogates of exposure. METHODS AND FINDINGS: In an outbreak cohort of U.S. Marines deployed to Liberia, we modeled pre- and post-deployment IgG against P. falciparum sporozoites by immunofluorescence antibody test, and both IgG and IgM against the P. falciparum circumsporozoite protein by enzyme-linked immunosorbant assay. Modeling seroconversion thresholds by a fixed ratio, linear regression or nonlinear regression produced sensitivity for identification of exposed U.S. Marines between 58–70% and specificities between 87–97%, compared with malaria-naïve U.S. volunteers. Exposure was predicted in 30–45% of the cohort. CONCLUSION: Each of the three models tested has merits in different studies, but further development and validation in endemic populations is required. Overall, these models provide support for an antibody-based surrogate marker of exposure to malaria.