Serology as a measure of exposure and duration of infection in malaria trials Malaria is one of the leading causes of pediatric morbidity and mortaliy worldwide. Global awareness about the burden of malaria has led to increased investment in malaria control and has even raised the prospect of elimination in some areas. In an era of rapidly changing malaria transmission and mass scale-up of control efforts, tools to accurately measure changes in the burden of malaria infection and disease over time and space are crucial. The gold standard for measuring transmission intensity is the entomological inoculation rate (EIR) which measures the estimated number of infectious mosquito bites per person per year based on the biting rate and proportion of vectors carrying parasites. Parasite prevalence (PR), or the fraction of people carrying asymptomatic infections in a cross-sectional survey, is often used to quantify the burden of infection and is also used as a proxy for transmission intensity. The number or incidence of febrile episodes due to malaria, severe hospitalized malaria cases, and malaria- attributable deaths are important measures of the burden of clinical disease. Each of these indicators - EIR, PR and incidence - measures the success of the parasite at a different point in the malaria parasite's life-cycle and the relationship between these measures is poorly understood. Therefore, comparing interventions that target different stages of the parasite's life- cycle is challenging. For example, interventions that reduce the number of mosquitoes (such as larviciding or indoor residual spraying) would be expected to reduce the EIR. However, the introduction of an effective drug may not be expected to reduce EIR but will reduce the duration of infections, incidence of severe disease, and death. How can the impact of these interventions best be compared? Antibodies to malaria antigens begin to develop from the first exposure to blood-stage parasites, but the magnitude and longevity of the antibody response is related to the duration of infection and frequency of re-infection. Therefore, seroconversion from malaria-na?ve to malaria exposed is related to the intensity of transmission and duration of infection. Recent studies have shown that serological techniques can be used to rapidly assess exposure at a population level in cross-sectional surveys. It has been shown that the increase in seroprevalence with increasing age is directly proportional to EIR. Inflections in the age-seroprevalence curve are hypothesized to reflect recent reduction in transmission in a population. These techniques have not yet been tested prospectively in the context of an intervention trial and compared alongside other measures of malaria burden. We propose to incorporate serological measures of malaria into a large, cluster-randomized malaria intervention trial. The Mvomero trial is designed to compare the impact of a vector control and a diagnostic intervention and to detect synergies between the two in a latin squares design. The work proposed will determine whether population age- seroprevalence and age-titer profiles reflect changes in population-level exposure to infection on short time scales in the context of an intervention trial and whether absolute antibody titers and seroprevalence can distinguish between interventions that reduce exposure to infection (ITN, larviciding, IRS) and interventions that reduce the duration of infection (diagnostic tests, effective drugs, early treatment and activ case detection).
Serological measures of malaria can relate epidemiologic parameters to entomologic parameters and are a powerful addition to the malaria surveillance toolbox. We propose to extend the understanding of how serological measures reflect the impact of interventions that target different stages of the parasite life- cycle in a prospective, cluster randomized, multi-intervention trial.