Malaria transmission structure in the Peruvian Amazon through antibody signatures to Plasmodium vivax

Abstract

Background: The landscape of malaria transmission in the Peruvian Amazon is temporally and spatially heterogeneous, presenting different micro-geographies with particular epidemiologies. Most cases are asymptomatic and escape routine malaria surveillance based on light microscopy (LM). Following the implementation of control programs in this region, new approaches to stratify transmission and direct efforts at an individual and community level are needed. Antibody responses to serological exposure markers (SEM) to Plasmodium vivax have proven diagnostic performance to identify people exposed in the previous 9 months.

Methodology: We measured antibody responses against 8 SEM to identify recently exposed people and determine the transmission dynamics of P. vivax in peri-urban (Iquitos) and riverine (Mazán) communities of Loreto, communities that have seen significant recent reductions in malaria transmission. Socio-demographic, geo-reference, LM and qPCR diagnosis data were collected from two cross-sectional surveys. Spatial and multilevel analyses were implemented to describe the distribution of seropositive cases and the risk factors associated with exposure to P. vivax.

Principal findings: Low local transmission was detected by qPCR in both Iquitos (5.3%) and Mazán (2.7%); however, seroprevalence indicated a higher level of (past) exposure to P. vivax in Mazán (56.5%) than Iquitos (38.2%). Age and being male were factors associated with high odds of being seropositive in both sites. Higher antibody levels were found in individuals >15 years old. The persistence of long-lived antibodies in these individuals could overestimate the detection of recent exposure. Antibody levels in younger populations (<15 years old) could be a better indicator of recent exposure to P. vivax.

Conclusions: The large number of current and past infections detected by SEMs allows for detailed local epidemiological analyses, in contrast to data from qPCR prevalence surveys which did not produce statistically significant associations. Serological surveillance will be increasingly important in the Peruvian Amazon as malaria transmission is reduced by continued control and elimination efforts.

Fig 1. Map of study sites.

Mazán and Iquitos are located within the Maynas province, in the department of Loreto. The ICEMR study surveyed seven communities from the Mazán district: Gaminatacocha (GC), Libertad (LI), Primero de Enero (PE), Puerto Alegre (PA), Salvador (SA), Urco Miraño (UM) and Lago Yuracyacu (LY). The CAM study surveyed three communities from Iquitos: Rumococha (RC), Santo Tomás (ST) and Quistococha (QC). Each circle represents a household location. Map generated with QGIS 2.16 (QGIS Development Team, 2016. QGIS Geographic Information System. Open Source Geospatial Foundation Project) based on public geographic data extracted from OpenStreetMap contributors (www.openstreetmap.org) under Open Data Commons Open Database License (ODbL) 1.0 (http://openstreetmap.org/copyright).

Fig 2. Prevalence of Plasmodium vivax measured by three tools in communities of Iquitos and Mazán.

A) Prevalence of P. vivax by light microscopy (LM). B) Prevalence of P. vivax by qPCR and C) Prevalence of P. vivax by antibody responses to SEM. Dots denote the mean and lines denote the 95% confidence intervals.

Fig 3. Age trend of seropositivity to P. vivax in the studied communities.

Each panel indicates the age seroprevalence detected by antibody responses to SEM in the villages of Iquitos: Rumococha, Santo Tomás and Quistococha, and the villages of Mazán: Gaminatacocha, Libertad, Primero de Enero, Puerto Alegre, Salvador, Urco Miraño and Lago Yuracyacu. Dots denote the proportion mean and 95% confidence interval of seroprevalence per age-group.

Fig 4. Spatial distribution of seropositive cases in the study sites.

The seropositive rate per household was calculated as the proportion of seropositive individuals per number of household members participating in the study. The figure depicts households in the villages of Iquitos: Rumococha (RC), Santo Tomás (ST) and Quistococha (QC), and the villages of Mazán: Gaminatacocha (GC), Libertad (LI), Primero de Enero (PE), Puerto Alegre (PA), Salvador (SA), Urco Miraño (UM), and Lago Yuracyacu (LY). Each point indicates a household location. Households are colored accordingly to their seropositive rate. Map generated with QGIS 2.16 (QGIS Development Team, 2016. QGIS Geographic Information System. Open Source Geospatial Foundation Project) based on public geographic data extracted from OpenStreetMap contributors (www.openstreetmap.org) under Open Data Commons Open Database License (ODbL) 1.0 (http://openstreetmap.org/copyright).

Fig 5. Visual summary of estimates of multilevel logistic regression models of P. vivax seropositivity and parasitaemia in Mazán and Iquitos.

A) Fixed effects for seropositivity to P. vivax. B) Fixed effects for P. vivax parasitaemia; no factors were significantly associated. Adjusted Odds Ratios and the 95% CI are represented by blue (Mazán) and orange (Iquitos) lines. Ref = Reference; *: p < 0.05.