G: Global Health
Oral Abstract Submission
Cody S. Nelson, PhD
Postdoctoral researcher
Duke University
Durham, NC
Disclosure: Nothing to disclose
Kelsey Sumner
Graduate student
University of North Carolina
Chapel Hill, NC
Disclosure: Nothing to disclose
Andrew Obala, PhD
Professor
Moi University School of Medicine
Eldoret, Western, Kenya
Disclosure: Nothing to disclose
Jane Mangeni, MPH
Lecturer
Moi University School of Nursing
Eldoret, Western, Kenya
Disclosure: Nothing to disclose
Background : Falciparum malaria transmission has failed to decline in proportion to control efforts in certain regions such as Bungoma county, western Kenya. One proposed strategy to eradicate malaria is ring testing and treatment, however it remains unknown whether infections spread locally or if asymptomatically-infected household members are a risk factor for clinical disease.
Methods : From April 2013 to June 2014, we enrolled 442 cases (RDT+ children hospitalized with malaria) and 442 matched controls; all household members of cases and controls were also enrolled and tested, of which 13.6% (n=608/4449) were RDT+. From each RDT+ participant, parasite gDNA was PCR-amplified at both Pf circumsporozoite protein (csp) and apical membrane antigen 1 (ama1) loci, amplicons sequenced on an Illumina Miseq, and haplotypes inferred using dada2.
Results : We identified 120 csp and 180 ama1 unique haplotypes (Figure 1). We evaluated genetic distance between infected individuals using three novel indices: sharing of parasite haplotypes on binary and proportional scales and the L1 norm. Case children median [IQR] binary/proportional sharing of both csp and ama1 haplotypes was significantly increased with members of their origin household (e.g. csp binary sharing: origin = 50.3 [0-87.5] vs. similar household = 0 [0-50.3]; p = 0.01; Wilcoxon Sign-Rank test), indicating that cases are more likely to share haplotype-identical parasites with members of their own household (Figure 2). We also computed population-level haplotype sharing indices for all pairs of case children and observed no association between genetic relatedness and geographic distance. In contrast, we identified a strong inverse relationship between haplotype sharing and temporal distance, which we exploited to identify the molecular signature of an outbreak (Figure 3).
Conclusion : Overall, these findings suggest that, although haplotype sharing is more common within households, temporal rather than geographic proximity predicts parasite genetic similarity. The observation that identical haplotype combinations are found nearly simultaneously across the study area implies that ring testing approaches may not effectively reduce transmission.
