Systematic sampling of adults as a sensitive means of detecting persistence of lymphatic filariasis following mass drug administration in Sri Lanka

Abstract

Background: Sri Lanka's Anti-Filariasis Campaign conducted 5 annual rounds of mass drug administration (MDA) with diethylcarbamazine (DEC) plus albendazole to eliminate lymphatic filariasis (LF) in all endemic districts between 2002 and 2006. Post-MDA surveillance has consistently documented Wuchereria bancrofti microfilaremia (Mf) rates below 1% in all sentinel and spot check sites since that time, and all implementation units easily satisfied WHO's target for school-based transmission assessment surveys (school-TAS) in 2013. However, more detailed studies have identified foci of persistent infection in the large coastal evaluation unit (EU) (population about 0.6 million) in Galle district. Therefore, the purpose of this study was to assess the sensitivity and feasibility of community-based TAS in adults (adult-TAS) and to compare results obtained by adult-TAS with prior school-TAS and molecular xenomonitoring (MX, molecular detection of filarial DNA in systematically sampled mosquitoes) results in this known problem area.

Methodology and principal findings: Two cluster surveys were performed in independent samples of 30 evaluation areas (EAs) in the coastal Galle EU in 2015. Each survey tested approximately 1,800 adults for circulating filarial antigenemia (CFA) with the Alere Filariasis Test Strip. The CFA prevalence for all persons tested (N = 3,612) was 1.8% (CI 1.4-2.2), and this was significantly higher than the CFA rate of 0.4% obtained by school-TAS in 2013. CFA prevalences in the two samples were similar [1.5% (CI 1.0-2.2), and 2.0% (CI 1.4-2.7)]. Antigenemia prevalence in sampled EUs was highly variable (range 0-11%), and it exceeded 5% in 6 EAs. The 30 EAs sampled in one of our adult-TAS surveys had recently been assessed for persistent filariasis by molecular xenomonitoring (MX). CFA prevalence in adults and filarial DNA prevalence in mosquitoes in these EAs were significantly correlated (r = 0.43; P = 0.02).

Conclusions: Community based adult-TAS provided a reproducible measure of persistent W. bancrofti infection in a large evaluation unit in Sri Lanka that has low-level persistence of LF following multiple rounds of MDA. In addition, adult-TAS and MX results illustrate the focality of persistent LF in this setting. Adult-TAS may be more sensitive than school-TAS for this purpose. Adult-TAS and MX are potential options for post-MDA and post-validation surveillance programs to identify problem areas that require mop-up activities. Adult-TAS should also be useful for remapping areas with uncertain LF endemicity for possible inclusion in national LF elimination programs.

Fig 1. Summary of community circulating filarial antigen (CFA) prevalence in 60 evaluation areas (Public Health Midwife areas) assessed by adult-TAS in the coastal Galle evaluation unit.

The antigenemia target of 2% (the upper CI for the prevalence estimate) for the EU is shown with a dotted line. The CFA prevalence was >2% in 7 of 30 evaluation areas in the first survey (panel A), and the prevalence > 2% in 8 of 30 EAs in the second survey (panel B).

Fig 2. The figure shows circulating filarial antigen (CFA) prevalence (%) by age and gender for the population surveyed in 60 evaluation areas (EA) in the coastal Galle evaluation unit.

Antigenemia was significantly more common in males for the total population (Males, 38 of 1537 or 2.5% vs. 26 of 2075 or 1.3% in Females, P = 0.007) and in most age groups.

Fig 3

The map shows locations with positive signals for persistent filariasis that were identified by adult-TAS antigenemia surveys (panel A) in 60 evaluation areas, and by adult-TAS and molecular xenomonitoring surveys (panel B) in 30 evaluation areas in the coastal Galle evaluation unit. Panel A: Blue and red circles indicate households (HH) where at least one resident had a positive filarial antigen test in two surveys, respectively. Panel B: Blue circles indicate HH with positive signals for antigen in the first survey and red triangles indicate mosquito trap locations with one or more pools positive for Wuchereria bancrofti DNA. Infection signals in mosquitoes and in humans were widespread near the coast, but the strongest signals were detected in evaluation areas near Balapitiya and within the Galle Municipal Council (greater Galle city limits). Solid lines show the boundaries of Medical Officer of Health administrative areas within the EU in Galle district.

Fig 4. Comparison of circulating filarial antigenemia (CFA) prevalence obtained during adult transmission assessment surveys (adult-TAS, 2015) and filarial DNA prevalence in mosquito pools (molecular xenomonitoring survey-2014) in 30 evaluation areas in the coastal evaluation unit in Galle district.

The association between human CFA prevalence and mosquito filarial DNA prevalence at the EA level was statistically significant (Spearman rank test, r = 0.43; P = 0.02).