Molecular xenomonitoring using mosquitoes to map lymphatic filariasis after mass drug administration in American Samoa

Abstract

Background: Mass drug administration (MDA) programs have dramatically reduced lymphatic filariasis (LF) incidence in many areas around the globe, including American Samoa. As infection rates decline and MDA programs end, efficient and sensitive methods for detecting infections are needed to monitor for recrudescence. Molecular methods, collectively termed 'molecular xenomonitoring,' can identify parasite DNA or RNA in human blood-feeding mosquitoes. We tested mosquitoes trapped throughout the inhabited islands of American Samoa to identify areas of possible continuing LF transmission after completion of MDA.

Methodology/principle findings: Mosquitoes were collected using BG Sentinel traps from most of the villages on American Samoa's largest island, Tutuila, and all major villages on the smaller islands of Aunu'u, Ofu, Olosega, and Ta'u. Real-time PCR was used to detect Wuchereria bancrofti DNA in pools of ≤ 20 mosquitoes, and PoolScreen software was used to infer territory-wide prevalences of W. bancrofti DNA in the mosquitoes. Wuchereria bancrofti DNA was found in mosquitoes from 16 out of the 27 village areas sampled on Tutuila and Aunu'u islands but none of the five villages on the Manu'a islands of Ofu, Olosega, and Ta'u. The overall 95% confidence interval estimate for W. bancrofti DNA prevalence in the LF vector Ae. polynesiensis was 0.20-0.39%, and parasite DNA was also detected in pools of Culex quinquefasciatus, Aedes aegypti, and Aedes (Finlaya) spp.

Conclusions/significance: Our results suggest low but widespread prevalence of LF on Tutuila and Aunu'u where 98% of the population resides, but not Ofu, Olosega, and Ta'u islands. Molecular xenomonitoring can help identify areas of possible LF transmission, but its use in the LF elimination program in American Samoa is limited by the need for more efficient mosquito collection methods and a better understanding of the relationship between prevalence of W. bancrofti DNA in mosquitoes and infection and transmission rates in humans.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Figure 1. The study was conducted in American Samoa which is composed of the eastern islands of the Samoan Archipelago.

(Swains Island and Rose Atoll not shown.)

Figure 2. Catch of the three most numerous mosquito species which had >1 positive pool overall as a percentage of those three species' combined total in each village.

The number above each bar is the combined total number captured of the three species. Ten traps were operated for 1–2 days in each village, except in Alega-Avaio and Amaua in which six and four traps were used, respectively. “Satala-Leloaloa Area” includes Satala, Anua, Atuu, and Leloaloa villages and “Leone Area” includes Auma, Leone, and Puapua villages.

Figure 3. Mosquito trapping locations in villages on Tutuila and Aunu'u Islands, American Samoa.

Filled circles represent traps which captured mosquitoes in which PCR testing detected W. bancrofti DNA.

Figure 4. Estimated prevalence of Ae. polynesiensis females containing W. bancrofti DNA from trap catch in each village.

Prevalences were estimated by maximum likelihood and confidence intervals by the likelihood ratio method . The total number of Ae. polynesiensis is shown above each bar. “Satala-Leloaloa Area” includes Satala, Anua, Atuu, and Leloaloa villages and “Leone Area” includes Auma, Leone, and Puapua villages.

Figure 5. Estimated prevalence of Ae. polynesiensis containing W. bancrofti DNA for each village versus total Ae. polynesiensis females tested (left) and mean pool size (right).