Entomological assessment of lymphatic filariasis transmission in "hotspot" and control districts after several rounds of mass drug administration in Ghana

Background
Lymphatic filariasis (LF) is a major health problem, which mostly affects individuals in tropical and subtropical regions despite global efforts to control and interrupt its transmission in endemic countries. An estimated 120 million are infected, with about 40 million disfigured and incapacitated worldwide. The main strategy for the control of LF by the Global Programme to Eliminate Lymphatic Filariasis (GPELF) is through mass chemotherapy. In West Africa, specifically in Ghana, mass drug administration (MDA) commenced in the year 2000 with endemic districts receiving at least eight rounds of treatment. In principle, transmission of infections should have been interrupted in all areas after this long period of treatment with reported therapeutic coverage of more than 65%. However, recent information gathered from the Ghana Neglected Tropical Diseases Programme Unit has revealed ongoing transmission in some districts despite their involvement in at least eight rounds of MDA. The main aim of the GPELF is to eliminate this disease by year 2020. However, the current elimination status in Ghana poses a serious challenge in meeting this goal. It is therefore important to investigate driving factors that could possibly be responsible for the observed ongoing LF transmission in endemic districts in Ghana having undergone several rounds of MDA. This will provide information that will add on to existing evidence for appropriate intervention or approach specific to each district.
Aim and objectives
The main aim of this study was to explicitly look at entomological and sociological factors which might possibly be contributing to persistent LF transmission in “hotspot” districts, together with the development and validation of a community-based vector collection system. The specific objectives were (i) to establish a system for collecting large numbers of mosquito samples for xenomonitoring, through the development of a community-based vector collection system; (ii) to determine the mosquito species composition in the various study districts; (iii) to determine the role of different species of mosquitoes in the transmission of lymphatic filariasis in the “hotspot” and control districts; (iv) to determine the role and variations in the cibarial armature of different mosquito species in the study communities; and (v) to undertake a questionnaire survey to determine compliance to MDA and possession and use of bednets and other vector control measures in the study districts.
Methods
This study was conducted in Ahanta West and Kassena Nankana West districts located in the Western and Upper East regions of Ghana, respectively. Both study areas were identified as “hotspot” districts in the country by the Ghana Neglected Tropical Disease Unit of the Ghana Health Service. This was due to high prevalence of LF in sentinel and cross check communities. Additionally, two control districts, Mpohor and Bongo, were also selected due to their zero microfilariae (mf) prevalence.
A 13-month (July 2015 - July 2016) collection of mosquitoes was concurrently conducted in all study districts. This involved the training of community vector collectors (CVCs) in the various mosquito collection methods, which included human landing catches, pyrethrum spray catches and window exit traps. Supervisors were further trained on how to package samples for shipment to the Noguchi research team. Sampled mosquitoes from the respective districts were later subjected to molecular analysis for the detection of Wuchereria bancrofti infections as well as determine the sibling species of the Anopheles gambiae complex. Mosquito dissections were also done to estimate various entomological transmission indices. Variations in cibarial armatures of various mosquito species were investigated by clearing of mosquito heads with chloral hydrate to make cibarial teeth visible for counting.
Questionnaires were administered in the various districts to obtain information on MDA compliance and vector control activities. Data were also obtained from the Ghana Neglected Tropical Disease Unit on the number of rounds and MDA coverage in the respective districts.
Results
A total number of 31,064 mosquitoes were collected from all the districts using human landing collections, pyrethrum spray catches and windows exit traps. Mosquitoes sampled were Aedes, Anopheles coustani, An. gambiae, An. pharoensis, Culex and Mansonia species. Molecular identification of An. gambiae complex showed An. gambiae s.s. in all districts. An. arabiensis and An. melas sibling species were identified from Kassena Nankana West/Bongo and Ahanta West districts, respectively. Furthermore, there was no difference in the shape and mean number of cibarial teeth of mosquitoes collected from hotspot and control districts in the Western and Upper East regions. In general, MDA coverage was ≥65% for all districts. However, MDA coverage in the Upper East region was <65% for Kassena Nankana West in 2003 and 2004/2005 in Bongo district.
Validation of mosquitoes sampled by CVCs showed no significant difference in the numbers sampled by CVCs and the research team in the dry (P = 0.258) and rainy (P = 0.309) season in southern Ghana. However, there was significant difference in the numbers sampled by research team and CVCs during the rainy (P = 0.005) and dry (P = 0.033) season in northern Ghana. Assessment of the cost-effectiveness of sampling mosquitoes for xenomonitoring activities using CVCs and research team was done. Results indicated that the cost of sampling mosquitoes was lower using CVCs compared to research team (USD 15.17 vs 53.74 USD). The highest recurrent and capital cost was personnel (USD 21,370.04) and transportation (USD 2,900.14) costs, respectively.
Furthermore, the assessment of W. bancrofti infection in mosquitoes as post-MDA surveillance tool using xenomonitoring was done. Results showed the sampling method human landing collections (27,739: 89.3%) recording the highest number of mosquitoes, followed by pyrethrum spray collections (2,687: 8.7%) and windows exit traps (638: 2.1%). Restriction fragment length polymorphism (RFLP) showed the high presence of An. coluzzii species in almost all districts. Dissections reported the presence of W. bancrofti in An. melas from Ahanta West district. Also, the annual transmission potential (ATP) for An. melas from the Ahanta West district was 7.4.
Conclusion/recommendations
Persistent LF transmission in “hotspot” areas in this study presents information that shows the importance of local understanding of factors affecting elimination of LF. However, the study shows that it is feasible to use CVCs to sample large numbers of mosquitoes with minimal supervision. It is also cost-effective to use CVCs to collect mosquitoes for xenomonitoring compared to a dedicated research team. The inclusion of CVCs in xenomonitoring activities promotes active community participation and ownership of vector control activities. Additionally, W. bancrofti infections are found and sustained in Ahanta West district in An. melas that uses the phenomenon of limitation for lymphatic filariasis transmission. This study also showed the possibility of using xenomonitoring as a post-MDA surveillance tool. We recommend that LF interventions should consider spatial heterogeneities and best approach to use in all endemic foci. Moreover, xenomonitoring should be considered in the decision-making processes to stop or continue MDA by stakeholders and programme managers. Also, mosquito traps and sampling methods should be safe, practical and convenient for CVCs to use with less supervision and the inclusion of vector control activities by programme managers and stakeholders in planning intervention programmes.