The Democratic Republic of Congo (DRC) is the largest lymphatic filariasis (LF) endemic country in Africa with over 49 million people at risk 1 2 3 . The challenge of mapping LF in the 2.3 m sq km of inhabited regions is further compounded by co-endemicity with Loa loa and the poor road infrastructure in the post-conflict environment 1 . Delimiting L. loa endemic areas using a high resolution mapping strategy, is critical in minimising the risk for severe adverse events (SAEs) associated with mass drug administration (MDA) with ivermectin. Nevertheless, scaling up treatment for soil transmitted helminths (STH) with albendazole and widespread distribution of long-lasting insecticidal treated nets (LLINs) will positively impact LF endemicity. Incorporating these factors into an overlap mapping strategy will constitute a valuable prerequisite for LF control and surveillance.

The Global Programme to Eliminate Lymphatic Filariasis (GPELF) initiated in 2000 aims to eliminate the disease as a public health problem by the year 2020. The recent publication of the Progress Report detailing the successes of the programme to date have been recorded, as well as, the Strategic Plan for the next decade of activities 1 . The principle strategy is to interrupt LF transmission with MDA using annual treatment with the drugs ivermectin and albendazole in countries in Africa co-endemic with onchocerciasis, and elsewhere with DEC and albendazole. Africa has a significant burden of LF caused by the parasite Wuchereria bancrofti with 35 endemic countries, and a further 6 countries where the process of the verification of the absence of transmission is underway 1 2 3 .

While GPELF is making progress overall 1 4 with significant economic saving being demonstrated 5 , Africa remains behind other regions with 16 countries still to start MDA implementation. Many of these countries are conflict/ post-conflict countries, characterised formally as Least Developed Countries (LDCs) and are among the poorest in the world with minimal human and financial resources 6 . These constraints pose problems for the national LF programmes with the potential to severely hinder the 2020 goal of LF elimination globally. The WHO Strategic Plan 1 identifies two priorities for the Africa Region to achieve the expansion of MDA; first to resolve the treatment challenges posed by the co-endemicity of L. loa due to concerns associated with SAEs when treatment with ivermectin is given in onchoceriasis programmes 7 8 ; second to explore how best to utilise vector control as a supplemental intervention for interrupting transmission, especially in filarial co-endemic areas.

The expansion of LF programmes in Africa will inevitably spread into L. loa and Onchocerca volvulus endemic areas. Due to the increased risk of SAEs, the current MDA regime of ivermectin and albendazole is not recommended where W.bancrofti prevalence overlaps with L. loa, or where all three filarial parasites co-exist unless those implementing onchocerciasis control adhere to strict guidelines developed by the Mectizan Donation Programme. However, recent maps on the distribution of L. loa produced from the rapid assessment procedure for loiasis (RAPLOA) 8 9 10 11 , and of Community Directed Treatment with ivermectin (CDTi) produced from rapid epidemiological mapping for onchocerciasis (REMO) 8 12 help identify high risk areas that may require fine scale mapping and/or alternative intervention strategies.

Importantly, the recent L. loa and CDTi distribution maps will also highlight the potential synergies and benefits between the LF elimination and onchocerciasis control programmes 1 12 . For example, it has been shown that CDTi treatment over 5-6 years with coverage of 65% or more could significantly reduce the prevalence and intensity of other filarial parasites, mainly W. bancrofti even though there appeared to be little impact on the adult worm 13 14 . Similarly, repeated doses of albendazole used in STH programmes could potentially impact on LF prevalence 15 . Conversely, the scale up of LF elimination programmes using ivermectin and albendazole across large regions of Africa could help to reinforce the achievements made by the onchoceriasis programmes by increasing the use of ivermectin in areas where onchocerciasis was not controlled as it was previously defined as hypo-endemic but now referred to as low transmission areas 8 . This will assist in reducing the potential residual human reservoir population of O. volvulus that could then cause re-infection in areas that have been cleared of the disease. The widespread distribution of albendazole as a component of LF programmes will also enhance the efforts of STH programmes 16 .

Additional benefits may be gained from malaria vector control programmes, especially in Africa where the filarial parasite W. bancrofti is transmitted by Anopheles species 17 18 . Malaria is commonly controlled using insecticide treated nets (ITNs)/LLINs, which act as baited traps for the mosquito when a person sleeps under it 19 . The use of ITNs/LLINs has shown to be effective in reducing LF prevalence in a W. bancrofti - L. loa co-endemic area in Nigeria 20 , and in W. bancrofti transmission areas in Kenya 21 , Uganda 22 and Papua New Guinea 23 . Currently, there are large scale malaria control programmes under way across Africa, which involve the mass distribution of ITNs/LLINs 19 . The rapid expansion of malaria vector control could also significantly impact LF prevalence, and distribution maps of ITNs/LLINs would help national LF programmes assess their utility, especially in countries yet to start MDA and those with L. loa endemic areas.

WHO 1 stated "Better tools are needed to assess risk as are guidelines on criteria for including or excluding people from treatment programmes". This paper seeks to address these issues by introducing the concept of micro-stratification overlap mapping (MOM) of filarial infections and effective interventions in Central Africa to define more precisely where and what strategies for LF programmes might be implemented in areas of L. loa co-endemicity, whilst also assessing the historic data available on LF in one of the most important countries in Africa facing this problem, the DRC.

The DRC is a priority country for GPELF since it is yet to start MDA for LF elimination, and there are several significant challenges for the national LF programme. First, DRC is considered to have the second highest population at risk of LF in Africa (est. 49, 140, 000) 1 . However, this risk is not well defined as no recent in depth studies appear to have been carried out, and the historic information available is largely from studies undertaken in the pre-independence era. This data suggests that the distribution of W. bancrofti is discontinuous 24 25 26 . Second, L. loa is endemic across large remote geographical areas, and closely associated with dense tropical rain forests which favour the vectors Chrysops sp. dimidiata and silacea 9 27 28 . Third, DRC is one of the poorest countries in the world and has been ravaged by civil unrest and conflict for nearly two decades 6 29 . It is now classified as a post conflict country, but these conflicts have left a scarcity of resources, poor infrastructure and numerous challenges to the health care system. Fourth, there is little vector control in place as an alternative intervention with less than 10% of households owning an ITN reported in 2007 19 30 31 . These factors coupled with its sheer geographical size of 2.3 m sq km, limited transport networks and the remoteness of dispersed communities, present major challenges for many health programmes.

The aim of this paper is to review and synthesise the current knowledge of the distribution of W. bancrofti in the DRC, and factors that will impact on the control and elimination of LF such as loiasis co-endemicity, onchocerciasis control programmes, STH deworming activities and malaria bed net distributions. We thus introduce a new term Micro-stratification Overlap Mapping (MOM), which we suggest is a prerequisite for planning any future LF programmes in countries where there is co-endemic loiasis and indeed other control programmes which impact on LF implementation. For this reason, MOM needs to be widely applied when any Preventive Chemotherapy Programme is planned given the synergistic impact of ivermectin and albendazole used together in LF elimination or separately in onchocerciasis or STH programmes respectively and where there is bed net distribution. This information is essential to optimise the future LF MDA implementation strategy to ensure safety, maximum cost effectiveness as well as impact.

This study highlights the lack of detailed and current data on the distribution of LF in DRC, a country considered to have one of the largest burdens of disease in the world, and second in sub-Saharan Africa, after Nigeria 1 . The national LF Programme in DRC is yet to commence MDA implementation, which is complicated by approximately one third of the country being endemic for loiasis 9 ; a contraindication for any MDA regimen which includes ivermectin. The reason for the lack of information and action may be attributed to the widespread civil unrest affecting many parts of the country for decades, which has left it severely under-resourced in public and private sectors, as well as the absence of efficient transports systems making access to many areas of the country difficult 29 32 .

The insidious nature of internal instability is a major barrier for disease control programmes, and has implications for the national LF programme in terms of accessing remote, rural endemic regions, readily and safely 32 53 . These problems are not unique to DRC as many of 16 countries still to start MDA in Africa are loiasis endemic and considered to be conflict or post-conflict countries 54 . This group of countries are among the poorest and most fragile in the world 6 , which raises the importance of an integrated effort between international partners, and the various national NTD 8 16 36 and vector control programmes 19 55 to ensure that resources are maximised, and the elimination of these diseases achieved collectively.

The review and mapping of historical data in DRC highlights gaps in our knowledge and the difficulties in fully defining the problem. The value of collating and analysing disparate data sources cannot be underestimated. This activity is recommended for countries planning to start MDA, as it will help to identify high risk areas and key risk factors that are crucial to control efforts even if resources are limited. We also advocate that the new mapping approach of MOM is essential in countries where there is co-endemic loiasis as well as other control programmes which impact on LF implementation. However, it may be important to evaluate if transmission is still ongoing in some areas where CDTi or deworming activities have taken place for several years, since reported coverage may not always be true or may not always have the predicted impact on the targeted diseases. Therefore, before excluding an area for LF control, it may be prudent to rapidly assess the impact of other interventions on the LF endemicity.

The simple reproduction and overlapping of historical data and maps, clearly shows the close association of LF distribution with the Congo River and its tributaries in different regions of the country. This provides key information about the ecology and transmission of W. bancrofti vectors, which is important in DRC where only one study has been carried out in the past 70 years 44 . In Matadi, Bas Congo, An. funestus was identified as the most important vector of W. bancrofti 44 , however, other entomology studies indicate a diverse range of Anopheles across the country 56 57 58 . An extensive survey found An. moucheti and An. wellcomi most abundant along the central Congo River system, and An. paludis in the surrounding hinterland 59 . In Katanga, the mosquito species were considered to be distinct to other ecological regions 56 , and in the forested Mayombe region, Bas Congo, An. gambaie s.l was found to be much more abundant than An. funestus and An. moucheti 60 , which may explain the lack of LF found in this region by Fain 48 .

The recent maps on the distribution of L. loa in Africa 9 27 are a useful resource for the national LF programme in DRC. The L. loa maps elucidate SAEs risk areas, where extra precautions and alternative intervention strategies may be required, especially if LF is found to be endemic. Comparisons between the historical and current L. loa maps indicate that its distribution has remained relatively stable for more than half a century, and been a long-standing public health problem, particularly in the Oriental Province. The problem extends across large areas of central Africa affecting many vulnerable people 9 , and poses a major obstacle to the LF and onchocerciasis elimination programmes in Africa. Whilst L. loa is not included in the list of WHO's 17 NTDs 61 its importance as an impediment to progress of programmes based on preventive chemotherapy cannot be understated.

Overlapping the C. silacea and C. dimidiata distribution data with the L. loa map indicates that both vectors may transmit the infection in the high risk areas. In Central Africa these Chrysops sp are found to be sympatric in tropical rain forests and rubber plantations, and shown to bite throughout the day and have different annual transmission cycles 26 28 . To eliminate LF in rain forest loiasis co-endemic areas will require new innovative strategies, possibly including Chrysops control as a novel approach. Although the development of villages, clearing of vegetation and the spraying of insecticides of Chrysops breeding sites have shown to produce a degree of control of loiasis transmission 62 63 64 65 , these methods are impractical. Therefore, it may better to use impregnated/baited traps or trapping methods similar to those used for Human African Trypansomiasis (HAT), which is transmitted by Glossina but shares some biological features of Chrysops biting patterns and habitats 66 . Recent data shown in the global atlas of HAT indicates geographical overlap in some high risk L. loa areas, and the use of MOM may elucidate areas where both vectors may be trapped for control purposes.

Alternatively, a combination of new drug regimens and integrated vector management (IVM) 67 may be an option in high risk LF areas and co-endemic high loiasis areas. The use of alternative drug strategies such as twice yearly treatment or higher doses of albendazole 15 or 4-6 week course of doxycycline 68 may be a better approach to reduce filaria loads, and reduce the risk of disease and SAEs in selected populations. These alternative drug strategies used in combination with ITN/LLINs and/ or indoor residual spraying (IRS) may also significantly reduce W. bancrofti transmission 19 . More synergies between malaria and LF programmes are essential, especially as the distribution of ITN/LLINs is increasing dramatically. Although no large-scale IRS activity is planned for DRC 55 , historical IRS activities using gammexane, show considerable reductions in An. moucheti, An. gambiae s.l and An. paludis densities in DRC 69 , and more recently with DDT in An. funestus in other countries 70 , which suggests that IRS could be targeted with significant impact, especially in Oriental Province. However, it will be important to monitor insecticide resistance of vector control activities 70 71 .

The extent to which W. bancrofti and O.volvulus overlap in L.loa > 40% high risk areas in Oriental Province is unknown, however, Woodman 72 in neighbouring southern Sudan illustrates that the three diseases are co-endemic across relatively large geographical areas. The use of MOM could elucidate different patterns of co-endemicity within such areas, to determine if certain interventions would be of risk or benefit. The three finer scale maps in this study show that different regions have different spatial and overlapping patterns with different risks and benefits. For example, in the Mayombe region, Bas Congo 48 L. loa was most prevalent, and only overlapped with W.bancrofti and O.volvulus in distinct areas. This region has high bed net/ITN coverage, which is of significant benefit; however, the overlap of the CDTi priority area in the L.loa and O.volvulus co-endemic area is a risk as ivermectin treatment could result in SAEs. In contrast, in the Bandundu (Banningville) region 47 , the overlaps between the CDTI priority areas and W. bancrofti and O.volvulus co-endemic areas will be of considerable benefit as ivermectin could reduce the prevalence of both diseases (as well as STHs) along the river with limited risk of SAEs. However, there is a low or no coverage of bed nets/ITNs in this region 30 31 .

On a broader scale, the country level maps highlighting the overlaps between different disease distributions and interventions provide important insights into the potential risks and benefits of multiple large-scale disease control programmes operating in one country. It clearly identifies high risk and vulnerable populations which need to be targeted with more effective or alternative innovative intervention strategies. Importantly, it also shows the potential large-scale benefits that the combination of the onchocerciasis, malaria, and future STH programmes could provide to the national LF programme and vice versa. Clearly, there are huge benefits of coordinated and overlapping mapping activities so Ministries of Health have a clear picture of the epidemiology for planning purposes.

This paper identifies a new approach to mapping that takes account of factors that will influence country LF elimination strategies and plans. The paper suggests that country plans require higher resolution mapping that has been used to date, before implementation of LF activities can be efficiently deployed. This is because 1) distribution of ivermectin through APOC projects will already have had an impact of LF intensity and prevalence 2) DRC as an example, has been up scaling bed net /ITN distribution which will impact over time on transmission of W. bancrofti and 3) recently available predictive maps of L. loa allow higher risk areas to be identified. This will allow LF implementation to be initiated with reduced risk where L. loa is considered non-endemic. We believe that using the proposed MOM approach is a prerequisite for planning the expanded distribution of drugs for LF programmes in countries co-endemic for filarial infections.

The authors declare that they have no competing interests.

DHM conceived the idea for the study. LKH identified data sources, developed the study design and collated, mapped and analysed the data. BCT assisted with the data collation, mapping and analyses. MJB contributed to the analyses and interpretation of results. All authors contributed to the writing of the manuscript and approved the final version.