Lymphatic filariasis in India: Epidemiology and control measuresS Sabesan, P Vanamail, KHK Raju, P Jambulingam
Vector Control Research Centre, Indian Council of Medical Research, Indira Nagar, Puducherry - 605 006, India
Lymphatic filariasis caused by Wuchereria bancrofti and Brugia malayi is an important public health problem in India. Both parasites produce essentially similar clinical presentations in man, related mainly to the pathology of the lymphatic system. Filariasis is endemic in 17 States and six Union Territories, with about 553 million people at risk of infection. The Government of India has accorded a high priority for elimination of this infection through mass chemotherapy programme (annual, single dose of Diethylcarbamazine citrate, i.e. DEC - 6 mg/kg of bodyweight, plus Albendazole repeated four to six times). This campaign has become a part of the National Vector-Borne Disease Control Programme in 2003 under the National Health Policy 2002 and aims to eliminate filariasis by 2015. We discuss here the epidemiology and current control strategy for filariasis; highlighting key issues, challenges and options in the implementation of the programme, and suggesting measures for mid-course corrections in the elimination strategy.
Keywords: Lymphatic filariasis, epidemiology, control strategy, elimination, India
Lymphatic filariasis (LF) caused by Wuchereria bancrofti and Brugia malayi is an important public health problem in India where about a third of the global population lives at risk of this disease.  In India, description of a disease that resembles filariasis was found in Chapter 12 of the 'Susruta Samhita', 6 th century BC. The description of signs and symptoms of this disease by Madhavakara (7 th century AD) in his treatise 'Madhava Nidhana' (Chapter 39) holds good even today. Several studies were carried out on the epidemiological aspects and population dynamics of parasites, vectors and human hosts over the years. ,,,,,,,,,,,,,,,
The National Filaria Control Programme (NFCP) was launched in India in 1955. The control strategy was selective chemotherapy with Diethylcarbamazine citrate (DEC) for 12 days at 6 mg/kg body wt. for parasite carriers detected from the night blood survey, and larval control of vector mosquitoes. The major constraint of the NFCP was that it did not cover the vast majority of the population at risk residing in rural areas and that the strategy demanded detection of parasite carriers by night blood survey, which is less sensitive, expensive, time-consuming and poorly accepted by the community. 
An effective interruption in transmission of B. malayi via mass drug administration (MDA) with annual single dose DEC (6 mg /kg body wt.) was demonstrated at the community level in Shertallai (now: Cherthala), Kerala for the first time during 1987-1990 , and the same strategy showed success for W. bancrofti in Pondicherry (now: Puducherry) during 1992.  The successive studies on these lines also brought similar results and this formed a basis for adopting the MDA strategy for interrupting transmission/ elimination of LF at the global level.  The World Health Assembly made a resolution in 1997 (WHA 50.29), calling for Elimination of LF as a public health problem by the year 2020. Following this, a Global Programme for Elimination of LF (GPELF) was launched in 1999 to facilitate initiation of National programmes in endemic countries including India and others. , A narrative review has been attempted here to highlight the findings of various epidemiological studies carried out over the years, and assess the effectiveness of control strategy and to suggest mid-course corrections, if any, appropriate for facilitating the ongoing LF elimination programme in India.
The literature search aimed at finding epidemiological studies published by the scientific community (PubMED, NLM, etc.) and reports/data related to programme implementation by Government Agencies viz. National Vector-Borne Disease Control Programme (NVBDCP), etc. till date. A criterion was adopted to look for the merits of the study or report that would be utilized for better achievement of the LF elimination goal.
All Geographical Information System (GIS) database was developed using ArcGIS 8.3 software (ESRI, Redlands, CA). District level digital map of India (Survey of India administrative boundary map: scale 1:50,000) was used for LF mapping by human infection prevalence, following the standard procedure. 
The most widespread LF infection is due to W. bancrofti (98%) and the remaining infection by B. malayi (2%), and they are commonly called Bancroftian and Brugian filariasis respectively.
Variants below the species level are of epidemiological significance in the genera Wuchereria and Brugia. These variants have been characterized on the basis of differences in the natural host range, time of appearance of microfilariae (mF) in the human peripheral blood circulation - microfilarial periodicity, and vector susceptibility.
Nocturnally periodic W. bancrofti (mF appear in peripheral blood circulation only during night): The most widespread form transmitted by Culex quinquefasciatus, a ubiquitous mosquito, breeds in almost all organically polluted water bodies.
Diurnally sub-periodic W. bancrofti (mF appear in peripheral blood circulation at any time, but in high count during daytime): Small foci in Nicobar Islands transmitted by Ochlerotatus (Finlaya) niveus, an outdoor day-biting mosquito, breeds in swampy saline water bodies.
Nocturnal periodic B. malayi (mF appear in peripheral blood circulation only during night): A single largest focus in the central coastal part of Kerala, and small isolated foci in six other States, transmitted by Mansonia mosquitoes (Ma. annulifera, Ma. uniformis and Ma.indiana, breed in association with floating hydrophytes in freshwater bodies.
The adult worms of both W. bancrofti and B.malayi live in the lymph canals and lymph nodes of man. The longevity of adult parasite varies from five to 18 years, , but a maximum lifespan up to 40 years has been reported.  The mF are produced from ova in the uterus of the female worm and sheathed mF begin to appear in the peripheral blood six months to one year after infection.  The mF remain in the arterioles of the lungs during the day and emerge at night (nocturnally periodic). They (when picked up by mosquito during blood meal) undergo development in the mosquitoes (intermediate hosts) to form the infective larvae.
The duration of development of mF to infective stage varies between species; W. bancrofti takes 10-14 days, whereas B. malayi 7-10 days. When the infective mosquitoes bite, the L3 larvae escape from the proboscis and actively enter the human host through the wound made by the mosquito for blood-meal. The shortest pre-patent period (from the entrance of L3 to the appearance of mF in the peripheral blood) is estimated at about nine months for W. bancrofti and three months for B. malayi. The appearance of mF depend upon the mating probability of adult worms, pre-patent period and the intensity of transmission.
A district-level endemicity map created for India in 2000 shows that of the 289 districts surveyed up to 1995 (62% of all districts), as many as 257 were found to be endemic.  Seventeen states and six Union Territories were identified to be endemic with about 553 million people exposed to the risk of infection; and of them, about 146 million live in urban and the remaining in rural areas. About 31 million people are estimated to be the carriers of mF and over 23 million suffer from filarial disease manifestations in India.  The state of Bihar has highest endemicity (over 17%) followed by Kerala (15.7%) and Uttar Pradesh (14.6%). Andhra Pradesh and Tamil Nadu have about 10% endemicity. Goa showed the lowest endemicity (less than 1%) followed by Lakshadweep (1.8%), Madhya Pradesh (above 3%) and Assam (about 5%). B. malayi is prevalent in the states of Kerala, Tamil Nadu, Andhra Pradesh, Orissa, Madhya Pradesh, Assam and West Bengal. The single largest tract of this infection lies along the west coast of Kerala, comprising the districts of Trichur, Ernakulum, Alleppey, Quilon and Trivandrum, stretching over an area of 1800 sq km. The infection in the other six states is confined to a few villages. Surveys undertaken recently in Kerala and a few villages in other states revealed either a reduction of foci or complete elimination of the parasite as well as the vector(s) in many villages which were known to be endemic for B. malayi infection four decades back. ,
The district-level endemicity map had some limitations as the survey data obtained at different points of time on the mF rate and disease rate were summed up for calculating the endemicity rate, since there had been no satisfactory definitions for "filarial endemicity" and "endemic area" available in the literature. The terms have now been defined as follows:  Filarial endemicity-incidence of human cases, with evidence of persistent local transmission of filarial infection, which results in either microfilaraemia or disease or both. Endemic area-a geographically defined area, with evidence of persistent local transmission of filariasis. It may be more meaningful to draw separate maps for mF and disease prevalence depending on the availability of data, and the spatial extrapolation if any is required, must be done very cautiously.
LF mapping by mF prevalence has been generated to depict the present scenario of human infection prevalence in India. The results of survey carried out in 443 districts out of a total 593 districts in India at different time points up to 2006 were considered for mF distribution mapping. , In 239 (54%) districts the survey was carried out between 1960 and 1990. In the remaining 204 (46%) districts the survey results were updated after 1990. Accordingly, the mF prevalence map at the district level is made up to the year 2006 and is shown in [Figure 1]. It is observed that the level of mF prevalence is not known in 150 (25.3%) districts. In other words no survey has been carried out at any point of time in these districts. Among the surveyed districts, 172 were found to be with over 1 % mF prevalence. Many of these districts (58%) were detected of this status after 1990. Maximum mF prevalence (12%) was recorded from Nicobar Islands during 1996. The National level average mF rate showed a declining trend from 1.24% in 2004 to 0.63% in 2008. 
Many studies showed that the prevalence of mF in females was lower than that in males in South East Asia, and particularly India. ,, Gender-specific estimates indicated that prevalence of W. bancrofti infection in males was 10% more than that in females.  The studies carried out in rural and urban areas indicated that the mF prevalence increased monotonically throughout childhood and young adult age classes to attain peak levels at approximately 20 years of age and thereafter declined sharply from approximately 20 to 40 years of age. ,,,
Depending on the clinical manifestations, the disease is described as acute or chronic. Acute filarial disease includes filarial fever, epididymo-orchitis, lymphangitis and adenolymphangitis (ADL). Chronic disease includes hydrocele, lymphoedema which can be reversible on elevation or specific exercise of the limb and elephantiasis (which is irreversible oedema of the limb with skin thickening, papillary and nodular growth).
Among all the manifestations in males, hydrocele appears to be the commonest manifestations. Females can be affected by all the manifestations except hydrocele. However, the prevalence of lymphoedema was significantly greater among females than males.  Studies carried out in India and abroad have shown that disease rate was age-dependent and steadily increased from age 10 onwards. ,,,,, Prevalence of chronic manifestations was significantly higher than acute manifestations in both the sexes. Prevalence of lymphangitis was significantly higher in males compared to females. Elephantiasis of limbs was the predominant chronic manifestation among females. A majority of lymphoedema cases and elephantiasis cases in either sex had only lower limb involvement. The prevalence of lymphangitis was significantly higher in 'mF' carriers than others, whereas the prevalence of elephantiasis was significantly higher in 'mF-negative' individuals. In children below 15 years, lymphangitis was the predominant clinical presentation and in adults above 20 years the most common chronic manifestations were hydrocele, lymphoedema and elephantiasis. , Since there has been no chemoprophylaxis, or any radical cure for chronic cases, early diagnosis and treatment is the best option for filariasis. 
Detection of mF carriers through night blood sample examination and line-listing of disease cases are being carried out under the National Programme in selected endemic foci for specific purposes.  However, there is no regular/active surveillance in the programme. We have several diagnostic tools for detection of parasites (Concentration techniques or the Nucleopore Membrane technique; Parasite DNA, using polymerase chain reaction (PCR) and circulating filarial antigens (Immuno-Chromotographic Test and ELISA Og4C3 kit). All these tools may be useful for evaluation of intervention measures, etc. but, may not be feasible for routine surveillance owing to logistic reasons and cost constraints.
Studies on the transmission dynamics of the W. bancrofti - C. quinquefasciatus complex in India have formed the basis for the development of epidemiologic models such as "EPIFIL" , and "LYMFASIM" .  These models have been quantified using longitudinal data from Puducherry, Southern India. , They are useful to address the issues of local relevance by quantifying the parameters describing LF population dynamics in areas where different vector-parasite complexes prevail. Further, these models will be supportive for decision-making on the duration of MDA programme in different epidemiological situations, viz. (i) initial endemicity (force of infection), (ii) coverage levels, (iii) compliance patterns (systematic versus random), and (iv) migration issues (of infected people and vectors; of native populations into endemic areas).
In view of achieving the global elimination of LF, the programme in India has been made a part of the NVBDCP in 2003, under the National Health Policy 2002, and set a target for elimination of LF by 2015. The strategy for achieving this goal is by annual MDA single dose DEC (6 mg/kg body wt.) for at least five years to the entire population of an endemic district (excluding children under two years, pregnant women and severely ill patients), and home-based management of lymphoedema cases and hydrocelectomy operations in identified Community Health Centres (CHCs) and hospitals.
MDA with DEC was launched as a pilot project in 13 districts of seven states in the year 1996.  The NVBDCP up-scaled the MDA to cover a population of 77 million in 2002 from 41 million in 1996-97. During the year 2004, a population of about 468 million from 202 districts was targeted for MDA. There have been several views on the use of Albendazole (Alb) for MDA towards the elimination of LF. [44-48] A large-scale trial on the feasibility and impact of co-administration of DEC and Alb in selected districts in the country was carried out in 2000-05, with the support of Indian Council of Medical Research (ICMR) Task Force. It recommended the co-administration (DEC 6 mg/kg/ body wt. and Alb 400 mg) strategy for all endemic districts. The entire population in all the known endemic districts was targeted for MDA by 2007.
Drug coverage and compliance
In the year 2004 about 276 million persons from 202 districts were given a dose of DEC against an eligible population of 378 million, showing a coverage rate of 73%. The MDA campaign in 2005 covered a population of 463 million using DEC alone and 17.34 million with DEC plus Alb combination.  The results of independent evaluations of MDA have indicated low levels of coverage. ,,, In 2006, the coverage was recorded as only 54.5%. In 2007, all the selected 250 endemic districts have been covered under MDA with coverage of 82% against an eligible population of 518 million. The coverage has in 2008 further increased to 85.92%. 
The drug coverage and compliance rates were evaluated in three districts of Madhya Pradesh.  While the drug coverage was between 29 and 68%, the consumption rate among the individuals who received the drug was in the range 61-77%. These findings imply that the present MDA implementation strategy could not achieve the required level (>85 %) of drug consumption in the present scenario.
Many issues, challenges and options of MDA have been highlighted during the recent past, and they include the identification of target area/community really under 'risk' of infection, drug dosage/distribution, community compliance, and need for site-specific approach in strategies etc. ,,,,
At present, the target for the 'elimination' of infection aimed at the district level is not in relation to 'spatial' and or 'demographic' structure in terms of 'risk'. Ideally, this could be decided and defined at the regional level, since there has been a wide variation in the geographical areas, population density and disease prevalence or amongest different areas within the districts.  In the present context our need is a map showing the transmission risk zone. , And this will facilitate a geographically targeted approach for LF elimination.
As per WHO guidelines any area (usually an administrative unit) in an endemic country with mF prevalence of 1% or more is to be covered, under the MDA programme.  The administrative units in India are "districts". The average population of a district is more than 1·7 million, living in an area of approximately 5340 km 2 . Paradoxically, many of these districts may not qualify for MDA, since the average mF prevalence at this level is < 1%. However, there may be several pockets within a district with varying degrees of prevalence. Because the level of LF infection is not uniform within a district, it may be that the use of an overall mF prevalence as the criterion for introduction of MDA is inappropriate. Indeed, the poor acceptance of drugs by the communities in such areas has revealed that MDA was not felt to be needed.  In reality, the problem areas (i.e. those with mF prevalence >1%) alone need the intervention. These areas are generally small and, to some extent, geographically defined. From a logistical point of view, we suggest that the intervention unit could be at the level of Primary Health Centre (PHC, approximate population of 30000, and cover an area of about 150 km 2 ), because the infection foci are usually clustered at village level within a PHC, and all healthcare delivery of the public-health system is tailored through the PHC. This approach will eventually resolve an issue of public-health ethics - i.e. to avoid giving drugs to communities in non-problem areas within the so-called endemic districts. Similarly, in urban situations, the slum areas are to be given special attention for intervention as the mF carriers and/or the probability of exposure to vector population and generally higher among the dwellers here. 
One of the main reasons for "non-compliance" to the MDA programme is the occurrence of side-effects reported by consumers. DEC is reported to be safe, and does not produce any chronic toxicity.  However, people harboring filarial infection are likely to experience side-effects as a consequence of the interaction between the drug and the parasite, particularly if mF counts are high. In others, the symptoms are usually nonspecific and self-limiting. There is no evidence yet that this message has reached the community at large. The people's awareness on MDA widely varies between different community settings. , An intensive information, education, communication and advocacy campaign involving professional bodies will be highly useful to achieve the desired community compliance.
The present regimen of drug advocated in MDA is not based on a systematic study, but only on the selective approach of a few studies.  Presently, an adult individual of 50 kg body weight is required to take DEC 300 mg i.e., three 100 mg tablets at a time. This dose definition based on body weight could be dispensed with for MDA purpose, since the action of DEC is immune-mediated.  Most of the mF from the bloodstream are destroyed by reticuloendothelial cells of the liver; yet, the precise mechanism of action of DEC remains a subject of debate. Therefore, an optimum single dose DEC for MDA, irrespective of the body weight i.e. uniform for all age classes has to be determined and administered. A single strip of two tablets, one each of DEC and Alb in blister pack could be presented to the community. This would enhance the compliance, besides facilitating a better drug delivery and a reduction in the cost.
Programme managers at the state and district level should be encouraged to adopt the principle of 'directly-observed treatment' through community-based approaches, which would substantially enhance the compliance. This could be done by involving well-trained drug distributors, planning the MDA activities well on time, arranging prompt side-reaction management, organizing appropriate social mobilization and advocacy tied with morbidity management.
The MDA envisaged in 'endemic' countries continuously for five to six years is to reduce the transmission to 'zero' or a near 'zero' level. Also, repeated MDA is expected to result in the 'elimination' of infection i.e., to a level when the LF is no longer a public health problem.  There exists an uncertainty in the statement of LF as a public health problem. One should realize that 'infection' and 'disease' are not synonyms.  All infected individuals do not become 'disease' cases. It is only the diseased individuals that could be viewed as the public health problem. The present MDA, though in the long run may ultimately lead to 'disease' prevention, its immediate objective could only be the interruption in transmission. And, this alone is possible to achieve by the year 2015. Therefore, the strategy and the objectives need to be redefined, so as to make it straightforward and achievable.
Among the important mosquito-vector-borne diseases, the lifecycle of filarial parasite is relatively long. In contrast to other parasites like malaria etc., it does not multiply in the mosquito-vector, nor do the infective larvae multiply in the human host, where each develops to a single adult worm, male or female. The parasite, therefore, does not cause explosive epidemics of disease. A large number of infective mosquito bites are required to produce a patent case.  Taking advantage of this fact, the transmission of infection could very well be reduced/prevented by keeping the vector at bay, through source reduction and personal protection measures.
Under the present LF elimination campaign, vector control is not given due importance. It is to be noted that an effective and sustained vector control has significantly brought down the transmission rate of W. bancrofti, in Pondicherry,  and of B. malayi in Cherthala.  An integrated disease vector control approach needs to be incorporated in the LF elimination campaign throughout the country, and more specifically in areas where persistent transmission occurs even after five to six rounds of MDA.
The objective of MDA is to eliminate LF from the entire country, and logistical convenience is an important aspect from the programme perspective. This necessitates a site-specific approach especially in situations like islands, and other geographically defined areas. For example, in Nancowry group of islands, where vector control is not feasible and the target community was not available at home during day time to receive/consume DEC through MDA,  DEC medicated salt (common salt medicated with 1-4 g of DEC per kg of salt) may be made available to them either free of cost or at a subsidized rate for a period of one to two years, which could alone eliminate LF from such situations. 
In case we continue to depend on the present strategy without further improvement in compliance, the 'low endemic' areas will continue to remain as low-grade transmission foci, and once the programme is stopped, the resurgence of the problem is very likely. Ideally, before it too late, it is wise to take stock of information on the impact of MDA in all endemic areas where the LF elimination programme is implemented. Once these areas reached a low level of transmission, DEC medicated salt and vector control as an adjunct may have to be considered for complete cessation of transmission, as we have many success stories with this strategy in small-scale field trials in India, and several island situations in China and elsewhere. ,, The optimum duration for this strategy, of course, has to be worked out independently for different community settings, considering the socioeconomic characteristics and other logistics. This approach would be sound in effectively liquidating the 'community microfilariae load', and enhancing the community acceptance, as it addresses the mosquito-biting nuisance, too.
The following mid-course corrections are suggested which would facilitate the present control/elimination strategy:
Elimination of filariasis using annual MDA is one of the most economical and beneficial disease control strategies undertaken so far in public health programmes. Now, the whole world is looking at the progress of the LF elimination programme in India as the population living at risk of infection is high, and hence the height of its achievement will greatly have a bearing at the global level.