Efforts to control lymphatic filariasis (LF) have a long history in the Pacific island countries and territories. Although only Wuchereria bancrofti is found in the South Pacific, the epidemiology of transmission is complex. In Micronesia and in the Melanesian countries of Papua New Guinea (PNG) and Vanuatu, the parasite is periodic. Transmission is by Culex quinquefasciatus in Micronesia, while the members of the Anopheles punctulatus complex transmit both malaria and Wuchereria bancrofti in PNG and Vanuatu. In New Caledonia, W. bancrofti is aperiodic and transmitted by the night and day-time biting Aedes vigilax. Where W. bancrofti is subperiodic, it may be transmitted by various Aedes vectors (Table 1). From Fiji to French Polynesia, 15 Aedes vectors are reported (six vectors are found in Fiji alone). Except for Ae. polynesiensis and Ae. vigilax, little is known about the ecology of these mosquitoes. Ae. polynesiensis is arguably the most important LF vector in the Pacific, in part because it exhibits a characteristic known as 'limitation', whereby the percentage of microfilaria (mf) which develop to stage 3 larvae increases with decreasing densities of mf 1. For this reason, Ae polynesiensis may pose the greatest challenge to LF elimination in the region.

The Pacific Programme for the Elimination of Lymphatic Filariasis (PacELF) was the first regional LF elimination programme established 2. Following the Global Lymphatic Filariasis Elimination Programme recommendations for stopping transmission, PacELF has targeted >80% of populations in endemic areas with diethylcarbamazine (DEC) and albendazole annually for at least five years. Evidence that the DEC/albendazole combination will be more successful than monotherapy with DEC against adult worms is supported by reductions in both adult worm antigen levels and in clinical reactions in infected humans receiving the drug combination 3. This combination therapy appears to be more effective in reducing the prevalence and density of mf for longer time periods than DEC alone 4.

Under PacELF, national MDAs were undertaken in the following countries where Aedes species are important vectors of W. bancrofti: American Samoa, Cook Islands, Fiji, French Polynesia, Niue, Samoa, Tonga, Tuvalu, and Wallis and Futuna (Tokelau is no longer considered to be endemic). By the end of 2005, five or more rounds of MDA had been completed in the Cook Islands, French Polynesia, Niue, Samoa and Tonga. Samoa is the only country to have completed its prevalence assessment after five rounds of MDA, with annual coverage ranging from 57% to 90%. In addition to these programmes where Aedes transmit filaria, MDA has been undertaken in countries where Anopheles (Vanuatu and Papua New Guinea) and Cx quinquefasciatus (Federated States of Micronesia, Kiribati, Marshall Islands and Palau) are the primary vectors.

Previous MDA based campaigns using only DEC in Samoa and French Polynesia suggest that caution should be exercised on deciding when to stop MDA. These campaigns often succeeded in achieving significant reductions in mf rates and intensities and LF associated morbidity 567. In Samoa, five extensive campaigns using DEC, including 12–18 month treatments in 1966 and 1971, and single annual doses in 1982, 1983 and 1986, reduced the mf rate from 21% in 1964 to 2.3% in 1987 (Figure 1). Mf rates declined to 0.14% in 1974, following the second DEC campaign, but rebounded to 2.1% within two years 5.

In French Polynesia, twice yearly DEC chemotherapy (6 mg/kg) was administered for 34 years beginning in 1955 to an average of 85% of the population on Maupiti island (excluding the years 1960–67 and 1970–74) 6. In addition, mosquito control using DDT (1955–1957) and larval source reduction (1955–1970) was implemented. Despite these efforts, a comprehensive survey in 2000 found that 0.4% of residents still had circulating mf and 4.6% had antigenaemia 6.

After cessation of the MDA campaigns in Samoa and French Polynesia, mf rates increased 58. While these extensive campaigns succeeded in reducing filariasis transmission, elimination of the parasite was not achieved. Mf resurgence was attributed to poor compliance, inadequate campaign durations and failure of DEC to completely kill or sterilize adult W. bancrofti in treated people (Chow CY: Filariasis vectors and their control in the South Pacific. In 4th Joint WHO/SPC seminar on Filariasis and Vector Control. Apia, W. Samoa, WPR/Fil/9. Manila, World Health Organization, 1974). A subsequent analysis of LF positive individuals on Maupiti suggested that continued transmission can be attributed to individuals who persistently failed to participate in the MDA programme (Nguyen, personal communication).

The historic MDA campaigns in Samoa and French Polynesia, described above, relied on DEC alone. The present GPELF (outside Africa) and PacELF strategies rely on the treatment combination of DEC with albendazole that is reportedly more effective against W. bancrofti. However, recent systematic reviews question whether there really is an improved effectiveness of DEC and albendazole compared to DEC alone. These analyses suggest that the addition of albendazole to DEC does not significantly improve the microfilaricidal 910 or macrofilariacidal activity of DEC 9. These disconcerting conclusions were based on a limited number of studies. A more recent study found significantly greater reductions in microfilaemia intensity and antigenaemia prevalence in persons treated with DEC and albendazole compared to DEC alone 11. Additional studies are clearly needed. There can be little doubt that administration of albendazole with DEC benefits the people receiving the drugs. However, concerns remain about the ability of MDA alone to succeed in eliminating LF without ongoing universal coverage of the eligible target populations, particularly in areas where efficient vectors such as Ae. polynesiensis exist. Hence, there is a need to consider complementary control strategies to ensure the success of the global LF elimination campaign. The only presently available alternative to filaricidal treatments (MDA-based campaigns or distribution of DEC-medicated salt) is vector control.

Vector control was the primary tool for controlling filariasis in the Pacific before effective antifilarial drugs were available and even after effective antifilarials became available vector control was preferred by Pacific island ministries and departments of health because MDA campaigns were considered too labour intensive 12.

Elimination of filariasis using vector control alone has already been successfully documented in the Pacific. Where Anopheles species are the vectors of malaria and filariasis, filariasis was eliminated from areas where indoor residual spraying (IRS) with DDT to control malaria was undertaken in areas of Papua New Guinea 13 and throughout the Solomon Islands 1415.

W. bancrofti was also eliminated from Australia by sanitation campaigns that controlled the major vector, Cx quinquefasciatus 16. Vector control also played a significant role in elimination of LF from Japan 17. Demonstration of the impact of polystyrene beads on Cx quinquefasciatus populations and hence on LF transmission in Zanzibar and India where pit latrines or soakage pits are major breeding sites suggests that control of Cx quinquefasciatus can augment MDA to suppress transmission 181920.

Aedes polynesiensis has been successfully controlled in a small number of trials of limited scope in the Pacific (Table 2). Mesocyclops aspericornis reduced the number of Ae. polynesiensis larvae in treated crab holes by 98% 21. The effectiveness of larval source-reduction campaigns against Ae. polynesiensis has been repeatedly demonstrated in French Polynesia 22 (Kessel JF: Combined control methods in filariasis. Manila, World Hlth Org, unpublished report, 1965 (FIL/WP/16.65), particularly when integrated with other measures 23 including removal of vegetation to facilitate discovery of breeding sites (Byrd EE, St Amant LS: Studies on the epidemiology of filariasis on Central and South Pacific Islands. In SPC Technical Paper, 1959, 125:52–55) and MDA campaigns 24. Such campaigns can be effective even though they are labour intensive (Hairston N: Assessment of filariasis in Western Samoa. Assignment report. Manila, World Hlth Org 1973) and despite the rapidity with which breeding sites reappear after campaigns are concluded 25.

However, not all vector control interventions are successful. Insecticide fogging campaigns had minimal impacts on Ae. polynesiensis biting rates, with reductions of less than 64% reported in three unpublished trials (Suzuki T, Stone F: Laboratory and field tests of insecticides against vector mosquitos of subperiodic filariasis in Western Samoa. Unpublished report to WHO, 1976; Chow CY: Filariasis vectors and their control in the South Pacific. In 4th Joint WHO/SPC seminar on Filariasis and Vector Control. Apia, W. Samoa, WPR/Fil/9. Manila, World Health Organization, 1974; Wharton JD, Jachowski LA. In: Zahar, King and Chow. A review and annotated bibliography on subperiodic bancroftian filariasis with special reference to its vectors in Polynesia, South Pacific. Manila, World Health Organization, 1980). There have not been any studies on controlling the salt marsh breeding Ae. vigilax in the Pacific islands. While runnels (ditches designed to allow tidal flushing of salt marshes) have decreased larval numbers in Australia 26, the impact of runnels on Ae. vigilax biting rates needs to be evaluated before widespread programmatic application can be considered.

Despite significant progress towards LF elimination, a number of challenges to MDA-based LF elimination programmes remain. Firstly, we do not know the exact level and duration of mf suppression required for elimination to be achieved. A target goal reduction of mf prevalence below 1% for transmission elimination may not be an appropriate target where Aedes mosquitoes are vectors. In Samoa, mf rates were less than 0.33% between 1972 and 1974, but LF prevalence rates rose soon after MDA was stopped. It is likely that foci of mf positive individuals capable of initiating resurgence of LF will remain after five or more MDAs with DEC and albendazole. Five years of MDA with DEC and albendazole in Egypt do not appear to have been sufficient to eliminate LF transmission in areas that had high baseline infection rates 27. Effective tools for locating such foci in countries with a low overall residual mf prevalence after five or more annual MDAs have yet to be developed.

A second significant challenge in the Pacific is population mobility. Migration is particularly common in many of the Pacific islands; more Cook Islanders live in New Zealand than on the main island of Rarotonga. Samoans frequently travel between Samoa and American Samoa for economic opportunities and to visit relatives and friends living in the neighbouring country. A significant proportion of internal migrants in Papua New Guinea are mf positive 28. Such migrants are likely to miss annual MDA treatments, thereby raising the possibility of reintroduction of LF.

A third challenge is the presence of individuals whose occupational or social behaviour places them at risk of infection or who consistently fail to participate in MDA, placing their communities at risk of ongoing transmission 2930. Merely increasing the number of years of MDA campaigns will not reach these individuals. A better understanding of their perceptions and priorities will allow tailoring of elimination messages and interventions so that they are locally appropriate and acceptable 31.

A fourth challenge is the potential for W. bancrofti to develop resistance to either DEC or albendazole. Albendazole resistance is already common amongst helminths of veterinary importance. Although there is currently no evidence of resistance to DEC or albendazole in areas where LF elimination programmes are under way, no reliable assay system is currently available to allow assessment of resistance in filarial nematodes. Resistance is more likely to appear in an MDA programme after several rounds of treatment when success appears to be in sight.

A fifth challenge is areas where LF and loiasis are co-endemic 30. MDA cannot be implemented in these areas until mf densities are reduced to levels where significant adverse reactions from MDA will not occur. This is due to severe adverse events associated with treatment with Mectizan^® (ivermectin) in community directed treatment in onchocerciasis if patients have high microfilarial loads of Loa loa.

These challenges can be met by integrating vector control with MDA for LF elimination 32. Vector control is the only presently available adjunct LF control measure with the potential for immediate implementation. Although treatment with doxycycline has been shown to eliminate mf and to have macrofilaridal activity, mass treatment is not practical due to the logistic difficulties of delivering daily doxycycline treatments of 200 mg for 6 to 8 weeks 33. In addition, doxycycline is also contraindicated for use in children under 8 years and in pregnant women.

National scale vector control programmes would have multiple potential benefits for LF elimination programmes. These include (1) the ability to suppress LF transmission without the need to identify all individual 'foci of infection'; (2) minimizing the risk of reestablishment of transmission from imported mf positive individuals; and (3) reducing the spread of any DEC or albendazole resistant W. bancrofti which might emerge. Furthermore, control measures targeting vectors will also decrease the risk of dengue or malaria transmission where, respectively,Aedes or Anopheles are the LF vectors. The reduction of nuisance in addition to vector mosquito biting is likely to enhance community support for ELF programmes. Implementation of vector control strategies as adjuncts to the MDA campaigns will better ensure the success of the elimination efforts and enhance the prospects for sustainable benefit.

Mosquito surveillance and control as adjuncts to MDA are already included in many country filariasis elimination plans even though they may not be actively implemented. Among PacELF countries, activities commonly mentioned include larval surveys for filariasis vectors, environmental sanitation to reduce mosquito breeding sites, use of mosquito nets and ultra-low-volume (ULV) spraying against adult mosquitoes. We need to support vector control efforts to ensure that the limited resources available to ministries of health are spent effectively. There are a number of vector control strategies whose efficacy against transmission of LF has been demonstrated. These include the use of insecticide treated mosquito nets where Anopheles are LF vectors 3435 and the use of polystyrene beads for control of Culex vectors that breed in pit latrines or soakage pits18. These strategies for those vectors are ready for widespread implementation. Mosquito nets should also suppress LF transmission whenever the vectors are night-biting (i.e., Anopheles, Culex, Mansonia, Ae. vigilax), but the degree of the impact needs to be confirmed in controlled trials before implementation on a country basis should be advocated.

Vector control methods for LF are more than just mosquito nets and polystyrene beads. Recommendations for research and control trials were presented at a "LF Research Forum" in 2003 36. Examples of priorities for vector control research included the evaluation of insecticide treated materials (including, but not limited to nets) on Culex and source reduction on Aedes vectors.

Dengue control campaigns are now primarily based on larval source reduction to limit transmission by the primary vector, Ae. aegypti. While important behavioural differences exist between Ae. aegypti and Ae. polynesiensis, there are sufficient similarities between the two species (both bite in the daytime and breed in containers) to allow LF programme managers to draw some lessons from the anti Ae. aegypti campaigns for Ae. polynesiensis control. Integration of LF with dengue control programmes in Aedes transmission areas makes sense. Similarly, malaria control programmes in Anopheles transmission areas would also suppress LF transmission though insecticide treated mosquito nets and/or IRS if LF elimination and malaria control programmes were integrated.

In addition, control of the night-biting but predominantly outdoor-feeding Mansonia vectors should include evaluations of the impact on the adult populations of vegetation removal from rivers and ponds (the larvae obtain air from plants via a modified siphon) as well as the use of fish to reduce the larval populations 37. Similarly, studies to measure reductions of Ae vigilax adults by environmental modifications of salt marshes will be required before vector control can be advocated for extensive geographic areas.

Before implementing a vector control strategy on a country-wide basis, the effectiveness of interventions needs to be validated at the population level, so that limited resources are targeted for optimal control. It is noteworthy that China, the first country to apply for formal verification of interruption of transmission used an integrated control strategy that included vector control 38.

It has been argued that vector control is not cost-effective for LF elimination. However, the cost of failing to achieve final elimination of LF will far exceed the short-term costs of implementing vector control for LF elimination where it is needed.

1. MDA alone should meet the criteria for elimination of LF in many areas, if adequate treatment coverage can be maintained. However, it may be difficult to sustain sufficiently high MDA coverage to achieve LF elimination where Aedes spp are the vectors or where Culex populations are abundant.

2. Vector control has been effective in controlling and even eliminating transmission of W. bancrofti either alone or when implemented in an integrated LF programme with MDA.

3. In addition to reducing the risk of the re-establishment of LF, Aedes and Anopheles control for LF will reduce the risk of dengue and malaria transmission, respectively.

4. Integration of LF with dengue and malaria control programmes where Aedes and Anopheles are the vectors, respectively, will enhance the sustainability and success of the LF elimination efforts.

5. Vector control is the only presently available adjunct LF control measure with the potential for immediate implementation.

6. There is still a need for larger scale trials of vector control interventions for some LF vectors.

7. However, there are vector control strategies that have proven effective in limiting the transmission of W. bancrofti that can be implemented immediately, including the use of insecticide treated mosquito nets and polystyrene beads for control of transmission by Anopheles and Culex, respectively.

The author(s) declare that they have no competing interests.

TRB conceived the main points presented in the manuscript and wrote the original manuscript draft. DND and KI provided unpublished evidence to support the main points presented in the manuscript as well as significant intellectual input into framing the major points presented. RS and WDM provided intellectual input into the arguments presented.