An Expanded Transmission Assessment Survey to Confirm the Interruption of Lymphatic Filariasis Transmission in Wallis and Futuna

Abstract.

Historically, the human prevalence of Wuchereria bancrofti infection in Wallis and Futuna (WAF) was among the highest in the Pacific and mass drug administration (MDA) against lymphatic filariasis (LF) either with diethylcarbamazine citrate (DEC) or the combination of DEC and albendazole had been implemented for decades. To determine whether LF antigen prevalence in WAF was lower than 1%, the infection threshold for elimination in an area where Aedes spp. are the principal vectors, we conducted the WHO-recommended transmission assessment survey in 2012. We present the results of a school-based survey, which targeted 1,014 students in all 13 elementary schools in WAF. From a fingerprick, the circulating filarial antigen (CFA) positivity was checked for grade 2–5 students using BinaxNOW filariasis test (immunochromatographic test). Of 935 children tested, three were positive for CFA in two schools. At the territory level, this was below the critical cutoff of nine cases, if the whole territory was considered as a single evaluation unit. The prevalence of CFA in WAF is less than 1%, reaching the goal for LF elimination set by the WHO. We were able to recommend stopping LF MDA and move to post-MDA surveillance to detect any recrudescence. This survey successfully paved the way for WAF to be validated as achieving LF elimination as a public health problem by 2020.

INTRODUCTION

Lymphatic filariasis (LF) is a mosquito-borne parasitic infection, which causes debilitating conditions such as hydrocele and lymphedema.¹ Globally, approximately 90% of infections are caused by the parasite Wuchereria bancrofti and the remainder by Brugia malayi or Brugia timori. The major vectors of W. bancrofti are mosquitoes of the genus Culex for urban areas, Anopheles for rural areas in Africa and selected areas of Asia-Pacific, and Aedes in the Pacific Islands Countries and Territories (PICTs).² In 2000, there were 120 million people affected by LF and 40 million estimated to be suffering from either hydrocele or elephantiasis, whereas presently 886 million people in 52 countries worldwide remain threatened.³ In 1997, the World Health Assembly (WHA) acknowledged LF as a global public health problem and adopted a resolution to eliminate LF by 2020 (WHA 50.29)¹. The recommended strategy for LF elimination in areas where onchocerciasis is not co-endemic is annual mass drug administration (MDA) of diethylcarbamazine citrate (DEC, 6 mg/kg body weight) and albendazole (400 mg per person) to the entire population at risk and provision of morbidity management and disability prevention to those with chronic manifestations of LF.²

According to the mathematical models on filariasis transmission dynamics,⁴ it was estimated that five or six rounds of MDA with adequate coverage levels (i.e., at least 65% of the total population) were sufficient to lower LF prevalence to a level that interrupts transmission.⁵ The levels differ by the specific vector species: that is, 1% in areas where the vectors are Aedes spp. and 2% in Culex or Anopheles spp.⁶ To assess the impact of MDA, the WHO recommends transmission assessment surveys (TAS), which are conducted by testing children born after the initiation of MDA to determine whether there is recent transmission.⁷ The TAS design is based on the lot quality assurance sampling (LQAS), thus testing the hypothesis whether the level of circulating filariasis antigen (CFA) is equal to or below the critical cutoff threshold rather than estimating a point prevalence with CIs.⁸ One of the important prerequisites of TAS is the sentinel and spot-check site surveys after the fifth MDA round, no sooner than 6 months after the MDA. The results of the assessment should show the prevalence of microfilaraemia to be < 1% in all sites to continue to implement the TAS.⁷

Wallis and Futuna (WAF) is a French territory in the South Pacific and one of 16 LF-endemic PICTs, which joined the Pacific Programme to Eliminate Lymphatic Filariasis (PacELF) in 1999,⁹ a regional counterpart of Global Programme to Eliminate LF.² Lymphatic filariasis in WAF is caused by W. bancrofti, and its principal vector is Aedes polynesiensis.⁹ Since 2002, annual rounds of MDA using DEC and albendazole were implemented by the Public Health Agency (PHA) of WAF targeting the entire territory and all age-groups, except for children younger than 2 year, pregnant women, and severely ill people.^2,9 The first MDA round was conducted in 2002, with a coverage of 60.2%, and following rounds were continued annually until 2007, with coverage ranging from 52.5% to 66.4%.¹⁰

No surveillance was conducted until 2012 when the PHA and the PacELF office of the WHO had a meeting to decide 1) whether PHA should continue or stop MDA after a decade of annual MDA rounds and 2) what would be the best form of assessing the impact of MDA rounds. Consensus was made on obtaining expert consultations to conduct its first transmission assessment survey (TAS 1) following the newly available WHO guideline.⁷ Hence, the objective of the study was to assess whether MDA has succeeded in lowering the CFA prevalence to a level that recrudescence⁷ is unlikely to occur among children in the territory who went through at least four to six rounds of MDA.

MATERIAL AND METHODS

Study area.

Wallis and Futuna is located at about two-thirds of the way from Hawaii to New Zealand (west of Samoa and northeast of Fiji) in the South Pacific (Figure 1). It is made up of three volcanic islands along with 20 islets, which are further divided into two island groups that lie about 260 km apart, namely, Wallis Islands and Futuna. The territory occupies a land area of 145 km² and is one of the smallest countries in the world, with a total population of 14,000.¹¹ Most of the residents are inhabitants of the two major islands of WAF proper,¹² which are two historical and existing administrative divisions.

Figure 1.

Sketch map of Wallis and Futuna showing the location of all elementary schools and schools with positive circulating filariasis antigen. Black dots are schools with immunochromatographic test card–positive case(s).

Study design and sampling strategy.

The whole WAF territory was considered as one TAS evaluation unit (EU), to enable the health authority to make stop MDA decisions at the territory level. Wallis and Futuna follows the French schooling system, and there are 13 “elementary” schools altogether (Figure 1), with five grades, starting from the first grade comprising children aged between five and 6 years to the fifth grade with children aged between 10 and 11 years. According to the algorithm for the choice of the TAS design in areas where Aedes spp. is the principal vector for LF transmission, the target population was to include all children in classes 1 and 2 across 13 schools, given that the net primary school enrollment ratio is ≥ 75% in WAF and the number of children in these two classes in the EU is < 1,000.⁷

To take advantage of this surveillance activity, older children included all 1,014 children in classes 2 to 5 in each of 13 elementary schools, given that most of the children in the first grade are 5 years old and the TAS guideline recommends 6- to 7-year-old children to be tested.⁷

Data collection procedure and specimen examination.

School principals of the 13 elementary schools were contacted through the Department of Education to explain the purpose of the survey, and a list of students from each school that included the student’s name, gender, age, and grade was obtained. As participation was voluntary, a letter to parents was drafted explaining the purpose of the survey in French, requesting their written consent and asking them to have their children bring their consent forms to the school on the designated survey date.

Before the survey, investigators were trained on blood sampling procedures and on how to use and interpret the tests as described in the WHO monitoring and impact assessment manual.⁷ Blood was collected from each participant by fingerprick and tested on-site using filariasis rapid tests. For detecting CFA, 100 µL of blood was collected in the capillary tube provided by the manufacturer. The collected blood was then transferred slowly to the sample pad of the immunochromatographic test (ICT) card. In total, l00 µL of the blood was required for a BinaxNOW filariasis ICT card (Alere, Scarborough, ME). The results of the test were read in the field exactly at 10 minutes as per the instructions of the manufacturer. A positive antigen (Ag) control was used to test the validity of the ICT cards before the cards were sent to WAF by the WHO.

Data management.

Data generated before the survey were managed using Excel (Microsoft, Redmond, WA). Transmission assessment survey data were recorded on paper questionnaires by trained investigators. The questionnaire form included basic demographic information and test results. The investigators reviewed the questionnaires before leaving the school to ensure its completeness. Data were entered in a database programmed with EpiData version 3.1 (EpiData Association, Odense, Denmark) and analyzed using STATA version 12 (StataCorp LP, College Station, TX).

Statistical analysis.

Transmission assessment survey analysis is based on the LQAS technique in a sample. However, if the population size of the targeted age-group is 1,000 or less in an EU where Aedes spp. is the main vector for LF transmission, as it was the case in WAF, it is recommended to conduct a census survey instead of sampling.⁷ In this case, the critical prevalence cutoff (1%) was calculated as the first integer smaller than 0.01 times the total number of individuals tested.⁷ If the total number of CFA positive is at or below the critical cutoff, the EU is considered to pass the TAS, which means LF transmission will not sustain and MDA can be stopped in the EU. If the total number of CFA positive is above the critical cutoff, the transmission is likely ongoing and MDA should be continued.

Ethical considerations.

Written informed consent in French was obtained from the parents or guardians of participating children. The field team also provided opportunities for students to opt out if they did not want to participate in the study in an environment where they felt free to express their wishes, even if their parents sent the signed consent form.

The study was reviewed and approved by the PHA of WAF and the WHO Western Pacific Regional Office Ethics Review Committee.

RESULTS

The survey was conducted in November 2012, in collaboration with PHA of WAF and WHO Division of Pacific Technical Support.

Demographic characteristics of the study population.

Of 1,014 children targeted for TAS, 1, 950 (93.7%) were present at schools but 11 refused the blood test. The overall survey participation was 92.6%, with the school-level participation ranging from 72.4% to 100% (Table 1). Of the 939 children tested, 477 (50.8%) were male. They were born between December 1999 and January 2006, and their mean age was 9.5 years.

Table 1.

School-level survey coverage and CFA prevalence in Wallis and Futuna

School Numbering*	Number of students registered	Number of students tested for CFA	Survey coverage† (%)	Number of invalid ICT‡	Number of positive ICT	CFA prevalence (%)
Futuna schools
1	28	27	96.4	0	0	0.0
2	54	52	96.3	0	0	0.0
3	108	105	97.2	0	1	1.0
4	100	95	95.0	0	0	0.0
5	52	52	100.0	0	0	0.0
Sum	342	331	96.8	0	1	0.3
Wallis schools
6	120	117	97.5	1	2	1.7
7	69	59	85.5	0	0	0.0
8	87	63	72.4	0	0	0.0
9	40	34	85.0	0	0	0.0
10	104	90	86.5	0	0	0.0
11	72	70	97.2	1	0	0.0
12	76	76	100.0	0	0	0.0
13	104	99	95.2	2	0	0.0
Sum	672	608	90.5	4	2	0.3
Total	1,014	939	92.6	4	3	0.3

CFA = circulating filariasis antigen; ICT = immunochromatographic test.

* Numbering of schools as shown in Figure 1.

† Coverage is defined as the proportion of students who were tested with immunochromatographic test (ICT) cards divided by the number of students registered.

‡ Invalid means when ICT failed to yield the result.

Geographical distribution of the circulating filariasis Ag prevalence.

In total, the ICT card was invalid for four, so they were excluded from further analysis. Three children of 935 were found to be circulating LF Ag positive. This was below the critical cutoff of nine cases (i.e., first integer < 0.01 × 935), and we concluded that WAF passed its TAS 1 at the territory level. The positive cases were a 10-year-old girl and an 8-year-old boy, both residents in Wallis who were registered at the same school, and a 7-year-old girl resident in Futuna. They were further treated with the combination of DEC and albendazole after the blood test.

DISCUSSION

Our survey of elementary schoolchildren from classes 2 to 5 indicates that at the time of the study, WAF was on track in achieving the global target of reducing LF antigenemia to less than 1% in Aedes-endemic areas.⁷ The implementation of MDA is likely to have been the main intervention as vector management has not been implemented for LF or for malaria as it is non-endemic for malaria.¹³ However, this is difficult to definitely conclude which rounds and from when to when of these rounds mostly had an impact because recent program data especially from 2008 to 2011 are lacking, and there have not been any post-MDA coverage surveys conducted in WAF. Also, not all the coverage levels from 2002 to 2007 reached the target of 65%. Nevertheless, it is also unlikely that LF transmission interruption would have occurred without continued rounds of preventive chemotherapy, considering that was the only intervention which had been implemented for decades.

Historically, in the 1800s, LF was common in WAF, and it was reported that half of the adult population suffered from elephantiasis; the rate of microfilariae was 40% in 1954, 20.4% in 1959, and 21.9% in 1977.^9,14–16 Monthly DEC distribution targeting the entire population officially began in 1978 by the French army and was continued until 1987, and then the distribution interval was decreased to every 6 months until 2001.⁹ Impact of these interventions was demonstrated as microfilariae prevalence levels had decreased from 5.3% in 1978 to 3.2% in 1985 in Wallis, and from 1.7% to 0.4% in Futuna over the same period of time.⁹

A more recent survey in 2001 supported by the PacELF adapted highly sensitive techniques such as ICT cards, showing that CFA prevalence ranged between 0% and 2.3% in five districts of WAF (Figure 2) (Public Health Agency, unpublished data). These data were considered as baseline by the initial monitoring and evaluation framework proposed by the PacELF, and named as “A” survey, which became the basis for commencing MDA rounds in the territory from 2001 to 2005.⁹ The country program had planned to conduct the “B” survey after three rounds of MDA in 2004 as midterm evaluation in sentinel sites same as those of the initial A survey, and the “C” survey on the completion of five rounds in 2006, covering randomly selected clusters across the territory (Figure 2).⁹ However, the initial B survey was in fact conducted after five rounds of MDA in 2006, enrolling 1,539 people where six were found LF Ag positive with per district CFA prevalence levels from 0% to 0.8% among the general population (Figure 2) (Public Health Agency, unpublished data). The country program further planned to proceed to the final D survey (Figure 2), if the CFA prevalence in the C survey is < 1% as an equivalent to TAS nowadays. However, neither the C survey nor the D survey materialized until our study was undertaken. The principal reason was partly because of frequent turnover of the responsible staff of the national program. Numerous attempts had been made to reestablish the partnership and to retrieve the relevant information which would enable the Regional Programme Review Group on neglected tropical diseases in the WHO Western Pacific Region (RPRG WPRO) to provide strategic guidance on the way forward, and finally, the recommendation of moving to TAS was made following the availability of the TAS guideline and the 2006 survey data⁷ to RPRG WPRO in 2011.

Figure 2.

Monitoring and evaluation framework proposed by the Pacific Programme to Eliminate Lymphatic Filariasis⁹ (A) and what was implemented by the national program (B) in Wallis and Futuna.

In recalling the initial monitoring and evaluation framework set by the PacELF,⁹ there have been major changes in the current TAS methodology. The program can now use TAS when 1) all IUs should have had at least five effective (> 65%) rounds of MDA and 2) in all sentinel and spot-check sites, the prevalence of microfilaria (Mf) is < 1% or the Ag is < 2% after the last effective round.^7,17 The latter enabled RPRG WPRO to make the recommendation of conducting TAS in 2011 for the national program of WAF, via considering the results of the 2006 survey as successful pre-TAS results, even if it was not a multiple cluster–based survey. This was not the standard practice as of 2006 under the PacLEF, when the multi-clustered C survey was required among all age-groups.⁹ In other words, it would have been possible for WAF to adapt TAS right away in 2006, if the current TAS guideline had been available at that time and the MDA rounds before the 2006 survey were all regarded as effective. In turn, the national program could have achieved the target of elimination of LF as a public health problem set by the WHO much earlier. This shows the importance of the partner’s role in the provision of technical support based on the up-to-date information and maintaining the national program’s vigilance on interpreting the program progress.

It might be of importance to explore whether replacing the cluster-based survey among all age-groups, so called as “C” survey as sentinel and spot-check site surveys, would have had any impact in making stop-MDA decisions. Stopping MDA before the interruption of transmission poses the risk of recurrence that may not be detected for years, possibly after the elimination goals have been achieved.¹⁸ On the other hand, the continuation of preventive chemotherapy in the absence of disease requires additional resources that could have been allocated to other public health programs. Survey methods to determine stopping points must be pragmatic for the national programs to implement,¹⁹ and adapting sentinel and spot-check site surveys is in line with this. However, such approaches may not be robust to pick up residual “hotspots” of transmission, which require special attention.¹⁸ For instance, the number of minimum clusters proposed by the PacELF was up to 30 across the territory or implementation unit (IU), whereas according to the current guideline, at least one sentinel site for each IU and at least one spot-check site per sentinel site should be identified. It may be worthwhile to review the program data in countries where TAS had failed would have had different results if randomly chosen multiple clusters were surveyed and especially in which settings.

As for the differences in the survey design of the D survey and TAS, it is notable that the threshold proposed was changed from ICT positivity 0.1% to 1%. This was mainly due to the high alpha error with the previous criteria, meaning that the prevalence is already lowered enough, but it is not possible to conclude to be lower than 0.1%.⁹ Also, the minimum sample size to test the hypothesis of the prevalence < 0.1% was up to 3,000 5-year-old children, which could be practically challenging compared with the current TAS sampling framework. In our study, our age-group was expanded to those who were older than 6–7 years, considering that the presently proposed age-groups in the TAS methodology, children aged 6–7 years, may have limitations for assessing residual LF endemicity. In the extreme, especially where the validation of LF elimination had been declared at the national level, even adult populations have been assessed to detect remaining filarial infections that had previously passed school-based TAS.²⁰ Several explanations have been proposed, such as 1) young children could be less exposed to mosquitoes than adults because they spend less time outside in the evening²¹ and 2) children may be more often protected by bed nets when they sleep than adults.²¹ However, relying on adults or older children in stop-MDA assessments has also disadvantages, as Ag-positive, Mf-negative adults in an ongoing MDA setting may be interpreted to mean the presence of reproductively viable worms.¹⁸ Therefore, including different age-groups in TAS could be considered as an alternative option, especially when antigenemia is more prevalent among adults in endemic areas, where Mf-positive infections have been found in adults but no Ag-positive children identified.^18,22,23

This serologic survey targeting elementary schoolchildren in WAF provides evidence for interruption of LF transmission in the territory. In light of these results, it is recommended stopping MDA in WAF and moving to the post-MDA surveillance phase to detect any LF recrudescence by reconducting the TAS: A final TAS (TAS 3) was recommended in 5 years (2016–2017) to confirm that the disease has been eliminated as a public health problem in WAF even after stopping MDA. The decision was adopted by the PHA, and the TAS 3 was conducted in 2017, which showed only one positive in the school of Wallis,²⁴ confirming that the transmission of LF is reduced below the elimination threshold. Finally, WAF was validated by the WHO in 2018 as having reached the goal of elimination of LF as a public health problem.^25,26 This is another success story of LF elimination in the world, and the territory joins six other PICTs which also have reached the goal of elimination of LF as a public health problem. Wallis and Futuna experience also reaffirms the effectiveness of the MDA strategy in LF elimination.²⁷

This survey is subject to several limitations. The rapid test had less than 100% sensitivity and specificity. The sensitivity of ICT cards is 95% and specificity ranges from 72% to 100%, with possible lower sensitivity in areas of low endemicity.²⁸ This may have resulted in false-positive tests and a slight overestimation of CFA-positive cases. A second limitation relates to data collection. One might question what exactly happened to the national LF elimination program during the period of 2008 and 2011. For instance, personal communication with the previous program manager before the survey implementation affirmed that there had been an MDA round in 2011 in Wallis only, but without actual numbers, it was decided that only the data from 2002 to 2007 to be officially deposited to the WHO’s PCT databank.¹⁰ This may have impacts on estimating the actual duration without MDA rounds. Last, although with a high participation rate of 90%, it is possible that the 10% of children who did not participate in the survey are not similar to the other 90% in terms of the infection status.

CONCLUSION

Since joining the PacELF, the National LF Programme of WAF has successfully implemented MDA rounds with the two-drug regimen, reduced LF prevalence, and has been able to implement the WHO standard TAS methodology as part of LF endgame surveillance strategy. This study mainly presents the results of the TAS 1 conducted by the country program, highlighting the first evidence of a reduction in the risk of LF across the whole territory, where the subsequent TAS showed the persistent low-level infection. The program achieved the global elimination target of LF by 2020, and the entire population in WAF is now considered to have very low or no risk of the infection. The timing of the TAS 1 could have been much earlier, suggesting the importance of continued vigilance in interpreting the survey results by the national program and partners.