This is the first detailed report on the effects of repeated rounds of MDA with DEC and albendazole on filariasis infection parameters in an area with Anopheles transmission. The study site had baseline filarial infection rates that were high in the global context but moderate for Papua New Guinea. Baseline infection rates were higher in older age groups and higher in males than in females. These trends probably reflect increased cumulative exposure in older people and differences in exposure between males and females. Since the gender differences were much more striking in adults, these could also be related to biological factors such as hormone levels.
The performance of the ICT antigen test and the Bm14 antibody test require some comment before we address the impact of MDA on filariasis parameters in this study. The ICT test detected filarial antigenemia in a high percentage of untreated Mf-positive subjects (detected by membrane filtration of 1 ml of venous blood), and this sensitivity was maintained after MDA. This is in contrast to a recent study from Kenya that reported decreased sensitivity of the ICT test in Mf carriers (detected by the counting chamber method with 0.1 ml of finger prick blood) after 2 rounds of MDA 
. The Bm14 antibody test was less sensitive than the CFA test in Mf-positive subjects in the present study, and it was also somewhat less sensitive than previously reported 
. However, both the CFA test and the Bm14 antibody test were much more sensitive than Mf detection for detecting filariasis activity in the study communities, and this is consistent with prior reports 
. Antibody rates in children <11 years of age were much higher than Mf or CFA rates, both before and after MDA. This supports the strategy of testing sentinel populations of young children for antifilarial antibodies as a means of assessing recent filariasis activity in communities 
Population MDA compliance rates were very good throughout the 3-year study period. However, this required a lot of effort, with multiple visits to the study villages and labor-intensive recruitment of village residents. It might be easier to achieve high compliance rates in a national MDA program that was not linked to collection of venous blood. MDA compliance rates were low in children <6 years of age. This may have reflected parents' concerns about blood tests in their young children. We believe that the national LF elimination program in Papua New Guinea will need to develop new strategies to achieve high MDA compliance in young children. Information campaigns should emphasize the dual benefits of MDA on LF and soil-transmitted helminth infections.
MDA dramatically reduced all filariasis infection parameters in the study villages. As in earlier studies, Mf rates in people and parasite DNA rates in mosquitoes fell more rapidly than CFA or antibody rates 
. While low residual filarial DNA rates in mosquitoes indicate the presence of Mf carriers in communities following MDA, this does not necessarily mean that significant LF transmission will continue in these areas. PCR can detect DNA from dead filarial parasites in mosquitoes 
, and most Mf taken up by anopheline vectors do not survive to become infective larvae (L3) 
CFA and anti-filarial antibody rates fell more rapidly after MDA in children <11 years of age than in the total study population; this may be because infection intensities and years of infection/exposure tend to be lower in young children than in older individuals. Although infection rates decreased in children after MDA, many young children had positive CFA and/or anti-filarial antibody tests after 3 rounds of MDA. Of course, these children had been exposed to the parasite for years prior to MDA. Children born after LF transmission has been interrupted should not have positive CFA or antibody tests 
. Surveillance activities to verify interruption of transmission should focus on testing young children. Mosquito monitoring provides a non-invasive means of detecting residual infections in communities if the number of young children available for testing is small.
This study provided interesting longitudinal data on effects of MDA on Mf clearance in individuals and on incidence rates for different filariasis parameters. Mf clearance rates in this study after one or more annual doses of DEC with albendazole were higher than those reported from clinical trials performed in Sri Lanka and Egypt 
. However, while all subjects in the clinical trials had high baseline Mf counts, all Mf carriers were considered in current community-based study. The current study also found that Mf clearance rates after MDA were lower in persons with high baseline Mf counts. The incidence data are very exciting, because they demonstrated that MDA significantly decreased the incidence of Mf, CFA, and antifilarial antibodies in the study population. This is the first study that has documented decreased filariasis incidence rates following MDA. However, incidence events observed after MDA-3 suggest that 3 rounds of MDA was not sufficient to completely eliminate LF transmission in this setting.
The impact of three rounds of MDA in the current study was at least as impressive as that recently reported from Egypt (Giza governorate) with the same MDA regimen, although this “high prevalence” study area in Egypt had lower LF infection rates before MDA (11.5% Mf, 19.0% CFA, and 3.07% mosquito DNA) and higher MDA compliance rates than our study site in Papua New Guinea 
. This suggests that the encouraging results reported from Egypt can be replicated in areas with very different epidemiological parameters. Changes in infection parameters following MDA must be considered in the context of the local mosquito vector. An. punctulatus
is a less efficient LF vector than Cx. pipiens
(the principal LF vector species in Egypt). We do not know the minimum requirements for sustained transmission of W. bancrofti
by An. punctulatus
. However, Tisch et al recently reported that LF parameters continued to decrease in villages in a different area of Papua New Guinea (in East Sepik Province, approximately 300 km from the Usino study site) for at least 5 years after Mf rates had been reduced to low levels by 5 rounds of MDA with DEC and ivermectin 
. It is possible that three rounds of DEC with albendazole (which reduced the Mf rate to 1.3% with a 97.9% decline in CMFL) would have been sufficient to reduce LF transmission rates to unsustainable levels in the Usino study area. The study villages received a fourth round of MDA in 2006; long term follow-up studies will be needed to determine whether four rounds of MDA have interrupted LF transmission in this area. For the time being, the results from Usino are quite encouraging. Taken together with earlier studies, they suggest that LF elimination should be feasible in Papua New Guinea and other endemic areas with Anopheles
transmission if MDA can be effectively delivered to endemic populations. Prospects for LF elimination should be even brighter if MDA can be integrated with distribution of insecticide-treated bednets