Lymphatic filariasis endgame strategies: Using GEOFIL to model mass drug administration and targeted surveillance and treatment strategies in American Samoa

American Samoa underwent seven rounds of mass drug administration (MDA) for lymphatic filariasis (LF) from 2000-2006, but subsequent surveys found evidence of ongoing transmission. American Samoa has since undergone further rounds of MDA in 2018, 2019, and 2021; however, recent surveys indicate that transmission is still ongoing. GEOFIL, a spatially-explicit agent-based LF model, was used to compare the effectiveness of territory-wide triple-drug MDA (3D-MDA) with targeted surveillance and treatment strategies. Both approaches relied on treatment with ivermectin, diethylcarbamazine, and albendazole. We simulated three levels of whole population coverage for 3D-MDA: 65%, 73%, and 85%, while the targeted strategies relied on surveillance in schools, workplaces, and households, followed by targeted treatment. In the household-based strategies, we simulated 1-5 teams travelling village-to-village and offering antigen (Ag) testing to randomly selected households in each village. If an Ag-positive person was identified, treatment was offered to members of all households within 100m-1km of the positive case. All simulated interventions were finished by 2027 and their effectiveness was judged by their 'control probability'-the proportion of simulations in which microfilariae prevalence decreased between 2030 and 2035. Without future intervention, we predict Ag prevalence will rebound. With 3D-MDA, a 90% control probability required an estimated >= 4 further rounds with 65% coverage, >= 3 rounds with 73% coverage, or >= 2 rounds with 85% coverage. While household-based strategies were substantially more testing-intensive than 3D-MDA, they could offer comparable control probabilities with substantially fewer treatments; e.g. three teams aiming to test 50% of households and offering treatment to a 500m radius had approximately the same control probability as three rounds of 73% 3D-MDA, but used < 40% the number of treatments. School- and workplace-based interventions proved ineffective. Regardless of strategy, reducing Ag prevalence below the 1% target threshold recommended by the World Health Organization was a poor indicator of the interruption of LF transmission, highlighting the need to review blanket elimination targets. Author summaryLymphatic filariasis (LF) is a parasitic disease caused by infection with filarial worms and is currently endemic in 72 countries. Mass drug administration (MDA) is used to interrupt LF transmission, in order to reduce LF prevalence below a target threshold set by the World Health Organization. When prevalence is below said threshold, continued LF transmission is believed to be unsustainable. American Samoa implemented seven rounds of MDA from 2000-2006 and, more recently, a further three rounds of MDA in 2018, 2019, and 2021. Despite these recent interventions, American Samoa has not yet reached elimination targets. In this study we use GEOFIL, a LF transmission model specific to American Samoa, to investigate the most efficient interventions for LF elimination-whether to continue with MDA or instead try targeted surveillance and treatment strategies. We find that, without further intervention, LF prevalence will rise, but further rounds of MDA have the potential to eliminate LF. Household-focused targeted surveillance and treatment interventions also have the ability to eliminate LF using far fewer treatments than MDA; however, they require testing a large number of people. Furthermore, we found that reducing antigen prevalence below threshold targets does not necessarily guarantee successful long-term elimination.