A Search for Snail-Related Answers to Explain Differences in Response of Schistosoma mansoni to Praziquantel Treatment among Responding and Persistent Hotspot Villages along the Kenyan Shore of Lake Victoria

Abstract

Following a 4-year annual praziquantel (PZQ) treatment campaign, the resulting prevalence of Schistosoma mansoni was seen to differ among individual villages along the Kenyan shore of Lake Victoria. We have investigated possible inherent differences in snail-related aspects of transmission among such 10 villages, including six persistent hotspot (PHS) villages (≤ 30% reduction in prevalence following repeated treatments) located along the west-facing shore of the lake and four PZQ-responding (RESP) villages (> 30% prevalence reduction following repeated treatment) along the Winam Gulf. When taking into account all sampling sites, times, and water hyacinth presence/absence, shoreline-associated Biomphalaria sudanica from PHS and RESP villages did not differ in relative abundance or prevalence of S. mansoni infection. Water hyacinth intrusions were associated with increased B. sudanica abundance. The deeper water snail Biomphalaria choanomphala was significantly more abundant in the PHS villages, and prevalence of S. mansoni among villages both before and after control was positively correlated with B. choanomphala abundance. Worm recoveries from sentinel mice did not differ between PHS and RESP villages, and abundance of non-schistosome trematode species was not associated with S. mansoni abundance. Biomphalaria choanomphala provides an alternative, deepwater mode of transmission that may favor greater persistence of S. mansoni in PHS villages. As we found evidence for ongoing S. mansoni transmission in all 10 villages, we conclude that conditions conducive for transmission and reinfection occur ubiquitously. This argues for an integrated, basin-wide plan for schistosomiasis control to counteract rapid reinfections facilitated by large snail populations and movements of infected people around the lake.

Figure 1.

Map showing position of study villages along the shore of Lake Victoria, Kenya. The map was constructed using the QGIS version 2.18.22 and 3.2.1; the base map layer was added from the standard terrain tiles provided from http://maps.stamen.com through the OpenLayers function. The tiles are ^© Stamen Design (San Francisco, CA), under a creative commons attribution (CC BY 3.0) license. The data for the locations were obtained manually using cellular phone global positioning system (GPS) functionality according to the WGS1984 mercator projection. This figure appears in color at www.ajtmh.org.

Figure 2.

Number of Biomphalaria sudanica (blue bars) and Biomphalaria choanomphala (black bars) collected at each of the four sampling times (April and August 2016, January and May 2017) from the six persistent hotspot villages on the top vs. the four responding villages on the bottom. Snail numbers are summed between collection sites at a village. Numbers over bars indicate number of snails infected with Schistosoma mansoni shedders/total number of snails found. Numbers shown within mouse figures represent the number of worms recovered from sentinel mice at that time point (summed across mice and collection sites at that village). An H in a closed circle indicates hyacinths were present and dredging was carried out. An H in an open circle indicates hyacinths were present, but dredging could not be carried out. An asterisk (*) indicates Kabuong Beach was dredged but not Mumbo Beach. This figure appears in color at www.ajtmh.org.

Figure 3.

Relative abundance of Biomphalaria snails at persistent hotspot and responding sites of Lake Victoria in Kenya. (A). Persistent hotspot sites did not contain more Biomphalaria sudanica snails than responding sites when time point, village, and hyacinths were taken into account. (B). Biomphalaria choanomphala was far more abundant at persistent hotspot sites compared with responding sites (P = 0.0091). Error bars indicate standard error of the mean. This figure appears in color at www.ajtmh.org.

Figure 4.

The number of Biomphalaria sudanica collected was greater when hyacinths were present than when hyacinths were not present (P = 0.00973). The presence of hyacinths did not influence Biomphalaria choanomphala abundance (not shown). Error bars indicate standard error of the mean. This figure appears in color at www.ajtmh.org.

Figure 5.

The total number of adult schistosomes collected from sentinel mice was greater in persistent hotspot than responding sites, but there was no statistical difference between them when village, hyacinths, and time point were considered. This figure appears in color at www.ajtmh.org.

Figure 6.

Total number of Biomphalaria sudanica collected at a village was not significantly correlated with the prevalence of Schistosoma mansoni in humans either before or after the drug treatment intervention (before: Pearson r = −0.0369, P = 0.460; after: Pearson r = −0.009, P = 0.49). Yellow = persistent hotspot villages; blue = responding villages. This figure appears in color at www.ajtmh.org.

Figure 7.

Total number of Biomphalaria choanomphala collected at a village was significantly correlated with the prevalence of Schistosoma mansoni in humans both before and after mass drug administration (before: Pearson r = 0.587, r² = 0.345, P = 0.037; after: Pearson r = 0.9034, r² = 0.816, P = 0.0002). Yellow = persistent hotspot villages; blue = responding villages (*two villages with same prevalence). This figure appears in color at www.ajtmh.org.