Reducing Baylisascaris procyonis Roundworm Larvae in Raccoon Latrines

Baylisascaris procyonis roundworms, a parasite of raccoons, can infect humans, sometimes fatally. Parasite eggs can remain viable in raccoon latrines for years. To develop a management technique for parasite eggs, we tested anthelmintic baiting. The prevalence of eggs decreased at latrines, and larval infections decreased among intermediate hosts, indicating that baiting is effective.

ingest them (3). Transmission often occurs at raccoon latrines when eggs are ingested with seeds found in fecal material (4). Larvae migrate through intermediate host tissues and can enter the central nervous system, resulting in death (3). Adult raccoons become infected when they prey on infected intermediate hosts (3). Raccoon population densities have increased in response to increased anthropogenic resources that are available in agricultural and urban ecosystems (5). Thus, raccoon latrines often exist near human habitats, increasing the risk for zoonoses (4).
Reported cases of human B. procyonis roundworm infections are rare (n = 18), and all have occurred in North America; however, prevention is a public health priority because of the severity of the resulting neurologic disease (6)(7)(8)(9)(10). Our objective was to develop a management technique that could interrupt transmission of B. procyonis roundworm eggs between raccoons and intermediate hosts, ultimately decreasing the environmental levels of eggs and potential for reinfection. We examined the effects of latrine removal and treatment of raccoons by using randomly distributed anthelmintic baits on the basis of B. procyonis roundworm prevalence at latrines and among intermediate hosts. By implementing a specifi c, protocol-based approach to disease prevention, supported by experimentally derived data, we hope to provide public health offi cials with an effective, spatially explicit, prophylactic method for reducing infection risk.
We conducted this study in Grant, Miami, and Wabash counties in north-central Indiana in portions of the Upper Wabash Basin. This area is 88% agricultural; only 8% of the landscape remains forested (11). Some contiguous riparian forest remains; however, most patches are <5 hectares (ha; 1 ha = 10,000 m 2 ) (11). Our experiment was conducted in 16 forest patches (1.91-8.80 ha). Eight treatment patches received anthelmintic baits, and 8 control patches did not. The range of patch sizes, levels of patch isolation, and raccoon densities in treatment and control patches were representative of the landscape (Figure).
In March 2007 (spring 07), we removed all visible latrines (n = 559) in the treatment patches. We located latrines by systematically searching all appropriate horizontal substrate and area at the bases of large trees throughout each forest patch (3). After manual removal, we used a torch to sterilize the substrate and surrounding soil associated with each latrine (online Technical Appendix, www. cdc.gov/EID/content/17/1/90-Techapp.pdf). At control sites, we sampled a minimum of 20 latrines (n = 198) by removing ≈2 g fecal material per fecal deposit at each latrine (12). We returned to our study sites 3 additional times for fecal sampling in October and November 2007 (fall 07), June 2008 (summer 08), and November 2008 (fall 08). During these subsequent visits, we sampled ≈2 g of fecal material per fecal deposit at a minimum of 20 latrines in all treatment and control patches. All samples were stored at −20°C until they were examined for B. procyonis roundworm eggs. Eggs were identifi ed by microscopic examination following centrifugal fecal fl otation in Sheather sugar solution (3). We identifi ed B. procyonis roundworm eggs on the basis of size and morphologic appearance and designated each sample as positive or negative. Prevalence was measured as the proportion of positive samples at each study patch during each sampling period. Differences between pretreatment and posttreatment prevalence and between treatment and control patches were determined by using log linear analyses performed with PROC CATMOD SAS version 9.1 (SAS Institute Inc., Cary, NC, USA) (goodness-of-fi t tests).
In spring 07, after the initial latrine removal from treatment patches, baits were distributed throughout treatment patches once a month for the duration of the study. Baiting densities were determined on the basis of average abundance of raccoons in each study patch (online Technical Appendix).
Prevalence of B. procyonis roundworm larvae within an intermediate host, white-footed mice (Peromyscus leucopus), was determined. A minimum of 10 mice were captured from each of the 16 study patches during each of 3 sampling periods: 1 pretreatment (summer 07), and 2 posttreatment (fall 07 and summer 08). After capture, mice were euthanized with carbon dioxide and refrigerated until examination for B. procyonis roundworm larvae. Brains were removed and examined separately by pressing them between glass plates, and larvae were examined under a dissecting microscope. We recovered larvae from tissues digested in acid-pepsin solution (3). Larvae were counted and identifi ed (3). Prevalence of infection was determined for mice within each study patch for each sampling period. Differences between treatment and control patches were determined by Fisher exact test (13).
We collected 1,797 fecal samples. Pretreatment sampling of latrines in spring 07 detected B. procyonis roundworm eggs at 757 (33%) of latrines sampled across all patches (Table). However, prevalence of eggs in treatment patches declined by >3-fold after baiting in all sampling periods (p<0.04). Our baseline pretreatment estimate of prevalence of infection among intermediate hosts did not differ (p = 0.426) between treatment patches (32%) and control patches (37%). Approximately 1 year after baiting activities began, we detected a signifi cant decline in the prevalence of B. procyonis roundworm larvae in mice between treatment and control patches (27% vs. 38%; p = 0.05; Table).

Conclusions
Current public health initiatives to prevent human infections with B. procyonis roundworms focus on education of human health care and veterinary professionals (6). Our practical approach decreased prevalence of the parasite, suggesting decreased transmission and possibly reduced risk for humans. Baiting strategies have effectively controlled rabies (14) and decreased prevalence of zoonotic parasites, including Echinococcus multilocularis tapeworms (15). Our baiting strategy combined with latrine removal effectively decreased egg levels at latrines and ultimately decreased prevalence among mice. Hegglin and Deplazes (15) demonstrated a long-term decrease in prevalence of E. multilocularis tapeworms among foxes (defi nitive hosts) after monthly baiting for ≈4 years and conjectured that this decrease was caused by decreased infections among intermediate hosts. Our study supports their hypothesis because we measured decreases in prevalence among intermediate hosts after baiting. The reduction of prevalence at latrines and among intermediate hosts suggests that our low-cost approach (online Technical Appendix) could have a lasting effect on transmission dynamics; however, further study to assess frequency of distribution and type and dose of baits for sustained prevalence is needed. Raccoon latrines are commonly found near homes (4), and implementation of baiting strategies, in conjunction with traditional raccoon management on public lands, could reduce the risk for transmission on nearby private properties.