Performance of rapid tests for canine visceral leishmaniasis diagnosis in Argentina

To diagnose dogs infected by Leishmania infantum rK39 rapid diagnosis test is widely used in the Americas, while dual path platform (DPP) was recently adopted by Brazil. In this study we assessed the performance of rK39-RDT and DPP tests in recent urban transmission scenarios of Argentina. The sensitivity and specificity were evaluated with a sera panel and field samples, taken as true infected those from parasitological and/or PCR positive tests. Since none of these tests can be taken as a gold standard, the performance was also evaluated using Latent Class Analysis, a statistical modeling technique which allows to estimating sensitivity and specificity defining a latent class variable as the reference standard. The sensitivity of both tests in the panel was around 92% (symptomatic dogs 96%, asymptomatic 83%), while the sensitivity in field samples of rK39-RDT was 77%, and DPP 98% (mean in symptomatic dogs 89%, asymptomatic 82%). The specificity was similar for both tests and samples, around 98%. Therefore, these tests are acceptable for program dog population-based studies, as spatial stratification, focus intervention and follow up, and they could be used for individual screening and confirmation of clinical presumptive diagnosis in polysymptomatic dogs. The inability to discriminate between immunity and actual infectiousness suggest that a combination with other non-immunological based tests will be required for highly sensitive/specific diagnosis in order to targeting control measures in individual reservoirs from public health perspective, as for individual management from animal health perspective.

Para diagnosticar perros infectados por Leishmania infantum, en las Américas se utiliza ampliamente la prueba rápida rK39, mientras que DPP fue adoptado recientemente por Brasil. En este estudio se evaluó el desempeño de las pruebas rK39-RDT y DPP en escenarios de transmisión urbana reciente en Argentina. La sensibilidad y especificidad se evaluaron con un panel de sueros y muestras de campo, considerando muestras infectadas verdaderas aquellas con pruebas parasitológicas y/o de PCR positivas. Como ninguna de estas pruebas puede considerarse estándar de oro, el desempeño también se evaluó mediante análisis de clases latentes, una técnica de modelado estadístico que permite estimar sensibilidad y especificidad definiendo una variable de clase latente como estándar. La sensibilidad de ambas pruebas en el panel fue de alrededor del 92% (perros sintomáticos 96%, asintomáticos 83%), mientras que la sensibilidad en muestras de campo fue rK39-RDT: 77%, y DPP 98% (media en perros sintomáticos 89%, asintomáticos 82%). La especificidad fue similar para ambas pruebas y muestras, cerca de 98%. Por lo tanto, estas pruebas son aceptables para estudios programáticos caninos de base-poblacional, como estratificación espacial, intervención de foco y seguimiento, y podrían utilizarse para el tamizaje individual y la confirmación del diagnóstico clínico presuntivo en perros poli-sintomáticos. La incapacidad de discriminar entre inmunidad e infectividad real sugiere que se requerirá una combinación con otras pruebas, de base no inmunológica, para un diagnóstico suficientemente sensible/específico que permita definir las medidas de control en reservorios individuales, tanto para salud pública, como para la gestión individual en salud animal.

Distribution and pathogenicity of Trypanosoma cruzi isolated from peridomestic populations of Triatoma infestans and Triatoma guasayana from rural Western Argentina

We assessed the distribution of Trypanosoma cruzi infection in peridomestic triatomines collected manually at a district-wide scale in rural villages around Olta, Western Argentina, and typed the isolated strains according to their pathogenicity to laboratory mice. Of 1623 triatomines examined, only 14 (0.9 percent) were infected with T. cruzi based on microscopical examination of feces. The prevalence of T. cruzi infection was 0.8 percent in Triatoma infestans, 2.3 percent in T. guasayana, and nil in T. garciabesi, T. platensis, and T. eratyrusiformis. Local transmission occurred in kitchens, store-rooms and goat corrals or nearby, though at very low levels. T. cruzi was detected by at least one parasitological method in 11 (79 percent) of 14 microscope-positive bugs. Hemoculture was the most sensitive method (67 percent) followed by culture of organ homogenates, histopathology or xenodiagnosis of inoculated suckling mice (55-58 percent), and culture of microscope-positive bug feces (46 percent). The evidence suggests that most of the isolated T. cruzi strains would be myotropic type III. Our study establishes for the first time that peridomestic, microscope-positive T. guasayana nymphs were actually infected with T. cruzi, and may be implicated as a putative secondary vector of T. cruzi in domestic or peridomestic sites.