Optimization of the most widely used serological tests for a harmonized diagnosis of Toxoplasma gondii infection in domestic pigs

Abstract

The intake of Toxoplasma gondii tissue cysts through raw or undercooked pork meat is one of the main infection sources for humans. Thus, surveillance is recommended to control and prevent infection in domestic pigs. However, the lack of comparative studies hampers the updating of their performance and the comparison of seroprevalence data. Therefore, the aim of this study was to develop and validate three in-house tests and accomplish a comparative analysis of the most widely used serological tests employed in pigs. A panel of sera from pigs experimentally infected with either oocysts or tissue cysts from type II and III isolates (n = 158) was used to develop and validate a tachyzoite-based Western blot assay. Then, this technique was used as a reference to develop and preliminary validate a lyophilized tachyzoite-based enzyme-linked immunosorbent assay and an immunofluorescence antibody test. Next, a comparative study of the three in-house tests and three widely used commercial ELISAs (IDScreen®, PrioCHECK™ and Pigtype®) was accomplished with the abovementioned sera together with an additional serum panel of pigs experimentally infected with oocysts from the type II isolate (n = 44) and a panel of naturally infected pigs (n = 244). The results obtained by the majority of the tests were regarded as reference, and data analyses included TG-ROC calculations and agreement tests. Finally, the kinetics of anti-T. gondii IgGs from experimentally infected pigs was analyzed. Excellent sensitivity (Se) and specificity (Sp) values (≥ 93%) and moderate to near perfect agreement (k = 0.63–0.91) were observed using sera from experimental infections without requiring further readjustment, except for PrioCHECK (100% Se, 73% Sp). However, the Se of IDScreen® (87%) and TgSALUVET WB (71%) and the Sp of PrioCHECK (72%) were slightly or notably reduced when sera from naturally infected animals were analyzed, which also influenced the kappa values (k = 0.30–0.91). Cutoff readjustments increased the Se and Sp values to equal to or above 97% for all tests, except for TgSALUVET WB, which can be used as a reference for initial validation of tests, but it is not recommended for routine diagnosis. Seroconversion was recorded from two weeks post-infection by most of the tests, with significantly higher IgG levels in sera from pigs infected with the T. gondii type III vs. type II isolate. Again, differences regarding the test employed were observed. Differences in the diagnostic performance among tests evidenced the need to harmonize serological techniques to obtain comparable and reliable results.