Abstract
Taenia solium cysticercosis was reported in official veterinary and medical statistics to be highly prevalent in pigs and humans in Madagascar, but few estimates are available for pigs. This study aimed to estimate the seroprevalence of porcine cysticercosis among pigs slaughtered in Antananarivo abattoirs. Firstly, the diagnostic performance of two antigen-ELISA techniques (B158B60 Ag-ELISA and HP10 Ag-ELISA) and an immunoblotting method were compared with meat inspection procedures on a sample of pigs suspected to be infected with (group 1; n=250) or free of (group 2; n=250) T. solium based on direct veterinary inspection in Madagascar. Sensitivity and specificity of the antigen ELISAs were then estimated using a Bayesian approach for detection of porcine cysticercosis in the absence of a gold standard. Then, a third set of pig sera (group 3, n=250) was randomly collected in Antananarivo slaughterhouses and tested to estimate the overall prevalence of T. solium contamination in pork meat traded in Antananarivo. The antigen ELISAs showed a high sensitivity (>84%), but the B158B60 Ag-ELISA appeared to be more specific than the HP10 Ag-ELISA (model 1: 95% vs 74%; model 2: 87% vs 71%). The overall prevalence of porcine cysticercosis in Antananarivo slaughterhouses was estimated at 2.3% (95% credibility interval [95%CrI]: 0.09-9.1%) to 2.6% (95%CrI: 0.1-10.3%) depending on the model and priors used. Since the sample used in this study is not representative of the national pig population, village-based surveys and longitudinal monitoring at slaughter are needed to better estimate the overall prevalence, geographical patterns and main risk factors for T. solium contamination, in order to improve control policies.