Field evaluation of the MM3-SERO ELISA for detection of anti-Fasciola IgG antibodies in milk samples from individual cows and bulk milk tanks.

We carried out a field evaluation of the MM3-SERO ELISA for the diagnosis of Fasciola hepatica infection, by analysing serum and milk samples from individual cows and samples from bulk milk tanks. The diagnostic performance of the assay was assessed with serum samples from all 257 cows in eight fluke-free herds, and 240 cows with natural fasciolosis (diagnosed in vivo and/or post-mortem). Assay performance for individual milk samples was determined by analysis of paired serum and milk samples from 947 lactating cows from 33 F. hepatica-infected farms. The diagnostic usefulness of the assay for bulk tank milk was evaluated by analysis of bulk milk from infected (33) and non-infected (35) farms. For serum samples, the sensitivity, specificity and diagnostic accuracy of the assay were respectively 99.2% (95% CI: 97.0%-99.9%), 100% (95% CI: 98.6%-100%) and 0.997 (95% CI: 0.987-1.000). The only two infected animals in which serum antibodies were not detected had very low parasitic burdens (with only 2 and 3 flukes observed). The performance of the MM3 SERO ELISA for individual milk samples was similar to that for serum samples, and the stepwise linear regression revealed a strong correlation between the results for the milk samples and the serum samples (R(2)=0.84; p<0.001). The agreement between results obtained with pairs of serum and milk samples was very high: there was matching classification in 96% (910/947) of paired samples (kappa=0.92; p<0.001). Individual milk samples may therefore be used, instead of serum samples, in the MM3-SERO ELISA, for reliable detection of seropositive cows. Testing bulk tank milk samples enabled detection of infected herds, even when the within-herd prevalence of infection was as low as 12%. We conclude that the MM3-SERO ELISA is a sensitive and highly specific test for serodiagnosis of bovine fasciolosis, and can be used with individual samples of either serum or milk. Use of the assay with bulk milk samples enables estimation of the within-herd prevalence of infection.