1.3 MARCO TEÓRICO
1.3.4 Control Interno
The distributions of PP values overall and within each herd were skewed to the right, with most animals having a low PP value and only a small number having a high value. This was seen across all herds, even in farms with 60-80% of animals testing positive. PP value and fluke burden are correlated (Charlier et al., 2008; Salimi-Bejestani et al., 2008), so the observed distribution may reflect the fact that even in herds where most animals are infected, most animals have only a light parasite burden. This is supported by other research on fluke numbers at post mortem (Byrne et al., 2016; Charlier et al., 2008) and using faecal egg counts (Chaparro et al., 2016). Some caution is needed in interpreting very high antibody levels because there are many other factors that may affect them,
particularly in milk. However, fluke antibody levels have been shown to be negatively associated with milk yield (Charlier et al., 2005; Mezo et al., 2011), so provide an indication of the relative effect that the parasite is having on the cow’s health. Animals with high and low parasite burdens may not be affected in the same way, and a quantitative assessment of antibody levels may give an indication of this, which is potentially useful in interpreting the results of co-infection studies.
The distribution was significantly different between the milk and serum tested cattle due to the higher maximum values seen in the milk test results, although the median values were very similar for both tests. The concentration of antibodies in serum is much higher than in milk, and this is reflected in the different dilutions used for the samples in the ELISA, so the results are not directly comparable. In addition, antibody levels in milk vary depending on the time of lactation and milk yield, which may contribute to the extreme high PP values. The finding that the bulk milk tank ELISA PP was higher than expected given the individual serum ELISA results was interesting. During the development of the bulk milk tank ELISA it was compared against serum ELISA results and faecal egg counts (Salimi-Bejestani et al., 2005a). We did not have paired milk and serum samples or paired milk and faecal samples for any of the farms in this study, but comparison with the faecal egg count data available from one of the study farms indicates that the serum ELISA has high sensitivity. It is known that the specificity of both of the milk ELISAs are lower than that of the serum ELISA (80- 88% compared to 96%), whilst the sensitivity of all three tests is similar (Salimi-Bejestani et al., 2007, 2005a). The larger range of PP values and more extreme high values seen in individuals from the milk tested farms will affect the bulk milk tank result and are likely to
36
lead to some false positives, since a positive herd result is interpreted as more than 25% of the herd being infected.
For the purposes of screening herds, all except one herd fell into the correct
positive/negative category. For epidemiological purposes, if a single test is used for all study subjects, the slight difference between the ELISA results should not introduce bias, however it does highlight the importance of being aware of potential differences of performance between tests.
The pattern of seasonality seen in condemnations of cattle livers at slaughter shows a peak in around January each year, with a trough in July (Skuce and Zadoks, 2013), and juvenile fluke are found only in autumn, indicating that this is when infections are occurring
(Charlier et al., 2008). The peak seen here is a little earlier, and this could be partly because the ELISA detects infections from two weeks post infection (Salimi-Bejestani et al., 2005b), before the fluke become large enough to cause condemnation of the liver, which may not occur until 1-2 months post infection.
The seasonal variability seen in ELISA PP in the current study is relatively small, with a difference between the predicted peak and trough of around 10 PP. This may reflect the lag period between decrease in liver fluke numbers (occurring due to treatment and fluke deaths without re-infection) and decrease in antibody levels, which remain elevated for some weeks after cure, by which time some animals may have become re-infected (Skuce and Zadoks, 2013). In high fluke areas, practitioners have observed that all grazing animals have a constantly elevated antibody level as a result of ongoing exposure, regardless of current infection status (Skuce and Zadoks, 2013). In addition there is considerable noise in the data due to the variation between individuals which may depend on other factors such as production stresses, food sources, grazing rotations and flukicide treatments.
Fluke antibody levels showed an increase with age, which approached significance. Other studies report increasing fluke burdens with age (Chaparro et al., 2016; Gonzalez-Lanza et al., 1989), and this is likely to reflect ongoing exposure without the development of protective immunity (Torgerson and Claxton, 1999).
Test performance and differences between tests, differing ages of cattle within a herd, the time of year that samples are collected and the variation in the effects of fluke infection between individuals should all be considered when interpreting the results of
37