Análisis de los suelos del centro histórico de la ciudad de Trujillo.

Abortion among the heifers that were seropositive using the N. caninum antibody ELISA were clustered in the fourth and fifth months gestation while those among seronegative heifers occurred between months 3 and 7 (Figure 3.1). Abortion among the seropositive heifers in this study occurred earlier than is typically reported in studies of abortion due to

N. caninum (Dubey and Lindsay 1996). In a similar study that followed the pregnancies of 18 naturally-infected heifers for two consecutive gestations, abortions occurred in the fifth (twice), seventh (twice) and eighth months (Stenlund et al. 1999). Paré et al. (1997) followed 254 adult cows in a herd with an abortion rate of 17.3% and found that the median survival time of fetuses was 147 days for the 30/146 N. caninum-seropositive cows that aborted, and 117.5 days in the 13/108 seronegative cows that aborted; in our study the median survival time of fetuses was 122 and 120 days of gestation for seropositive and seronegative heifers, respectively.

The risk of abortion among seropositive heifers was 0.65 (95% CI=0.42-0.87), considerably greater than in the study reported by Stenlund et al. (1999) in which the corresponding risk was 0.17 (95% CI=0-0.38). The relative risk of abortion among

N. caninum-seropositive heifers was 23.6 times greater than that of the seronegative heifers, demonstrating a very strong association between serological status for N. caninum and risk of abortion. In a study in California, congenitally-infected heifers had a 7.4-fold higher risk of abortion in their initial pregnancy than non-infected heifers (Thurmond and Hietala 1997a). Paré et al. (1996) reported a 2-fold greater risk of abortion among seropositive cows compared with seronegative cows, and another study following four Dutch dairy herds found a 2-3-fold increased risk of abortion in seropositive cows (Moen et al. 1998). Both of these studies were of multiparous cattle and it was not reported whether they were congenitally or post-natally infected. In our study, results from regular testing of the heifers for N. caninum antibodies from approximately 4 months of age suggested that those classified as seropositive for the purposes of this study were congenitally infected. We propose that such animals may have a less effective immune response, dependant on the level of fetal immunocompetence at the time of initial infection, than cattle that are infected post-natally, and the consequent risk of abortion may be significantly higher.

The results presented here provided strong evidence of an association between serological status for N. caninum and risk of abortion. However, it was not possible to ascertain the cause of abortion or to definitively conclude that N. caninum infection was responsible for any of the abortions that occurred. BVD was ruled out as a cause of abortion and there was little evidence of active BVD infection within the study group, as only three heifers seroconverted over a 12-month period. The heifers were managed as a group and pasture-fed. Immunosuppression due to mycotoxins, which has been hypothesised to be associated with abortion due to N. caninum in some herds (Bartels et al. 1999), was considered unlikely in this situation.

Changes in antibody titre in the seropositive heifers suggests either recrudescence of infection or new exposure to the parasite during pregnancy occurred. New exposure would have been a risk for all heifers, resulting in more abortions in seronegative heifers or at least an increase in the number of seropositive heifers, so this is unlikely to have occurred. None of the seronegative heifers that aborted had a detectable IFAT titre (the lowest dilution tested was 1:200) during the study nor was seroconversion evident from ELISA test results (Table 3.1), suggesting that abortions that occurred amongst these heifers were not due to N. caninum.

One non-aborting heifer may have been misclassified as seronegative in this study, as all previous and subsequent ELISA results for that animal were positive. Inclusion of this heifer in the seropositive group would reduce the relative risk of abortion amongst seropositive vs seronegative heifers to 22.2 (95% CI=7.9-62.3). A further 16 heifers that were classified as seronegative had sporadic positive ELISA results in their first 29 months of life, evident from samples collected before, during and after this study, and none of those heifers aborted. Of these occasionally seropositive animals, 7/16 had two positive ELISA results and for four of these animals the positive results occurred in consecutive samples, suggesting that some horizontal transmission or recrudescence of latent infection occurred.

Cox et al. (1998) and Dubey et al. (1996c) reported that N. caninum IFAT titre did not always rise to high levels during abortion outbreaks. It has also been observed that IFAT titres can fall from as high as 1:4,000 to 1:200 within 2 months (Reichel and Drake 1996).

In most abortion outbreaks, there has not been the opportunity for serial blood collection prior to the event so the temporal pattern of change in titres has not been elucidated as it has been in the current study.

The wide variability in the peak IFAT titres of affected animals and the consistently rapid decline in circulating antibody levels (Figure 3.2) illustrates the difficulty of diagnosing N. caninum as a cause of abortion, even when detected close to the time of occurrence. If a loss of pregnancy is detected some months after it has occurred,

N. caninum serology will be of little diagnostic value, as a positive titre only indicates previous infection and a negative result does not rule-out N. caninum as the cause. Attributing cause is not possible in an individual animal but consideration of serological results among groups of aborting and non-aborting cattle can be useful when investigating herd outbreaks (Thurmond and Hietala 1999). The detection of high titres in non-aborting cattle (Figure 3.3) further confuses the diagnostic picture and supports the use of herd serology, not individual animal samples, where this is possible.

IFAT is widely regarded as the ‘gold standard’ when comparing serological tests for the diagnosis of N. caninum infection in cattle (Reichel and Pfeiffer 2002), although this was recently challenged by Frössling et al. (2003), as the IFAT had a lower sensitivity than an ELISA in that study. The IFAT is often conducted at an initial dilution of 1:200. The fact that 2/4 samples from a seropositive heifer (H65; Figure 3.2) taken around the time of abortion in our study were negative on IFAT, while all ELISA results for this heifer had an S/P ratio >0.8, suggests that the standard IFAT testing conducted for commercial purposes may lack sensitivity, particularly in cases where the immunological challenge may have occurred several months earlier. However, it is possible that this heifer may have aborted from another cause. An alternative testing regime using an ELISA, and reporting quantitative results in the form of an S/P ratio accompanied with a guide to interpretation, should be considered (Álvarez-García et al. 2003).

The increase in IFAT titres among most of the 11 aborting and seropositive heifers suggests a recrudescence of infection may have occurred in these animals prior to abortion. The increase in titre in 4 / 6 seropositive but non-aborting heifers (Figure 3.3) indicates that these heifers were also immunologically re-exposed to N. caninum antigen. Stenlund et

al. (1999) followed 18 naturally-infected heifers through their first two pregnancies and found a consistent increase in antibody titre at 5-6 months gestation among both aborting and non-aborting cattle, that then decreased prior to parturition. In our study, the peak in IFAT titre tended to occur later in gestation than that. The time of the increase in IFAT titre evident in our study did not vary significantly between aborting and non-aborting heifers, but this comparison was only made between group sizes of 10 and 5, respectively, which provides little power for comparison. These results support the finding of Williams et al. (2000) that the time of recrudescence during pregnancy determines fetal survival. This may be a result of fetal immunocompetence or the differing effects of Th1- vs Th2- type immunological responses in pregnant animals at different stages of gestation.

In another study, nine cows that were naturally and persistently infected with N. caninum

were housed and sampled intensively during pregnancy to monitor abortion and serological responses (Guy et al. 2001). A marked increase in maternal antibody occurred in one cow at 17 weeks gestation and a fetus was aborted. Five cows produced congenitally-infected calves and had an increase in maternal antibody between Weeks 22 and 36 gestation. The remaining three cows showed no changes in maternal antibody and did not produce congenitally infected calves.

Infection of a pregnant cow with N. caninum will activate the Th1-type response and sway the balance of CMI away from the Th2-type response which protects the pregnancy. The Th1-type cytokines can have a direct embryotoxic effect and damage the placental trophoblast (Raghupathy 1997). In cattle that have a latent N. caninum infection, the significant down-regulation of specific cell proliferation and IFN-γ responses that occur around mid-gestation may be a factor contributing to recrudescence in persistently-infected animals (Innes et al. 2002).

Progesterone is known to bias a T-cell response towards a Th2 phenotype (Piccinni et al. 1995) and the presence of prostaglandin E2 will bias the priming of naïve T-cells

towards a Th2 phenotype (Kalinski et al. 1997), so fluctuations in maternal hormone levels may also interact to regulate immune response during pregnancy. Oestrogen concentrations in plasma increase in pregnant cows at approximately the fourth month of gestation (Hoffman et al. 1997), and some studies suggest that this suppresses CMI and

enhances the formation of systemic antibodies (Styrt and Sugarman 1991). Thus, the increased maternal antibody titre could possibly reflect either an increased release of parasites from tissue cysts due to suppressed immunity, a hormone-triggered antibody production, or both. Alternatively, the rise in antibody titre may be due to antigenic stimulation from an N. caninum-infected fetus.

Under-reporting of abortions during the first trimester is likely, regardless of cause, as it is less likely that an expelled fetus would be noticed, particularly among heifers under conditions of pastoral management. Regular pregnancy testing of animals confirmed as being pregnant is most useful for diagnosing abortion and has been used in other studies (López-Gatius et al. 2004). Pregnancy diagnosis before about 160 days gestation (Figure 3.1) in our group of heifers would not have correlated well with calving results and so early pregnancy diagnosis may not be accurate in predicting the number of heifers to calve in herds where animals are seropositive. Early pregnancy diagnosis does, however, provide a basis for diagnosing early abortion and allows the economic benefit of removal of non- pregnant animals.

It is likely that the pattern of abortion due to N. caninum differs according to whether the infection is endemic or epidemic. In congenitally-infected heifers, previous exposure (probably in utero) primes the immune system to respond and a heifer’s immune response to a re-exposure or recrudescence of infection causes the abortion (Quinn et al. 2002). In naïve animals, abortions occur over a wider range of gestational ages according to time of exposure. It is unknown whether the parasite is lethal to the fetus during the phase of tachyzoite proliferation, through damage to the fetus or the placenta, or whether the maternal cell-mediated immune response is detrimental to the maintenance of pregnancy. Categorising cows simply as seropositive or seronegative may be misleading, as variation in antibody levels may provide useful information; some animals have high antibody levels for their entire life-time, some have detectable levels for only a short time period, and some alternate between seropositive and seronegative states repeatedly over their life-time. These differences in serological state may be associated with different reproductive outcomes, the highest rates of abortion and vertical transmission occurring among cows that are consistently seropositive. Cattle challenged with N. caninum 6 weeks prior to mating were protected against vertical transmission when a second challenge occurred on

Day 140 of gestation (Innes et al. 2001a). This suggests that there is a difference in immune response and ability to transmit the infection in utero between cattle that are infected as adults and those that are congenitally infected.

The very high (0.61 or 0.65 depending on the inclusion of one heifer) risk of abortion during a first pregnancy in heifers that were seropositive for N. caninum in this study suggests that testing and culling of seropositive heifers may be indicated on some farms. However, to be effective, a test with high sensitivity and specificity is required and, as has been discussed, interpretations of serological test results vary. If the serological status of the dam is also known then this is a useful preditor of the serological status for N. caninum

of the offspring (Anderson et al. 1997) and may be considered in the decision-making. The apparent reliability of pre-colostral testing to determine the serological status of vertically- infected potential replacement heifers before significant investment is made raising them commend this as an approach to reducing the number of heifers within a herd with a propensity to abort.