CLINICAL PATHOLOGICAL MONITORING OF DAIRY COWS IN THE TRANSITION PERIOD

Ideally, due to the variation in methods, procedures, sampling sites and population of animals sampled among different laboratories, each diagnostic laboratory should establish a set of reference intervals for all the parameters and all the species analysed. Even if the transference of existing RIs may be possible in some circumstances, this may be not applicable when all the information concerning possible sources of variation are not available. In accordance to the guidelines recently published by the American Society for Veterinary Clinical Pathology (ASVCP) (Friedrichs et al., 2012), reference intervals for healthy high producing dairy cows around 3 and 30 days after calving were established as shown in the study presented in chapter 5. The adoption of separated RIs for these two different moments of lactation was mandatory due to the remarkable metabolic, hormonal, and immune changes occurring during the post-partum period but was also supported by the statistical analysis as shown in the study. The statistical analysis confirmed also some known differences concerning the variation occurring in the early post-partum period, such as the higher degree of negative energy balance (higher NEFA and BOHB), increased inflammatory status (higher ceruloplasmin and haptoglobin, and lower paraoxonase-1), and the increased oxidative stress (higher dROMs) near parturition. The establishment of RIs may be based on an a priori study, characterized by the determination of inclusion and exclusion criteria preceding the sampling or, as happened in this case, through the selection of data from an existing database, with an a posteriori method (Friedrichs et al. 2012). In the present case, the availability of a database derived from a large field study allowed us to select a sufficient number of reference values (results obtained from a selected reference individuals selected according to the inclusion and exclusion criteria) belonging to cows that were clinically healthy for the whole lactation period and that had production and fertility consistent with the normal standard of dairy herds of our geographical area. This allowed also to reduce the variability of results, leading to the establishment of narrower reference intervals compared to the previously adopted reference intervals in many cases. According to the aims, in the same work, the possible effects of herd, days of sampling and parity were evaluated, evidencing only rare variations mainly regarding the samples obtained 30 days post-partum. Even if the low number of samples in each subgroup did not allow to establish specific RIs for the affected variables, some important information can be derived from these results. In particular, the differences observed among the herds one month after parturition highlighted that during a period less influenced by huge metabolic and hormonal variations as occurs one month after parturition, also slight differences in dietary and management strategy may reflect in changes of some metabolic and inflammatory variables, with a higher magnitude compared to the peri-parturient period. Moreover, the metabolic, hormonal, and inflammatory variations occurring during transition, point out the importance to find possible markers that early indicate the presence of a subclinical condition in cows. Based on the published literature available, this is the first time that RIs for the main acute phase proteins (APPs) in cows (paraoxonase-1, haptoglobin, ceruloplasmin) (Giordano et al., 2004; Eckersall and Bell, 2010) and for some markers of oxidative stress (dROM and thiol groups) were provided. APPs are promising early markers of inflammation due to their rapid increase or decrease during this condition (Bertoni et al., 2008; Huzzey et al., 2011). Markers of oxidative stress are of relatively recent adoption in dairy monitoring but, as stated in the general introduction of this thesis, their role may be crucial to understand the relationship between metabolism and inflammation during the transition period (Sordillo and Mavangira, 2014). The RIs established in the present thesis were generated for laboratory methods that are commonly available in routine practice. Commonly applied biochemical parameters, but also recently adopted inflammatory and oxidative parameters, were all performed on automated biochemistry analyzers, helping the standardization of the procedures and reducing the cost of analyses, a primary issue for dairy industry. In practice, these RIs may be useful for practitioners to interpret results from cows in different lactation stages, especially in our geographical area. Moreover, applying a transference method (schematically represented in figure 8.1), every laboratory could adopt these RIs, provided that the laboratory’s animal patient population and laboratory methods of the ‘receiving’ laboratory are appropriate (Friedricks et al., 2012). Concerning the second aim of the thesis, even if a prognostic role of the investigated parameters in predicting the occurrence of retained placenta (RP) was not completely confirmed (based on the low number of affected cows enrolled in the study), some peculiarities in the biochemical, hematological and inflammatory patterns associated with the peri-partum in dairy cows with and without retained placenta had been shown in chapter 6. Concerning differences between healthy and diseased animals, the combined results from the two works evidenced a lack in peripheral neutrophils increase and lower neutrophil and monocyte concentrations at parturition in cows that retained placenta compared to not affected cows. Thus, although a similar pattern of NEB was present in the two groups, as shown by the presence of metabolic parameters within reference intervals before and soon after parturition in both the groups, a lower mass of phagocytes was present at parturition in blood of cows that subsequently developed RP. This deficiency occurs when one of the mechanisms responsible of the disruption of the feto-maternal link that involves the local presence of these cellular elements should be working efficiently (Beagley et al., 2010). In addition to the alteration of the normal leukocyte functions, found associated with retained placenta in different studies (Gunnink, 1984a, Gunnink, 1984b, Gunnink, 1984c, Gunnink, 1984d; Kimura et al., 2002b), the insufficient number of peripheral leukocytes may reflect a decreased availability of cells for migration into the endometrium. Due to the primary role of the aspecific immune response during transition, attempts to increase the number and functionality of leukocytes at parturition, in order to improve the ability of the cow to prevent clinical production diseases are in course (Kimura et al., 2014). It is well known that disease prevention requires the adoption of a multidisciplinary team approach involving the farmer, the veterinarian, the nutritional and animal breeding consultants (Mulligan and Doherty, 2008). The present thesis highlights the important role that the veterinary clinical pathologist may have in the strategic prevention and monitoring of dairy herd diseases. According to this topic, the study reported in chapter 7 confirmed the utility of the monitoring of blood level of specific metabolic parameters to identify cows at increased risk of disease. In this study, the use of cut-off values for NEFA and BOHB was applied in conjunction with clinical signs in order to identify animals requiring intervention. The absence of significant differences of these metabolites across time in both groups with low and high incidence of diseases, once more highlights the limited power of the adoption of thresholds to identify animals requiring intervention to prevent production diseases. Different thresholds, also called decision limits, concerning NEFA and BOHB have been established experimentally and by consensus, in order to discriminate between individuals or groups of cows with and without subclinical diseases (Opsomer, 2015). However, the analytical variability among methods and instruments used in different laboratories become even more relevant when these thresholds are applied by practitioners which often use cow-side tests that are characterized by lower sensitivity and specificity compared to the gold standards (Oetzel, 2004). The execution of laboratory tests in dairy herds is often hampered by economical and practical factors that limit the monitoring to only few parameters (namely NEFA and BOHB). Based on the results achieved in this thesis, it could be suggested that, in order to find possible subclinical conditions associated to metabolic, immune and oxidative disturbances during the transition period, a wider panel of analytes, from a consistent number of the herd individuals, should be applied. Moreover, each animal should be considered individually, taking in mind the possible sources of variability deriving from the effect of parity, distance from parturition, season of production, and management of the herd. The combination of this knowledge with those derived from clinicians, nutritionists and farmers is the gold standard in the final monitoring of the health of dairy cows. In conclusion, the generation of RIs specific for 3±1 days after parturition, characterized by the start of lactation, and 30±3 days from parturition, when dairy cow metabolism is reaching a more stable phase, is of primary importance in the interpretation of laboratory results, since the application of non specific RIs may lead to the incorrect classification of animals as diseased or healthy, in absence of clear clinical or productive issues. All these considerations need to be kept in mind for the proper use of laboratory data in the management of the dairy herd.