Subclinical Endometritis in Dairy Cattle Subclinical Endometritis in Dairy Cattle

Subclinical endometritis is recognized as a cause of poor reproductive performance in dairy cows. Inflammation of the endometrium persisting after postpartum uterine invo lution has been related with prolonged calving-conception intervals and low fertility in dairy cows. The subclinical nature of this condition makes it necessary in the use of endometrial cytology or biopsy for diagnosing it. There are some controversies among authors in relation to the postpartum period from which a physiological endometrial inflammation should be considered a pathological subclinical endometritis. Therefore, depending on the sampling period after calving, different studies establish a different degree of polymorphonuclear leukocyte infiltration as cutoff point to diagnose subclini - cal endometritis. Controversies also exist regarding the pathogenesis of the disease and its consequences on the fertility of dairy cattle. The aim of this chapter was to review the current knowledge on this uterine pathology.


Introduction
One of the main factors affecting reproductive performance of dairy cattle is postpartum uterine disease. Metritis and endometritis have been associated with delays in restarting ovarian activity postpartum, prolonged intervals from calving to first service, increased number of days open, decreased conception rates, and increased culling rates [1][2][3][4][5]. Affected animals are easily identified when they show clinical signs indicative of uterine disease. Though symptoms of systemic illness are often absent, a purulent or mucopurulent vaginal discharge warrants further investigation, and therefore, clinical metritis and endometritis rarely remain undiagnosed.
Fourteen years ago, Kasimanickam et al. [6] found that many clinically normal postpartum cows had subclinical endometritis (SE). Those authors evaluated endometrial cytologies collected from 228 healthy cows at 21-33 days postpartum and related the cytological findings with the subsequent reproductive performance of cows. They used a receiver/response operating characteristic (ROC) curve to determine a threshold percentage of polymorphonuclear leukocytes (PMN%) in the cytological smears above which fertility was significantly reduced, and therefore, subclinical endometritis was diagnosed based on PMN% threshold. Since that pioneer work, many other studies have investigated the etiology, prevalence, and impact on reproduction of SE in dairy cows.

Etiopathogenesis
Subclinical endometritis is the inflammation of the endometrium without clinical signs and often without evidence of infection [7][8][9]. Alteration of the inflammatory response postpartum could be at the origin of this condition.
There is no doubt that uterine pathogens may negatively affect reproduction both by causing direct endometrial damage and by producing toxins [10,11]. Bacterial endotoxins are known to have numerous effects on reproduction: (a) they may affect estradiol and progesterone secretion and alter follicular growth and the normal development of the corpus luteum [10-12], (b) may interfere with LH production and cause ovulation failure [13,14], (c) may increase PGE 2 secretion and prolong the life span of corpus luteum [15], and (d) may induce embryo mortality [16].
In cows with metritis and clinical endometritis, recognized pathogens such as E. coli, Trueperella pyogenes, Fusobacterium necrophorum, or Prevotella spp. are commonly isolated from the uterus [17]. In the case of SE, in contrast, several studies [7-9] showed that bacterial populations isolated from the uterus of cows diagnosed with SE did not differ from those of healthy cows. Prunner et al. [18] found that presence of Trueperella pyogenes in the uterus postpartum was a risk factor for development of clinical endometritis, but neither Trueperella pyogenes nor E. coli were associated with SE. Results of the cited studies suggest that common pathogens associated with metritis and clinical endometritis do not have a significant role in the SE pathogenesis. It has been suggested that SE may be a response to unspecific uterine infections [19] or a prolonged inflammatory process that persists after bacterial elimination.
In several studies, cows with clinical and subclinical endometritis were shown to have increased endometrial mRNA expression and elevated serum concentrations of pro-inflammatory mediators as compared with healthy cows [20][21][22][23][24]. Situations of prolonged inflammation after elimination of bacterial contamination may occur when an exacerbated production of eicosanoids concurs with a low production of anti-inflammatory substances, originating a delayed restoration of homeostasis in the affected tissues [25]. It has also been suggested that an unbalanced production of pro-inflammatory/anti-inflammatory cytokines during the first week postpartum could play a determinant role in the subsequent development of SE. A high ratio of pro-inflammatory/anti-inflammatory cytokines during the first week postpartum could lead to an excessive inflammatory response [26], whereas a low ratio of pro-inflammatory/anti-inflammatory cytokines might impair activation of inflammation and clearance of bacteria and lead to development of endometritis [22,27].
On the other hand, diet fat levels and the type of fatty acids present in diet may affect cellular immune function [28]. Linoleic acid-enriched diets fed to dairy cows during the transition period [29] induced a pro-inflammatory status during the first week postpartum. Several studies have demonstrated that excess of adipose tissue and high serum concentrations of non-esterified fatty acids constitute risk factors for postpartum pro-inflammatory diseases in dairy cows, such as metritis or mastitis [30][31][32]. Innate immune response is activated when an aggressor agent is recognized by toll-like receptors (TLR). Different aggressor agents are recognized by specific TLR, which might also be activated by certain molecules in the absence of aggressor agents. For instance, lipopolysaccharides present in the cell wall of gram-negative bacteria are recognized by TLR4, which may also be activated by some fatty acids (lauric, palmitic, and oleic) [33]. Thus, an inflammatory response might be induced without the existence of infection.
In addition, the oxidative stress may contribute to an abnormal inflammatory response during postpartum [33,34]. Increase of oxygen metabolism during postpartum would increase ROS production rate [33,35]. Studies carried out in bovine endothelial cells evidenced that oxidative stress increased lipid hydroperoxide formation which enhanced a pro-inflammatory phenotype of these cells [36][37][38].
Independently of the cause of inflammation, the inflammatory status of the endometrium may have a major impact on reproduction. A direct negative effect of SE on embryo quality and survival has already been described [39,40], which would affect conception rates. In addition, results from various studies suggest that SE may be associated to altered patterns of prostaglandin E 2 and F 2α synthesis [41,42] which could compromise luteal function and pregnancy.
On the other hand, certain cytokines are known to play essential roles on the physiological regulation of ovarian function [43]. Cytokines are involved in regulation of follicular growth, ovulation, luteal formation, and regression [44,45]. Inflammatory mediators, such as cytokines released in SE, may perturb this regulatory function.

Diagnosis
Cows with subclinical endometritis, by definition, do not show any clinical sign of endometritis, and therefore, the diagnosis of this condition requires the use of endometrial cytology, biopsy, or any other method able to evidence the presence of endometrial inflammation.
Ultrasonography has been used as a method to diagnose SE based on the presence of intrauterine fluid and on the evaluation of uterine diameter. A small amount of fluid in the uterine lumen and/or thickened uterine walls can be considered signs of endometrial inflammation. However, in various studies ultrasound was found to be less sensitive than endometrial cytology [6, 46,47] for SE diagnosis. Presence of intrauterine fluid and a thick uterine mucosa may be normal findings in physiological situations such as estrus or early postpartum [48], and perhaps the evaluation of fluid characteristics could improve the sensitivity of ultrasound diagnosis [49]. Mariño et al.
[9] found a significant relationship between presence of abnormal intrauterine fluid and SE diagnosed by biopsy but not by cytology.
Doppler ultrasonography might be useful for the diagnosis of endometritis in cattle, but it is still an unexplored tool. Debertolis et al. [50] found significantly increased blood flow in uterine arteries of cows to which acute endometritis had been experimentally induced. Whether patterns of vascular flow may differ between healthy uterus and those with SE still has to be investigated.
Endometrial cytology is considered the most reliable method for the diagnosis of SE [46], and therefore, it is the one most frequently used. Samples for cytology can be obtained by two main techniques, cytobrush and uterine lavage.
The cytobrush technique consists of connecting a cytobrush to the plunger of an insemination catheter [6] and, protected by the catheter, introducing it into the uterus as for doing artificial insemination. Inside the uterus, the cytobrush is pushed out of the catheter, gently rotated against the uterine wall, guarded back inside the catheter, and removed from the uterus. The brush is rolled onto a microscopic slide and stained. Cytology samples can be obtained from the uterine body or from one of the uterine horns. Mariño et al.
[9] compared cytology and biopsy findings between the two horns of 100 bovine uteri collected postmortem and observed that samples collected from the left horn were more representative of both uterine horns than those collected from the right one.
The uterine lavage technique consists of infusing sterile saline solution into the uterus with a catheter, gently massaging the uterus to allow fluid distribution within the lumen, and recovering some of the fluid by aspiration using the same catheter. The collected fluid is centrifuged, the supernatant discarded, and the sediment smeared onto a microscopic slide. Regardless of the collection technique, cytological smears are fixed and stained using conventional stains (e.g., Diff-Quick).
Kasimanickam et al. [51] did a comparative study of the two sampling techniques and concluded that cytobrush had some advantages over uterine lavage: it was less time-consuming, was easier to perform independently of the uterine size, did not produce endometrial irritation, and induced lower degree of cell structure distortion and lower presence of erythrocytes. One disadvantage of cytobrush is that the sample is collected from a specific area of the endometrium, whereas uterine lavage provides cells from the whole endometrial surface.
Recently, Pascottini [52] described a new method for sample collection that consisted of using a paper tape rolled around the top of an insemination catheter. With this method the author observed less contamination with erythrocytes and a better preserved structure of epithelial cells than when using cytobrush. Moreover, this system would allow taking a sample for cytology at the same time of doing insemination.
Concerning the threshold used in different studies for the diagnosis of SE, the cutoff PMN% reported by the different authors has varied between 4 and 25% ( Table 1) depending on the postpartum period at which the diagnosis was done.

Cytobrush
Week 3-5 Week 5-7 ≥7 week PMN% Uterine contamination at parturition or in the following days is unavoidable and normal, with 80-100% of animals having bacteria in the uterine lumen in the first 2 weeks postpartum [17]. Uterine contamination elicits neutrophil migration from peripheral blood to the uterine lumen and the subsequent phagocytosis of contaminating organisms by neutrophils. Prunner et al.
[18] observed that uterine bacterial growth density increased from calving to 15 days postpartum and decreased from day 21 onwards and the PMN% in cytological samples decreased from calving to day 9, then increased around days 15-21, and decreased thereafter, but at each sampling period, the proportion of PMN strongly depended on bacterial counts. Kasimanickam  It has been suggested that the stage of the estrous cycle might have an effect on the proportion of PMN present in the cytology and, therefore, on the diagnosis of SE. During the follicular phase of the estrous cycle, there is an increased infiltration of PMN in the endometrium elicited under estrogenic influence [65]. Several studies [9, 52,57] have found that the PMN% in cytological samples taken with cytobrush was not affected by the stage of the estrous cycle. However, when the SE diagnosis was done by biopsy, a higher degree of inflammatory infiltration could be observed in the follicular phase of the cycle [9]. This was because cytology only detects PMN infiltration in the superficial epithelium, whereas biopsy allows identifying inflammatory cells in deeper layers of the endometrium.
Another possibility to evidence the existence of endometrial inflammation is the use of urinary test stripes, which detect the presence of leukocytes in urine [66][67][68]. The diagnosis of endometritis can be done using uterine lavage fluids, or an endometrial cytobrush can be immersed in saline solution during 30 sec and then the strip introduced in the solution for 2 sec. It is a qualitative colorimetric test that showed a variable correlation with cytology. Santos et al. [66] reported a sensitivity of 96% and specificity of 98%. However, Cheong et al. [67] observed 77% sensitivity and 52% specificity. This test is rarely used in commercial dairy farms probably because it is not specifically designed for the diagnosis of endometritis.
Uterine biopsy is commonly used in human medicine as it is considered the gold standard for evaluating the human endometrium [69]. In domestic animals, uterine biopsy has been used since the 1960s to investigate causes of infertility in mares [70], and it is a routine diagnosis method today [71]. However, in dairy cattle uterine biopsy is rarely performed by practitioners, and it is almost exclusively used for research purposes. The limited use of biopsy in clinical practice may be related with inconveniences associated to sampling time, requirement for laboratory skills, laboratory costs, and time to report [71] and also to the risk of inducing endometritis and the subsequent negative effects on fertility [72].
There are few studies using uterine biopsy for the diagnosis of SE in dairy cattle [73], and when biopsy and cytology findings were compared, the two diagnosis methods showed poor agreement [8,9,47]. The histopathological examination of biopsy samples gives detailed information about the degree of inflammation, distribution of the inflammatory infiltrate, or the lesions that may exist, whereas cytology only assesses the superficial layer of the endometrium [47]. Thus, it is not surprising that direct comparison between biopsy and cytology results showed low agreement. Evaluation criteria for SE diagnosis on biopsy samples, as for cytology, should be established based on detrimental effect on subsequent reproductive performance rather than on the presence of an arbitrary number of inflammatory cells [47].

Prevalence
The reported prevalence of SE in postpartum dairy cows has varied between 7 and 53% ( Table 1). Such disparity among studies in SE prevalence may be due to differences in (i) postpartum period in which the diagnosis was made, (ii) PMN% established as threshold above which an endometrial cytology was considered positive for SE, and (iii) the method used to take the cytological sample, i.e., cytobrush, uterine flushing, or cytotape. In general, SE prevalence tended to be higher when the sample was collected by uterine lavage, the diagnosis was made before 30 days postpartum, and the cutoff PMN% applied was >5%.

Effects of SE on productive and reproductive performance
There are some discrepancies among authors concerning the effects of SE on reproductive performance of dairy cows. Whereas some authors [59,74,75] did not find significant effects on reproduction, many other studies described a variety of negative effects on fertility ( Table 2). The disparity of results may not only be due to the different diagnosis criteria (e.g., postpartum period for SE diagnosis, threshold of PMN applied, etc.) used in the different studies but also to the numerous confounding factors that may have a negative effect on reproduction (e.g., poor heat detection, inadequate nutrition, insufficient cow comfort, old cows, poor semen quality, other diseases, etc.).
Subclinical endometritis has been related with the repeat breeder cow syndrome with controversial results. Whereas in some studies [78,79] the prevalence of SE in repeat breeder cows was reported to be close to 50%, in another study [80] the observed prevalence was lower than 15%. The PMN% thresholds used in those studies differed from 3% [78] and 5% [80] to 10% [79]. In addition to the potential effects on reproduction, endometritis may negatively affect milk production [81]. Clinical and subclinical endometritis have been related with a decrease in milk production of 0.6-1.03 kg/cow/day, reduction of milk fat and protein, and with increased somatic cell counts in milk [64,76,82]. Nevertheless, some authors [83] question these effects.

Treatment of subclinical endometritis
Antibiotics and prostaglandins F 2α , combined or individually, have constituted the most common treatment for clinical endometritis postpartum. Haimerl et al. [84] and Lefebvre and Stock [85] did a critical evaluation of the scientific literature that in the last 20 years reported the use of PGF 2α alone or combined with antibiotics for the treatment of clinical endometritis in postpartum dairy cows. Both groups of researchers concluded that there was not enough clinical evidence that using PGF 2α in endometritis postpartum had a beneficial effect. And the only antibiotic that seemed to be effective for clinical endometritis was cephapirin (a firstgeneration cephalosporin).  In the case of SE, the treatment with PGF 2α and/or antibiotics was tested only in a few studies that cannot be easily compared as included different hormonal protocols for synchronization of estrus or ovulation, animals in different postpartum periods, and different diagnosis criteria for SE. In the studies of Kasimanickam et al. [51], Galvão et al. [86], and Denis-Robichaud and Dubuc [87], intrauterine infusion of cephalosporins was tested as treatment of SE, and Kasimanickam et al. [51], Galvão et al. [88], and Lima et al. [89] tested the use of prostaglandins. Kasimanickam et al. [51] and Denis-Robichaud and Dubuc [87] obtained higher pregnancy rates at first insemination in cows treated with intrauterine cephapirin than in control cows, whereas Galvão et al. [86] did not observe any positive effect on reproduction when cows diagnosed with SE were treated with intrauterine ceftiofur infusion. Concerning the use of PGF 2α in cows with SE, Kasimanickam et al. [51] and Galvão et al. [88] observed positive effects on reproductive performance, whereas Lima et al. [89] did not find any effect. It should be pointed out that the magnitude of the positive effects observed in some of the cited studies was dependent on other factors such as existence of ovarian activity at the time of treatment [87] or body condition [88]. The scarce number of studies done so far and the different results obtained do not allow us to draw a conclusion about the efficacy of using PGF 2α and/or cephalosporins for the treatment of SE.
Because in many cases of SE there is no uterine content or positive bacterial culture, treatment with antibiotics or prostaglandins should be expected to be unsuccessful. However, there is an inflammatory response that very likely is the cause of the negative effects of SE, and therefore, the use of nonsteroidal anti-inflammatory drugs (NSAID) would be fully warranted. Priest [5] tested the use of the NSAID carprofen, three doses administered at 3-day intervals between 21 and 31 days postpartum, in cows diagnosed with SE when the cytology showed >14% PMN at 14 days postpartum. The treatment did not reduce the incidence of SE at day 42, but increased pregnancy rate as compared with untreated control cows. However, in a subsequent study [90], cows were treated with carprofen at 1 or at 3 weeks after calving, and the treatment did not improve milk production, indicators of health or reproductive performance.
Uterine lavage with sterile saline solution is a common treatment for endometrial inflammation in mares. Uterine lavage favors the elimination of inflammatory products, such as nonfunctional PMN, and induces uterine contractions that facilitate the evacuation of any content. In addition, elimination of nonfunctional PMN favors migration of new functional PMN that is able to counter the infection [91]. In bovine, potential usefulness of uterine lavage has been described in connection with treatment of repeat breeder cows, either as the only treatment or combined with prostaglandins and/or antibiotics, assuming that many repeat breeder cows may suffer chronic endometritis [92]. In cases of SE postpartum, uterine lavage with physiological saline at day 30 postpartum was associated with a reduction of the PMN% in cytological samples obtained at day 40, but its effect on reproduction was not evaluated [93].
Other protocols that have been used for the treatment of metritis or clinical endometritis, such as intrauterine infusion of dextrose [94], ozone [95], or N-acetylcysteine combined with amoxicillin and clavulanic [96], have not been tested in cows with subclinical endometritis. Nevertheless, the effect of those substances is mainly antibacterial or mucolytic, whereas in SE there is no mucopurulent secretion and, in most cases, no pathogen bacteria.

Conclusions
Subclinical endometritis is a uterine inflammation probably originated by the alteration of the inflammation regulatory mechanisms. The inflammatory status may abnormally persist after elimination of postpartum bacterial contamination, which may be associated with an unbalanced production of anti-and pro-inflammatory factors. Prevalence of subclinical endometritis in dairy farms may reflect the immune status of cows, which in turn would be indicative of the metabolic status of cows in transition and, eventually, of the nutritional management of farms.